Report No. 34580-IN India Inclusive Growth and Service Delivery: Building on India's Success Development Policy Review June 15, 2006 Document of the World Bank Report Team and Acknowledgments This Report was preparedby a Team ledby RinkuMurgai, Lant Pritchett, and Marina Wes, underthe overall guidance of Sadiq Ahmed, Michael Carter, Ijaz Nabi and Kapil Kapoor. The peer reviewerswere GauravDatt, Manuela Ferro, Ricardo Hausmann, Ajay Shah, and Michael Walton. The team receivedhelphl feedback from Shankar Acharya, Suman Bery, andN.C. Saxena, and from key government officials. The DPR Team received valuable comments on the draft report from Government o f India.A large number of comments were also received from other Bank colleagues. Major contributors to the report were Brian Pinto, Farah Zahir, and Gaobo Pang on fiscal policy, Priya Basu on the investmentclimate for industryand services and the financial sector, Dina Umali-Deininger on agriculture and rural development, Clive Harris on infrastructure, Ahmad Ahsan on labor markets, Geeta Sethi and Yamini Aiyar on decentralization, Philip O'Keefe on Social Protection, Sumir La1on the political economy o freform, Christine Allison on humandevelopment, Maitreyi Das on caste and gender, R.RMohanon tribals, Salimah Samji onrural livelihoods, Joel Turkewitz on procurement and Smita Misra on rural water supply. Arindam Nandiand Arpita Chakraborty provided research assistance. A backgroundpaper on market integration was prepared by ICRIER inNew Delhi. Bhaskar Naidu providedstatistical support. Administrative support was providedby Shiny Jaison and Rita Soni. This report was editedby Bruce Ross-Larson and his team at Communications Development Inc., Washington DC. i TABLE OF CONTENTS Introduction................................................................................................................................... x Chapter 1 Overview of India's development progress and pressing challenges . .................... 1 India's growth patterns: a star performer................................................................. 2 Recent growth experience .............................................................................. 5 Cluesfrom thepatterns of growth ................................................................. 9 Core public services: implementationi s everything.............................................. 10 Urban and rural water supply ..................................................................... 13 Health........................................................................................................... 14 Elementary education .................................................................................. 16 A booming economy with growing gaps............................................................... 17 18 Differences across househoIds and individua1s........................................... Differences across states and regions.......................................................... 24 Guide to the rest ofthe report ................................................................................ Introduction to Part I:Improved accountability for better services 29 Chapter 2. Service Delivery: A Diagnostic ............................................................................... ....................................27 31 A framework for understanding service delivery outcomes.................................. 31 The state o f service delivery and diagnosis ........................................................... 36 Service delivery: access and satisfaction ..................................................... 36 State of service delivery: providers.............................................................. 38 State of service delivery: assets ................................................................... 41 Improvingservice deliveryperformance: more than business as usual ................44 Increased budgets: business as usual with more money.............................. 46 Greater electoral competition...................................................................... 49 Chapter 3 Improving internal accountability: . Piecemeal reform and "recapture" ............................................................. ....................................................................... 50 52 Strengthening accountability for service delivery: Three Rs of information ........53 Improved compact: unbundlingthe roles o f government ...................................... 55 Usingtechnology to strengthen services byprocess engineering.......................... 59 Handling the political economy o f internal and administrativereform.................61 Chapter 4 Local governments and service delivery inIndia . ................................................. 64 The three F's of decentralization: funds. functions. and functionaries ..................65 65 Functions: blurred delegation ..................................................................... fund^............................................................................................................ 68 Functionaries ............................................................................................... 69 Status of the three F's inPRIdecentralization....................................................... 69 Aligningthe three F's for more effective services from local governments .........70 Functions...................................................................................................... 70 Funds............................................................................................................ 73 Capacity building for accounting and accountability with social inclusion ..........75 Functionaries ............................................................................................... 74 Accounting: capacity buildingfor financial management ........................... 76 Strengthening accountability ....................................................................... 77 Strengthening the "synapse" between communities and governments for better local governance .................................................................................................... 78 ... 111 Reservations ................................................................................................. 78 Local governance and user and community groups .................................... 80 82 Handlingthe transition........................................................................................... Convergence for better local governance .............................................................. . . Chapter 5: Community andPrivateSector as Providers:InfrastructureandMore A focus on infrastructureservices.......................................................................... ...........85 83 85 Three modes o f external reform: communities. contractingout. and PPPs..........87 Unbundling. delegation. autonomy........................................................................ 90 Enforceability: Subsidies and user fees ................................................................ 91 External accountability and information................................................................ 94 Political economy o f external reforms of service provision.................................. 96 The role of the center in promoting-or delaying.. reform .......................... Introductionto PartI1 101 Chapter 6.Easingconstraintsto ..............................................................................................................98 sustainrapidgrowth ......................................................... 103 Infrastructure as a bindingconstraint................................................................... 105 The fiscal deficit as a bindingconstraint ............................................................. 109 Continuingfiscal adjustment ..................................................................... 109 112 Prospects andrisks............................................................................................... Increasingpriority developmental spending ............................................. Chapter7 Acceleratorsto equalizethe benefitsof growth . ..................................................116 120 Labor laws............................................................................................................ 120 The financial sector.............................................................................................. Chapter 8.Reformingagricultureto combatruralpoverty .................................................123 126 Agricultural productivity and growth .................................................................. 129 Development pathways for growth...................................................................... 131 Intenszjkation ............................................................................................ 131 Diverszjkation............................................................................................ 132 Nonfarm linkages ....................................................................................... 132 Exit ............................................................................................................. 132 Matching agricultural strategies with production potential ................................. 132 Strategiesfor high-potential irrigated areas ............................................. 133 Strategiesfor moderate- to high-potential areas....................................... 136 Strategiesfor marginal areas .................................................................... 137 The political dynamics o f rural reform ................................................................ 139 Thepolitical rationalefor agricultural subsidies ...................................... 139 The complexity and sensitivity of reforms .................................................. 140 Chapter 9 Bringingup laggingstates . Negotiated solutions................................................................................... .....................................................................................141 143 IntegratingIndia: goods and labor ....................................................................... 143 Unleashing the industrial potential o f lagging states........................................... 145 Reducing infrastructure bottlenecks .......................................................... 147 Improving economic governance ............................................................... 150 Introducing law and order ......................................................................... 154 Improving the efficiency of land markets................................................... 154 155 Transforming the rural economy o f lagging states .............................................. Improving access toJinancefor business .................................................. 156 Upgrading rural infrastructure .................................................................. 157 iv Ruralfinance .............................................................................................. 157 Fiscal federalism and the Twelfth Finance Commission recommendations .......159 Chapter 10 Achievingequitable growth . ................................................................................ 165 Equalityo f access to markets and assets.............................................................. 166 Addressing caste and gender bias in labor markets .................................. 166 Facilitating access tofinancefor thepoor ................................................ 168 Providing social protection .................................................................................. 174 Looking aheadto a more coherent social protection system............................... 179 Improving safety nets ................................................................................. 180 Expanding social security coverage .......................................................... 181 References .................................................................................................................................. 183 StatisticalAnnexes ..................................................................................................................... 194 Tables Table 1. 1: The average pattern i s for one decade's growth stars to fall to earth inthe next. but Indiahas defiedthat trend.................................................................. 4 Table 1. 2: India's growth rate has accelerated to 7-8 percent inrecent years. drivenby the continued service sector expansion and a recovery inindustry.................. 6 Table 1. 3: While some indicators o f health care rose. the share o f children who were fully immunized. were deliveredinapublic hospital. andwere attendedby a visiting nursedeclined.................................................................................... 15 Table 1.4: Differences inincome and output among Indian states are associated with large differences inhuman development indicators o fhealth and education. 21 Table 2. 1: Three relationships inthe "long route" o f accountability. illustratedwith the case o fpolicing ............................................................................................... 35 Table 2. 2: Examples o f low learning achievement inprimary schools from around India reveals shockingly low competence................................................................ 44 Table 3. 1: A sample o f outcome indicators developed by the Ministryo f Health and FamilyWelfare reveals the devil inthe details............................................... 58 Table 6. 1: Fiscal trends inIndia. 198.5186-2004/05 ..................................................... 109 Table 6. 2: Projections o f selected variables under scenario 1...................................... 118 Table 6. 3: Projections o f selected variables under scenario 2 ...................................... 119 Table 8. 1: Average yields o f major crops inIndia and other major producing countries. 1998-2000 ..................................................................................................... 127 Table 8.2: Variation inagricultural performance, input use, and infrastructure across Indian states .................................................................................................. 129 Table 8. 3: Agricultural systems inRajasthan, by soil conditions................................. 131 Table 8. 4: Tailoring development pathways to agricultural production systems ......... 133 Table 8. 5: Policy and investment implications of a differentiated agricultural strategy Table 8. 6: Impact o f irrigation on agricultural production and employment ...............135 ....................................................................................................................... 136 V Table 9. 1: Investment performance o f India's lagging states....................................... 146 Table 9. 2: Infrastructure index ranking o f Indianstates............................................... 147 Table 9. 3: Numbero f inspections a year and required management time inIndia and comparator countries..................................................................................... 152 Table 9. 4: Aspects o f formal borrowing inAndhra Pradesh and Uttar Pradesh. ...,......159 Table 10. 1: Impact o fparticipationinself-help groups on women's attitudes and Table 10. 2: Implementationo f the BPL targetingsystem, by state, 2000..................... behavior......................................................................................................... 172 176 Table 10. 3: Coverage rates o f social insurance inIndia's organized and unorganized sectors, 2004 ................................................................................................. 178 Figures Figure 1. 1:Indiahas been inthe top 10percent o f all countries ingrowth performance since the 1980s............................................................................................... 2 Figure 1.2: An Indianat the 90th percentile o f the income distribution has much lower income thanthose at the 10thpercentile inOECD countries. ....................... 3 Figure 1.3: For everyworker inthe organizedprivate sector there are four unemployed-and the number o f organized private sectorjobs i s falling. ...7 Figure 1.4: Wage growth has been disproportionate at the top end o fthe labor market. ..8 Figure 1.5: Private investmentinIndia has been low-at least untilrecently. ................. 9 Figure 1. 6: Facedwith inadequate public services, both elites and noneliteshave developed ways to augment education, water, and health services inthe private sector................................................................................................ 12 Figure 1. 7: While many major cities indeveloping countries maintainnearly 24-hour water supply, most Indian cities have service only a few hours a day. ....... 14 Figure 1. 8: Inthe urban areas o f eight major states, more than halfo fthe children in elementary school are inprivate schools. .................................................... 17 Figure 1. 9: Growth accelerated innearly all states inthe 1980s, but gaps widened dramatically inthe 1990s............................................................................. 19 Figure 1. 10: Inequality across states o f Indiaremains low compared with developing countries, but higher than integrateddeveloped regions. ............................20 Figure 1. 11: Conditions vary widely across districts in states. The worst districts in Maharashtra have IMRhigher thanthe average o f the poorest states. ........22 Figure 1. 12: Parts o f Indiahavepoverty rates similar to those inmany Sub-Saharan African countries, while other regions o f Indiaresemblericher Latin American countries. ..................................................................................... 23 Figure1. 13: The poorer sections ofthe wealthier Indian states are very nearly as poor as those inthe poorer states.............................................................................. 24 Figure 1. 14: Growth, ifit is fast enough, can reduceincome poverty, even with inequality increasing. .................................................................................,.25 Figure 1. 15: Despite a decade o f rising inequality, Indiaremains a country with low income inequality by global standards.......................................,.................26 vi Figure2. 1: Relationshipso f accountability ....................................................................34 Figure2.2: Whilephysical access has improvedgreatly, satisfaction with services provided is very low across all sectors ........................................................ 37 Figure 2. 3: Ina nationwide survey usingunannounced visits less thanhalfo f teachers were both present and engaged inclassroom activity.................................. 38 Figure2. 4: Public PHC doctors inNew Delhiscore less well on their competence than doctors inTanzania or Indonesia................................................................. 39 Figure 2. 5: Inpoorer neighborhoods the competence and effort o f all providers i s low- but effort is PHCsis always low and "off the chart" inpoorneighborhoods ...................................................................................................................... 40 Figure2. 6: Allocation o fthe total estimated outflow dueto "retail" level corruption across sectors-the health sector i s more than a quarter o f all retail corruption.. ... .. . . ........................................................................................41 . ... Figure2. 7: Some states get more water connections for their money............................ 42 Figure2. 8: In1999-2000 public sector workers hadwages 68 percent higher than equivalent workers inthe private-formal sector-up from 48 percent higher in1993-94.................................................................................................... 47 Figure2. 9: Enormously cost effective options are available-but the systemic pressures are for more spending, not more and better services from the same spending ...................................................................................................................... 49 Figure3. 1: Increases incitizen satisfaction with services inBangalore following introduction o f a citizen report card (CRC) ................................................. 53 Figure 4. 1: Ownrevenues contribute little to Panchayat revenues inKarnataka, FY 2000-01 ............................................................................................................................. - 66 Figure4. 2: No alignment inthe devolution o fthe three F's ........................................... 70 Figure4. 3: Reservations for women and SC/STs is having some impact...................... Figure4.4: Weak formal linksbetween GPs and user groups at the village level .........79 81 Figure 5. 1: Access to pipedwater has declined insome states between 1991 and 2001 87 Figure5.2: PPPs are making a growing contribution to financing o finfrastructure in India............................................................................................................. 89 Figure5. 3: Indiahas overtakenBraziland China inattractingPPI................................ 90 Figure5. 4: Comparingthe sources of finance: Either users or taxpayers pay-but sometimes for inefficiency and shortfalls create neglect o f assets .............. 92 Figure5. 5: Most fundingprograms still adopt traditional supply driven approaches.. 100 Figure6. 1: Infrastructure stocks inIndiaand China, 1998 and2003...........................106 Figure6. 2: Total public debt and interest paymentsinIndia and selected countries...111 Figure7. 1: Manufacturingemployment inIndia and selected Southeast Asian countries, 1980s.......................................................................................................... 122 Figure8.1 : Labor productivity and share o fthe labor force inagriculture inIndian 1 126 Figure8. 2: Food grain price subsidies, by state, 2001/02 ............................................ 128 Figure8. 3: Agricultural production systems: a nationalperspective ........................... 130 vii Figure 9. 1: Dispersion o fprices across markets inIndia. by commodity. 1994 and 2004 144 Figure 9. 2: Growth incasual wages inIndianstates. 1983-2000 ................................ ......................................................................................................................................... 144 Figure 9.3: Decadal growth o fper capita gross state domestic product and population in India. 1961-99 ........................................................................................... 145 Figure 9.4: Access to power inIndia's lagging states .................................................. Figure9. 5: Percentage o fhabitations not connected byroads. by Indian state............148 149 Figure9.6: Percentage ofthepopulation with accessto sewerage facilities. Figure 9. 7: Cost and time requiredto start a business inIndia's lagging states...........149 byIndian state............................................................................................ Figure9.8: Cost andtime requiredto register a property inIndia's lagging states......149 151 Figure9.9: Time requiredto enforce a contract inIndia's lagging states .................... 154 Figure 9. 10: Access to finance inIndian states. 2003.................................................... 155 Figure 9. 11: Growth incredit and industrial output inIndian states. 1991/92-1999/2000 ......................................................................................................................................... 155 Figure 10. 1:Distributiono foccupations inIndia. by caste status. 1983-2000 .............167 Figure 10 .2: Alternative paths for scaling up rural livelihood initiatives. dependingon Figure 10.3: Coverage andtargetingo fmajor social protectionprograms ...................174 the objective and mode o f empowerment .................................................. 175 Figure 10.4: Differencesinimplementation o f social protectionprograms. by state.... 176 Boxes Box 2. 1: The costs o f deferred road maintenance........................................................... 43 Box 3. 1: Launch o foutcomes budget............................................................................. 57 Box 3. 2: National Urban RenewalMission: An integrated reform effort to improve urbanservices.................................................................................................. 61 Box 3. 3: Twelve lessons from reforms ina difficult sector............................................ Box 4. 1: UrbanDecentralization and the three F's: a status review............................... 63 75 Box 4. 2: An impact evaluation of cornmunitisation inNagaland................................... 78 Box 4. 3: Strengthening accountability and encouragingpolitical inclusion inurban settings ............................................................................................................ 83 Box 5. 1:D o water subsidies inIndia reach the poor? ..................................................... 93 Box 5.2: PPPs inLatin America: an assessment............................................................. 96 Box 5. 3: Power Privatization inDelhi............................................................................ 97 Box 5.4: Delhi Jal Board Reforms.................................................................................. 97 Box 6. 1: Lessons from the 1990s.................................................................................. Box 6. 2: Other issues for growth .................................................................................. 103 105 Box 6. 3: International experience with fiscal adjustment ............................................. 114 Box 8. 1: Priorities inthe TenthFive-Year Plan for agriculture 128 Box 8.2: Issues for effective watershed management at the national, state, and local levels ............................................................................................................. 138 Box 9. 1: Natural Resources inIndia's poorer states...................................................... 152 Vlll ... Box 9. 2: State financial accountabilityissues............................................................... 160 Box 10. 1: The Indira Kranti Pratham Project inAndhra Pradesh ................................. Box 10.3:The National Rural Employment Guarantee Act ........................................... Box 10. 2: Examples ofprivate sector initiatives to improve rural livelihood...............169 172 178 Annexes Annex 9. 1: Rajasthan: closing the development gap ................................................... 162 Annex 9. 2: Bihar: toward a development strategy........................................................ 164 ix India: DevelopmentPolicy Review Introduction Ihavenopromises tomake,butIhavepromises tokeep Prime Minister Manmohan Singh, August 15 2004 A review o f India's current development challenges andpolicies couldperhaps best begin from a long-run perspective. Relative to the often dire expectations expressed at its birth, Indiahas been successfbl, and increasingly so. Indiahas maintained electoral democracy, not only holding free and fair elections for the nation's government, but now over 3 million people hold elective office inits local governments. The specter o f famines has beenbanished, absolute poverty cut by more than half, illiteracy dramatically reduced, health conditions vastly improved. India has become, inpurchasingpower parity terms, the fourth largest economy inthe world. India's economic growth since 1980has been among the most rapid-and India managed its one incipient crisis inthe early 1990s and avoidedthe catastrophic losses elsewhere. While many feared that exposing India's economy to global competition would reveal India's economic weakness, it has rather revealed strengths (often unexpected strengthsinnew areas no planner would have dreamed of). India i s increasingly taking its place on the global stage and ininternational forums as a 21'` century superpower. The generations o f politicians and policy makers who have beenpart o f leading India to where it i s today can bejustifiably proud o fthe transformation. But achievements create new challenges, and this report focuses on two ofthose: improving service delivery, particularly to the poor, through greater accountability and expanding the benefits o f rapid growth-across sectors, regions, andpeople. Part I:Increasing accountabilityfor improved service delivery While the challenge is expressed indifferent ways-"improving quality o fpublic expenditures" or "increasing institutional capacity" or "more effective implementation" or "better service delivery"-all agree that India's core publiclyprovided services, such as water supply (both rural and urban), power, education, policing, sanitation, preventive andpromotive health services, road construction, and implementation o fpoverty programs, face enormous pressures. These new challenges are rooted infive achievements: 0 Achievement. Economic progress and rising incomes means that citizens are demanding more, andhigherquality, services and have more income to seek those from the private sector ifthe public sector provision i s inadequate. o Challenge. This revolution o frising expectations for reliablepublic services means that governments are under pressure to deliver services like never before. 0 Achievement. Massive investments inasset creation success have made the goal of physical access (to a road, to a clinic, to a school, to a market) achievable. X o Challenge. The achievement ofaccesscreates the challenge o fthe "access/satisfaction" gap. People have access to the physical facilities but are largelydissatisfied with the services that are actually available as frequently staff are missing, facilities lack keyinputs, assets are not maintainedandare dysfunctional. According to one survey while 70 percent o frespondents had access to the Public Distribution System, less than 10percent were fullysatisfied with the services delivered-a finding that cuts across all services. 0 Achievement. The flourishing o f electoral democracy and its expansionto include newlypowerful social groups-has meantthe competition for control o f the public sector is intense and more and more citizens and groups feel empoweredto participate. o Challenge. This political situations createsthe temptation o fusingthe politics "jobs, contracts and subsidies" to provide patronage to narrow groups andprivate benefits to insiders rather than improved service delivery as the means o f satisfying constituencies and supporters. 0 Achievement. Government agencies have managed to accomplish logistical goals and, for the most part, are organizationally viable, so for example, the education program o f Sarva Shiksha Abhiyan i s being implementedand the target o f universal enrollment i s close to being achieved. o Challenge. These same organizations needto move to the next level o f performance inservice delivery-from quantityto quality, from physical targets to outcomes. This often requires different approaches (more demand driven and less technocratic, more focused on external accountability and pleasing the "citizen" and less on compliance with internal procedures) that are require organizations to undertake difficult changes. 0 Achievement. There has been a revolution in aspirations and expectations o f upward mobility among all segments o f Indian society. o Challenge. This empowerment creates new demands andpressures on services to be more inclusive as the increased connectivity o f rural India and the increased fluidity economically and socially means the aspirations o f all are mounting. This collection of successeshas ledto a situation inwhich the capability o fthe public sector, inits current configuration, has not keptpace with the services demanded by citizens. For instance, India, inspite o f its rhetorical commitments to universal health and universal education andwater provision i s defacto quite privatized. Only five countries inthe world have a lower portion o f health spendinginthe public sector-the proportion o fhealth spendinginthe private sector is twice as highinIndia as inthe "free market" USA. Inurban areas inmore than halfo f the major states a higher fraction of children are inprivate elementary school than inthe largely "privatized" system in Chile. Nearly all new capacity created inirrigation recently has beenprivate (e.g. tube wells), not public. EveninDelhi's water supply more than halfo fthe total cost o fwater is borne privately (as people invest assets or time incoping with the irregular water supply). In fact, many knowledgeable observers believe that the configuration o f circumstances are leading organizational and institutional quality to notjust not keep pace with economic xi progress, but to deteriorate absolutely. The interaction o f a "jobs, contracts and subsidies" pressure from politics with weakening public sector capacity, elite withdrawal from demand for publicly provided services with a "top down" input and asset creation orientation for public policy can indeedcreate a situation inwhich the public sector's ability to deliver on core responsibilitiesdeteriorates. This is not a pessimistic assessment, it is rather a realistic assessment about the magnitude o f the challenges faced (which are often created by previous success) but this Report i s optimistic about the capability of India, its citizens and government, to face and surmount these challenges. Tackling the new challenges requires confidence from past achievements but not complacency about the magnitude o fthe challenges to be faced and the Government has put institutional reforms o fthe various agencies and service providers right(1y) at the top o f its agenda. Infact, one should be less optimistic about the likelihood o f success ifone begins with a falsely rosy scenario that provides an inadequate assessment of the true magnitude of the challenges, and the deep causes of these challenges, faced inreaching the next level o f focus on the quality o f service provision. There are many successesacross states, sectors and agencies that point towards viable new approaches for creating greater accountability o fthe state and o f service providers to deliver. These approaches, while disparate inmany ways and adapted to the contextual features o fthe sector and place, tendto have five features incommon: 0 Clearer delegation of responsibility o fproviders for outputs and outcomes- moving away from merelyresponsibility to compliance with internalprocesses in the use ofinputs. 0 An unbundling o fthe roles o fgovernment betweenthe general responsibility for a sector and theproduction o fthe outputs-moving away from situations inwhich line agencies are bothumpire (responsible for setting standards, creating and disseminating information, monitoring compliance, evaluation) andplayer (responsible for day to day management o fproviders). 0 Greater autonomy to the providers (both organizational and frontline) inhow they achieve their goals and insulation from top down or narrowly political micro- management. 0 Increased external accountability which requires greater transparency andbetter flows o f information and social mobilizationlempowerment to make that information effective. 0 Greater enforceability so that citizens and communities as the direct "clients" of service providers (both public andprivate) can have greater voice (over the responsible level o f government) and choice across providers (as an effective mechanism to exercise power). Internal reforms ofpublic sector actors. While the public sector faces systemic challenges, it also shows promise. There are many examples o f institutional innovations inservice deliveryat the Center (such as the creation o fthe National HighwayAuthority of India) and inthe states ( e g Bhoomi for landrecords inKarnataka, Education Guarantee Scheme inMadhya Pradesh). There are ways these successes can be xii promoted. For instance as one example, many successes are driven by "reform champions" withinthe public sector as many examples o f successful "internal" improvements inservices happenwhen civil servants are given a clear political mandate for reform, adequate time, and autonomy. One o f the most womsome features o fthe current system is the shortening duration o ftop civil servants intheir posts as it both detracts from their capability to carry through and reflects an undesirable politicization of the civil service. Information for greater external accountability. There is now widespread recognition that ifcitizens are to create pressure for better services they needbetter information, and information that meets the three Rs: reliable, relevant, and regular. There are a variety o fpromising initiatives. The government has recently passed a Right to Information law, that creates enormous scope for citizens to know what is happening-but it is only a promise that will require support to become a vehicle for service improvements andnot merely a means for political grandstanding. The use o f survey information via "citizen report cards" that began inBangalore shows the need for persistence-while "once off' efforts can generate agitation and fleeting press attention-it i s only when people take it to heart as a regular indicator can it be used to drive reforms. There are also a variety of efforts to make information available from the bottom up through existing channels- such as the Gram Sabha or user committees invarious sectors-such as the nongovernmental organizationPratham's efforts to make information on learning performance available invillages and districts. Merelyputtingthe letter "e" infront o f something doesn't magically make it better and "e-governance" can be as muchcaught up in"jobs and contracts" as governance without an "e." Butthere are examples ofthe use o fthe occasion ofmaking functions of government "e" to undertake what would otherwise be organizationally and politically impossible fundamental business process changes-including eliminating functions and limitingthe potentially abused discretiono f government officials. Decentralization to PHs. More than a decade ago India's 73`d and 74thconstitutional amendments launched an effort to improve local governance through devolving responsibilities to democratically elected bodies heading autonomous units o f local government. But decentralization alone i s no panacea--services can get better or worse-- as with all other reforms implementationi s everything. Currently inmost states decentralization is "unbalanced" inthat political decentralization has moved ahead and India now has more than 3 million citizens serving inlocally elected bodies. Moving forward with a well-designed decentralization requires aligning the "three Fs": funds, functions, and functionaries inways that make it possible to have technically effective services with both "accounting" (the capability to provide reliable budget control and reporting) and "accountability" (the ability o f citizens and communities from the bottom up to holdelected officials andproviders responsiblefor outputs andoutcomes). This "activity mapping" of the three Fs cannot be done by slogans or ideology, as a functional activity mapping will vary from sector to sector, vary within sectors (e.g.hospitals versus aanganwadz), and even vary byfunction (e.g. asset creation versus operation), but can be informed by general principles. Two general tendencies do emerge. First, decentralization creates an opportunity to unbundle responsibilities across tiers of xiii government inorder to create checks and balances inthe interests o fthe tiers o f government-so that one tier reports on the performance of another-in order to assist citizens ingetting the informationthey need to create accountability for performance. Second, decentralization, increating new lines o fresponsibility also creates opportunities to strengthen the "demand side"-the mobilization o f communities from the bottom up to demand better performance with better information and greater scope for voice and choice. Non-stateproviders in service delivery. Public-private partnerships(PPP) canplay an increased role inthe provision o f infrastructure services, and India i s inthe position to buildon successesinthe transport and communications sectors. However, a concerted effort to buildcapacities, particularly at the state level, will be important. The role o f PPP, as another way o fpromotingbetter services, is not limitedto infrastructure. In health, ineducation, even inthe implementationo fpoverty programs there are promising ways to use the empowerment generated by allowing people to make their own choices by channeling funds to people first rather than providers. Part 11: Rapid and inclusivegrowth While Indiahas been growing very rapidly-even with some acceleration inthe last few years--international experience shows that a recipe for slow growth is complacency about pushing ahead with reforms when growth is high. Most countries with rapid growth in any one decade show marked deceleration o f growth rates inthe next. Indiahas so far been able to avoid that, as adroit response to an incipient slow-down inthe 1990s restarted another episode o frapid growth. The reformmomentum has been sustained across changes ingovernment so that the basic direction o freform continues. While there are many areas o fpolicy that require attention, and any analysis o fthe "investment climate" or o f specific sectors will reveal a long list o f desirable changes, we want to focus on the few "binding constraints"-those areas that, without attention, are the most likely to be obstacles to sustained rapid growth; relieving these constraints wouldbe a growth accelerator. India's successeshave not been shared equally andit is obvious that, particularly since the reforms o f the 1990s, some states have done muchbetter than others; even inthriving states some cities have thrivedwhile rural areas have stagnated, and, given the pattern o f economic growth, people with skills have done muchbetter than those with less education. This Report follows the framework of the World Bank's recent World Development Report on Equity and Development which emphasizes that the primary concern i s equity o f opportunity-which is a notion o f fairness andjustice-and absolute well-being-elimination o f economic and social deprivation not narrow measures o f inequality o f outcomes. The government o f India's justifiable concern with the inclusiveness o f economic growth can be addressed by focusing on expandingthe regional scope o f economic growth, expanding access to assets and thriving markets and expanding equity inthe opportunities for the next generation o f Indian citizens no matter who they are or where they live. While reforms that are growth accelerators are important, even more pressingi s the need for equalizing accelerators-actions that xiv promote more rapid growth inthose areas, sectors and groups where it is needed the most. Infrastructure, for example, has impacts on economic growth o fbothtypes. Some investments ininfrastructure are neededto maintainrapid growth-ports and airport modernization, improved highways. But infrastructure is also important to equalize growth-investments that raise productivity and farmer incomes inagriculture, infrastructure that helpjobs move to people, as well as the infrastructure that is neededto connect rural Indiawith the benefits o f a growing economy. Potential Binding Constraint: Infrastructure. A major concern for maintaining rapid growth inIndia i s the provision o f adequate infrastructure as there are some suggestions that, while the "high skill" services sector will continue to thrive, a greater ability o f India to engage inmanufacturing that is intensive insemi-skilled labor is desirable. Services are, to a large extent, either reliant on the infrastructure sector that has made the most progress (telecommunications) or able to cocoon themselves from service inadequacy. Butengagement inmanufacturing, particularly export orientedmanufacturing that typically requires integration of supply chains, or inhigher value added agriculture will require substantial improvements inbasic transport infrastructure (roads, ports, airports, railways) and inmore reliable power andwater. Butwhile everyone agrees that infrastructure i s a constraint, the question is how to addressthe problem-this report argues this is not merelya financing problem o fmobilizing additional billions but rather the emerginginfrastructure constraints require systemic reforms that not only "fix the pipes" but also "fix the institutions that fix the pipes." Indiamust invest around 3-4% more o f GDP on infrastructureto sustain growth o f around 8%, address existing gaps and meet policy-driven coverage goals. The private sector can play an important role inproviding such levels o f resources, including through PPPs, although improving implementationcapacity andthe quality o f investment will be as or more important than increasingthe quantumo f funds available. A greater emphasis must be placed on the actual delivery o f services rather than completion o f infrastructure: this will include shiftingresources to maintain existing assets andmaking service providers more accountable to consumers andtheir owners. Inall sectors reforms are underway, but these must be accelerated and difficult issues such as rationalizing user fees for services cannot be avoided. Potential Binding Constraint: Fiscal Deficit. Indiahas maintained a very highdeficit byinternational standards for many years now. This creates three concerns. 0 While for many reasons the burdeno fthe deficit and debt inIndiahavebeen surprisingly c`light'7-as Indiamaintains muchhigher deficits and debt with much lower risks o f internal or external crisis-no one has repealed the laws o f gravity. Gettingback on and staying onthe FiscalResponsibility and Budget Management Act targets for deficit reduction i s essential to avoiding increased risk o f crisis (especially if external factors such as interest rates were to turn less favorable) and associatedhigher domestic interest rates that could easily choke-off an expansion or cause a much higher fiscal burdeno f debt. xv 0 The UPA government elected in2004 has many areas inwhich they believe additional expenditures are desirable and is launching many potentially desirable initiatives: rural infrastructure (Bharat Nirman), employment (National Rural Employment Guarantee Act), education (Sarva Shiksha Abhiyan), rural health (National Rural HealthMission), urbaninfrastructure (National Urban Renewal Mission). But those new initiatives must be financed with some combination o f higher taxes (or user charges), or cutting existing expenditures. The existing deficit leaves the government nofiscal space even for highlyproductive new spending which can sustain the current rapid growth. This means a focus on raising revenues inways that donot choke offgrowthandcuttingout unproductive government expenditure to make space for better, more effective, expenditures. 0 Large deficits interact with an unreformed bankingsector to reduce the ability of the private sector to obtain bank financing as risk averse bank managers can simplyhold government debt and are under few pressures to innovate and find new customers and expand their portfolio. Inclusivegrowth. Government policy i s looking beyondjust maintainingrapid growth to making growth more inclusive by addressing imbalances inthe patterno f growth: too few jobs, too little growth inagriculture, lagging states and regions, and groups and people left out o fprogress. The foundation o f these efforts is empowerment, which i s key to expanding to expanding the equity o f opportunity. But while social protection remains one element o f inclusive growth, an empowerment approach does not seek to protect peoplefrom the market but rather it sees policy andpublic sector activity as facilitating success in the market. This makes empowerment foundational both for greater accountability and service delivery as well as for expanding the benefits of economic growth. EqualizingAccelerator: Labor Regulations. Ill-designed labor regulations can leadto unnecessary conflict between legitimate concerns: protecting workers injobs and creating newjobs and rising wages for workers. It is a major concern that for every person employed inthe "organized" private sector (only around 8 millionpeople ina country o f a billion) there are four unemployed (roughly 36 million on a current daily status basis). Moreover, according to official government sources, the numbero f organized private sector jobs has beenfalling inrecent years, and the "elasticity" of employment inthe manufacturingsector to growthinoutput has been negative inrecent years (rising output and falling employment). Current labor regulations seem to be protecting workers injobs by "protecting" other workers from havingjobs. This i s not to say a regime of "automatic hire and fire" i s the right solution (though this i s o f course the regime the vast majority o fworkers already face). Reforms ought to protect the interests of all workers by creating an environment for labor relations inwhich the legitimate social interests infairness for workers can be met. At the same time the labor regulations should not exclude other workers fromjobs by creating unnecessary and arbitrary regulations that discourages firms from hiringworkers, forcing firms to stay inefficiently small, and inducing firms to use capital at the expense o f labor. Properly done, labor regulationreform could be an equalizing accelerator. xvi Equalizing Accelerator: Financial Sector. On one level it appears that the demand for investment and not the supply o f financing i s the key issue: interest rates are relatively low, banks are holding substantially more government debt than they are required to, and, inaggregate terms, Indiahasbeen(at least untilquiterecently) exporting capital as its national savings exceeded investment. However, what the aggregate picture masks is that the financial sector reforms havemoved very far inequity markets and, with booming stock markets drivenbyportfolio investment large firms havebeen able to finance expansion easily-while massively reducingtheir leverage. But the combination o f a closely regulatedpublicly owned banking sector, the interaction o f a large deficit with fears o f makingnew loans, and few incentives for innovation means that the sector has been slow to expand its lendingto the private sector. Access to finance for small and mediumscale enterprises (which are animportant engine o fgrowth andproductivity) who are too small to be o f interest to equity markets or portfolio FDIare stymied. Lagging States. Duringthe 1990san interestingdynamic developed. Ifone divides states into the richer, middleincome, andpoorer states, what one finds is that while all groups of states enjoyed an acceleration o f growth by about 2 percentage points inthe 1980scomparedto the 1970s, following the reforms o fthe 1990sthe middle income states saw growth accelerate by another 2 percentage points while the poorer states stayed at exactly the same level o f growth. Hence, the "lagging" states are lagging, not inthat they have decelerated on average (though Bihar did) or because their incomes are declining but inthe sense that they have not been able to yet create the investment climate conditions necessary to accelerate to the more rapid rates o f growth experienced inthe middleincome states. The differentials ingrowth do producewider disparities across the country, and while the Finance Commission allocations are progressive, compensating fiscal policy i s not the longterm solution. There is policy playinthree distinct policy arenas. First, it is important continue to pursuethe policyreforms that leadto a completely integratedIndiaingoods and labor markets by eliminating barriers and distortions to goods movingacross states. Second, with integrated labor markets and largewage differentials there are incentives for either "people to move to jobs" (either permanently or temporarily) or "jobs to move to people." But for jobs to move to people the lagging states must address the deficiencies inthe investment climate inorder to attract industry: law andorder, transport connectivity, regulatory burdens. Third, the lagging states needto focus on the rural areas-biggest bang for the reform buck inrural areas i s the expansion of finance and infiastructure-for which, implementation is everything. Particularly inthe lagging states simplybuildingmore rural branches o fpublic banks inthe presence o f an unreformed banking sector or pouring more fhndingthrough line agencies inthe absence of service delivery reforms are unlikely to unleash a booming rural economy. Agriculture and the rural economy. One ofthe major concerns with the pattern o f growth i s the slowdown inagriculture and the effect that has on the rural economy and particularly on employment and wages (as rural and urbanmarkets for casual labor are increasingly linked). Butwhile subsidies have stabilized over the long sweep since the 1980ssubsidies have doubled while public investmenthas fallen by more than halfand public support for increasingyields has had diminishing impact-even though yields on xvii many crops inIndia are still halfthose o f comparator countries. A workable strategy for acceleration ofproductivity growth inagriculture has to tailor the rightmix o f actions to promote four directions for agriculture inthe overall rural economy with the specific potentials o fregions: 0 Intenszjkation: increased cropping intensity o ftraditional crops via irrigation, HYVs, agro-chemicals, mechanization(Green Revolution) 0 Diverszfication: shift to new more profitable crops (hits, vegetables, higher value cereals, medicinal plants) and livestock 0 Non-farm linkages: emphasis on value addition: trading, ago-processing. input supply 0 Exitfrom agriculture, recognizingthat as productivity increases fewer peoplewill need to be employed inagriculture, andthat the long-term strategy inareas with limited agricultural potential i s off-farm activities. The traditional approach of a relatively heavy handedpromotion o f intensification, primarilyinfood crops, with public investments coupledwith input and output subsidies, has been successful inmany ways, but i s largely played out and is often promoting agricultural practices that are neither economically nor environmentally desirable. A new strategy shouldrely muchmore on a mix o fpolicy o f supporting input and output liberalization with land administration and tenancy reform, and support to R&Dand extension that i s tailored to specific needs that complements the needed investments in infrastructure. Promoting a new strategy for agriculture and the rural economy would assist in addressing the problem o f economically and environmentally unsustainable pattern o f farming inducedbythe lack o frestrictionson ground water usage, subsidized power, and price distortions. But it would bena'ive to pretendthat there are not very deep and difficult problems inthe political economy o freforms inthe subsidies givento farmers. Demonstrating that subsidies do not reach the poorest i s one step, but to make progress the very real concerns o f the middle and larger farmers need to be addressed head-on as while some are benefiting from access to power andwater and doing well, there are many other farmers facing difficult circumstances evenwith power subsidies. Fair access to markets. There has been a seismic shift from a view that the poor needto beprotectedfvom markets to the view that abetter route out o fpoverty is to help themto strengthen their own livelihoods strategies through empowerment and fair access to marketsfor credit, labor, land andproducts. The Self-Help Group (SHG) movement has shown that with efforts at mobilization and empowerment even the poorest women can manage their own resources well and benefit from economic opportunities. The expansion o fmicro-credit is an important tactic inan overall strategy for rural growth. The experience o fprivate firms with expandinginto credit and into marketing (e.g. e- choupals o f ITC) shows that there is no intrinsic contradictionbetween commercial viability and fair access for the poor. A keydisadvantaged group are tribals who suffer from bothgeographic and socio- cultural exclusion. A central factor affecting tribal livelihood possibilities i s secure access xviii to and control over natural resources. The Panchayats Extensionto Schedule Areas (PESA) Act, 1996,which attempts to redress inequities by transferring rights to ownership o f minor forest produce to Gram Sabhas has had limitedimpact largely due to unevenimplementation and inconsistencies with state government provisions. Draft legislation to confer forest rightsto forest dwellingtribals i s a progressive step in restoring social justice but its success will depend largely on resolving implementation constraints. As unskilled labor is the key asset o fthe poor, it is a key concern. There are still social cleavages inthe labor market across caste and gender. Women make only 55 percent o f men's wages, even incasual labormarketsandvery little o fthe difference can be accounted for bythe usual determinants o fwages. Inspite o f far-reaching changes, some occupations continue to be castebased, even within the public sector. Indiahas longhad caste based reservations injobs that untilrecently were the most coveted-regular salaried work inthe public sector - and there i s evidence to suggest that reservation policies are helping SC/STs to overcome occupationalbarriers. However, while there are positive impacts o f some types o f reservation, the mere extension o freservation to larger andlarger spheres without accompanying reforms creates realrisks o ffreezing uprather thanfreeingupopportunity. To maketheplayingfield more evenacross castes and gender, the most significant reform that India can undertake is to reform labor laws. Perhaps paradoxically, the very dualism created between formal sector jobs and the casual and self-employed workers, created inlarge part by existing labor regulations, facilitates discriminations, such as gender and caste-bias, inhiring. Equity in the creation of human capabilities. Investments inhumandevelopment are important, notjust to makethe most o f livelihood opportunities but also for ensuringthat inequities are not perpetuated across generations. However, there are large gaps inhuman development outcomes across states and betweenrich andpoor inmost states. Many of the core public services addressedinpart Iare related to health, nutrition, and schooling. Implementingreforms incurrently lagging areas is particularly important for addressing the inclusiveness o f growth-the health, nutritional and educational outcomes for children today will determine their economic capability for decades to come. Social Protection. Of course, even with the best o f access to markets and opportunities there are needs for the government to engage in social protection to assist the poorest and helpincopingwith the risks andvulnerabilities that citizens face. Well-designed social protection systems can promote not only equity but also dynamic efficiency bymitigating market failures and enhancing opportunities for the poor. Indiahas a longtradition of social protection. However, its social protection system has onlyjust begunto adjust to developments inpoverty andvulnerability inrecent decades. While it spends a significant amount on social protection, this remains largely focused on social assistance and formal sector social insurance programs, with very limited efforts at insurance-type interventions for the unorganizedsector. A strategy is called for which relies on a more balanced mix o f efficient and accountable social assistance programs, expanded contributory systems where feasible (withprovision for non-contributory programs for certain groups), andprograms which perform an insurance-like function such as xix workfare. The rebalancing of the social protectionwould also result over time in greater reliance on cash rather than in-kindbenefits over time. The new National Rural Employment Guarantee Act (NREGA) embodies the use o f an employment guarantee ina bold effort to provide social protection andtransform the rural economy. The international (and Indian) experience has been that a well designed employment program can be a usehl component o f a social protection strategy but the devilis inthe design and implementation details, two inparticular. First, keepingwages low is a key design element o f employment schemes as it encourages self-selection among potential beneficiaries and prevents fiscal costs from "rationing" jobs (explicitly or implicitly). Inthis context the decision to set wages at the state agriculture minimum wage will likely prove problematic as, at that level o fwage, demands may exceed available funding, an issue that may have to be addressedbased on the experience with the first round o f districts. Second, the desirability o f employment programs depends to a large extent on the value o fthe assets they create. The success o f the proposed use o f greater PRIinvolvement inimplementationo fthe Act i s apromising initiative andcould, ifdonewell,constituteamodelforotherprograms-but the accountability mechanisms for GPs will needto be strengthened. Organization of the Report A guide to the report itselfis inorder. Chapter 1is anarrative o frecent issues that justifies the attention to the two major themes: service delivery reforms for greater accountability and sustaining rapid and inclusive growth. Part Ibegins with chapter 2, which presents a diagnostic that motivates the attention to accountability and a framework o f common principles for reform efforts. The next three chapters discuss various ongoing efforts to improve services: internal reforms and external accountability are discussed inchapter 3, how to make decentralizationwork to improve services is discussed inchapter 4, and the use o fpublic-private partnerships-particularly in infrastructure, although the principles can be applied inthe social arena as well--is discussed inchapter 5. Part I1discusses actions to achieve more inclusive growth. Chapter 6 discussespotentially bindingconstraints to continued growth-infrastructure andthe fiscal deficit. Chapter 7 discusses potential equalizing accelerators-reforms that bothraise growth and expand inclusion. Chapter 8 is focused on agriculture and the rural economy (even inprosperous states). Chapter 9 treats the lagging states and regions while Chapter 10addressesthe issues o fequity and opportunity for people. A final word. This is the WorldBank's Development Policy Review, which has four implications. First, ina county as large as Indiawith reforms going on inmany sectors and states drivenbothby government and by outside groups and advocacy we cannot provide detail on each area. Rather we try and present some factual evidence about the issues and context, an analyticframework and commonprinciples for thinking through options, andpromising directions forpragmatic solutions. Second, this document is intendedas a synthetic document drawing almost exclusively onnew research and analysis done within the World Bank and, while it draws on and i s informed byresearch and findings on Indiafrom all sources, i s not intended as a comprehensive review o f all literatures on the topics discussed. Third, this document makes no pretense to represent xx the voice o f any group but the World Bank-not "the poor", not "the global community"-these are the views o f the World Bank based on the particular expertise, experience, and evaluations o f its staff. This i s a view from a window not a mirror. Fourth, this makes no pretence to finality or certainty but is rather one view, our best judgmentsbased onthe available information and evidence, andhopes, at best, to be one additional voice ina vibrant ongoing dialogue. Taking o f fon the phrase Amartya Sen has recently popularized, we would hope nothing inthis report is accepted by any Indian without argument. xxi Chapter 1.Overview ofIndia's developmentprogressandpressing challenges Any observer o fIndiatoday is struck bythree puzzlingobservations: a Economic optimism is rampant and seems to outstrip the changes in fundamentals. India's economic progress has been awe-inspiring andhas awakened the world to India's potential as a global actor. People now speak o f a fiture "tripolar" world with India, China, and the United States as the superpowers (Virmani 2005). Eventhough India's growth, at 7 percent, is already among the highest inthe world, the constant talk is o f "catching" the phenomenally rapid growth rates o f China. But at the same time, by many individual measures o f infrastructurecapacity, corruption, education, and regulatory environment, Indiaremains a difficult economic environment. a The economy is booming, but thepublic sector seems to be deteriorating. India's sterling economic performance has beenaccompanied by a curious inversion. In past decades people would fret about economic performance, but marvel at India's institutional strengths inthe public sector-a vibrant democracy, an extraordinarily talented top-tier bureaucracy (the "steel frame" o fthe Indian Administrative Service), and a set o f organizations that could provide law and order, revenue collection, and a modicum o f services ina sprawling poor country. Today, these concerns are almost inverted:it is easy to be optimistic about India's economic prospects, but there i s growing concern that the basic institutions, organizations, and structures for public sector action are failing-especially for those at the bottom. a It is the best of the world, it is the worst of the world-and the gaps are growing. The top students from the IndianInstitutes o f Technology are notjust globally competitive, they have set the global standard. Yet, many, ifnot most, children in India finish government primary schools incapable o f simple arithmetic. While there is increasing "medical tourism," where people travel to India for high- quality, low-cost medical treatments-the typical Primary Health Center doctor in Delhi is less competent than doctors inTanzania. India's Supreme Court i s justly world renowned, but local courts are backlogged and ineffective and localpolice are frequently a miasma o f corruption and brutality. Similarly, ineconomic performance, while parts o furban India compete for business insoftware engineering andbiomedical research, parts o f rural Indiahave poverty rates comparable to borderline "failed states," such as Haiti and Nigeria, and have child malnutrition rates higher than any country inthe world. This Development Policy Review starts from the premise that the goal o f development i s to improve humanwell-beingina sustainable way, with particular emphasis on the less well-off. The discussions o f development and poverty reduction have been enormously enriched by the contributions o fAmartya Sen, which emphasize that development is 1 about muchmore than increasing material well-being, but also includes expansions in capabilities andinbothpositive and negative freedoms. A more consistent attention must be given to assessing development progress not simply as a measure o f an aggregate of economic activity but as an assessmento fthe inclusiveness o f economic growth, with emphasis not only on the distribution o f gains but also on the security, vulnerability, empowerment, and sense o f fullparticipation the people may enjoy insocial civic life. India's growthpatterns:a star performer India's economy i s growing rapidly by historical and global standards. Although different authors break India's post-war experience into various periods, all agree that growth beganto accelerate inthe 1980sandcontinuedafter the reforms o fthe early 1990s, perhaps at a modestly faster pace. This acceleration has taken India from being a below- average growth performer inthe 1960s and 1970s to one o f the most rapidly growing economies inthe world inthe 1990s (figure 1.1). Virmani (2005) shows Indiaranked ninthamong 107countries inpercapitagrowth of gross domestic product (GDP) inthe 1992/93-2003/04 period. Part of the reason there i s so much attentionto China's growth now is not because China is doing well and India not so well, but rather because China i s one o fthe very few countries growing faster than India. Indeed, Indiai s doing very well on growth and China spectacularly so. Figure1.1: Indiahasbeen inthe top 10percentof all countriesin growth performancesince the 1980s. Distribution of GDPper capita growth rates across countries, 1960sto 1990s Chino 7.02 Chlna 5.47 India 4.47; 19503 13903 1 2 3 4 By decode Note: Box plots show the median, loth, 25th, 75th, and 90thpercentiles. Source: Author's calculations from Aten, Heston, and Summers 2001. 2 India's rapid growth has been accomplished, particularly since the reforms o f the early 1990s, inwhat appears to be an economically sustainable way: although the persistent deficit is a concern, there i s low inflation (by international standards), there have beenno serious macroeconomic crises, andreserves are large and expanding as there are persistent current account surpluses (at least untilvery recently). While initially there were some fears in India o f the implications o f India's integration into the global economy, now there are mainly fears of India. Although Indiangrowth has beenamong the highest inthe world, especially inrecent years, Indiaremains a very poor country, because it started from a very low base. Rapid growth inIndia, even ifit does not differentially benefit the poorest o f the poor, does benefitpeople who are quitepoor by any global standard. Evenpeople at the 90th percentile ofthe income distribution inrural India have less than a fourth the income o f a very poor person inDenmark (figure 1.2). The gap between the rich and poor within urban Indiais much smaller thanthe gap between a rich urban Indian and a very poor Dane. This is not to say that inequality within India i s not a serious problem (much o fthe report is devoted to that issue), but that a country with global aspirations such as India should also gauge itself against a global standard. Applyingthat global standard, even growth that benefits some o f the richer Indiansmore than poorer Indians dramatically reduces global inequality. Figure 1.2: An Indianat the 90thpercentileof the incomedistributionhas much lower incomethanthose at the 10thpercentilein OECD countries. Range of monthly income and expenditures across countries, by region and income group (in 1993purchasing power parity equivalents) 0 250 500 750 10010 1250 1500 1750 2000 22% 2500 Note. Blue bars indicate the range inincomes. Orange bars indicate the range of expenditures. Source World Bank 2005d. 3 While there is a sense that rapid economic growth has put the "middle class" on a frivolous consumer binge, it i s worth pointing out that the real middle o fthe Indian income distribution (the 50th percentile) did benefit from growth inthe 1990s-in that a thirdo fthembought abicycle (so 35 percent ofhouseholds own abicycle, although still only 3.5 percent own even a motorcycle), bought a fan (still less than one infive own a television), and switched away from wood as a cooking fuel. A reality check on the gloss of the "middle class" is that in 1998, at the actual middle o fthe asset distribution inIndia two-thirds o f households usedwood for cooking fuel (Chaudhury and Hammer 2005). Broadly based economic growth bringsnot frivolous consumption, but enormous gains in basic well-being. These gains inthe basics from economic growth are also reflected inIndia's poverty statistics. There has been considerable debate about the poverty estimates from the latest available survey (the 55th round, 1999/2000)--to which this report will addnothing, as the issues are well andwidely known (see Deaton and Kozel2005 for a summary) and the next round o f data will be available by early 2006 when the issues can beresolved (or at least re-debated with new information). But whatever the issues around the 1999/2000 estimates, it is clear the steady and rapid economic growth since the 1980s has ledto spectacular reductions inpoverty compared with previous periods. In 1979 (the last "pre- reform" estimate) headcount poverty was hovering at 50 percent-higher than the estimate for 1951,so that inalmost 30 years from Independence to 1979 essentially no progress had beenmade. Indecided contrast, the 1999/2000 official poverty rates of the Planning Commission were at halftheir 1979 levels. Table 1. 1: The average pattern is for one decade's growth stars to fall to earth in the next, but India has defied that trend. Change in growth rate, decade to decade, among countries with toppeflormance, 1960s to 1990s (in percent) Growth slowdown of Period Top ten growth perjfiomzers of All others previous decade's stars previous decade versus other countries 1970svs. 1960s -3.4 -1.o -2.4 1980svs. 1970s -3.4 -1.0 -2.4 1990svs. 1980s -1.9 +0.2 -2.1 Note: Negativei s a slowingingrowth. Source: Author's calculations. Giventhe centrality o f economic growth for improvements inwell-being for the population at large, one key theme o fthis report is the importance o f sustaining, and perhaps even accelerating, economic growth. Indiais sufficiently far from the global economic leaders that there is no risk o f growth decelerationbecause India has "caught up": evenat India's current rapidgrowth rates itwould be almost 40 years (whenany students readingthis will beretiring) before India attains the level o fU.S. per capita GDP in 1950,and 60 years (beyond the lifetime o fmost readers) to reach the current U.S. level. But, as table 1.1illustrates, poorer countries cannot take rapid growth for granted- the past pattern is for rapid growth to slow-a pattern Indiahas defied once and can, with 4 effort, defy again, but complacency about the need to push ahead with policy reformjust to maintain rapid growth could lead to slow growth. Recent growth experience H o w does one apply the lessons o f India's success (andthat o f other countries) to creating more success inthe years ahead? This report aims to identify the most important constraints on economic activity, those policy changes with the biggest growth gains for the reform effort. While there are many areas o fpolicy that require attention, and any analysis ofthe "investment climate" or of specific sectors will reveal a long list of desirable reforms, indesigning a growth strategy policymakers should focus on the "binding" constraints-those areas that, without attention, are the most likely obstacles to sustained rapid growth. Relievingthese constraints would be a growth accelerator. The aphorism "In a desert one should find camels not hippopotamuses" i s a good guide * to analyzing the patterno f economic growth. Ifrain is scarce, then one should find animals adapted to a scarcity o fwater, not those reliant on water. Inan economic environment where, say, transport infrastructure i s scarce, one should find "infrastructure camels''- those industries and firms that are thnving shouldbe less than usually reliant on infrastructure. Similarly, ifthe cost o f finance i s high, one should expect to find "finance camels"-industries and firms that are not particularly reliant on capital intensity. This section does not attempt to give a comprehensive review ofIndian economic growth, as there are many such reports from government (Go12005) andnon- government sources (Virmani2004a, 2005; Acharya 2002; Panagariya 2004), butrather focus just on those features o f growth that are indicative o f constraints on growth. What are the features o fthe recent Indian growth experience that give indications o f what mightbe the next set o fimportant constraints to growth inthe mediumterm? 0 The service sector i s booming, agriculture lagging, andmanufacturing performance i s a mixedbag. 0 Job creation inthe private organized sector has been very weak. 0 Wage expansion has beenmore rapid for those with already highwages. 0 Investment, both domestic and foreign, has been relatively weak. Thriving services, lagging agriculture, mixed performance in industry. The service sector grew at an average annual rate o f9 percent inthe 1 9 9 0 ~ ~ contributing nearly 60 percent of the overall growth o fthe economy (World Bank 2004a). While the most visible growth has been information technology andbusiness process outsourcing services, other sectors-including telecommunication, financial services, community services, and restaurants-have also grown considerably faster than GDP over time. Consequently, India's services share o f GDP i s higher than the average for low-income countries. 'This phrase is from Ricardo Hausmann. Our analysis i s motivated by the "binding constraints" approach outlined inHausmann, Rodrik, and Velasco (2005), which makes the point that a growth strategy i s strategic inthat it makes hard choices about which particular avenue to pursue rather than a check-list o f all possible reforms that could have an impact o n growth. While it i s empirically difficult to quantify and rank the most bindingconstraints, examining the pattern o f growth can provide clues to what important constraints to growth are likely to be. 5 Table 1.2: India's growth rate has accelerated to 7-8 percent in recent years, driven by the continued service sector expansion and a recovery in industry Real GDPgrowth, by sector, in India, 1992-93 to 2005-06 Percentage changefrom theprevious year (atfactor cost) Sector 1992-93 to Tenth 2002- 2003- 2004-05 2005-06 2002-03 Plan 03 04 (quick (advanced target estimates) estimates) Agriculture 2.4 4.0 -6.9 10.0 0.7 2.3 Industry 6.3 8.9 7.0 7.6 8.6 9.0 Mining and quarrying 4.4 8.7 5.3 5.8 1.o Manufacturing 6.8 6.8 7.1 8.1 9.4 Electricity, gas, water 5.4 4.8 4.8 4.3 5.4 supply Services 7.6 9.3 1.3 8.2 9.9 9.8 Trade, hotels, transport, 8.5 9.1 12.0 10.6 11.1 and communications Financial services 7.8 8.0 4.5 9.2 9.5 Community, social and 6.1 3.8 5.4 9.2 7.9 personal services Construction 5.5 7.1 10.9 12.5 12.1 Total GDP at factor cost 5.9 8.0 3.8 8.5 7.5 8.1 Source: Central Statistical Office. Perhaps the most worrying feature o f the recent growth performance has been the performance of agriculture: average agricultural growth over the past three years has been 1.3 percent-against a Tenth Plan target of 4 percent per year. Agricultural growth appears to be decelerating: from 3.2 percent in 1980-92, to 2.4 percent in 1992-2003, and to 1.3 percent over the past few years. As two-thirds o f India's people depend on rural employment for their main source o fincome, this is directly affecting many households. The deceleration seems to be general across all crops and appears to reflect a broadly based deceleration inproductivity growth (Go12005). 6 Industry has grown rapidly, at 6.3 percent between 1992 and 2002103but this lags well behindservices growth and, at 22 percent in2004/05, the share ofindustryinGDP remains very low inIndia. The contrast with China is stark: inChina the share of industry inGDPwas 51percent. Nagaraj (2005) shows that China's industrial growthrate is close to one-and-a-half times that of India, andestimates suggest that, although impressive Total Factor Productivity (TFP) growth has been achieved inIndia, India's level o f industrial TFP is between about 33 percent and 40 percent o fwhat it couldbe; and it i s only halfthat of China.2 Similarly, while India's expansion o fmanufacturedexports has been impressive (inthe first half of the 1990s India's manufacturing exports grew 30 percent faster than world trade inmanufacturing), China's manufacturing exports grew 57 percent faster than India's. India's world market share inmanufactures is still very small and still a small fraction of China's, which has increased from about 3 percent in the mid-1990s to over 6 percent in2002. Labor markets: little growth informal jobs. A second major concern about the structure o f economic growth inIndia i s that the number o fprivate sector jobs that have been created is small compared with the rapid growth o f output. In2003 only 8 million were employed inthe organized private sector, out o f a labor force of roughly 390 million people. Eventhough the labor force grew that year by 12 millionpeople employment in the organized private sector fell by200,000. The lack o f employment growthis particularly stark inmanufacturing where, according to one study, the "employment elasticity" between 1994 and 2001 was negative, as employment fell even though output rose by 5.3 percent per year. In2003 for every worker inthe organized private sector there were four unemployed(figure 1.3). Figure 1.3: For every worker in the organized private sector there are four unemployed-and the number of organized private sector jobs is falling. Number of workers employed in the organizedprivatesector or unemployed, 2001-03 Organizedprivate sector Unemployment,Current DailyStatus Source: Go1Planning Commission2005b, tables 8.1 and 8.2. *See Rodrik, Subramanian, and Trebbi (2004) and Rodrikand Subramanian (2004). Estimates are based o n regressions of TFP based on the deep determinants of TFP. 7 Wageincreases concentratedat the top. While wage growthhas been impressive and helped to drive down poverty, there are some signs that excessive wage growth inthe top 10percent o fjobs may reflect the increasingscarcity o fhighly skilledlabor inthe service sector (figure 1.4). Over the past decade, annual growth rates o f20-30 percent inthe information technology sector have depleted existingtalent pools, producing high-skilled labor shortages and drivingup salaries that are thought to have quadrupled over the past five years for Indian I T engineers (Oxford Analytica, November 14,2005). All three o f India's leading software companies (Tata, Infosys, and Wipro) are reportedly looking to develop educational institutions to increase the number o fhigh-skilled workers. Increasing returns to secondary education point to the general scarcity o f skills at that level, while highunemployment o f educated labor points to a possible mismatchbetween the skills o f the labor force with tertiary education and the skills demanded bythe market. Figure1.4: Wage growthhasbeendisproportionateat the top end of the labor market. Increase in real daily wages, by decile, 1983 to 2000 Deciies of wages 1 2 3 4 5 6 7 8 9 10 19831993194 W 1993194-199912000 Deciles of wages Source:Narain 2005. Weak investment with low reliance onforeign sources: As outlined inthe Tenth Plan Mid-Term Appraisal, the domestic savings ratehas actually exceeded the plan's targets. The current account balance was also positive until(very) recently. Thus, India's growth pattern i s unlikethe old-fashioned "two gap" thinkingthat growth depended on investmentand would be constrained by either low domestic savings or the inability to attract foreign savings. India's investment has been less than available savings and has actually been, in essence, exporting capital as burgeoning o f foreign exchange reserves do not contribute to additional physical investment. Untilvery recently, India's record in attractingprivate investmenthas also not beenvery impressive; at around 15 percent of GDPover the past decade, private investment inIndiais below the average for low- income countries. Inward foreign direct investment to India has remained modest, averaging $5 billion annually over the past few years, incontrast to the $40 billion annually to China. Foreign direct investment inIndiais 1percent o f GDP, comparedwith 4 percent inChina and between2 percent and 3 percent inmany emerging market 8 e c o n o m i e ~Some of this lack of investment has been reversed very recently, as the most . ~ recent data show investment increasing to more than 30 percent in2005/06, driven by increases inprivate investment. Nonetheless, it is too early to say whether that upward shift will b e sustained (figure 1.5). Figure 1.5: Private investmentin India has been low-at least until recently. Investment as a share of GDP, 1985-86 to 2004-05 30.0% T 25.0% -0 h 20.0% 15.0% 10.0% 5.0% 0.0% 1985-86to 8th Plan 9th Plan 2002-03 2003-04 2004-05 1989-90 Private Investment GovernmentCapital Spending -0- Gross Fixed Capital Formation Cluesfrom thepatterns of growth This evidence suggests further examination of several questions inthe remainder of the report: 0 I s the more rapid growth in services than industry due to a tight constraint o f infrastructure, as the service sectors are less reliant on movingphysical inputs and outputs and hence are "infrastructure camels'' that can easily cocoon themselves from infrastructure deficits? 0 I s the low job creation inmanufacturing the result of regulations that raise the cost of labor to firms by m u c h more than they benefit laborers? 0 I s it not the cost o f finance, but rather access to it, that limits investment? One needs to distinguish between portfolio investment, which is very strong, and physical investments. Capital inflows remainbuoyant, and the surplus on the capital account was $32.6 billion in2004105 (versus $20.9 billion in 2003104 and $12.1 billionin2002103).The fastest increase has been inforeign portfolio investment, with India accounting for one-fourth o f all portfolio flows to emerging Asia in2004. Foreign direct investment inIndia remains low and i s picking up only slowly, hindered by a difficult business climate as well as by caps incertain sectors (IMF-2005a). So, the stock market has soared as people bidup prices for existing f m s (and their assets). 9 Part I1(chapters 6 and 7) o f the report explores four issues as the most likely factors constraining aggregate growth inIndia inthe mediumterm: (a) infrastructure, (b) the fiscal deficit, (c) labor laws, and (d) the financial sector. Of course, there are many other areas of policy that require attention-such as poor governance, unclear and insecure property rights, and unsustainable use o fnatural resources. While reforms inthese areas are desirable, indeed critical, for a long-term development strategy, this report focuses primarily on the main obstacles to sustained growth inthe mediumterm. Inthe Indian context, one additional question that merits W h e r examination is whether the accelerators of growth will have an equalizing or disequalizing effect across states, regions, and people. Givenconcerns about the distribution o fbenefits from growth, as discussed below, there should be a focus on equalizing accelerators. Corepublicservices: implementationis everything There are heated ideological debates about theproper role o f government insome economic spheres, but there is near universal agreement that the government has a responsibility to its citizens incertain core areas: law and order, infrastructure, education, health, water resources, sanitation, social protection, and safety nets. Debates may continue over whether this responsibility i s best discharged using direct production o f services by the state or by other modes (e.g., contracting with private parties to build roads versus "force account," providing medical insurance versus care inpublic facilities) andhow services inthese core areas are to be financed (e.g., how much general taxation versus user fees). Few would argue, however, that the state should avoid responsibility for the quantity and quality o f services inthese core areas. But there is a growing sense among politicians, civil servants, and academics that the ability o f India's existing institutions to deliver on those responsibilitiesi s deteriorating-even as the economy booms. 10 It is a strikingfact that economic renewal andpositive growth impulses are now occurring largely outside thepublic sector ....I n the governmental or public sector, on the other hand, we have seen a marked deterioration at all levels- not only in terms of output, profits andpublic savings, but also in theprovision of vitalpublic services in the fields of education, health and transport. BimalJalan, TheFuture of India: Politics, Economics and Governance, 2004. While thefunctions of the state in India have steadily increased, capacity to deliver has declined over theyears due to administrative cynicism, rising indiscipline, and a growing belief widely shared among thepolitical and bureaucratic elite that the state is an arena wherepublic ofice is to be usedfor private ends. Naresh C. Saxena, Improving delivery ofprogrammes through administrative reforms in India, 2005. Theperformance of India's public institutions has become a matter of serious concern bothfor the quality of the country's democracy and the well being of its people. Devesh Kapur and Pratap Bhanu Mehta, Public Institutions in India, 2005. India's structures for service deliverywere designedto create the basic assets needed, and on that level they have been largely successful. If"access" to core services is measured crudely as proximity (to a school, to a water source, to a clinic), then the expansion in physical assetshas lead ever largerportions o fthe population to have acce~s.~ in But 2004 the newly chosen prime minister signaled the mindset shift that has to occur if India's publicly delivered services are to keeppace with its booming market: from promises made to promise kept. Keepingpromises will beneither easy nor quick, because it requires hndamental shifts inthe way government and its line agencies typically go about their business. The shift is from an emphasis on outlays to an emphasis on outputs and outcomes, from a focus on asset creation to a focus on asset operation, from compliance with procedures to a drive for better performance. Since this document i s a Development Policy Review, it is worth noting this shift i s also away from policy design (e.g., stated objectives, chosen instruments, and sectoral or schematic expenditure allocations) toward policy implementation. Inspite of India's rapid growth andthe accompanyingrapidreduction inthe headcount ratio o f income poverty, there are other elements o f a multidimensional approach to poverty inwhich problems remain stubborn andprogress i s slow. A recent World Bank report documented the persistence o f extremely highlevels o f malnutrition inIndia: among the middle-income group more than half o f children were underweight (less than two standard deviations below the norms o fweight for age)-which is twice the level in This does not meanthat all access problems havebeensolved (e.g., rural access to infrastructure is still poor inmany areas). 11 Africa (World Bank2005b). Infant and child mortality is a key indicator of well-being, andprogress onthis indicator slowedprecipitouslyinthe 1990s. Figure 1. 6: Facedwith inadequate publicservices, both elites and non-elites have developedways to augment education, water, and health services in the private sector. Elite and nonelite strategiesfor copingwith dysfunctional or unavailablepublic sewices Elite coping strategies 1 Weak public services 1 Non-elite coping strategies Inthefaceofthesestubbornproblems, astrikingfeatureoftherecentIndianexperience is a de facto shift from public to private provision o fmany core services. The attempt to discharge the governmental responsibility for core services exclusively throughthe direct productiono f services inthe public sector has led to a completely ad hoc, unregulated privatizationo f services because o f a systemic failure inthe accountability ofpublic sector for the quality o f services. There are distinct "elite" and"non-elite" strategies for coping with public sector inadequacy, which are illustrated for water supply, health, and education inthe next few paragraphs andinfigure 1.6. Paradoxically, it appears that the route to privatization (and an ad hoc, unplanned, and unregulatedprivatization at that) is a blanket commitment to public sector producers-if that commitment to public sector producers comes without an equally strong commitment to public sector production. When peopleturn to providers who are not full-time employees o fthe government to meet core public sector responsibilities, their actions cannot be caricatured as "pro- privatization"; infact they are simply pro fulfillment o fpublic sector goals-no matter what the instrument-an outcome orientation that does not begin with any preconceived notion that core public responsibilities are best delivered by state functionaries, NGOs, private sector, community groups, or local governments. 12 Urban and rural water supply Figure 1.7 shows that the daily supply o fpipedwater inurban areas i s much lower in Indiathan inother developing-country cities around the world. Moreover, it appears that the availability o f water has beendeclining-while water inBangalore was available for 20 hours a day inthe early 1980s, it is down to 2.5 today. InChennai water was available 10-1 5 hours a day inthe 1980s, but only 1.5 hours a day today. People cope with the declining availability ina variety o f ways. A recent study o f the costs o f coping with inadequate water supply inDelhifound that the true total costs o fwater supply are already "privatized"-on average the private coping costs are 262 rupees per month (when capital costs are included) versus a monthly water billo fonly 141rupees per month (Misra 2005). The differences inhow the elite and nonelite cope are stark: while those inauthorized colonies pay the price for declining public service incapital expenses for pumps, boreholes, and overheard storage tanks, those inthe slum areaspay the cost withtheir only asset: time. They spend almost two hours a day to collect water. 13 Figure1. 7: While many major cities in developingcountriesmaintainnearly24- hourwater supply, most Indian cities have service only a few hoursa day. Average hours per day of water service in selected Indian and comparator cities 226 244 Hrslday eLPCD 220 10 223 33 10 341 6 Source: Ministry of UrbanDevelopment and Water and SanitationProgrambenchmarlung study, ADB utilities book, verified with relevant utilities. For irrigation and rural drinkingwater, there i s a similar recourse to the private sector. Tubewells are widely used for irrigation, as groundwater now supplies 70 percent o fthe irrigated areas, and boreholes account for 80 percent o f domestic water supplies. Over 80 percent ofnewly irrigated areas have come through the expansion o fprivate irrigation. As Shah (2005) puts it succinctly: "We needto recognizethat self-provision o fwater is the best indicator o f the failure o fpublic water supply systems. Tubewells proliferate in canal commands because public irrigation managers are unable to deliver irrigation on demand.Urbanhouseholds want their own boreholes becausemunicipal service is inadequate and unreliable." Health A recent household survey for the Reproductive and ChildHealthProgram found continuing progress insome dimensions o fhealth care, but serious signs o f slowing in others. Comparingthe data from districts from household surveys in 1998/99 with those same districts in2002/03 one revealedthat, although there are some improvements in antenatal care, many dimensions o f service provision outputs have worsened (table 1.3). The percentage o f children delivered inpublic institutions fell by 5.5 percentage points to less than one infive births, while the percentage delivered inprivate facilities nearly 14 doubled. Women with timely visits from Auxiliary Nurse Midwives, never very high, fell even further to less than one ineight. Shockingly, the prevalence o f children with full immunizations for childhood diseases, appallingly low to beginwith, fell from 52 to 44.6 percent. According to the World Health Report 2005, immunization rates for inIndia are 81 percent for BCG, 71 percent for DTP, and 67 percent for measles, which are substantially lower across the board than inBangladesh (95 percent, 85 percent, and 77 percent, respectively) and are similar to those o f a much poorer country such as Cambodia (76 percent, 69 percent, and 65 percent). Table 1.3: While some indicatorsof healthcare rose, the share of childrenwho were fully immunized,were delivered in a publichospital, andwere attendedby a visiting nurse declined. Percentage of children receiving selected health services, 1998-99 and 2002-03 Change (in percentage points Health service 1998-99 2002-03 of thepopulation) Polio3 66.1 57.0 -9.1 Fullimmunization 52.0 44.6 -7.4 DPT3 64.6 57.5 -7.1 Deliveredat a public institution 24.0 18.5 -5.5 Measles 58.1 55.2 -2.9 Auxiliary Nurse Midwives m a h g home 14.1 12.7 -1.4 visit within two weeks Deliveries assistedbyprofessionals 39.6 47.5 7.9 Full antenatal care 31.9 40.3 8.4 Any antenatalcare 64.5 74.0 9.5 Delivered at a private institution 9.4 21.5 12.1 Note: Basedonhouseholdsurvey and districts with householddata inbothperiods. Source: Implementation Completion Report of the Reproductive and Child HealthProgram I. Although Indiahas a constitutional commitment to universalprovision o f health care, the ratio of government to total expenditure inhealth is only 21 percent-half the public spendingratio o fideologically "free market" countries such as the UnitedStates (45 percent) or Chile (44 percent). Ifone takes expenditures as a measure, Indiais one o fthe most "privatized" health systems inthe world-of the 195 countries listed inthe World Health Report 2005, only five (Cambodia, Guinea, Myanmar, Sudan, and Togo) had a ratio o fpublic to private spendinglower than India. This highprivate spending inIndia i s not the result o f an announced public policy to reduce services, but rather a coping strategy to deal with the failure o fthe public sector to actually provide promised services adequately. A recent survey that tracked public facilities inRajasthan continuously for a year found that, while use o fpublic facilities i s low for all income groups, the coping strategy o fthe poor i s to visit facilities less and to rely more frequently on traditional healers (Banerjee, Duflo, andDeaton 2004). 15 Elementary education Ineducationthe government ofIndiahaslaunched amajor initiativeofSarva Shiksha Abhiyanto expandenrollments, increase local capacity for planningand action, and ultimately raise learning achievement. Internalreports suggest massive reductions inthe numbero fchildren out o f school, which is estimated to have fallen from 25 millionto around 13 millionby 2005. Resolution o f how much o fthe reported increase is due to expansion o f government schools and how much due to increased demand and expansion of private schools will haveto await the householddata from the National Sample Survey. But whatever the trend, the share ofprivate schooling inurban areas is very high. Figure 1.8 reports that inthe urbanareas o f five states nearly two-thirds o f enrolled children are inthe recognizedprivate schools, and inanother five states more than 4 in 10 urban children are inprivate school^.^ Unfortunately, the data do not distinguish between private aided schools (which are very widespread inKerala, for instance) andprivate unaided institutions. Note that eveninsome states typically known for better than average public services andreasonable governance (e.g., Maharashtra, Tamil Nadu,) the private share i s extraordinarily high, while it i s low where public sector performance mustbe even worse (e.g., Bihar, Jharkhand). Again, there is no question o fthe government's constitutional, rhetorical, andlegal commitments to universal elementary education. Nonetheless, the quality of services provided leads people, given the choice when alternatives exist (as inurban areas; inrural areas the private sector share is much lower), to opt out o f government schools, even though this usually means paying the full cost of education themselves. 'What makes the highratios inthe official DISE data even more strikingi s that enrollment inthe unrecognizedprivate sector is widely acknowledged to be large and growing insome states. One study in seven districts o f Punjab found 26 percent of all children enrolled were enrolled inunrecognized schools. While the DISEdistrict report for Jaunpur, U P reports only 15 percent o f enrollment inprivate schools, an independent study infour blocks found over 40 percent inprivate schools. A study inHaryana infour districts found seven unrecognized schools for every ten recognized schools so the "recognized only" figure would be far too low. 16 Figure 1. 8: In the urban areas of eight major states, more than half of the children in elementary school are in private schools. Percentage of students in recognized private schools, by state ki 0.75 9\0 4\o / 0 - Chile's share of private schools v m I 4' 6 (decades after with full privatization) a, 1 i s 45.5% ._ > m h 0.50 s\o ,k.......,p\*o ............. ................................................ Source; Author's calculations with District Information System for Education 2004 State Report Cards. This i s not to say that private sector participation inany o fthese sectors is a negative thing-in fact, without aprivate sector outcomes couldhavebeenmuch, muchworse. Butthe advent o fprivate sector services is not the result ofan official policy nor, inmany instances, o f a "public-private partnership." Instead, the private sector is a coping mechanism for failing public services. This report discusses instances o f using the private sector, not to escapepublic sector responsibilities, but to meet core public sector responsibilities. How to strengthen the basic institutions o f service delivery incore areas o f government responsibility, one o f the most pressingchallenges facing India today, i s addressedinpart Io fthe report. A booming economy with growing gaps While the economy o f Indiahas been growing rapidly, with broadly based gains in absolute incomes and many outcomes, it i s also the case that progress has differed widely across states, across regions within states, and among different people and social groups. Genuineconcern for these gaps does not suggest a return to the "politics o f envy" and a focus on the redistribution o f income (which, at least rhetorically, drove many o f the policies that ledto neither rapid growth nor income redistribution). Rather, this report follows the shift to an emphasis on equity and on the expansion inopportunity-equal and expanded opportunities to obtain better services and to participate inthe growth process. The fact that some states are growing faster than others does not mean the instrumentsof fiscal federalism are the solution-rather it is a signal that there are opportunities for the lagging states to accelerate their growth as well. The fact that some regions are not prospering implies the need for strategies for these regions. The fact that 17 some people might beprospering-morethan others does not suggest limiting opportunities to those, but rather aggressively expanding the equality o f opportunity for all to access labor markets, assets, and product markets on fair terms and to expand the equality o f opportunity for investing inthe health, nutrition, and education o fthe next generation. Differences across states and regions India's average growth performance conceals very different growth experiences across states. Before the 1980s, growth rates were low-at most 2 percent a year over the decade-in all states other thanthe GreenRevolution states o f Haryana andPunjab and the state o fMaharashtra. Growth was also extremelyvolatile, particularly inthe lower- income states. The first stirrings o freform inthe 1980s signaled a change inorientation of government from direct control on private activity to de facto lifting o f controls and restrictions (Virmani 2004b). This ledto a markedacceleration in growth innearlyevery major state, more so inthe slower-growing states, so that the dispersion ingrowth rates across states decreased. Deeper and more comprehensive reforms introducedinthe 1990s were accompanied by a significant shift ingrowth pattems-in part, because average growth accelerated even further. But the truly dramatic shift was the large gap that emerged ingrowth rates between states. Sharp differentiation across states since the early 1990s reflects acceleration o f growth in some states but deceleration inothers. High-growth performers were a mixedgroup, spreadbetween rich states (e.g., Gujarat, Maharashtra) and middle-income states (e.g., Karnataka, Kerala, and West Bengal) and fairly well distributedregionally. Growth slowed down inthe richer northwestern states o f Haryana and Punjab, reflecting a marked slowdown inagricultural growth. More worryingly, growth failed to pick up in states such as Bihar, Orissa, andUttarPradeshthat were initially poor to start with, with the result that the gap inperformancebetweenIndia's rich and poor states widened dramatically duringthe 1990s (figure 1.9). 18 Figure 1.9: Growth acceleratedinnearly all states in the 1980s, but gaps widened dramatically in the 1990s. Decadal growth ofper capita gross domesticproduct in Indian states, by income group or region, 1970sto 1990s Average Decadal Growth Rates of Per Capita GSDP 6.0 1 5.5 5.0 4.8 -e h E 4 0 0) 5 .c 3.0 2.0 0 1 0 0.0 Low Medium North-High West-High m1970s mi98os 1990s Change in Growth Rates of Per Capita GSDP across decades 3.0 1 2.5 2.4 2.0 $- 2 1.0 L 5 0.0 B b -in m7Osto8Os raostows -1.o -2.0 Note: Low-income states include Bihar, Madhya Pradesh, Orissa, Uttar Pradesh, and Rajasthan. Medium- income states include Andhra Pradesh, Karnataka, Kerala, Tamil Nadu, and West Bengal. North-High includes Punjab andHaryana, and West-High refers to Gujarat and Maharashtra. Source: Author's calculations. Increased dispersion ingrowth rates i s not a failure, but a success o fpolicy. The liberalization policy o f 1991 allowed states to play a larger role indeterminingtheir development paths and attracting investment (Ahluwahlia 2000). The increasinggap betweenthe middle-income andpoorer states is not because growth slowed, on average, inthe poorer statesbutbecause growth accelerated inthe middle-income states. The comparisonbetween India and other regions is interesting. Evenwith the increase, the current gap inincome betweenthe richest and poorest Indian states is much smaller than that o f other large federal countries. Figure 1.10 shows the standard deviation inper capita output among internal provinces, states, or countries within India, Indonesia, China, Brazil, the United States, and the European Union(EU12)-all regions with more than 200 millionpeople. While the variation inoutput across states increased inIndia, it is still substantially smaller than inBrazil, China, and Indonesia-but strikingly larger than the cross-state differences inthe United States or the cross-country differences inthe European Union. 19 Figure 1.10: Inequality across states of India remains low compared with developing countries, but higher than integrated developed regions. Standard deviation of GDPper capita in India and comparator countries, 1980s and 2000s 0.7 1980s 1990s 0.6 0.5 0.4 0.3 0.2 0.1 0 Indonesia China Brazil India EU us Note: Left bars indicate standard deviation in 1980s; rightbars indicate standard deviation in2000s. Source: Authors' calculations using data providedby B.Milanovic. For details on conversion o fprovincial GDPper capita to constant equivalents o f purchasingpower parity, see Milanovic 2005. At the same time, accepting that differences ingrowth performance across stateswill be a natural outcome of a liberalizedpolicy environment does not mean that the low growth rates inIndia's poorest states should also be accepted as a matter o f course. Most socioeconomic indicators are explained overwhelmingly by income (and education, which itselfi s drivenby income). Iflarge differences ingrowth rates betweenrich and poor states were to be sustained over long stretches o f time, these could eventually translate into vast differences inmaterial well-being. Low-income states already rank well below the other states on a number o f social and economic indicators (table 1.4). Unless the poor states improve their performance, it will become increasingly difficult to accelerate poverty reduction and development inIndia. 20 Table 1.4: Differencesinincomeand output amongIndianstates are associated with large differencesinhumandevelopmentindicatorsof healthand education. Ranking of India 's poorest states by GSDPper capita and human development indicators Rank Literacy Infant Under-weight Sex ratioe bY rate' mortality' childrend (0-6 years) GSDP per HDI capita rank" ~ i j ~ Rank Di# Rank DijJ Rank Di# Rank Bihar 13 13 -17.9 13 -5.3 9 -7.4 11 15 5 Orissa 12 10 -1.8 9 -13.4 11 -7.4 12 26 3 UttarPradesh 11 12 -8.0 12 -19.1 13 -4.7 10 -11 8 Madhya Pradesh 10 11 -1.3 8 -18.5 12 -8.1 13 5 7 Rajasthan 9 8 -4.4 11 -12.8 10 -3.6 9 -18 10 West Bengal 8 7 3.8 5 18.9 3 -1.7 7 33 2 Andhra Pradesh 7 9 -4.3 10 1.8 8 9.3 4 34 1 Karnataka 6 6 1.6 7 16.1 4 3.1 5 19 4 Tamil Nadu 5 2 8.1 2 19.4 2 10.3 3 15 6 Haryana 4 4 3.2 6 10.8 5 12.4 2 -108 12 Gujarat 3 5 4.6 3 5.0 7 1.9 6 -44 11 Punjab 2 1 4.6 4 10.5 6 18.3 1 -129 13 Maharashtra 1 3 11.9 1 23.9 1 -2.6 8 -14 9 All-India (avg.) 65.4 67.6 47 927 Kerala 7 1 25.5 1 51.3 1 20.1 1 33 3 Note; GSDP refers to Gross State Domestic Product. Diff.refers to the difference between state-level indicator and All-India average. a: HDIranking refers to the HumanDevelopment Index methodology in the UNDPHuman Development Report 2001. Ranking across 16 major states, including Assam, b: 2001, percent o fpopulation 7 years and older, c: 1998199, per 1000live births, d: 1998/99,percent o f children under 3 years of age, e: 2001, girls per 1000boys in0-6 years group Source; 2001 National HumanDevelopment Report; 2001 Census, 1998199National Family and Health Survey, 1999 Sample Registration System. Evenwithin relatively prosperous states, there are regions that do not share the general dynamism of the state. There are a number o f districts inprosperous states that have infrastructure and human development indicators comparable to the worst-off districts in the poorest states. For example, inone district o f Maharashtra over 80 percent of households use clean cooking fuel, while at the same time less than 10percent of households living inanother district inthe same state have access to this basic provision. Access to toilets, low inmost states with the notable exception o fKerala, ranges from over 70 percent of households inone district to only 10percent inanother inthe state o f Gujarat. Figure 1.11shows the average Infant Mortality Rate (IMR) and the IMRinthe two best and two worst districts for each o fthe major states. There are large gaps inIMR across districts-the worst districts inMaharashtra have IMRcomparable to or higher than the average o f the poorest states. 21 Figure 1.11: Conditions vary widely across districts in states. The worst districts in Maharashtra have IMR higher than the average of the poorest states. Range of IMR in districts within Indian states, 2002-03 160 7 120 i 1 Ms'hangi'i P b KYIIY ':i 40 1 1 Source: Based on Reproductive and Child Health survey data for 2002-03, inRamand others 2004. One important dimension of the income gap within states i s the rural-urban divide. Duringthe 1990s,inmost states, improvements inurbanincomes outpaced rural incomes, widening the gulfbetween rural and urban India (Sen and Himanshu 2005). As with states, the concern is not that rural incomeshave not grown at all, but rather that urban incomes have grown sharply and rural Indiahas been left behind. A second dimension o fwithin-state income disparitiesi s the presence of "backward" regions within otherwise prosperous states. Karnataka, amiddle-income state, has regions (Eastern and Northern) with poverty rates that are comparable to rural areas o fthe richest andthe poorest states inthe country. Maharashtra, India's highest-income state, is home to booming and prosperous Mumbai, but at the same time nearly 50 percent o f the population inits Inland Eastern region i s close to or below the poverty line. Improving incomes inrural India i s critical to reducingpoverty, because nearly three-fourths of the poor continue to reside inrural areas. 22 Figure 1.12: Parts of India have poverty rates similar to those in many Sub- Saharan African countries, while other regions of India resemble richer Latin American countries. Percentage of thepopulation below theIndianpoverty line, by country and Indian state SSA-Highest Onssa-R Bihar-R Assam-R SSA-Medium Mahar-R MP-R Kamataka-R India Nepal AP-R El Salvador Bihar-U TN -R Cambodia Bangladesh WB -R UP-R Gujarat-R Mongolia LAC-Highest UP-u Rajasthan-R China-Rural Lao PDR Onssa-U Indonesia-Rural MP -U Moldova, Rep SSA-Lowest Mahar-U Assam-U WB -u TN-U Rajasthan-U Karnataka-U AP -U LAC-Medium Kerala-R HP-R Kerala-U Uzbekistan Philippines Tajikistan Turkmenistan Gujarat-U Jarnmu-R Pakistan Haryana-R Yemen, Rep Haryana-U LAC-Lowest Punjab-U Indonesia-Urban Punjab-R Sn Lanka Indonesia Georgia Azerbaijan Ukraine Jammu-U Egypt,Arab Rep Turkey China-Urban Hp-u- 10 20 30 40 50 60 Source: Author's calculations based on data in the World Development Indicators 23 The gaps inthe levelof income and inthe differences between rural andurbanareas implyenormous differences across regions ofIndiainpoverty rates--from thebest inthe developing world to the worst. Figure 1.12 shows that rural areas o fAssam, Bihar, and Orissa have poverty rates worse than many Sub-Saharan African countries, while rural areas o f Kamataka, Madhya Pradesh, and Maharashtra have povertyrates that are only marginally lower. Incontrast, urban areas o fHaryana and Punjabhave poverty rates similar to the richer Latin America countries and Turkey.Urbanareas o fmiddle-income states and the rural areas o fbetter-off states have poverty rates comparable to the average Latin American country and are similar to national averages for China and the Philippines.The enormous diversity ofpoverty levels within Indiamakes treating "poverty inIndia" as a phenomenon as ifone lumped Africa and Latin America into a grouping and tried to discuss poverty in"LAFRICA". Differences across households and individuals Wideningdisparities inthe midst o f abooming economy are most evident when comparing outcomes across households andindividuals. The 1990s were a decadewhen economic inequality increasedvisibly, markinga significant departure from trends in previous decades when inequalityhadeither been decliningor remained stable. There is certainly a perception that inequality has increased sharply, very likely drivenby the observation that rich Indians have done extraordinarilywell duringthe boom o fthe 1990s. According to one study, in 1999/2000, the gap inper capita income betweenthe 99th and 99.5th percentile (almost P$7,000 inpurchasing power parity equivalents) was almost four times as large as the gap (P$1,750) betweenthe medianperson and the 95th percentile (Banerjee and Piketty2005). The super-rich at the 99.99th percentile, with growth inincomes o f over 285 percent between 1987/88 and 1999/2000, enjoy annual incomes of around P$160,000 per person. Although the unprecedentedprosperity o f the super-rich i s critical for perceptions of inequality, because it i s largely confined to the top 0.lpercent o fthe population of taxpayers, it is unlikelyby itselfto explain shifts inthe overall income distribution. There are three dimensions to the increase inincome inequality. As discussed earlier, improvements inmean urbanincomes outpacedrural incomes, widening the gulfbetween rural and urbanIndia. Interstate inequalitieswidened as rich states grew faster thanpoor states. Within-stateurbaninequality also rose, reversing a previouslydeclining trend (Sen andHimanshu2005). The combined impact o fthe growingrural-urban divide and rising within-urban inequality has resulted inlarge within-state disparities across people. In fact, correcting for cost-of-living differences across states, the large differences across people within rich states implies that the poorer sections o f more wealthy states are very nearlyas poor as those inpoorer states (figure 1.13). Figure 1.13: The poorer sections of the wealthier Indian states are very nearly as poor as those in the poorer states. 24 Distribution of per capita expenditures, by state, 1999-2000 Note: The box plots show the 25th, median, and 75thpercentiles and upper and lower adjacent values o f the logarithm o f per capita expenditures adjusted for cost-of-living differences across states. Source: Authors' calculations using 1999-2000 Schedule 1.O unadjusted 30-day consumption distribution and multilateral prices indexes for cost-of-living adjustments across states from Deaton and others 2005. Risinginequality by itself does not necessarily implythat poverty reduction duringthe 1990swas lower than it mighthavebeen otherwise. There is a complex relationship between growth, inequality, and poverty, which depends on the underlyingprocesses that lead to growth and changes ininequality (Bourguignon 2004). Microeconomic evidence points to a number o fways inwhich inequality can negatively impact growth, through its effects on the functioning o f institutions, resource allocation, and collective decision- making processes (for a detailed discussion, see World Bank 2005a). Insuch cases, policies that redistributesome o fthe misallocated resources or otherwise correct market failures can improve bothgrowth and the distribution o f income. Nonetheless, even policies that worsen the distribution o f income can be good for the poor, iftheir effects on growth are large enough (figure 1.14). For instance, one interpretation o f China's recent experience with worsening inequality inthe 1990si s that poverty went down because inequality was allowed to go up in order to provide incentives for investment and innovation, which ledto extremely high growth (World Bank 2005c).` Figure1.14: Growth,ifitis fast enough, can reduceincomepoverty,even with inequalityincreasing. 6For the opposite viewpoint, see Ravallion and Chen (2004) who argue that inChina, time periods (and provinces) in which inequality rose most were periods (provinces) of lowest growth and poverty reduction. They conclude that it is not the case that the rise ininequality has permitted higher growth rates; rather inequality has put a brake on the rate o f progress against poverty. 25 Change in inequality in China, 1981-2001 %population living beiow $1 a day 70 - 42 - 2 -E 40 - 60 - 5 38 - 36- 5 0 - 3 , m z 3 4 - ::I *. Z i 3 2 - 40 - x , 0 3 0 - .- . e . . * ._8 2 8 - = 30. 5-Z 26- ._ 24 - + z ;i; 22- . ., India+China, , , , 2 0 ~ ~ " " ' " " ' 0 1980 1985 1990 1995 2WO 2005 Source: Chen and Ravallion 2004 (LHS graph); Ravallion and Chen 2004 (FWS graph). Comparedwith other countries, India remains a society o f low inequality inincome (figure 1.15). Evenso, rising inequality i s of concern for other reasons. Rapid but unequally distributed growth can create tremendous social and political pressures, possibly undermining consensus on economic policy, making reforms unsustainable, and stalling aggregate growth and poverty reduction. This concern i s especially relevant in India, which is undergoing a shift inwhich rewardsto skills are becoming more unequal, butthese changes are occurring ina society that has hada legacy o f social stratification andexclusion. Figure 1.15: Despite a decade of rising inequality, India remains a country with low income inequality by global standards. Gini coefJicients of inequality in income and consumption, by regions and country groups 700 .... .. Notes: latest data for each country for the 1995-2002 period. Source: SIMA Although Indiahas low income inequality, by other measures o fwell-being it is a deeply unequal society. Gender disparities inIndia, particularly inhealth and education 26 outcomes, are among the largest inthe world. Ina typical non-South Asian country, child mortality rates betweenmale and female children are roughly equal, but inIndia, a girl born inthe early 1990swas 40 percent more likely to die betweenher first and fifth birthdays than a boy o fthe same age (Filmerand others 1998).Moreover, there is disturbing evidence o fworsening gender gaps inchild malnutrition (although gender gaps ineducational outcomes have decreased), particularly inrural areas o fnorthern and eastern states where nutritional status has beenimproving substantially more for boys thangirls (Tarozzi andMahajan2005). India's richnorthwesternstates ofHaryanaand Punjab are well known for their extremely skewedjuvenile sex ratios, which are much higher thanmostparts o f the world and havebeen worsening since 1981.Many other measures o f gender disparity (e.g., child mortality, enrollments, probability o f seeking healthcare) also do not appear to be lower inhigh-income states. Gender disparitydoes not seem to b e simply a phenomenon o fpoverty or a problem that economies "grow out of." Another key axis o f stratification inIndia is caste or ethnicity. By most indicators of outcomes and opportunities, Scheduled Castes and Scheduled Tribes (SC/STs) are particularly disadvantaged, compared with the rest o fthe population. Differences in outcomes are due inpart to the fact that SC/ST households tend to bepoorer, to have parentswith lower levels of education, and to live inremote andmarginal environments that have difficult terrain, poor infrastructure, and lower accessibility to the market economy. Indeed, controlling for observed householdcharacteristics and location o f residence, recent data from the Reproductive andChild Health Survey suggest that there i s no significant independent effect o f caste or tribal status on health outcomes such as infant mortality rates. This result, however, should not be interpreted to mean that membership ina social group does not affect health outcomes. L o w income levels are themselves inpart due to inferior returns to productive assets, reflecting patterns o fpast discrimination. Caste-based inequalities can be particularly insidious. There i s evidence to suggest that individuals internalize their chances o f success or failure and transform theminto lower aspirations and expectations, andhencetendto reproducethe inequalities over time and across generations (Rao and Walton 2004; H o f fand Pandey 2004). Guideto the rest of the report From this overview there emerge two important themes that are addressed inthe remainder o fthe report. Part Iaddresses the questiono fimproving service delivery: how can India's performance incore areaslike schools, roads, water, sanitation, health, poverty programsberaised so that its transformation o f services make it as muchthe global admiration as its economic performance? This Report takes the view that to improve service delivery one has to go deeper than sectoral budget allocations or scheme by scheme discussions o fprogram design-when systems are failing, it is not enough to fix the pipes, one needs to fix the institutions that fix the pipes. The dual questions are: how can Indian citizens hold service providers accountable for quality services and how canproviders be equipped (with budget,resources, capacity) to meet those demands? 27 Part I1moves on to the discussion o fcreating more inclusive growth. To have inclusive growth, one must have growth, and sustaining rapid growth i s central to achieving nearly everything else. But growth can be made more inclusive inseveral ways. A focus on reforms that are "equalizing accelerators"-actions that increase growth and makethe benefits more equally distributed. Coming to grips with the poor performance o fthe agricultural sector, which directly and indirectly, has a central role inpoverty reduction. Addressing the lagging states and regions-to continue rapid growth it will have to spread to those states that havenot yet accelerated beyondtheir performance in the 1990s. Expandingopportunities for peopleby increasingaccessto markets andto assets-along with effective assistanceand social protection. One theme that runs throughout the report, bothwhen discussing services and the public sector and incomes andthe market, i s empowerment. Indiahas already embarked, irrevocably, on two transformations that, when successfully completed, will make India perhaps the great power o fthe 21Stcentury. The first i s the transformation from viewing the population as passive recipients o f government initiative to citizens that are socially, politically, and institutionally empowered, not to be acted upon, but to be the actors in improvingtheir satisfaction with the services they expect from the public sphere. The second i s from viewing the market as somethingpeople needto be protected from to something all people should be empowered-through assets, skills, and capacity--to have fair and equitable accessto, with social protection as an integral support. 28 Introductionto Part I:Improved accountability for better services This Report takes a different approach than many reviews o f service delivery, which proceed on a sector by sector, scheme by scheme basis. For instance, the Planning Commission has produceda Mid-Term appraisal o fthe TenthFive Year Plan, which is an excellent overview and analysis o fthe current state o f affairs (Government o f India, Planning Commission, 2005b). Since this Report cannot hope to compete with the reporting on available information, depth o finsider expertise, and encyclopedic scope o f this excellent work, how canthisReport, inthe few dozen pages available, contribute to the ongoing dialogue about reforms o f service delivery inIndia? We hope to emphasize the broad themes inthe diagnostic that cut across all sectors and emphasize the common concepts that animate reformproposals across sectors as diverse as rural water, airports, education, andhealth. The first step to helpfulprescriptionis accurate diagnosis-which needs to go beyond identifying and listing symptoms to understanding the system. Chapter 2 lays out a diagnosis o fthe current state o f service delivery inIndia and argues that, while there are many budgetaryandtechnical reasons why services are weak, a weakness that cuts across manyindividual schemes, programs, and sectors is that therelationships o f accountability betweencitizens, the state, and service providers are not well articulated. Chapter 2 suggeststhat successful reforms will improve service delivery by addressing one o fthe five keyweaknesses inthe current links o f accountability: 0 Unbundling. Don't ask a ministryto be simultaneouslyumpire, bowler, and batter and expect a good game-the roles o f government as "policymaker" who sets the rules o f the game and as one particular organization that provides services needto be clarified. Creatingclearer roles and responsibilities amongst various actors facilitates accountability for hlfillingthose roles. 0 Delegation. The combination o f two aphorisms: "you get what you pay for'' and "what gets measured gets done"-implies that ifone only pays for and measures inputs (wages, assets) then one will get inputs. Unless there is clarity about the delegation o fresponsibility for outputs and outcomes there cannot be accountability for outputs and outcomes. 0 Autonomy to enabled providers. A stress on accountability is not an "attack" on providers, quite the opposite. When providers have clear goals and are given adequate resources the next step i s allowing them sufficient autonomy to accomplish those goals. One cannot holdpeople accountable for not doing something with nothing. 0 Information. Regular, reliable, relevant information is the key to any accountability relationship. This is true inside an organization (so that monitoring systems are essential), betweenparts o f the government (so that a line agency can achieve goals cost-effectively), and between citizens and the state (transparency) and the citizen and providers (informed consumers). 29 0 Enforceability. Ingovernment there are amazing dedicated and competent professionals strivingto do their best to fulfill their responsibilities. Unfortunately, there i s also malfeasance o f all kinds (corruption incontracts, absenteeism from posts, lack o f effort). Ifactions have no consequences this undermines the efforts o f good providers. An ideathat should come easily to a nationinfluencedby a creation story that concludes that no one really knows how the world began i s that there is no uniqueway to implement even a common set o fprinciples. There is no one best system, there is no unvarying recipe for service delivery success: countries havemanaged to create functional schooling with systems that are largely private (Holland) and systems that are largelypublic (Germany), with systems that are largely centralized (France) and systems that are largely local (USA). But, systems have to work coherently as systems-there has to be a strategic choice about the direction for reform and then that has to be implemented. So, even ina single sub-sector one state o f Indiamight choose to pursue service delivery improvementsbyrelying more on local governments while another mightmove to greater reliance onnon-state providers (e.g. Public-Private Partnerships) while yet another might choose some type o f internal administrative reform. Three chapters o fpart Iarticulate how different types o freform can implement the broad principles for improving services through greater accountability. Chapter 3 addressesinternal andadministrativereforms, drawing on case studies o f successful reform initiatives inIndia. Chapter 4 discusses how to make local government systems work to improve services. This isnot advocacy for local governments butrather a discussionofthe question: ifa state were to choose greater reliance on local governments (Panchayati Raj Institutions) as a mechanism for service delivery, how should the implementation o fthat be designed? Chapter 5 discusses, with aparticular emphasis on infrastructure services, the various modes o f engagement with non-state providers-communities and user groups, contracting out, and public-private partnerships. Again, this is not saying these are the only possible paths for service delivery reform, but they are one approach that has proved successful inmany contexts, and drawing on those lessons will help make these successful. Whatever reform i s beingproposed, whether it be devolution to PRIs or contracting out or community level committees or "e-governance" one has to ask the same sets of questions ofhow the proposed arrangements will, for instance, create reliable regular and relevant information and make that available to involve stakeholders ina way that will lead to improved services. 30 Chapter 2. Service Delivery: A Diagnostic Iamconvincedthatthegovernment,ateverylevel, istodaynotadequately equippedand attuned to deal with this challenge and meet the aspirations of thepeople. To be able to doso, we require the reform ofgovernment and ofpublic institutions ... No objective in this development agenda can be met f w e do not reform the instrument in our hand with which we have to work, namely the government andpublic institutions. Clearly, this will be my main concern and challenge in the days to come. Prime Minister Manmohan Singh, June 24,2004. Inone of hisfirst speeches, India's new primeministersignaledthe shift that hasto occur ifIndia'spubliclydeliveredservicesaretokeeppacewithitsboomingmarket:from promises made to promise kept. He also emphasized that keepingpromises will be neither easy nor quick, becauseit requires fundamental shifts inthe way govemment meets its responsibilities for services and its line agencies typically go about their business ofproducing services. One of the most important priorities for the government of Indiais not so much a new "policy" agenda as a shift inthe way that policies are carried out across the board: a shift from outlays to outputs and outcomes, from asset creation to asset operation, from compliance with procedures to a drive for better performance, from policy designon paper to policy implementation on the ground. Part Io f the report discusseswhy this fundamental shift i s necessary and the ways in which this complex and difficult cross-cutting effort canbe successful. Indoing so, the report tries to berealistic about the dire straits o f service deliveryinmany sectors and states without beingcynical or defeatist. While the challenges faced inadequately meeting core public responsibilities are large, as has been demonstrated, the capacity o f India and Indians to address those challenges i s larger. A framework for understandingservice delivery outcomes Ifthe Indian state has a weakness, it is this: most of the institutions and rules- courts, bureaucracies, police-are so riddled withperverse incentives structures that accountability is almost impossible. Pratap Bhanu Mehta, TheBurden of Democracy, 2003. There are many voices inIndia who believe that accountability i s the key "binding constraint" to improved services-but also believe, as the above selection illustrates, that the existing system itselfis structured to avoid accountability. Inthis chapter we focus on services such as education, policing, rural and urban water supply, health care (promotive, preventive, and curative), law enforcement (including civil law), tax collection, major irrigation works, land records, poverty programs, road construction and maintenance, and transport. These services, while different inmyriadways, share three characteristics. First, they are core public services-there is no question that there is, and 31 will remain, a fundamental public responsibility to deliver them. Whether the government chooses to discharge that responsibilityby direct production through its own agencies or whether it chooses to rely on nongovernmental or private sector agents for production (e.g., toll roads, electrical power production, contracting out) isjust a change inthe government's mode o f engagement. The government cannot and should not take a "hands off' attit~de.~Second, these services are locally delivered and require thousands or millions o f interactions between service providers and citizens every day and hence are transaction intensive. Third, the services discussed are more or less discretionary inthat the service provider needs to take local and even individual specific conditions into account-a doctor actually needs to prescribe treatment on the basis o f the condition o f the patient. What determines how well these core, transaction intensive, andpotentially discretionary services are delivered? Start from the range o fprimordial interactions o f citizen and service provider: a girl arrives at school-what determines whether she will learn? A pregnant woman arrives for a check-up-what determines whether she will be treated appropriately? Someone turns on the tap-what determines whether water will come out? A personentitled to rations through the Public Distribution System arrives at the shop- what determines whether she will get what she i s entitled to? A smallholder shows up to register her land title-what determines whether she will get fair treatment? While the exact details are different, ineach case inorder for services to be effective there needs to be a frontline provider who: 0 Knows what the service delivery goal is. 0 I s technically capable o f meetingthat goal. 0 Has at her disposal adequate assets and recurrent inputsto carry out the task. 0 I s motivated to use her capability and available assets and inputs to meet the service delivery goal. When one observes failure inservice provision, it is often easy to point to the proximate cause o f failure as a lack inone o fthose four categories: lack o fknowledge, lack o f capability, lack o f assets or inputs, or lack o f motivation. But an explanation o f widespread service delivery failure that is adequate to bethe grounding of institutional reform needs to be able to say why the proximate determinants were lacking and needs to be able to answer the question of why at a systemic level, for two reasons. First, as the old saying goes: "just because the tire is flat doesn't mean the hole inthe tube i s on the bottom"-an exclusive focus on the visible proximate determinants might miss the much more serious systemic problems. "Business as usual" attempts that focus on proximate determinants andtreat logistics rather than fundamental problems will not be effective. Consider the case inwhich the proximate problem i s that drugs are out of stock. Shipping This distinction is important because the word "privatization" carries metric tonnes o f ideological and political weight and at the same time clouds discussions. Reliance on contractual arrangements with individuals who are not full-time employees o f the state is not "privatization" inthe sense that the government "leaves it to the market." Infact, as chapter 1emphasized, the reliance on weak institutions o f service delivery has implied more de facto privatization of the sector than inmany countries that rely much more on private sector providers as implementers o f government policy. 32 more drugs will not solve the problem ifthe drugs being shipped currently are being sold on the black market. Second, no system creates good services by attacking or underminingthe frontline providers o f those services. A focus on proximate causes o f failure can lead to a counterproductive tendency to put "blame" for failure on the individual provider, rather than to identifythe systemic issues that trap even the best intentioned and capable people into dysfunctional systems. Many, ifnot most, frontline providers o f services are dedicated individuals who struggle against the odds to fulfilltheir duties. So, for instance, the education system cannot thnve bybeing"anti-teacher," butrather it mustbe positivelypro-good teacher-so that good teachers are giventhe mandate, resources, capability, and motivation to perform.' To go beyond a sector-by-sector analysis of the proximate determinants (e.g., health care is poor because drugs are not instock, learningachievement is low because learning materials are scarce, water reliability is low because o f too little maintenance) to a systemic analysis o fthe causes o f success and failure, one needs a framework and a few perhaps seemingly abstract concepts. Delivering services through public sector institutions involves a longroute o f accountability. First, ina democracy such as India citizens must be able to use the political processes to hold their state accountable for the services they receive, an accountability relationship called voice, although it could as easily be labeledpolitics. Second, the state, acting as an agent for its citizens must be able to holdthe organizational providers (be they line agencies, departments, public sector bodies, nongovernmental organizations, for profit firms) accountable for the services they provide using public sector resources, an accountability relationship called compact-this mightbe an actual contractual arrangement(say, betweenthe government and a contractor to build a road) or it might be internal to the government (as betweenthe financing agencies, such as the Ministryo f Finance or Planning Commission, and line agencies). Third, the organizational provider (again, o f whatever type) must be able to hold the frontline providers who are delivering services accountable, a relationship o f accountability internal to the organization called management (World Bank 2003a). * By thesame token, exploring the weaknesses inthe public sector current delivery o f services i s not "anti- public sector," but rather, it ispro-public sector. Deliberate denial, which allows those who benefit from the current system to thrive and leads to an inability to come to grips with the depth o f the problems in public sector delivery, has been a more effective privatization policy than privatization. A brutally honest diagnosis is the first step to a cure. 33 Short route Table 2.1: Three relationshipsin the "long route" of accountability, illustrated with the case of policing Management: Elements of an Voice: Compact: organizational provider to - accountability relationship citizens to state state to organizational provider frontline professional j Citizens delegate an array o f Police forces are given tasks j Withinthe police force via legislation, government j individuals (frontline Delegation orders, judicial instructions, j providers) are given specific and order and ensuring rule and political directives. ; roles (inspectors etc) and j tasks by their superiors. ___________________.-------~--------~----------------------------------- . - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - r - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - . ; Citizenspay taxes (direct Police forces are provided j Policemen are paid a salary budgets (capital and I andbenefits. Financing recurrent). ; Politicians and policymakers choose what actionto take choose what actions to take (constablesisubinspector) with regardto the police based on incentives created choose what actions to take based on the incentives through delegation, financing from day to day based o n created through delegation, information, and incentives createdthrough financing, information, and enforcement o f the state. delegation, financing, enforcement o f citizens. information and enforcement ------------------------------------~-----------------------------------. .__________________ o f the organization. j Citizens receive information Information on the j/ about the quality o f policing performance o f individual services via direct o f police forces via internal policemen is generated j observation o n their own reports from agency heads. internally by direct Informing j interactions with police, observation by their j personal observations on superiors, performance j safety, the media, official reports, official complaints- j reports, etc. as well as possibly _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __-_____----______-__--- information from outside. _ _ jj Citizens can hold the state Politicians and policymakers Individual policemen can be accountable through control police forces through rewarded or punished for j elections (with crime andor appointments, transfers, their performance through policing as one issue) or budget allocations, etc. promotions, favorable Enforcing jj other avenues ofpressure in postings, sanctions or even j a democracy (lawsuits, cases against then. j public agitation). Department has specific control over the police force with limitedpolitical 1 interference (from home ministry). There is, o f course, the "short route" o f accountability infigure 2.1,inwhich "client power" i s expresseddirectly via a market relationship. Ina market relationship, the same five relationships o f accountability are present betweenbuyer and seller-the buyer orders (delegation), the buyerpays (financing), the buyer gains information from the observation and use o f what is delivered, and the buyer can enforce by choosingwhether to make a repeat purchase (as well as, more rarely, legal recourse inthe case o f fraud, nonperformance, etc.). This is not to say that this i s easy or always works well-for instance, inhealth care the buyer (patient) has a difficult time knowing whether the 35 provider is doing the right thing or not, which creates incentives for the provider to over- treat (discussed below). Effective service deliveryrequires balanced relationships o f accountability inall dimensions. Systems that work, work as systems. While there are many ways o fpiecing together a system of relationships that creates accountability (chapters 3,4, and 5 discuss examples o f different paths for systemic reform), all the pieces must fit into a coherent whole. Even the clearest delegation to frontline providers, inthe absence o f adequate financing o fthe necessary assets, inputs, and wages, leaves them unable to perform. Similarly, increasing informationwithout creating any means o f making that information effective inenforcement simplyleads to frustration and cynicism. Increasing enforcement, for instance, on frontline providers through rewards and discipline without clear delegation and reliable information leads to confusion and resentment rather than motivation. So the strength o fall o f the elements o f the relationships o f accountability- delegation, financing, performing, informing, and enforcing-frame the choices made by agents, and one weak link can undermineaccountability. The state of service delivery and diagnosis The broadpicture that emergesfrom our analysis of thepublicfeedback is that the quality of governance in many Indian states is appalling. Conventional technocratic answersfor improving sewice delivery or a call to allocate morefinancial resources may not be an adequate response to thisproblem. The range of reforms required may cover a wide spectrum-from political leadership and administrative structures and incentives to service-specific changes including alternative delivery options. S. Paul and others, "State o f India's Public Services," Economic and Political Weekly, February 2004. Service delivery: access and satisfaction What does a typical citizen encounter whenhe/she seeks services from public sector producers? A recently publishednationwide survey o f citizen access, use, and satisfaction with public services illustratesthe gap betweenthe access createdby investments inasset creation and citizen satisfaction with the quality and reliability o f services (Paul and others 2004). For instance, while inthis survey 78 percent o f households reportedusing government (or government-aided) schools, only 16percent reported being "fully satisfied" with the behavior o ftheir child's teacher-including shockingly low values, such as only 1percent fuIly satisfied inPunjab, 3 percent inOrissa, 5 percent inHaryana, and 6 percent inRajasthan (figure2.2). Similarly, while 72 percent o fhouseholds used the Public Distribution System (PDS), less than one inten were fully satisfiedwith the quantity, quality, or fairness o fthe system. Problems with access and satisfaction tendto bemuchworse inlagging states. The same survey placed Assam, Bihar, Orissa, Rajasthan, and West Bengal bnthe lowest tier ina ranking o f the quality o f service 36 delivery across 16major Indianstates. Moreover, poor households inthe top tier o f states rank access to all five public services-drinking water, transport, education, health, and the Public Distribution System-better thaneven the nonpoorhouseholds inthe bottom five states. While there are millions o f dedicated civil servants-teachers, health workers, policemen, engineers, registrationofficials-attempting to do theirjobs well inspite o f the systems that work against this, it also cannot be denied that all too often attempts to seek services from the public sector encounter workers who are absent, incompetent, indifferent, and outright corrupt. The next few paragraphs resent evidence from various sectors and places inIndiathat illustrate these problems. While the data inchapter 1suggested that people are coping !Y with inadequate services inmanyways, including by large-scale recourse to the private sector when possible, this section emphasizes the specific aspects that add up to low service quality: absence, incompetence, indifference, and corruption o f staff, as well as underutilization and undermaintenance o f assets. Identifyingthe aspects of low satisfactionis key to a di,agnosis. Figure 2.2: While physical access has improved greatly, satisfaction with services provided is very low across all sectors Percentages of thepopulation having access topublic services and expressing satisfaction with those services 60 40 20 0 Physical Access Full satisfaction Source: Paul and others 2004, tables 2-6. That is, the specific examples given here are not meant as an exhaustive survey o f all o f the evidence, nor are they individually meant to be probative about the state of services-which of course varies widely across India and across sectors inIndia. Rather, the specific examples illustrate the specific issues that the weight of the evidence and experience suggests are generic. 37 State of service delivery:providers Absence. Since the pioneeringPROBE report on education infive Northern states raised the issue o f teachers' widespread absence and lack o f attention to classroom activity, these findings have been replicatednationwide, extended to health, and confirmed over time. Figure2.3 shows an estimate from a large-scale nationwide surveyin2003 o fthe fraction o fteachers present inthe school and engaged inclassroom activity during random visits from a research team. On average across India, only about half o f teachers were bothpresent on the school grounds and engaged inclassroom activity. Inthe worst-typically the poorer-states only one inthree teachers were engaged inteaching duringschool hours, while even the best performing states it was only two out o fthree. In the same study, things were evenworse inthe healthsector, as on average 40 percent o f healthworkers were absent altogether. One recent study inRajasthan went further and carried out a continuous facility surveyinwhich each o f 143 public facilities was visited weekly during regular hours for an entire year. This studyreplicated the basics o f previous findings-finding 45 percent o f doctors absent from Primary Health Centers- but also found that at subcenters and aidposts the doors were closed 56 percent o fthe time (and field visits do not account for this, as only 45 percent o fthe time could the researcher find the health worker inthe village). Worse, the patterns o f absences and facility closures were essentially unpredictable, so people could not even count on facilities beingopen on certain days or at certaintimes (Banerjee and others 2004a). Figure 2.3: In a nationwide survey using unannounced visits less than half of teachers were both present and engaged in classroom activity Percentage of teachers engaged in classroom activity, by state p 100 In all these states less than half are B % 80 present and teaching Source: Kremer and others 2004 Incompetence and low effort. Two World Bank researchers have camed out a painstaking evaluation o fthe quality o fmedical care inDelhi (Das and Hammer 2004a, 2004b). They first identifiedproviders by asking people who they went to for care, so that they could generate not an official list o f who was registered, but a roster o f those from whom 38 people actually sought care. They measuredthe competence o fthose medical providers bypresenting these providers withmeticulously prepared "cases" o ffive common and important disease conditions-diarrhea, viral pharyngitis, tuberculosis, depression, and preeclampsia-to see ifthe people providing medical care knew which questions to ask and examinations to make, how to interpret those to make an accurate diagnosis, and how to recommend the appropriate therapy. The findings are shocking. Averaged over all five conditions, a public sector MBBSdoctor ina Primary Health Center (PHC) inone o fthe three poorer neighborhoods had a greater than 50-50 chance o frecommending a positively harmful therapy. What i s perhaps even more shocking i s that a comparison o f Delhiwith a national random sample inTanzania and inIndonesiao f the equivale,nt o f MBBS doctors shows that (a) the "less than fully qualified" people providing medical care are strikingly incompetent, (b) the typical MBBS doctor ina Primary Health Center (not hospitals) inDelhii s less qualified than the typical provider inTanzania and substantially less competent than doctors inIndonesia, and (c) even hospital-based public sector MBBS doctors only about reach the Tanzanian level-and are still below that o f Indonesia (figure 2.4). Intreating diarrhea, a basic health problem that 70 percent o f providers report facing "almost every day," the typical provider recommended a harmful treatment three-quarters o fthe time. Figure2.4: PublicPHCdoctorsinNew Delhiscore less well on their competence than doctors in Tanzaniaor Indonesia Source: Das, Gertler, and Leonard2005. The same study o f medical care providers inDelhi also examined practitioners "effort" bydirect observation oftheir actual clinical practice. The striking finding from observing effort was that, while the private, non-MBBS providers were not very competent in practice, they did what they knew, while the public MBBS doctors didnot. Inthe hypothetical vignettes used to measure competence, about 30 percent o fpublic sector doctors asked the right questions-but less than 10percent did so inobserved practice. In 39 contrast, private non-MBBS doctors knew to ask the right question only 20 percent o f the time, but achievedthat same level inpractice. This low effort becomes even more striking when the public doctors inPrimaryHealth Centers inthe poorer neighborhoods inthe study are examined-there both competence and effort was below even that o f non-MBBS doctors-and bothwere much worse than inrich neighborhoods (figure 2.5). The contrast with the private sector is instructive: since private doctors are directly accountable to the patient, they put ineffort, although they tend to over-prescribe medicines that are ineffective (at best) simply to please the client. One does not want to extrapolate from a single city, because Delhi's Primary Health Center doctors mightbe muchbetter or much worse than inother parts o fthe country. We don't know, and that is itselfan important fact. Figure 2.5: In poorer neighborhoods the competence and effort of all providers is low-but effort is PHCs is always low and "off the chart" in poor neighborhoods The "effortdeficif'in PHCs: " ' ' ` ` ' ' ` effort-in-practiceis less than . . . . . . . .. . . Source: Das andHammer (2004b) Corruption. Although India is among the most rapidly growing economies inthe world, it i s not rated as particularly outstanding when it comes to the control o f corruption. Inthe Governance Indicators o f the World Bank, Indiai s 108th o f205 in"control o f corruption," and inthe recently published Corruption PerceptionIndex India i s tied for 88thplace with countries such as Benin, Mali, and Tanzania. There are many forms o f corruption, and the "retail" corruption associated with services is perhaps not the largest or more important, but it i s the case that corruption has penetrated deeply into the provision o f services (figure 2.6). Transparency Internationalrecently completed a study of the corruption that people face intheir day-to-day interactions with the public sector in a variety o f sectors (Transparency International 2002). Their estimate was that 26 crore was spent on this type o f petty, "retail" corruption inIndia. Inthese estimates, 27 percent 40 of that total goes to the health sector-more than police, taxation, and land administration. Figure 2.6: Allocation of the total estimated outflow due to "retail" level corruption across sectors-the health sector is more than a quarter of all retail corruption E 27% Taxation & Land Admin, 17% Telecome & Rail, 5% Source: Adapted from Transparency International 2002. State of service delivery: assets Ineficiency. A study o f the rural drinkingwater sector inIndian states compared the incrementalinvestment expenditures with the increase inconnections realized; findings varied considerably across states (figure 2.7). These differences might reflect inefficiencies ininvestment planning and execution or other factors, such as assets becomingnonoperational ina short time because o f inadequate maintenance or changes inrawwater availability. 41 Figure 2. 7: Some states get more water connections for their money Expenditure impact coeficients for rural water supply, by state, comparing changes in rural coverage with total government expenditure between 1999-2000 and 2004-05 All India Andhra Pradesh Karnataka Kerala Maharashtra 9 Orissa Ls In Punjab Rajasthan Tamil Nadu Uttar Pradesh Uttaranchal West Bengal 0.0 0.5 1.o 1.5 2.0 2.5 3.0 Ruralwater supply expenditure impact coefficient Note: The government o f India's definition of a fully covered rural habitation i s one receiving 40 liters per capita per day from a source within 1.6 kmor 100melevation. Source: World Bank, forthcoming(c) Low services deliveredfrom assets. By and large, targets andreported goals still focus on the capital stock and its growth, implyingthat the objective is not access to a quality service, but access to physical infrastructure. A good example comes from the water sector. Here, India is well on its way to achieving the MillenniumDevelopment Goals for water, if access to infrastructure i s used as the criteria. It i s likely that inurban areas, access to what the government o f India defines as a safe water source will be 100percent by2007 andwill reach 100percent inrural areas by around2012 (World Bank2005a). However, the definition adopted for the MillenniumDevelopment Goals calls for "sustainable access to safe drinking water" (the service), not simply access to infrastructure. This approach ignores whether consumers are really getting a service from the asset andto what extent this service is sustainable. The emphasis on creation o f infrastructure assets is evident inthe way authorities focus mostly on the designstandards o fthe system interms o f delivery capacity inliters per capita per day. Many different sources provide data on the number o ftaps, but few provide reports on the quantity, quality, and availability o f water distributed. There is often little correlation between the delivery capacity o f a system andthe actual quantity andquality of service consumers receive. Households often haveto revert to secondary sources whenwater from primarysources is insufficient inquantity and/or quality. More than 18percent ofruralhouseholds depend onmorethan one source to meet their needs, an indication o f the nonreliability o fthe primary source (World Bank 2005a). Similar issues about assessingperformance are seen in other sectors: for example, data on the quality o f the roadnetwork i s not widely available to policymakers or stakeholders. 42 Box 2.1: The costs of deferred road maintenance Maintaining India's existing highway network will require around Rs 70 billion, or three times the existing level of spending. While this would imply a substantial reallocation of resources, the evidence suggests important benefits to increasing spending on road maintenance: 0 Insome states Rs 1spent onroadmaintenancegeneratesnetbenefits ofRs7. Road user costs increase by over 50 percent ifroads are inpoor, as compared to good, condition. Increasing the resources devoted to road maintenance will both reduce the costs to road users that arise when roads are inpoor condition as well as reducing the costs of rehabilitation needed because of a growing backlog I of maintenance. Source: World Bank 2003c. I 1 I Under maintained assets. India spends only around one-third o fwhat i s required on road maintenance, implyinga deterioration o f the existingroad stock evenwhile new capacity i s being added. The situation i s particularly bad for national highways, where actual expenditure in2002 was less than one-quarter o fwhat was required (World Bank 2004b). There i s substantial evidence that there are major benefits to be obtained from spending adequately on maintenance (box 2.l), butthis evidence hasbeenaround for many decades and still there i s an overemphasis on asset creation rather thanmaintenance, Low quality of sector outputsfrom providers and assets contributes topoor outcomes. This lack o f attentionto the actual delivery o f quality services leads to low quality of outputs. Inroad construction, indistribution o fbenefits through the Public Distribution System, inhealth, and ineducation, the problem is not so muchthat the "policy" is wrong-the policy objectives are fine, and on paper the design seems adequate-but that the results on the ground do not match the objectives or designs. The result i s that inthe same country that i s gaining a reputation as a place for "medical tourism," the typical public doctor inthe capital city gives substandard care; inthe same country that boasts technical institute graduates who set notjust the national but the global standard, the typical Indian child i s woefully undereducated. This is notjust an issue o fthe availability of schools. Evenwhen there are schools, the learningachievement i s often low, inpart because the accountability o fthe public sector providers i s limited(table 2.2). 43 Table 2.2: Examplesof low learningachievement in primaryschools from around India reveals shockinglylow competence Study and reference population Example of learning achievement Pratichi Trust 2002. Three districts inWest Bengal. Only one of 14 (7 percent) childreninclass 3 and 4 who were not privately tutored could write their own name. Banerjee and others 2004. Urban schools in Less than half o f children inclass IV (47 percent) Vadadora and Mumbai. reached Math Standard Icompetence. Baseline diagnostic test districts of Andhra Pradesh. Only 12percent of children inclass I1to V could do single digit subtraction; only 54 percent could answer a question that required counting to three. Baseline test inJaunpur district inUttar Pradesh o f 72 percent could do n o numerical operations and 51 children age 7-14 ingovernment schools. percent could not read simple sentences. Improvingservice deliveryperformance:morethan business as usual Efficiency in the civil services was always very narrowly defined; it was in terms of contemptfor politics and adherence to rules, but never in terms of increasedpublic satisfaction. Over theyears, whatever little virtues the civil services possessed-integrity, political neutrality, courage and high morale-are showing signs of decay ... a model in whichpoliticians will continue to be casteist, corrupt and harbourers of criminals, whereas civil servants would be eficient, responsive and change-agents is not a viable mode. In the long run administrative andpolitical values have to coincide. N.C. Saxena. "Improving Delivery o fProgrammes through Administrative Reforms in India," 2005. How does one explain the current configuration inIndia-a reasonable performance in asset creation, staffing, and public spending, but a huge gap between assets created and satisfaction with quality o f services delivered? The political scientist James Scott described "bureaucratic highmodernism" as the constellation o f (a) the beliefin"the state" as the progressive social force, (b) the use o f the "modern" civil service bureaucracy as the instrumento f implementation, and (c) the beliefthat social problems are technicalproblems solvable by using government resources to create assets that meet "needs" in a cost-effective way. Indiahas been a classic case o f this approach-as problems havebeen addressed with "schemes" and "missions" that are top down, with narrow objectives, and have been carried out by a civil service with exclusively internal accountabilitymechanisms. Overlaid on that "high modernist" approach has been a broadening and changing electoral politics, which has reduced the tendency o fparty politicians to take an "encompassing" view o f the nation, the society, andthe future and insteadto focus on regional, identity-based, short-horizon politics. While any name i s necessarily also an oversimplification, one way o f characterizing the current Indian predicament with service deliveryis that the system i s "captured highmodernism." 44 Using the basic accountability framework above, there are four key elements to captured highmodernism: weak voice, weak citizen cohesion, blurredcompact, andweak client power. Together, these four elements explain many o f the symptoms: Weak voice. Accountability to the citizens and voters i s systemically weak because delegation is weak (voters do not have a clear idea o fwhat it is the government can accomplish),Jinancing i s weak (there is little connection between financing and delegation or performance), information i s weak (the typical citizen has little or no reliable, relevant, timely, benchmarked information on performance inservice delivery), and enforceability i s weak (because so many other factors outside o f the politicians control affect electoral outcomes). A key question i s what kinds o f changes to the system would make it more attractive for a well-meaning politician or policymaker to engage inreforms that would improve services? Weak citizen cohesion. One tremendously important aspect inIndia increating voice for effective services is the temptation for groups to organize only to improve their benefits fi-om the state, not services more generally. The politics of caste and other identity politics often work so that the benefits o f winning elections are not to improve services but to control access to providerjobs or contracts. Blurredcompact. A major feature o fthe institutional landscape i s that two roles o f the state are blurred: one i s the state as a steward to ensure adequate services, and the other is the state as an organization that produces services. The result i s that the exigencies (and temptations) o fbeing a provider overshadow the responsibilities o f being a steward. For instance, is the Ministry/Department ofHealth (at the center or state) responsible for improving health conditions inthe population or merely the operator o f one provider o f some health services? Weak clientpower. Since neither organizational nor frontline providers depend directly on the served citizens (either as individuals or communities) for their financing and since the served citizens have little capacity to enforce their preferences, the citizens' information about provider performance plays little or no role inthe prospects o fthe organization or frontline providers. As a result, "client power" plays almost no role inaccountability. With "captured highmodernism," the state and its apparatus are treated not so much as a means o f generating public goods, but rather as a means o f generating private benefits for those who control the state. Inparticular, the power to grant contracts, choose beneficiaries, and fill government jobs conveys the potential for enormous benefits. Whenthis power is exercised inthe absence o f any clear standards and external accountability to service delivery, the benefits o f the public sector are for those inthe sector. If problemwithservicedeliveryissystemicandtheresultofallofthesedimensions the o f low accountability inthe public sector, then tinkering solutions will not suffice. There 45 are many reforms that, althoughthey mightbepopular and might have some limited impact, will not be capable o freally improving services. Increased budgets: business as usual with more money Imustcautionthatoutlaysdonotnecessarilymeanoutcomes.Thepeopleof thecountry are concerned with outcomes. Theprime minister has repeatedly emphasized the need to improve the quality of implementation and enhance the eficiency and accountability of the delivery mechanism. Minister o fFinanceBudget Speech, February 28,2005. Nothing about the pattern o f expenditures over time or across states suggests that budget outlays are the major determinant o f either public sector outputs or development outcomes. As pointed out inthe previous India Development Policy Review (World Bank 2003b), real spending inmany sectors has increased substantially-while there is little or no evidence that service delivery has improved commensurately. Nonetheless, many discussions simply assume that, ifthe goal is improved infrastructureor better schooling or improved policing, the answer i s to spendmore. Often the "cost" o f meeting targets, such as the MillenniumDevelopment Goals-for water supply coverage, for rural road coverage, for schooling, for immunization-is calculated simply assuming that business as usual, spending more inthe same structure, will improve outcomes. Insome instances, almost perversely, the policygoals o f sectors are often stated as "increasing spendingto X percent o f GDP," rather than taking outcomes and outputs as goals with outlays, properly, merely as a means to a goal. Ifthe existing systemtreats the "benefits" o f public spending as gains to the public sector itself, then puttingmore money into this system without any accompanying reform will simply reinforce the existing systems. 46 Figure2.8: In 1999-2000 public sector workers hadwages 68 percenthigher than equivalent workers in the private-formalsector-up from 48 percenthigher in 1993- 94 Ratio of weekly earnings in thepublic sector to those in theprivateformal sector, 1993/94and 1999/2000 2.51 2.1 .-P v) 2 E m E 1.5 P L $ 1 % 0.5 0 E 0 Public sector to Publicsector to Publicsector to Public sector to private-formal, private-formal, private-formal, private-formal, 1993194 I999100 PSM corrected, PSMcorrected, 1993194 1999/00 Source: Glinskayaand Lokshin2005. Currently, services inIndia are not "low cost." As elaborated at length inthe previous Development Policy Review (World Bank 2003b), a very serious problem with services i s that, particularly with the implementationo f the FifthPay Commission, Indiahas movedtoward a high-wage civil service. The weekly earnings o f public sector workers in 1999/2000were Rs 1,240 compared with Rs 224 for informal/causal wage workers. These enormous pay gaps are reflectedinpeople's perceptions. For example, ina recent focus group discussion, women inAndhra Pradesh ranked different people's position on a "ladder o f life." Their view was that government officials occupied the top step ofthe ladder-higher than"businessmen," higherthan"money-lenders," andhigherthan"big farmers." O f course, public sector workers are employed indifferent occupations and have more skills and qualifications, so the absolute comparison is perhaps misleading. Buteven amongequivalentjobs the public sector exceeds the private: a factory worker in a wholly central govemment-owned establishment makes 2.5 times more than in a wholly private establishment, regular public sector teachers earn several times more than the average among private teachers (or than teachers hired directly by communities). Infact, one of the drivers o fhigher wage inequality inthe "liberalizing" 1990s was that public sector wages grew much faster than private sector regularjob wages or informal casual work (figure 2.8). Real wages inthe public sector increased by 44 percent over this period-increasing the public sector premiumfor (observationally) equivalent workers increased from 48 to 68 percent (Glinskaya and Lokshin 2005). Highwages with little accountability for actual service delivery makepublic sector agencies an obvious target for patronage hiring, which results at times inmassive overstaffing. "The Mumbai MunicipalWater Corporation has 35 workers per thousand 47 connections, whereas well-functioning utilities have about 3 per thousand. The UP ImgationDepartment employs an astonishing 110,000 people" (World Bank, 2005e). The overstaffing often comes at very low levels o f the organization. According to one analysis: "in most states about 70 percent o f all government employees are support staff unrelatedto public services-drivers, peons, clerks." Another problem with the "business as usual with more money" approach is that innearly every sector there are promising actions that are potentially muchmore cost effective for achieving sectoral goals than simply expanding the existingsystem. But the existing system o f accountabilities is not built to look for cost-effective innovations and then replicate them when found-if anything the opposite is true, as costs to the public are revenues to the public sector. A recent studyreported on a rigorous examination o fthe impact o f an innovative educational technique on learning achievement. Ineach school the bottom 20 students ingrades 3 or grade 4 were taken out ofthe classroom to work with a tutor (called a Balsakhi) for two hours o f the four-hour school day. The tutor is almost always a young secondary school graduate from the community who is giventwo weeks of training and hired for the year at a salary o fRs 500-750 permonth. Byusing the randomizedplacement o fthe program into schools, the researchers were able to estimate the impact o fthis tutoring on boththe students who receivedthe tutoring andthe other students, who for two hours a day had a much smaller class size. The programwas very effective: it increasedthe test scores o fthe students who participatedby around 0.75 of a standard deviation. Moreover, the program was very cost effective-it would only cost Rs 386 for each student to increase by one unit inlearning achievement. Achieving this same magnitude of achievement byanadditional year o f schoolwould cost between Rs 5,500 andRs 7,650-since the gain i s small and the cost i s large. This same study estimated the gains from reducingclass size and found them consistentlyto be very near zero. Even ifwe take the most generous possible estimate o f the learning gains from class size, it would take over Rs 20,000 to get the equivalent gain as the tutoring program producedwith only Rs 386. Other innovations include the use o f a time and date-stamped camera to encourage teacher attendance-this simple innovation innonfonnal schools runbyanongovernmentalorganization-reduced teacher absenceby 20 percentage points and raisedlearning achievement. But the point i s not about these particular innovations. The questioni s whether the systemis looking for cost-effectiveways of improving learning achievement or whether it is focused on simple indicators-like class size-that may only have marginal and highlycostly impacts on learning. 48 Figure 2.9: Enormously cost effective options are available-but the systemic pressures are for more spending, not more and better services from the same spending Cost-effectiveness of educational options to raise math and verbal achievement gg h 2 ,r 14.4 15 :$5 & g % $ 2 10 5 % % E -"t" 8.6 $ 2 0 0 s s Camera Computer Additional Additional Class size .-m $m - S r n mg .r E l Balsakhi E program monitoring assisted year of year of reduction, 50 teacher learning schooling schooling to 30 (32 attendance (Grade 3) (Grade 4) Source:Adapted fromBanerjee and others 2004b; Duflo and Hanna 2005. Paradoxically, many sectors inIndia are simultaneously overfhded and underhnded. They are overfundedinthat unit costs per output are very high-rupees per roadmile constructed or rupeesper household connected to improvedwater supply.Moreover, as we have seen, giventhe low attentionto asset management, the service flow from the assets is low and the cost per unit o f effective service i s often extraordinarily high.In such a situation o fhighunit cost and low effective delivery, the system is likely to be underfundedrelative to what it would take to accomplish outcome objectives. This is because the cost per output gain is so high.InBangladesh, for instance, there have been rapid gains inInfantMortality Rate inthe past decade, eventhough expenditures are much lower than inIndia. Clearly, ifpublic sector resources are beingcaptured as private benefits (for patronage jobs, for services not rendered, or for services renderedto ineligible beneficiaries), then pouring more money into existing institutions runs the risk o f reducingpressures for reform andinfact increasingthe desirability o f "capturing" the sector. Greater electoral competition Thestate is adjudged to be successful, the more opportunities it can create through its own spending,for large numbers ofprivate individuals: if the number of government jobs expands,for instance, even when not required, this is aGudged to be apolitical success, regardless of the opportunity costs thisform ofjob creation imposes on others. Thestate existsprimarily to satisfi theprivate interests of collusive interest groups. ...The net result has been that almost never has that state been governed by apublic philosophy; it is rather a high stakes competitivegame in which individuals or groups seek advantages alongparticularistic lines. Pratap Bhanu Mehta, TheBurden of Democracy, 2003. 49 One of the puzzles o fIndia is that active, functioning, highlycompetitive, electoral democracy has not served as an effective accountabilitymechanism for improving service delivery. India i s the world's largest democracy and, seemingly against all odds, is one o fthe few continuously democratic developing nations, an achievement o f which all Indians arejustifiably proud. But, as discussed inchapter 1, service delivery outcomes insome states ofcontinuously democratic Indiarivalthose of"conflict" economies or places ruled by autocrats. Authoritarian states on the "right" (Indonesia) and "left" (Vietnam) have far outstrippedIndia's achievements, notjust economically, but in outcomes inhumandevelopment. At its worst, with "captured highmodernism" the market provides the services the state is intendedto provide (education, water, health) or those services are not provided at all (security, contract enforcement)--while the state i s a "market" for control over the private benefits o f allocating patronage jobs and giving contracts for asset creation. Since no one partycan credibly promise to reform this system and base their claims for power on improved services to the general public, electoral competition proceeds by making appeals to voters along other dimensions. Thus, inspite o f a situation inwhich the poor vote with more frequency than the rich, in which voters can choose among a wide array o fpolitical parties, and inwhich an "anti- incumbency bias" prevents any particular party from becoming entrenched, few see "more democracy" inthe narrow sense o f electoral competition as a solution for improved services. Of course, "less democracy" i s also not a solution. Somehow there mustbemore accountability withindemocracy. Butthat will require some change inthe ability of citizens to use real information to holdpoliticians and the state accountable for services. Piecemeal reform and "recapture" The discretionary powers of the bureaucracy and the attendant opportunitiesfor rent seeking behavior provided by the existing system arepowerful reasonsfor thefunctionaries to opposeany change which reduces theirpower. Report of the I991 Vaidyanathan Committee on Pricing of Irrigation Water, GoI, Planning Commission, 1992. The problem inIndia i s not a scarcity of reform initiatives. Since there are enormous numberso fhighlycapable and well-meaning people inthe public sphere-as politicians, as civil servants, as frontline providers-there are an enormous number o f successful programs and projects innearly every state o f India. Even inthe worst-functioning states of India, there are many examples o fparticular programs or projects that succeed against the odds. The problem is that most reform initiatives depend on a single reform "champion" (either individual or organization) and either exist outside the system altogether-and hence have difficulty coming to scale-or are dependent on a particular person-and hence are "reabsorbed" into the system after the champion moves on. Efforts at improving services often emphasize instruments that privilege technological quick fixes, internal management changes, and the creation ofparallel channels o f deliverythat bypass weak public institutions. 50 The failures of the public delivery systeminIndia today are an outcome o f a systemic breakdown in accountability relationships within this institutional framework. Deep, systemic problems require deep systemic solutions. This doesn't meanthese solutions are "big bang," attempting to change everythingat once, but it does meanindividual efforts, however incremental, needto be part o f a larger vision and strategy for systemic reform. What are the elements o f a strategy for improving services? There are five elements that cut across promising reform initiatives that, not surprisingly, are the counterparts o f the current weakness: Clearer delegation-the absence o fdelegation that would support accountability is seen inthat there i s only internal accountability for process and inputswithout any clear monitoringandresponsibilityfor outputs and outcomes. Stricter unbundling-a distinct separation-ofthe roles o fthe government as the entity that sets goals, gives financing,,enforces the "rules o fthe game" and the role o f government as a direct producer or services. When the umpire i s a player everyone knows who will win. More autonomy for providers to use flexible means to reach their goals without undue political interference indecisions. Greater external accountability through better information. Better enforceability inthe system-whether between citizens and their elected officials or directly on frontline providers-without that link many initiatives founder. The answer is not "privatization" or "decentralization" or "e-governance"-any o fthese can be a way forward. The important thing i s choose a designo f decentralizationor o f administrative reform or o fpublic private partnerships that creates a system that increases accountability by improving those five elements listed above. Our premise is that it is not so muchwhich one chooses, but making a decisive choice and thenimplementingit well is important. "Decentralization" i s not the only way forward, nor appropriate for all sectors, but a well-designed decentralization i s a way forward. Chapter 5 discusseshow to make decentralization work. "Public-private partnerships" are not the only way forward, nor appropriate for all sectors, but a well-designed partnership i s a way forward. Chapter 6 discusseshow to makepublic-private partnerships work. "Administrative reform" i s not the way forward, nor i s it appropriate for all sectors, but a well-designed administrative reform i s a way forward. Chapter 4 discusses these reforms. 51 Chapter3. Improvinginternalaccountability: "Don't fix the pipes,fix the institutionsthat fix the pipes" The existingprimary school system is crying outfor radical reform. Amartya Sen, 2002 Policymakers and observers o f contemporary India acknowledge that strengthening service delivery inIndiarequires, as Amartya Sen argues, notjust new schemes or better logistics or managerial fixes but radical reform. Dr.Manmohan Singh, inone o f his first public speeches as primeminister, emphasized the urgent needto reform the instruments (government and public institutions) ifIndiais to respond effectively to the challenges o f development. The portrayal o fthe systemic problems inchapters 1and 2 should not lead to cynicism or pessimism-there are continually new reform initiatives, and many successes.Politicians and top policymakers can make reform happen. Two studies on successful reform inIndia conducted by the World Bank (2005a, 2005b)" draw lessons fi-om recent experimentsimplementedbythe government and civil society organizations andprovide some insightsto how reform canbe achieved. Followingthe five elements o f strategies outlined inchapter 2, three ongoing initiatives have demonstrated some success: Strengthening theflow of regular, reliable, relevant information aboutperformance on outputs and outcomes, both internally and externally. Reorienting internal monitoring and evaluation toward outputs and outcomes-not only expenditures and process-strengthens internal accountability (of both the compact and management). When citizens have better information about how well governments are performing their responsibilities, citizenvoice canimprove enforceability. Unbundlinggovernment's roles as the overall steward of the sector and as the provider of sewices. Separation creates clearer lines o f accountability. When roles are properly distinguished and goals clearly specified, it i s possible to give much greater autonomy to civil servants. Process engineering (often taking advantage of new technologies) that creates greater management control of serviceprovision. But simply adding new technologies to archaic systems can be dangerous-the biggest successes occur when the introduction o fnew technology is usedas an opportunity for fundamental change. Reforming institutions i s a long-term process, and radical reform does not suggest that "big bang" or "stroke of the pen" efforts areneeded (or would succeed). Success requires the initiation o f a sequence of reform measures that build to radical institutional reform in lo section This draws heavily on two recent studies on service delivery successes conducted by the World Bank:Amplifjiing Voiceand Client Power (20050 andReforming Services in India: Drawingporn Lessons of Success (2005g). 52 the long term. This pragmatic approach has been dubbed strategic incrementalism. But strategic incremental reforms also differ from piecemeal reform measures, which tendto be ad hoc and personality driven. Inpolitical environments that are resistant to radical reform and tend to undermine and reabsorb piecemeal reforms, strategic incremental reform privileges pragmatic reform initiatives as part o f a larger, sequenced chain o f events. The emphasis is on strengtheningthe weakest link inthe chain, which inturn can work to develop the conditions necessary for long-term, institutionalized reform (World Bank 2003a). The challenge India faces is to identify the weakest link inthe chain o f accountability, then start there to develop initiatives that contribute to a larger reform agenda for institutional change. Strengthening accountability for service delivery: Three Rs of information Providing citizens with access to regular, reliable, and relevant information has proved significant instrengthening citizen enforceability on civic agencies. Information i s better ifitisregular,allowingcitizenstotrackprogressfromyeartoyear.Knowingthatthere were transmission and distribution losses o fX percent or that water was available Y hours a day or that X percent o f children were immunized is interesting. Butregular reporting o fprogress toward a maintained goal i s muchbetter. This section examines two such initiatives, developed by civil society organizations, which helped strengthen the linkbetweengreater information andmore effective services. Citizen report cards. One o fthe most successhl experiments instrengthening public services through information flows can be found inthe work o f the Public Affairs Center (PAC), a Bangalore-based nongovernmentalorganization (NGO). In 1994 the PAC introducedthe concept o f report cards on government performance to the citizens o f Bangalore. Three report cards havebeenpublished, leadingto a dramatic improvement in the quality o f services provided by city agencies. PAC's experience highlightsthe importance o fregular information flows. Any sufficiently well entrenched system can shrugoff a one-time demonstration o fits poor performance; only when information is expected to be regular can it help drive improved system performance. Interestingly, performance increases inBangalore have continued, risingby almost 15 percentage points from 1999 to 2003. Progress has differedby sector. Public hospitals, for instance, improved from 25 percent to 34 percent satisfaction from 1994 to 1999 butjumpedto 78 percent in2003. Figure 3.1: Increases in citizen satisfaction with services in Bangalore following introduction of a citizen report card (CRC) 53 -g 50 6o1 49 iF .!4 40 c ," 30 m zg c 20 10 0 1994 1999 2003 Year Source: Paul 2006. Education report cards: A toolfor public mobilization. Pratham, a national NGO working on primaryeducation, has recently started an interestingexperiment to strengthen information flows on the status of primary education. In2004 it conducted rapid assessment surveys in 18 states on the status o f schools andthe learning ability of students. These surveys providedthe preliminarybasis for developing an education report card at the community, block, district, and state levels. The effort has been institutionalized inthe form o fthe Annual Status o f Education Report (ASER), with the first nationwide report publishedin2005. The report includes positive news about access (93 percent o f children aged 6 to 14 were enrolled) but, as expected, less positive news about functioning and learning achievement (cited inchapter 2). The way this program explicitly addresses the three R's-reliability, regularity, andrelevance-is interesting as an illustration of promising initiatives. The program uses local partners, outside government, to enhance the reliability o f the information. The first ASER report says "ASER will be conducted on an annual basis until 2010,'' emphasizingthe importance of regular information. Any one-off claim about levels o f learningachievement will generate some interest but also controversy about how learning was measured. But repeated measurements on the same standard can be used to buildpressure by comparing performance not only across statesbut also within each state. The ASER project also goes to great lengths to make the informationrelevant to governments, communities, and parents. The program's strategy for dissemination and community mobilization includes the preparation o f annual regional reports and brief summaries at the district level. ASER aims to provide real-time feedback to parents about the performance o ftheir children onbasic literacy andreading skills. Results from these surveys will beused as a tool to mobilize communities to advocate for strengthening the primary schooling system. This i s not the only or necessarilythe best approach. A rigorous impact evaluation o fASER i s beingconducted through a collaborative program developed by MIT-Poverty Action Lab, the World Bank, and Pratham. The impact evaluation team recently conducted a baseline survey to initiate the process. Results suggest that this experiment has indeed generated widespread public interest ineducation and discussions on learninghave increased dramatically invillage meetings-but it i s too early to know what the impact on learning will be. This type o f initiative for generating regular, reliable, relevant information for (and with) citizens about the quality o f services and their outputs and outcomes holds great promise. 54 Right to Information Act. The government o f India recentlypassed an important piece o f legislation: a right to information act. This i s an important, ifnot historic, step forward because it commits the government to an unprecedented level o ftransparency. Formerly nearly all information generated within the government was aimed at internal accountability, so that reports were generated to monitor performancewithin government. The right to information will change the extent to which citizens and groups can use this same information for external accountability. The more the government creates information that is reliable, regular, and relevant, the more useful the legislation can be. Improved compact: unbundlingthe roles of government Improvingthe compact requires a clear separation o fthe roles and responsibilities o f service providers and the tiers o f government. Each tier should have clearly delegated and financed roles and responsibilities and the requisite autonomy to perform effectively. Several reformmeasures show some promise inthis direction: 0 Autonomy for accountability. 0 Outcomes budget. 0 Separation o fregulatory and operational functions. 0 Agency restructuring. Autonomyfor accountability. One o fthe most critical factors responsible for the current breakdown o f accountability relationshipsinthe compact is the frequent transfer o f civil servants, often for blatantly political motivations. There is no shortage o f thinkers (policymakers, government agency staff, politicians, academics) proposingways to improve the public sector. One reform suggested nearly unanimously is to lengthenthe time o fservice inposts o fcivil servants. Doingso has three benefits. First, most reform initiatives depend on a reform champion andthis person i s very often a career civil servant. Behind the success o f many reforms inIndiaone frequently finds a capable, energetic civil servant who was left inplace for years. Ifvested interests and the bureaucracy perceive that a reform champion can or will betransferred soon, their ability to "slow roll" reform is strengthened. Second, although civil servants serve governments and governments are controlled by democratically elected politicians, there is growing concern that the ability to transfer civil servants as reward or punishmentfor actions that displease politicians underminesthe integrityo f the civil service. Thirdis a pragmatic factor: when average postings last only a year or two, the ability o f career civil servants to really master a subject is called into question. The temptation is to not invest inthe substance butjust keep the machinery rolling untilthe nextjob. Reducing transfers inthe civil service i s thus a priority issue on any reform agenda. Karnataka i s one o f the few states that have successfully moved toward resolving this problem (World Bank 2005g). Karnataka, which initiated reforms in2001,owes its success partly to the political commitment o fthen ChiefMinister S. M.Krishna. H e announced a blanketban on general transfers for 2001-02 and, despite opposition, maintained this ban through the next year. More important was the fact that critical institutional changes were introduced along with the banto regulate the transfer process. 55 These changes included creating cadre management authorities (CDMAs) to screen transfers for all regular civil service posts; raising minimumtenures across groups o f civil servants; notifying cadres exempted from transfers; imposing a qualitative ceiling on transfers; and creating a transfer database. Despite basic political commitment and institutionalized reform measures, Karnataka has yet to tackle the more politically contentious issue of senior Indian Administrative Service (IAS) transfers. A difficulty with givingmore autonomy and flexibility to providers ingovernment programs is that the current payment method o f salary andbenefits only, combined with lifetime employment security, provides very weak incentives for performance. This is not to criticize government workers or suggest that all or even most would shirk their duties, but clearly-as evidenced by data on absenteeism-some do. Treating these workers the same as those who perform well can undermine morale, the sense o fprofessionalism, and the intrinsic motivation o fothers inthe organization. Intriguinginitiatives are proposed innew schemes. The Accredited Social HealthActivist (ASHA)being created underthe NationalRuralHealthMission, launchedinApril 2005, reflects a more innovative approach. The ASHA will not hold a regular government post; instead, she will be eligible for a range o fhonorary payments linkedto her success in achieving higherlevels o f service coverage. For example, she is asked to identify pregnant women, counsel them on safe deliveries, and accompany them to qualified institutional facilities for delivery.The ASHA will receive a cash payment for every institutional delivery completed with her participation. This payment will be linkedto cash incentives to the mother, to compensate for the increased cost to the family o f institutional delivery. Demand-side incentives are also being more widely introduced. The national Reproductive and Child Healthprogram includes a scheme to provide vouchers to pregnant women living below the poverty line, vouchers they can use to pay for qualified delivery services at accredited providers both government and nongovernment. Interestingly, government facilities will receive additional revenue from this scheme, but the additional revenue i s based on their success inattracting patients, which will require improving how women perceive the quality o fpublic sector services. Outcomes budget. Another crucial weakness inthe current institutional structure that has contributed to the breakdown o f accountability relationships is the absence o f clearly articulated outputs and outcomes on the basis o fwhich performance can be assessed. Recently, the government initiated an important reform measure to address this weakness through the announcement o fthe outcomes budget inAugust 2005. The outcomes budget aims to make line agencies more performance orientedby increasingthe clarity o f delegation. It does so by makingexplicit the objectives, outcomes, and outputs expected from public expenditures and providing financing for those objectives. To be effective, this effort will have to be accompanied by increased information about how well the objectives were met. Enforceability will need to be exercised by usingperformance as the basis for determining subsequent budgetary allocations. 56 Box 3.1: Launch of outcomesbudget The government recognizes that announcing an outcomes budgetis merely one step on a long road. It highlighted inthe announcement the additional steps necessary: "Converting outlays into outcomes i s a complex process, which differs from ministry to ministry and program to program. Some o f the important steps inthis conversion process are as follows: 0 Outcomes to be specifically defined inmeasurable and monitorable terms; intermediate outputs should also be defined wherever required. 0 Standardizing unit cost o f delivery. 0 Benchmarking the standarddquality o f outcomes and services. 0 Capacity building for requisite efficiency at all levels, interms o f equipment, technology, knowledge, and skills. 0 Ensuring flow o f right amount o f money at the right time to the right level, with neither delay nor "parking" o f funds. 0 Effective monitoring and evaluation systems. 0 Involvement o f the communityharget groups/recipients o f the service, with easy access and feedback systems. Efficient conversion o f outlays into outcomes would, therefore, require making delivery systems effective with appropriate structures andprocesses, strengthening financial management systems, increasing the use o f information technology, and meaningfully involving all the ministries, parastatals, state governments, local bodies, Panchayat Raj institutions, and self-help groups, incritical decision-making and implementation processes." Source: GoI, Ministry o f Finance, 2005. The creation o f an outcomes budget i s an important step forward; however, developing quantifiable outcome and output indicators is a complex and difficult process (box 3.1 articulates some o fthe steps). The keyto a successhl outcomes budget lies inclear and concise identificationo f outcome indicators. Indicators also need to be tangible and realistic. Here the outcomes budget for 2005-06 falls short. Some very precise, quantifiable indicators havebeen articulated, but their benefit i s offset by vague indicators for other aspects o fthe same service. The outcomes budget for the health sector illustrates some o fthese weaknesses, where the objective o fthe fundingi s described as "funding o f the institutions" (table 3.1). The success o f the outcomes budget as an initiative to improve services lies ineffective monitoring andinformation flows on departmental performance and inthe ability o f line departments to articulate quantifiable, tangible, and realistic indicators-which as reforms elsewhere inthe world have shown, is necessary, but difficult and time consuming. 57 Table 3.1: A sample of outcome indicators developed by the Ministry of Health and Family Welfare reveals the devil in the details no. program Name Of scheme/ Objective/outcome Quantifiable deliverables I NationalVector- Reduce the incidence of Annual bloodexamination rate Borne Disease malaria (ABER): 10%ofpopulation covered Control Program- under the program Malaria Establishment of at least one Drug Distribution Center/ Fever Treatment Depot ineach village inhigh-risk areas XI Hospitals and Fundthe institutions for Provision of referral services Dispensaries providing secondary and Delivery of secondary and tertiary tertiary health care health care services services Source: GoI, Ministryof Finance, 2005. There is growing international experience with budget reforms as a tool for improving services. Because such reforms are difficult, mistakes will be madebut as one expert recently advised: "India should make its own mistakes, not repeat those o f others." Drawing on the experience o f countries that have worked out the kinks will behelpful. There are two generic issues. The first response to pressure for outputs is simplyto distort the reporting o f outputs-a widespread problem inIndia and elsewhere. Avoiding this outcome reinforces the need for reliable monitoring systems that are not subject to manipulation. A second danger i s to respond too narrowly ifoutputs are defined ina way that does not fully reflect the desired outcomes. For example, it i s very hardto define the output o fhospitals and dispensaries ina way that leads to clear, quantifiable deliverables (see table 3.1). Moving to simple output measures can distort incentives. Agency restructuring and unbundling: separating regulatoy and operationalfunctions. Overlaps inthe regulatory and operational functions o fpolicymakers and providers tend to diffuse accountablity relationships, particularly for infrastructure-dominated activities such as the provision o f water, sanitation, and electricity. Often, the politics o f patronage results inthe provider losing its autonomy and acting as an extension o f the policymaker. Similarly, providers may acquire political influence and capture the policymaking process." Insuch circumstances, reform instruments that aim to separate the roles o fthe policymaker andprovider can go a long way toward improving the compact. Regulatory functions could include delegation o f responsibilities and finance and information about performance and enforcement. One o fthe most successful reforminstruments through which this separation has beenachieved i s public-private partnerships (PPP), discussed at lengthinchapter 5. ' Butunbundlingis important inall sectors. Dothe peopleincharge o fthe health ministries think o fthemselves as stewards for the health o f the population or asjust one among manyoperational providers tasked with runninga complex organization whose purpose i s to provide services? While runningan orgnization that provides health service i s a complex and difficult task-which can overwhelm even the most capable-it i s just "Foramoredetaileddiscussion, seeWorldBank(2003a, pp.159-79). 58 one small part ofthe health system. But when bundled, the pressingcan drive out the important. Moreover, as emphasized inrecent work by V. Kelkar, one would not expect the heads of major private sector firms to be put incharge o fregulatingtheir industries. Butwhen regulatory andoperational functions are not unbundled,the temptation to regulate the sector to privilege the provider organization (say, by imposingburdens on competitors or protecting the budgetfor exclusively public sector providers) is often too tempting. Weak internal managerialprocesses coupledwith low performanceincentives have contributed inno small measure to weakening accountability relationships. Reform measures aimed at redressing this weakness by restructuring agency processes have proved succesful inimproving services. From an analysis o freforms inmultiple statewide agencies, including the registration department inMaharashtra and the Karnataka State Road and Transport Corporation, as well as city agencies inHyderabad, Bangalore, and Surat, the World Bank (2005b) highlights requirements for successful agency reformprocesses: 0 A strong management teamthat can support the initiatives o fthe leadIAS officer. 0 Re-engineered intra-organizationprocesses. 0 Empowerment o f senior management through the creation of centralized monitoring systems. 0 Improved interagency coordination. Efforts to restructure agency processes without these components have failed. This i s highlighted inthe report's cross-state comparison o f the experience ofrestructuring the stamps andregistrationdepartments inMaharashtra and Karnataka. InMaharashtrathe experience was extremely successful, but the one inKarnataka failed because the reform process didnot attempt to radically re-engineer managerial systems or to gain staff support for the restructuring initiative. And the tenure o f the U S officer leading the reform was cut short. This experience highlights the importance o fmanagerial autonomy and radical restructuring. Tinkering at the edges o fthe agency without radically changing intra-organizational systems cannot leadto sustainable reform. Using technology to strengthen services by process engineering The concept o fe-governance has takenthe developingworld by storm. For many, e- governance i s the panacea for governance reform because it simplifies transactions between government and citizens and in so doing encourages accountable and transparent transaction processes. But simply putting an "e" before a word does not make it better. The key lesson from the use of technology (including information and communications technology) to improve governance i s this: unless the change intechnology i s accompanied by a systemic revamping o fprocesses to eliminate the opportunity for service providers to abuse their discretion, e-governance will not be successful (Shah 2005). Computerizing dysfunctional systems produces computerized dysfunctional systems. The experience o fIndianstates with implementinge-governance substantiates this position. 59 Successful examples such as the e-sewa inAndhra Pradesh and the Bhoomi land registration process inKarnataka have been much discussed. But there are also many failures, which can be equally instructive. In Gujarat an attempt to reduce corruption and delay inthe collection of entry taxes at the state border beganwith promise, but since the incentives o f the agents to divert revenues into their own pockets (abetted bypowerful others) were not checked by a complete re-engineering o f the process, discretion crept back into the system. Similarly, an attempt to computerize the registration department in Andhra Pradesh didnot produce the expected benefits; substantial bribeswere still collected, even inthe Computer-Aided Administration o f Registration Department transactions. Thus computerized transactions do not helpcreate the right incentives for good performance; instead, a significant revamping o f organizationalprocesses coupled with computerizedtransactions can go a longway toward creatingtransparent and accountable governance systems. PRAJA, a Mumbai-basedNGO that collaborated with the Brihan-Mumbai Municipal Corporation to set up an online complaint and monitoring system (OCMS), has been particularly successful inthis regard. Organizational restructuring and the introduction o f new technologies along with the OCMS system contributed to the success o fthis initiative. Another critical area inwhich e-governance can be helphl, ifundertakenina context o f business reengineeringprocess, i s procurement. Inrecent years, there has been increasing recognitionthat procurement i s a critical component for effective service delivery since budgets get translated into services inlarge part through the operation o f the procurement system. A constructive strategy for improving procurement outcomes combines a simplification and standardization o f rules, increased transparency, and enhanced opportunities for participation by groups affected by decisions about what to purchase and the quality o f the goods, works, or services that are delivered. Utilizing modem technology, inthe form o f e-government procurement, canbe a vital tool inimproving procurement outcomes since it supports simplification, transparency, and internal and external monitoring and management when done as part o f a business reengineering process. Procurement reformhas proved to be especially effective inreducing costs and improving the value for money when it takes place alongside complementary reforms in budgetingand financial management that allow agencies to retain savings obtained from improved procurement and reduce the costs engendered by late payments to suppliers. Countries as disparate as Uganda, the Philippines, South Africa, and the U.K.have generated large savings and significantly improved service delivery through sustained efforts to improve public procurement. Indiahas already begunto take important steps towards establishing a public procurement system that contributes to improved service delivery. Kamataka andTamil Nadu have ledthe way inintroducing comprehensive laws with simplified rules to guide procurement practices. States, such as Andhra Pradesh have introduced e-government procurement systems and have generated cost savings o f almost 10%through their application. Inaddition, the Right to Information Law has established the foundation for greatly enhancedtransparency inregard to all aspects ofpublic purchasing. 60 Box 3.2: Jawaharlal NehruNationalUrban Renewal Mission: An integrated reformeffort to improve urban services The proposed Jawaharlal NehruNational UrbanRenewalMission (JNNURM)marks an important step toward improving urban services through strategic institutional reform. The JNNURM comprises a set o f reform instruments aimed at providing incentives for cities and states to work according to integrated plans that lead toward broader institutional reform. Reform measures include efforts to strengthen voice by creating formal urban institutional mechanisms for citizen participation. This involves establishing rigorous disclosure processes throughthe enactment of a public disclosure law to ensure that financial plans o furban local bodies are prepared with citizenparticipationand that performance information i s regularly disseminatedto all citizens. JNNURMalso aims to improve the compact through reform instruments that strengthen accountability relationships between policymakers and service providers. Reforminstruments include drawing up PPP models for development, management, and financing o f infrastructure development, and strengtheningthe role o f urban local bodies by ensuring the adequate devolution o f funds, functions, and functionaries. JNNURMalso aims to introduce reforms infinancial andaccounting practices withinstate agencies through information technology applications such as geographic information systems and management information systems, as well as modem accrual-based double-entry accounting inurban local governments. While the proposed reformmeasures represent a well developed, integrated effort at strategic, incremental reform, success will largely depend on the extent to which reformmeasures promotedby state governments are developed as part o f a larger integrated vision for institutionalreform. Isolated reformmeasures, even inareas suggested by JNNURM, may not succeed incausing long-term change. Handling the political economy of internal and administrative reform One o fthe most important lessons to learn from the cases discussed above is that strategic reform measures canbe implemented even inpolitical environments that are largely resistant to radical reform. The challenge lies increating the momentum necessary to initiate reform and inscaling up successful reform initiatives. This i s a difficult task. For reforms to be successful, they require strong political commitment and public support. However, the everyday reality o fpolitics inservice provision makes attaining this support somewhat problematic. Electedpoliticians and top policymakers inIndia, from the national to the state to the local level, are enormously capable, energetic, and ambitious. When services fail inIndia it is not becausethese people keep making mistakes or because the peoplewho control the state cannot innovate. Rather, the incentives for these people to innovate are few. The costs o frunninga reform program inthe current political environment are high.Every viable system has entrenched interests, and often these interests oppose change. Success i s uncertain; many reforms do not work as planned immediately but require sustained commitment, which i s difficult to maintain ina politicized climate. The electoral benefits of reform often seem dubious relative to the immediate benefits o fjobs and contracts 61 allocated to supporters. Evenfor a more than typically public-spirited politician or top civil servant, the case for taking on reform is weak. Another factor constraining reform initiatives i s the difficulty o fmobilizing sufficient public support. As discussed earlier, as the services deliveredby the public sector deteriorate, people increasingly opt out o f government provisionto rely either on the private sector (education, health) or their own investments (tube wells, water storage tanks). Thus there are few incentives for more privilegedcitizens to mobilize infavor of radical reform. Similarly, for a parent paying for achild to be ina private school, the private incentives to spend time andenergy to improve the public sector are diminished. This is particularly the case for reform measures that involve heavy capital investments and user charges. For instance, reform efforts aimed at providing pressurized piped water 24 hours a day, 7 days aweek, must cope with the fact that the elite inurban areas already have on-demand pressurized water available; by investinginboreholes, water storage tanks, andpumps, the rich ensure that their supply is continuous, even though the public supply flows only a few hours a day. Inthis case payingmore per liter for better service is unattractivebecause they have sunk costs inmitigating government failure. A recent World Bank report (200%) on the water sector (including irrigation) emphasizes the risks o fthe increasing pressures on the sector. It also recognizes that the political problems ininitiating reform are difficult to surmount and that simply repeatingthe conventional wisdom of technocratic solutions i s unhelpful.The report opts for "principled pragmatism"-"principled" because gettingthe principles o f reform correct is essential, and "pragmatism" fiom experience and political necessity. The report neatly summarizes some aphoristic rules or principles o f "principled pragmatism" or "strategic incrementalism" that reforming governments should keep inmindduringthe reform process (box 3.3). 62 Box 3.3: Twelve lessons from reforms in a difficult sector A recent World Bank report outlined the difficulties that India faces and will increasingly face with water resources. The report acknowledged that successful solutions will be difficult and that the art o fpolitical economy of reform is as important as the science o fhydrology. It specified 12 lessons: 1. Water is different. 2. Initiatereformwhere there i s a powerful need and demonstrated demand. 3. Involve those affected, and address their concerns with effective, understandable information. 4. Reform is dialectical, notmechanical. 5. It's implementation, stupid. 6. Develop a sequenced, prioritized list o freforms. 7. Be patient andpersistent. 8. Pick the low-hanging hit-nothing succeeds like success. 9. Keep your eye on the ball-don't let the best be the enemy o f the good. 10. There are no silver bullets. 11. Don't throw the baby out with the bathwater. 12. Reforms must provide returns for the politicians who are willing to make changes. Source: World Bank 2005e. Both o fthe reviews o f the cases on which this chapter draws demonstrate that positive changes and innovations are possible inIndia andpoint to positive lessons for success in service delivery. But review o fthe successful cases also highlights a major risk o f incrementalreforms that do not produce systemic change or complete process engineering. The risk o f internal reforms is that they depend too much on a single champion insidethe system(or a recent insider). These reforms may accomplish a great deal, but they are too easily subsumed by the systemwhen the champion moves on. 63 Chapter 4. Localgovernments and service delivery inIndia The UnitedProgressive Alliance (UPA) Government will ensure that allfunds given to statesfor implementation of poverty alleviation and rural development schemes by Panchayats are neither delayed nor diverted. Monitoring will be strict...devolution of funds will be accompanied by similar devolution of functions andfunctionaries as well...The UPA Government will ensure that the Gram Sabha is empowered to emerge as thefoundation of Panchayati Raj. National Common MinimumProgram, UnitedProgressiveAlliance, May 2004 The government o f Indiahas stressedthe urgentneedfor hndamental reforms inpublic institutions owing to their failure to adequately fulfill core public responsibilities. The national and state governments have signaled their commitment to accelerate the ongoing decentralization o f government responsibilitiesto Panchayati Raj institutions (PRIs) as one way to achieve fundamental reform. Many recent initiatives o fthe UPA government-Bharat Nirman,NationalRuralEmployment GuaranteeAct, National Rural Health Mission-have articulated important roles for PRIs inimplementation. Yet all would agree that decentralization i s no panacea, that the PRIshave weak capacity, that they are often as ineffective and as corrupt as the line agencies, that local elite capture (replacing state and national elite capture) i s a risk, and that simplyplumpinggreater responsibility onto the PRIs without systemic reforms and capacity buildingi s unlikely to lead to dramatic improvements inoutputs and outcomes. Decentralization is not the solution to any problem but rather an opportunity to undertake needed reforms that create the greater accountability that leads to more effective services. Butit is an opportunity that can easily be lost. (Using private providers to fulfillpublic goals i s discussed in chapter 5 and other reforms inchapter 3.)I2 Properly designed and implemented decentralizationmay effectively address the service delivery needs o flocal communities bymore accuratelyreflecting local priorities for services and preferences for delivery systems and content. This has beenthe primarycase made for a trend in government organization and better governance that has gained substantial currency. It has also been argued that the equalization potential is clearly greater in a centralized public sector. The more money the central government has to distribute, the greater i s the potential to equalize. Inother words, local government jurisdictions with large or predominantly poor populations would be likely to simultaneously face higher service deficits and a smaller resource base from which to addressthose deficits. Technical and administrative capacities are also likely to be lacking. It is therefore entirely possible that geographic or regonal divergence may increase and service deficits may actually grow with decentralization. However, interms l2 This chapter does not advocate for decentralization generally nor attempt to answer the question of whether a decentralized governance structure i s the only or even the right institutional arrangement for strengthening accountability inservice delivery. Rather, it addresses the question "how can decentralization be done to improve service delivery?"-which is very different from "is decentralization the most attractive reform to improve service delivery?" 64 of expenditures on public services and benefits to the citizens, it is not clear whether the current system i s effectively equalizing andwhether it is a viable option for improving service delivery. The design o f administrative decentralization inIndia hinges on the "three Fs": funds, functions, and functionaries. India's experience with decentralization can be characterized as unbalanced, with political decentralization (Panchayat elections) running ahead o f administrative decentralization. Functions have been devolved but PRIs have little control over funds (which come to PRIs tied to specific items) or functionaries (who remain appendages o fthe state government). This limitedand unbalanced decentralization i s unlikelyto improve services significantly because it does not improve the delegation, financing, information, and enforcing elements o f system accountability- and insome ways could make them weaker. This chapter reviews the status o f administrative decentralization andpresents an analytical framework for determining which activities will be conducive to improvements (to which tier o f the PRI) through decentralization and how to align the three F's to improve services. The last section addresses transitional issues inbuildingcapacity for accounting, transparency, and social inclusion for accountability inthe PRIsystem. Although this chapter discusses mainly rural decentralization, most o fthe same principles apply to urban areas and urban local bodies. The three F's of decentralization:funds, functions, andfunctionaries The 73rd and 74th amendments to the Indian Constitution, adopted in 1992, provide the legal framework for decentralization. The amendments mandated state governments to transfer some powers and responsibilities to Panchayats at the rural level and to urban local bodies (ULBs) so they could function as autonomous institutions o f self- g~vernment.'~ amendments also provided for regular elections to these local bodies The and reserved one-third o f the seats for women and a proportional number for Scheduled Castes/Scheduled Tribes (SC/ST) populations. Elections have been heldinmost states, resulting inthe inclusion o fmore than three million people inthe formal political system. However, few states have moved muchbeyondpolitical decentralization to devolve funds, functions, and functionaries to local governments. Funds To provide for adequate resources, the 73rd amendment recommended that state governments endow PRIs with taxation powers and enhance their resources through increased transfers o fdiscretionary funds from the state and central governments. In addition, all state governments were required to create state finance commissions, parallel to the national finance commissions. A comprehensivereview (World Bank 2004c) of fiscal decentralization inKarnataka andKerala (two states generally considered to have pursueddecentralization more aggressively) found that: j3 The 73rd amendment mandated the creation of a three-tier (district, intermediate, and village) local government system inrural India. These tiers are referredto as Zilla Panchayat (ZP), block or intermediate Panchayat, and GramPanchayat (GP). 65 Internal revenue mobilization at the Panchayat level is weak, and Panchayats are largely dependent on transfers (from the state and central government). Transfers usually come as tied grants, and Panchayats have limitedexpenditure discretion. State governments are reluctant to devolve finances, which, compounded with the dire state fiscal position overall, ensures that state governments continue to dominate local expenditures inkey areas (health, education, irrigation, etc.). l4 Figure 4.1: Own revenues contribute little to Panchayat revenues in Karnataka, FY 2000-01 70 1 Bagalkot Bidar Mandya Udipi Sample Districts Source: World Bank 2004 Weakrevenue efforts. Ina well-designed decentralized fiscal system, local governments need powers to mobilize revenues. But even insuch states as Karnataka and Kerala, Panchayat own source revenues (revenue through tax collections and user fees) amount to less than 1 percent o f the state domestic product. InKarnataka, for example, GPs raised an average o f only 19percent o ftotal resources from own source revenues in2000-01. In Kerala own source revenue contributed only 17percent or Rs 57 per capita to GP resources in 1998-99. A study conducted for the Twelfth Finance Commission reported that internal revenue mobilization o fPRIs constituted only 4.2 percent o f their total revenue.l5This i s extremely low, especiallyby international standards. Analysis suggests that tax collection remains low because o fboth design and implementation issues. For example, administrativeprocedures for collection are inefficient. Inmany cases Panchayats have not been legally endowed with a broad or productive tax base and, not surprisingly, local officials are unwillingto impose taxes, preferring to lobby the system for greater transfers. l4Although this study is limited to two states (both known for their far-reaching efforts to strengthen decentralization), subsequent analysis suggests that these findings reflect the status o f Panchayat finances across India. l5As reported inthe Government o fIndia (2005a), chapter 8, p. 138. 66 PRLs have limited expenditure discretion. Evenwhere subnational unitshave own source revenue, intergovernmental transfers constitute an essential component o f local government revenue. Internalrevenuemobilization i s very weak inIndia, so Panchayat dependence o n transfers i s extremely significant. For transfers to be effective, local governments must have expenditure discretion over the bulk o fthe funds received. This enables them to tailor resource allocations and expenditures to suit local preferences and priorities and encourages cost-consciousness because the trade-offs are made at the local level. It does not preclude conditional grantsto PFUsby upper tiers o f government, to encourage expenditures inareas o fnational priority, although such grants should ideally be only a small part o f local government resources. Underthe present system, however, Panchayat resources are dominatedby earmarked transfers designedby higher levels o f government, leavingPRIs with almost no expenditure discretion. A World Bank study (2004~)shows that in2002-03 central and state schemes contributed to as much as 54 percent o f the block Panchayat plan outlay in Karnataka. Because the funds come tied to schemes, even the micro-allocation o f resources is determinedby guidelines prepared by the state and central governments. Of the total funds devolved to Panchayats inKarnataka, 50 percent were earmarked for salaries, 20 percent for transfers, and the remainder for inputpurchases.I6The fact that expenditures are tied (and that the fiscal systemi s not transparent) implies that state governments can easily withhold PRI expenditures. InKarnataka, for instance, the share of plan allocation to rural local governments intotal planexpenditures was reduced from 8.3 percent in2001-02 to 5.1 percent in2002-03 owing to fiscal deficits inthe state. Inrecent years, partlybecauseoffiscal tightness at the state level, Centrally Sponsored Schemes (CSSs) have come to dominate plan expenditures even intraditionally state services such as education, health, and water and sanitation (in2002-03,58 percent o f the planexpenditure for elementary education came from the center). CSSs are also an important source o f funding for subjects that fall within the domain o fPRIs. A recent task force set up bythe government identifiedas many as 154 CSSs-to the tune o f approximately Rs 395 billion-that deal with subjects devolved to PRIs.I7Fundsreceived through CSSs are tied to specific guidelines (with greater or lesser flexibility even at the state level) and program implementationi s generally undertakenthrough line departments. Evenwhen PRIs are included, funds are tied and PRIshave little or no discretion over their use. For instance, Sampoorna Grameen Rozgar Yojana, a CSS for wage employment, contributes approximately 50 percent of the rural development ministrybudget (both at the center andinmost states). The program is implemented through PRIs, andfunds for the scheme flow directly from the center to the ZP. However, use o f these funds is tied to specific guidelines determined at the central level and PRIs have no discretion over them. Stategovernments continue to dominate expenditures in key areas. The typical Indian state devolves some functions without devolving the concomitant funds required for PRIs l6This study was conductedbefore the October 16,2004, government order that recommendedlarge-scale financial devolution to the PRIs. 17Government of India, Ministry of Finance, 2004. 67 to fulfill their roles (figure 4.2). As a result, the state governments continue to dominate. A survey inBelunDhamashinGP inWest Bengalreveals that inkey sectors such as primary education, primary health, irrigation, and agriculture, line departments are responsible for as much as 67 percent o f expenditure even though West Bengal's 1973 PRIAct mandates the devolutionof these functions to the PRIs (CMIE 2005). The story i s similar across India. Oommen (2005) found the expenditureo f PRIs as a percentage o f total combined expenditure (union, states, and local bodies) is only 4.7 percent. This contrasts starkly with more developedcountries, where local governments normally account for 20-35 percent o f total government expenditure (inDenmarkand Finlandit is as highas 45 percent and 41 percent). The system o f fiscal decentralization does not follow some basic principles-specifically, significant expenditure responsibility and discretion, andrevenue-raising powers. Panchayats have weak revenue-raising capacities and limitedexpenditure discretion, and funds do not follow functions. Because devolution o f funds i s limited, they have mandates without resources. Thus accountability relations are obfuscated. Functions: blurred delegation For citizens to enforce accountability, the tiers o f government must have clearly delegated functions or roles on the basis o fwhich they can be monitored and their performance judged. Effective devolution o f functions is thus critical to the design o f a decentralized system. As a general principle, PRIs should have a key role mainly in service delivery activities such as rural drinkingwater, rural roads, rural electrification, water tanks, education, and health. They should not take on responsibilities such as operatingparastatals inthe productive sector. The devolution o f functions to PRIs is determined by a list o f29 subjects identified inthe XIth schedule o f the Constitution. Although most state governments have devolved a significant number o f functions to Pus, functional devolution is incomplete and ad hoc and encourages concurrency. Incomplete devolution of functions. Onpaper, many state governments have chosen to devolve subjects to PRIs "wholesale"-without unbundlingthem into specific activities and subactivities. Yet public expenditures are budgeted and managed at the level o f activity or subactivity (schemes or budget items). For example, most states assign responsibility for basic education to local governments, but leave the key activities and subactivities necessary to deliver basic education, such as buildingschools or hiring teachers, to the state line agency-rendering the devolution to PRIstoothless. 0 Ad hoc devolution. There has been little rational consideration, on the basis o f principles o fpublic finance and accountability, o f which services to devolve to which tiers. As a result, the higher tiers of government tend to have responsibility for most services. 0 Overlapping roles. Most devolved services end up ina concurrent list, with different tiers o f government sharing responsibility for them, allowing for the continued, usually dominant, role o fthe state government. For example, one state has 72 schemes for tribal welfare, involving all five tiers o f government-from 68 center to village. InWest Bengal, the 1973 Panchayati Raj Act assigns concurrent responsibilities for all subjects to block and district Panchayats. Functionaries Across India, Panchayats have limited powers to hire staff (most state Panchayat acts provide minimalresources for GPs to hire secretarial staff) so they depend on state governments to depute functionaries from line departments. However, the mere devolution o f functionaries i s not sufficient because PRIs do not have any administrative powers over them (this holds true eveninKerala and Karnataka, where functionaries havebeen devolved inline with funds and functions). Deputed staff members are accountable only to the state government, which retains the power to hire, fire, and compensate. Consequently, PRIs are given functions but have little or no effective control over the people expected to carry them out. Naturally, because the state functionaries' careers and compensation are linked to state services, they are much more concerned with that than with responding to local governments. Status of the three F's in PRI decentralization Some consider that decentralizingto improve services and governance i s a failed experiment,while others believe it has yet to be seriously tried. Ina well-designed decentralized system, the tiers o f government have clearly allocated functions (so that citizens know who i s responsible), financing is commensurate with functions (so that governments can carry out their tasks), and functionaries are accountable to the body responsible for the service (so that the long route o f accountability from citizen to government to provider is unbroken). The current system almost could not be more different, though some states have clearly made more progress than others. Although PRI officials are elected, PRIs serve not as truly autonomous levels o f government independentlyresponsible to voters for functions over which they have control but, at best, as implementers o f activities, plans, and schemes made elsewhere (figure 4.2). They are delegated functions but only concurrently with other tiers o f government that control funds and functionaries. Inessence, the weak alignment o f funds, functions, and functionaries has not created conditions for accountable governance. 69 Figure 4.2: N o alignment in the devolution of the three F's -a, Functions Funds Functionaries Source: www.panchayat.nic.in. Aligning the three F's for more effective services from local governments A well-designed decentralized system o fgovernance requiresthat functions are clearly delegated to individual activities consistent with a budgetingsystem; finances follow functions; performance i sjudged through well-designed monitoring and evaluation mechanisms and information flows down to citizens; and citizens can enforce accountability both with policymakers (voice) and betweenpolicymakers and providers (compact). We tackle these issues ina different order--functions first, then functionaries, and only then funds. Functions A key stumbling block inmoving ahead with decentralizationis the development o f "activity mapping"-an exhaustive list o f activities that governments undertake, with each activity allocated to a specific tier o f government (center, state, district, block or its equivalent, GP).'~Only ifstate finance commissions, for instance, agree on an activity mapping of functions could they allocate -funds on some service-related basis. Hence the first and most difficult step i s to identifythe roles andresponsibilities o fthe tiers o f local governments insufficient detail. Functionaries and funds can then be allocated on the basis o f this mapping. A forthcoming World Bank studyrecommends mapping activity in three steps: unbundle into subsectors and functions within subsectors; apply technical first principles to align levels ofjurisdiction with functional assignments, based on the specifics o fthe service; and apply first principles o f accountability. The first step is to unbundlesectors into detailed subsectors and from subsectors into identifiable activities. The study identifies at least six activities inany subsector: This section draws heavily on a forthcoming study on functional allocation and decentralization. 70 0 Setting objectives and standards o f service. 0 Planning. 0 Creating assets. 0 Operating (nonstaff). 0 Operating (staff). e Monitoring and evaluation. A key factor leadingto accountability failure is having "the fox guardthe chickens." When the same organization i s responsible for all steps inprovision (setting standards, planning, creating assets, operating, and monitoring and evaluation), the organizational incentives to generate transparent information suitable for external accountability are very limited. Decentralizationto PIUsoffers an opportunity to use the multipletiers o f government to encourage a creative tension. This enhances a citizen's ability to create checks and balances through expressing "voice." So, for instance, ifone simply devolves responsibility for all aspects o frunninga Primary Health Center from the state to the district, the scope for improvement i s limited.But ifthe state sheds operational responsibilities while maintaining responsibility for setting standards and monitoring and evaluation, it can use the arm's lengthrelationship with the district to produce reliable, regular, relevant information about the performance o fPHCs across districts. The second step is to use sectorally relevant technical first principles. There are four basic principles: Economies of scale: match politicaljurisdiction to technicallyfeasible catchments area. A "catchments area" is the geographic space fi-om which the service provision unit(school, clinic, well, etc.) typically draws itsconsumer. This is determinedbythe relationship o f transport costs to the value o f the service. Economies o f scale will varybothbysubsector andby activity. Inhealth, the catchments area for personal preventive care i s small (requiring frequent, individual low-value transactions), while the catchments area for the surgical unit o f ahospital mightbe a district or even a state. Evenwithin an activity the economies o f scale varyby function-so standards for personal preventive care can be set at the state or district level even ifthe operation o fthose services has little or no economies o f scale. Equity: set activities so that interjurisdictionalfairness can be achieved. Note that this need not have any implications for who is responsible for operation o f the services. Externalities: match the scope of the spillover impacts to thejurisdiction responsible. A smalljurisdiction might easily assume responsibility for personalpreventive health activities (e.g., antenatal visits). But highlyinfectious communicable diseases may cause epidemics that spill over from one jurisdiction to the next-so the scope for controllingHIV/AIDS or avian flu is likely better served by state or central tiers. Heterogeneity of demand: maximize local choice overpriorities. Because needs and demands are large, resources are scarce, and conditions vary widely, evenwithin districts or blocks, allowing communities to choose where demands differ (while still meetingbasic standards) i s crucial. 71 The third step is to apply first principles o f accountability. Each relationship o f accountability involves delegation, financing, information, and enforceability. The questionis, at what tier o f government are those relationships potentially the strongest? Again, the answer is not the same for all subsectors or functions. Rather, it depends on three characteristics: Degree of discretion. The more discretionary the activity, the greater the need for local information and enforceability. A discretionary activity i s one inwhich the successful delivery o f a service requires the frontline provider to adapt to local conditions insteado fproviding the same thing every time. So, within health, immunizations are not discretionary-the health worker gives the same treatment to each child-while curative care is highly discretionary-the care provider must respond to each patient's symptoms. Degree of transaction intensity. The greater the transaction intensity, the greater the potential for local control. Transaction-intensive activities require repeated transactions at the local level. So, ineducation, textbook design is not transaction intensive (it i s done once and thenusedby millions), while classroom instruction involves millions o f individual teacher-student interactions every day. Ability to observeperformance. For information and enforceability, activities should be allocatedto the level inwhich the provider's performance is best observed. This depends to a large extent on how much technical know-how is requiredto determine whether the service was adequate. Everyparent can know whether their child's teacher was present and can, know with some questioning, what happened inthe classroom, butjudging the adequacy o f a curriculum or the validity o f a testing instrumentrequirestechnical skills. When applied, these first principles will give different answers for different subsectors andfunctions. Within each sector, subsectors will have different requirements: the appropriate scale for managingwater supply, for instance, will be different for small villages and for urban areas (because economies o f scale are different); and the appropriate scale for managingwatershed issues depends on hydrological conditions (surplus, deficit) and links (because the spillover impacts o f onejurisdiction on another differ). Moreover, when first principles are applied, the activities, even within subsectors, are allocated to different tiers. The analysis o f first principles inthe World Bank study (forthcoming(b)) on rural service delivery suggests three strong conclusions: e Thesame tier of government should not be responsiblefor operation andfor (all) monitoring and evaluation. External accountability (through either political voice or choice) requires information that no organizationalprovider has the incentive to create. The existence o f five tiers o f government gives scope for a creative tension in which higher tiers monitor the performance (on compliance with process, financial accounting, and outputs and outcomes) o f lower tiers (or vice versa). A simple transfer o fbundled subsectors i s a lost opportunity to strengthen accountability. e The capability and commitment of higher tiers of government to set standards for outputs and goalsfor outcomes and to monitor performance and evaluate the impact of alternatives should be strengthened by decentralization. Monitoringsystems are 72 key to an effective decentralized system because they create incentives for performance. Good monitoring systems require that outcomes are clearly benchmarked and standardsmaintained. However, citizens also needto judge the performance o f local governments (especially on functions such as operations and asset creation, which are locally observable). Thus monitoring must also be undertaken at the GramSabha level through mechanisms such as social audits-but monitoring compliance with the process elements o f local accountability is also a higher-tier function. There is enormous scopefor increasing local control of asset creation and operations. Many subsectors o f service delivery involve operations that are discretionary, transaction intensive andlocally observable and for which there are no technical constraints to local operation (e.g., rural water supply, primaryeducation, personalpreventive health, minor curative care). These are prime candidates to be taken onbythe lowest tier o fthe PRI-the GP. Again, this decision should bethe result o f analysis: not all education (e.g., universities) or all health care (e.g., tertiary hospitals), or even all aspects o frural water supply (e.g., largepipelines) canbe delegated to the GP, just as not all aspects of education, health, or water supply can be delegated to the state level, as inthe current system. Funds Funds must follow functions, so that Panchayats can fulfill their mandates-through both increased own revenues and improved transfers. Improving the tax system anduser charge collection levied by GPs i s perhaps the most important area for reform inthe current system. There is an urgent needto broadenthe tax base for GPs, improve the design and collection o fproperty taxes, and increase the capacity to generate user fees for services. The Twelfth Finance Commission has attempted to address these problems through recommendations that include the needto improve evaluations o f the tax base, incentives for revenue collection at the local level, and property tax assessments. Transfers to PRIs could be improved inthree ways: Movefunding designfrom schematic to thematic. Typically, revenues are micromanagedbythe design o f individual projects and programs, often with multiple schemes for the same purposes (there are more than five CSSs for water supply). These can bebundledina theme (such as water supply), so that PRIs can choose to allocate funds to their highestpriorities. This would allow PRIsto have more open menus inthe allocation o f funds within a sector. Give GPs more untied block grants. There have been some recent efforts inthis direction. For instance, inOctober 2004, a government order inKarnataka transferred all schemes that deal with subjects identified inschedule XI to PRIs. Approximately Rs 34 billion(about 30 percent o fthe state planbudget) was devolved to PRIs, of which Rs 15.88 billion was devolved directly to GPs.19At the national level, the ~~ Some of these funds have been brought back to the state accounts because of concerns related to revenue deficits inthe state. 73 Twelfth Finance Commission recommends a quantum increase (Rs 200 billion) in financial allocations to PRIs. This is a significant increase incomparison with the EleventhFinance Commission, which recommended an ad hoc annual grant o fRs 16 billion or Rs 8,000 over five years for Panchayats. 0 Define Panchayat expenditures. One specific detail with largeramifications i s that all transfers to Panchayats are defined as revenueexpenditures (according to the accounting laws inIndia). The government has adopted the Twelfth Finance Commission recommendation that all states should have zero revenue deficits by 2006 or face penalties that affect their consolidated fund allocation. This makes allocations to PRIs-evenwhen they finance investments-more difficult and gives states (abetted by the interests o f their line agencies) an excuse to not devolve revenues. Functionaries The issue o ffunctionaries is somewhat more complicatedthanthose o f functions and funds because o flegal constraints. Ideally, PRIsshouldhave their own cadres of employees with powers o f enforceability (understood as the ability to hire, fire, and reward). However, even ifthey did, the provisions o fArticle 311o fthe Constitution would extendto Panchayat employees, creating conditions o fweak incentives and poor performance, similar to those at the state and center. One way out could be to authorize PRIs to hire employees on a contractual basis. Such contracts should be renewable subject to satisfactory performance. Inaddition, PRIs shouldbe able to outsource technical expertise on a needs basis (from bothhigher tiers o f government and the private sector). States such as Karnataka, where GPs are allowed to contract engineers from a district pool of engineers, have experimented with this idea. These experiments have had some measure o f success and could be extended to PRIs across the country. 74 Box 4.1: Urban Decentralization and the three F's: a status review The 74th amendment provides the legal framework for urban decentralization inIndia. A recent World Bank review (2005) o f the status o f decentralization inurban areas argues that the devolution of funds, functions, and functionaries to urbanbodies has been relatively weak. Some key findings: Funds. Urbanlocalbodies (ULBs) suffer from poor-quality financing with weak revenue-raising powers and limited discretion. Revenues collected by ULBs account for a mere 3 percent o f all revenues collected inIndiaas a whole. Dependencyontransfers isthus significant. But transfers are not stable, timely, or predictable-and inthe final analysis, they are not equalizing. ULBshave limiteddiscretionover transferred finds. Almost 40 percent o f these funds go to salary payments. Other finds come tied to schemes designed by the state and central governments (ULBs get grants from at least 30 CSSs on a myriad of urban governance issues). Just as with PRIs, fiscal deficits at the state level have led to delays and reductions intransfers to ULBs.This inturnhas contributed to mounting local fiscal arrears. Budgetary expenditures undertaken by ULBs are thus very low, amounting to a mere 0.8 percent o f GDP (for FY 1997/98).20As a result, state governments continue to dominate. For instance, inKamataka, duringthe 10thplanperiod, ULBs spent Rs 47 per capita on urban water supply while the state government spent as muchas Rs 128 per capita. Functions. The constitutional amendment identifies 18 functions that state governments could assignto ULBs.Inpractice, functional responsibilities to localbodies are confinedto infrastructure-related activities such as water supply, sewerage, and solid waste management. Insome instances, ULBs have been assigned functions related to education and health care. The limited functional responsibilities o f ULBs tend to be compromised because o f the proliferation o f state boards and other agencies with overlapping roles and responsibilities. The problems of concurrency are further exacerbated by the lack o f coordination and consultationbetween these bodies. Functionaries. Like their rural counterparts, ULBs depend on state government officials to fulfill their administrative responsibilities. Senior municipal officials are appointed from the I A S or the State A h n i s t r a t i v e Service, limiting the ability of localpoliticians to hold staff accountable. Overstaffing i s a problem, because state governments control the number and distribution of employees across categories. Current debates inIndia consider urban and rural decentralization separately and for the most part emphasize rural decentralization. This i s problematic: there are many complementarities between rural and urban decentralization because they draw from the same revenue source-the state government. For instance, greater tax discretion at the urban level (since urban centers are naturally better sources o f tax revenue than rural areas) could free resources for rural areas. Thus when designing a model for decentralization, particularly fiscal decentralization, rural and urban decentralization shouldbe viewed in conjunction. Source: World Bank 2005g Capacity building for accounting and accountability with social inclusion The constitutional amendments have created a political space for PRIs and elections are held regularly inmost states (ifonly byjudicial insistence in such reluctant states as Bihar and Andhra Pradesh). But many argue that both PRIs, especially at the lowest level, and ULBs lack capacity and that decentralizationwould leadto even more elite capture because local politics are exclusionary. As this chapter stresses, no one should *'Compared with their urban counterparts across the world, the expenditure responsibility o f local governments inIndia i s extremely low. Evenin countries with few social service responsibilities such as Morocco and Greece, budgetary expenditures by ULBs inFY 1998 accounted for 3.3 percent and 2.1 percent of GDP, respectively (World Bank 2005g). 75 expect that decentralizationwill improve services overnight or that achievingbetter services will not require major investmentsincapacity, indeepeningdemocracy, and in promoting greater community capacity to organize for collective action and to demand services. Buildingcapacity for local governance therefore goes beyond local governments and will involve three major strands: 0 BuildingPRIcapacity for accounting-planning, budgeting, accounting, and auditing effectively. 0 Buildinggreater accountability, bothinternal and external. 0 Strengtheningthe "synapse" between households and communities and local governmentsby strengthening community capabilities and promoting greater inclusion inthe processes o f local government. Accounting: capacity buildingfor financial management We need to build the capacity of decentralized bodies and their elected representatives and this is best done through effective transfer of funds andfunctions to them. Prime Minister Manmohan Singh, August 15, 2004 The potential problem o fweak capacity among local PRIrepresentatives is well recognized. The capacity issue is essentially a "chicken and egg" problem: capacity cannot bebuilt ina vacuum. UntilPRIshave sufficient control over thematic and untied resources for financial management they will necessarily lack capacity. At the same time the weak capacity argument is often usedby state governments andline departments as an excuse to reduce or withhold the allocation o f the very resources needed to build capacity. Thus as the prime minister emphasizes, the best solution for capacity buildingis to make it necessary to have capacity. Efforts have beenmade byboth government and civil society organizations to fillthe capacity gap. However, a consistent implementation and support planfor PRIs has yet to evolve. Capacity inputs aim to strengthen the sectoral capabilities o f PRIs (partly because of the schematic nature of fimding) rather than their capacities to perform functions generic to local governments, thereby limitingany spillover impact. Strengthening generic capacity includes capacity for financial management (budgeting, planning, auditing), procurement, and perfonnance monitoring. Capacity-building initiatives could include the creationo fresource cells staffed with accountants, engineers, and other technical experts at the block level and the provision o f inputs inbasic accounting and procurement procedures. This would buildthe back office support necessary for PRIsto effectively use thematic, untiedfunds. Inessence, capacity-building initiatives need to be part o f a sequenced devolution process that strengthens the power and resources available to Panchayats while strengthening their capability to manage these resources. 76 Strengthening accountability Besidesperiodic democratic elections, the 73rd amendment provides for two additional mechanisms that aim to strengthenvoice at the grassroots and encourage a more inclusive polity: reservation o f seats for women and for SC/ST (discussed below) and the Gram Sabha. Gram Sabhasprovide the space for citizens to articulate their needs andpriorities and to monitor the actions o ftheir elected representatives (including whether programs are well targeted and reach intendedbeneficiaries). While it i s difficult to generalize, there i s a widespread perceptionthat the Gram Sabhas have provenvery weak: on most occasions they are not even heldandwhen heldthe quality o fparticipation is extremely poor. For instance, in2002-03 as many as 75 percent o f Gram Sabhas inHimachal Pradesh (generally a quite h c t i o n a l state) had to be adjourned for lack o f a quorum- and a quorum requires only 10percent attendance. Rao and others (2005) collected data from more than 4,000 households across four south Indian states (generally advanced on the PRI agenda) and found (a) a quarter o fthe GPs surveyeddid not even hold Gram Sabhas; (b) only 20 percent o fhouseholds had ever attended a Gram Sabha; and (c) Gram Sabhasreflected social inequityprevalent inthe village, because women and the landless were much less likely to attend and speak inmeetings. Rao and others found that despite low attendance, where Gram Sabhas are heldthey do yield some positive results. Participation is associated with a better chance o f the landless, illiterate, and SC/STs obtaining Below Poverty Line ration cards (probability o f an 8-10 percent increase). This analysis suggests that the Gram Sabha (when held) can encourage political participation and social inclusion. The challenge lies inensuring that they are heldand instrengtheningparticipation inthem. The analysis byRao and others found that literacy rates are positively correlated both with individuals hearing about Gram Sabhas andwith individuals attendingthem. Ktlshna's (2005) study on determinants o fpolitical participation yielded similar results. Along with education, Krishna found a significant correlation between information and political participation. The experiment by Pratham (a national NGO) with providing real- time feedback to parents about the performance oftheir children on basic literacy and reading skills has generated widespread public interest in education, and the number o f discussions inGramSabhas on learning has increased dramatically. Pratham's experience has important policy implications, because it highlights the important role that facilitating reliable, relevant, and regular information flows can play instrengthening Gram Sabha participation. Similar experiments inother sectors have the potential to contribute greatly to the quality o f Gram Sabhaparticipation. 77 Box 4.2: An impactevaluationof communitisationin Nagaland "The genius of communitisationis its simplicity. But in is simplicity is the dificulty of unlearning all that governance has become today-people alienating, unfriendly, centralized behemoths. Thecall to communities is a call not to decentralize but to recreate a new vision of relationship between people, governingstructures, and institutions based on an inclusive, egalitarian and democratic vision.'' Executive Summary In2002 the Government ofNagaland launchedanambitious effort to transfer ownership o f resources, delegation o f governmental powers to village committees, devolve control o f employees (via control o f salaries) and give control over assets-including responsibility for maintenance in three sectors: elementary education, health, and power. An impact analysis undertaken in2004 foundbroadly positive impacts inallthree sectors. In education, the study reported higher teacher attendance (with attendance rates over 90 percent in 18 o f 28 schools studied), higher student attendance, lower drop-outs, and higher passing percentages. Inmany villages students transferredback fromprivate to public schools as they improved. Source: Government o fNagaland, 2004 Strengtheningthe "synapse" betweencommunitiesandgovernmentsfor betterlocal governance Thus far, we have discussed the importance o f and the mechanismsthrough which to design an effective decentralized governance structure. However, the extent to which design translates into practice depends on the role that PRIs, as political institutions, play instrengthening voice andpoliticalrepresentation-access and inclusion o fthe marginalized. PRIs will be effective only ifthey are able to broaden the democratic base (bytackling political exclusion) andthereby improve citizens' capacities to influence the provision o fpublic services. Participationo f SC/STs and other backward classes inthe decision makingprocess at the grass root level i s a critical element o fbridgingthe gaps in decentralization. This canbe achieved formally through measures such as reserving seats and informally by bridgingthe gap between community organizations and local governments. Reservations Mandatory provisions for reserving seats for women and SC/ST were introduced inthe 73rd amendment as a means o f addressing traditional forms o f social exclusion at the village level. It i s no surprise that reservations have not undone social exclusion ina few short years nor that elite capture remains aproblem. Butrecent empirical work suggests that reservations (for bothwomen and SC/ST) are having some mixedimpact. Chattopadhyay and Duflo (2004) examinedthe impact o f reservations on women's political participation through a cross-state comparative analysis o f GPs inone district each inWest Bengal and Rajasthan. InWest Bengal reservations led to increased participation o f women--from 6.9 percent to 9.9 percent-in the Gram Sabha inGPs 78 with a woman Pradhan (head of the council o f elected ward members).They also found that inthe villages studiedwomen and men ask substantially different questions: 31 percent o f women asked about drinkingwater versus only 17 percent o fthe men. These differences lead to differences inoutputs-and not always inpredictable ways: inWest Bengal there is more investmentinwater and roads and less inschools inGPs with women Pradhans. Many expected that reservations for women would lead to no changes at all because discussions would be dominated by men anyway. However, this study demonstrates that, despite weak preconditions for leadership, the very presence o fwomen inthe political systemencourages women to participateinpolitical processes at least insome states, thereby contributing to more inclusive politics. Figure 4. 3: Reservationsfor women and SC/STs is having some impact Women's participation in gram sansad's is higher by 3% There are significantly more investments in drinking points in GPs with women Pradhans water in reserved GPs 1 30 Reserved Unreserved Reserved Unreserved a Participationin Gram Sansad a Number of dnnkingwater facilities Source: Chattopadhyaya & Duflo (2003) Source: Chattopadhyaya & Duflo (2003) SC/ST households are more likely to access government schemes to build housekoilet or for private water/electricity where GP Pradhan is SC/ST l2 1 9.8 .n SCiST Reserved GP SCiST Unreserved GP Source: Rao et al (2005) Analysis o fthe impact of reservations for SC/STs finds similar outcomes. Rao and others (2005) found that on average, SC/ST households are approximately seven percentage points more likely to access government schemes inGPs where the Pradhan is an SCIST. These results need to be viewed with some caution, because there i s evidence to suggest some private appropriation of public goods by the political elite. The same study found 79 that politicians benefit from a higher incidence o f overall targeting and from the provision o f toilets andpublic works programs. Thus, although reservations do encourage some level o f equity inaccessto public goods, political capture and private appropriation remain serious problems. Local governance and user and community groups The pushtoward decentralized government reform inIndiainthe early 1990s was accompanied by a parallel movement that privilegedanother kindof local organization as a key actor inrural development: user andcommunity groups.** It i s arguedthat the rural poor can be empowered both individually and collectively through the creation o f social capital. User groups are typically formed vertically-i.e., they are sector specific (watershed development, rural water supply, education, microcredit), often outside formal government structures, andpart of a scheme- or project-specific design (created bya line agency or society). There are few serious links (often only a dotted line) between the user group and the local government. Thus two institutional arrangements (the PRIs and the user groups) coexist, creatingmultiple institutional arrangements for service delivery at the village level (figure 4.4). Most observers today believe that the two approaches to local governance need convergence, drawing on the strengths o fboth approaches while avoiding the weaknesses o f each. ~ 21We use the term "user groups" generically to refer to any association(community-based organizations, self-help groups, etc.) of individuals who come together to implement a specific project. 80 IntermediatePanchayat Key v ---+ Strongformalrelationship Weak relationship css Centrally Sponsored Schemes SHG Self HelpGroup SSS State Sponsored Schemes UG UserGroup SSA SarvaSikshaAbhiyaan SGSY SwamajayantiGram Swarozgar Yojna EGS Employ~nentGuaranteescheme MHRD MinisbyHuman Resource Dwelopment DPED Depamnentof PrimaryEducation MORD Ministry of RuralDevelopment WDD WatershedDwelopment Source: World Bank, forthcoming(b) The strengtho f decentralization through PRIs lies inthe fact that they are the legally grounded representative institution at the grassroots. However, inthe absence o f effective checks and balances, these institutions are susceptible to elite capture, political exclusion, and corruption. The user group approach i s advantageous because it encourages an inclusive decision-making process at the local level by directly targetingthe poor. It enhances project effectiveness because it contributes to a greater sense o f ownership and commitment to project objectives. Crucially, recent analysis suggests that membership in user groups contributes positively to participation inthe Gram Sabha, thereby directly affecting the voice element o f accountability relationships (Narayana 2005). InIndiathe user group approach hastended to beviewed inparallel or oppositionto PRIs.Privileging user groups over GPs has had the following consequences: 0 It encourages a schematic approach to development at worst and a sectoral approach at best, undermining accountability. Information flows remain restricted to user group membersrather than open to citizens at large. 0 InsomecasesanexclusiveemphasisonusergroupscanunderminetheGP.Many state governments tend to privilege user groups over GPs and devolve finances directly to the user group, leaving GPs starved for funds.22 22 Andhra Pradesh's Janmabhoomi programbest exemplifies this phenomenon. While it has many positive features that illustrate the strengths of the user group approach, it is also the case that where resourcesfor 81 0 User groups can be manipulated so that rather than puttingbounds on elite capture they simply reallocate among elites. Ina field study o f 18 villages in2 districts, Powis (2003) found that filling powerful positions withinuser groups (such as chairperson o fthe water user association) are often used as opportunities to relocate important local leaders who are unable to find a space inthe formal political system. These local elites thenuse their positions to usurp powers and resources meant for the GP and often accessresources for schemes that even GP membersare unaware of. 0 Capacity buildingis often project or scheme specific and has limited spillover impacts inother sectors. A recent study on local organizations inthree sectors (women's development, drinking water and sanitation, and watersheds) by Alsop (2005) highlightsthe dilemmas. The study finds that PRIs have a limitedrole (usually restrictedto monitoring) and are often ineffective. The presence o fuser groups in such an environment encourages an informal transfer o f functions betweenlocal organizations, where user groups emerge stronger and take over functions such as financing, capacity building, and monitoring and evaluation that are formally mandated either to support organizations (NGOs/line departments) or to PRIs. At the same time, although user groups take on a multitudeo f hctions, their sustainability i s questionable. Given the role that PRIs can play inensuring sustainability andinreducingweaknesses associatedwith user groups, andthe concomitant role that user groups can play ininstitutingchecks andbalances on PRIs and enhancing project effectiveness, social inclusion, and community ownership, convergence between these two arrangements could be apowerful instrument to enhance developmental effectiveness. This suggests that effective service delivery requires different kinds o f local organizations to operate intandem. Convergence for better local governance Strong and effective decentralization needs both strong user groups and strong GPs. Convergence o f these two institutional arrangements can lead to more effective development outcomes. One way to achieve this convergence is through the mechanism of subcommittees at the GP level. For example, the government o f Karnataka (in collaboration with the World Bank) has formed Village Water and Sanitation Committees (VWSCs) as subcommittees o fthe GP. The Gram Pradhan i s the chair and all GP members and selected community stakeholders are memberso f the committee. There are significant risks associated with this approach. For instance, Pradhans can choose not to cooperate (as insome villages inKarnataka). However, with the right kindo f facilitation and capacity building, the VWSC could pressure the Pradhan to perform effectively (World Bank 2001a). Subcommittees could also operate as umbrella organizations for user groups at the village level. Other means o f encouraging convergence could include transferring financial resources to user groups through GPs and introducing information- sharing and feedback mechanisms at Gram Sabhas. ~~ rural development were dropped straight to user groups, Panchayatsremained starved for funds (only 5 departments have officially devolved funds to Panchayats). See Jayal(2004). 82 Box 4.3: Strengthening accountability and encouraging politicalinclusion in urban settings The means by which urban communities mobilize for collective action and its impact o n urban governance structures such as ULBs is relatively understudied. A recent study by Jha and others attempts to fillthis gap by examining the dynamics of localleadership, collective action, andformal governance structures inurban slumsinNew Delhi. The study reveals that urbanslumdwellers participate inthe formal political system through intermediaries inthe form o f informal local leaders. The form o f this leadership depends on many factors, including the demographic profile o f the slum. More homogenous slums transport traditional leadership structures from the villages into the slums while newer, heterogeneous slums depend on leaders who gain authority through political connections and education. Ingeneral, they found that although slum communities have access to politicians (as muchas 30 percent o ftheir sample had interacted with local politicians), they interact with political structures through their Pradhan, who for the most part represents their needs. Thus slums with a higher degree o f community organization have greater voice and access than those with weaker spaces for collective action. This study has criticalpolicy implications because it identifies forms o f collective action that could be enhanced to ensure better local governance inurban settings. Source: Jha and others 2005. Handling the transition The political economy o f local governments is complex. The ideas behindthe creationo f local governments and assigning functions have been accepted-and established inthe Constitution. The question i s one o fhow far each State will choose to go inpractice, some, like Kerala, have moved very far while others, like Bihar are just havingPRI elections in2006 (though with a `blank slate' this may allow the government to move very fast). Inpart the difficulties have beentechnical: aligning the "three Fs" to create clear delegation between the state and PRIs, between the PRIs and line agencies, and across the tiers o f the PRIs i s a challenging exercise. But inaddition to the technical difficulties there are two issues that need to be addressed to make PRIreform viable. One objection to placing the entire responsibility on the PRIs is "what ifthey fail?" Many suggest that while having greater responsibility at the local level, ifwell designed, may leadto improvements insome places there are fears that ifone relaxes the "control" from the State or Center that key functions (like primary schooling or immunizations) may fail altogether or that low capacity areas might get evenworse. Certainly ina country as large andvaried as India this is a very real risk. But a well designed assignment o f functions, funds, and functionaries by creating clear delegation o fresponsibilities can also create the possibility o f a "fail safe" mechanismwhereby a higher tier o f government can assume responsibilities (and receive the funding) inareas inwhich clearly designated criteria are met. An unconstrained authority o f higher tiers (states over districts, districts over GPs) to intervene will lead to confusion and conflict (andpoliticization o f decision making)-and there is not guarantee the state will do a betterjob. But having some clearly definedmechanismto avoid catastrophes will help ease the transition. 83 Perhaps the main obstacle to moving more authority andfundingto the PRIs is internal to government-existing line agencies at the State level are often less than enthusiastic about losing control o f funds and functions-and functionaries even less so. The transition on employees i s very difficult, and almost certainly will have to take place with the creation o f a new set o f employment relationships andthese becoming the n o m with gradual expansion rather than changing the status o f existing employees (which may itself be unconstitutional). On the budget side there will also have to be a transitional period which provides some guarantees to line agencies to avoid disruptionwhile incremental funds flow through new systems (as is beingdone with many o f the government's new initiatives such as NRHMandNREGA). 84 Chapter 5: Communityand Private Sector as Providers:Infrastructure andMore Inadditionto administrativereforms aimed at improvingperformance ofexisting agencies (chapter 3) and a move towards greater reliance on local governments (chapter 4) a thirdpotential path for better service provision is greater reliance onproviders outside the traditional public sector. This can come intwo forms, either greater reliance on communities and user groups (as is increasingly the case inruralwater) or in"public- private partnerships" that are arrangements between the government and either not-for- profit or for-profit private providers. This chapter discusses these efforts, particularly in the context o f infrastructure, but also drawing on examples from other sectors such as health and education. Any discussionofhow to improve services ininfrastructure must unfortunatelybeginby a warning against a "one fits all" or "miracle cure" approach. While the different "infrastructure" sectors have some common features, they are as much different as alike technically and economically. The scope for effective competition (which depends on economies o f scale and scope, extent o fnetwork linkages), desirability and feasibility of user charge recovery, and possibility o f direct cooperation o fusers varies enormously. Evenifincreased competition and private sector entry has worked wonders in telecommunications--it does not meanthe same approach will work for power or highways. Even ifgreater community engagement is the most promising way forward in improving rural water supply-it doesn't mean the same mechanisms would work in urban settings, or for highways. Rather than organize this chapter ineither the usual "sector by sector" approach or by the types o f reform, this chapter will discuss how the promising reforms are all exploring different ways o f coming to grips with the complex and difficult issues o fincreasing accountability outlined inchapter 2: unbundling of roles, clear delegation of goals, autonomyfor serviceproviders, informed external accountability, and enforceability (though both consumer choice and citizen voice). The chapter finishes with a discussion o f the most difficult element o f all reforms: the political economy. A focus on infrastructureservices Infrastructure i s important for manyreasons. Sustained rapid overall growth inIndiais at risk ifthe modernization o f economic infrastructure cannot keep pace with the demands-improvements inairports, ports, power andtransport (rail andhighway) are a potential constraint to sustained, job creating growth. Infrastructure i s about more than just sustaining rapid growth, it i s also central to makinggrowth more inclusive-while airports are crucial for linkingIndia to the world, rural roads are central to linking India to the world and ensuring rural areas gain and irrigation andwater control infrastructure are essential to progress in agriculture. Moreover, infrastructure is about more thanjust sustaining economic gains: access to reliable and clean drinkingwater is fundamental, adequate sanitation (inbothurban and rural areas) i s a must. But, as stressed inchapter 85 2, the issue is notjust "access" or asset creation; it is satisfaction with services through the operation o f assets and maintaining those assets. Infrastructureis an example o f the difficulties facing service delivery inIndia today and examples fromjust three sectors: roads, power, and water supply illustrateboththe challenges and ongoing efforts. Roads. At the national levelreforms to project monitoring and management have reduced cost over-runs from 60% inthe early 1990s to around 20% at present (IMF, 2005). Notable progress has beenmade inrecent years with the National Highway Authority o f India's (NHAI) Golden Quadrilateral (GQ) project, which links Delhi, Mumbai, Chennai and Kolkata. The North South-East West (NS-EW) highwayproject i s slated for completion by December 2008. Ifso, it will be a creditable achievement. The Comptroller and Auditor General o f India estimated that even the NHAIhadonly completed 29% o f its plannedprogram by the target completion date o f June 2004. Completedworks suffereddelays o f up to 28 months and there were cost over-runs o f 22% on average. One suspects that these problems inthe states are more severe andlittle evidence is available on the extent to which road improvements have been realized at the state level. Indiaincreased road investment from around US$ 1bnper year inthe 1990sto around $4 billionper year at present. But this has not been matched by improvements in implementation performance. A lack o freadiness o f construction sites, the overall procurement process and contract management, capacities o f implementing agencies and the capacity o f the construction industryinIndia are all factors inthe implementation difficulties that result inonly a fraction o f investments delivered on time and at cost-but with incentivesto initiate new works. Power. Despite some successes and the implementationo fimportant reforms to date, the most critical problems inthe sector remain difficult to resolve. Losses of electricity due to theft and technical factors remains stubbornly high, at around 32-35% o f net generation on average, and muchhigher insome states. Inefficiencies inthe collection o f bills means that aggregate technical and commercial losses are around 40-60% depending on the state. Financial losses inthe power sector havebeenreduced somewhat, from around 1.5% o f GDP to 1.2% o f GDP, but are still excessively large. Payment discipline imposed by central power utilities and independent power producers, the development o f the regulatory framework and the actions o fregulatory agencies at the central and state levels, and incentives provided by the center to states to improve performance of their power sectors, have all played a role inthe modest improvements that have taken place to date. However, the financial performance o fthe sector is not improving quickly enough. There has beenvariable progress on implementation o fthe Electricity Act (2003) which i s mandating steps by state governments and regulators to increase competition and accountability inthe sector. Some states have made active strides inthe separation of transmission and trading and the development of open access regulations. However, there has beenlittle movement on the phase-out o f existing cross-subsidies although the National TariffPolicy i s expected to help inthis regard. 86 Water Supply and Sanitation. Adequate water inquantity and quality is a core government responsibility-but, as documented inchapter 2, is far from being met in terms o f sustainability and satisfaction. And access to piped water has actually declined insome states (figure 5.1). More spendingalone through existing agencieshasproduced mixed results. As shown inchapter 2, the more effective states produce two to three times as many water connections perrupee expenditures as the less effective states. This suggests there are enormous improvements possible inthe efficacy o fpublic spending. Figure5. 1: Access to pipedwater has declinedinsome statesbetween 1991and 2001 Percentage ofpopulation with access topiped water, by state, 1991-2001 60 40 20 0 -20 -40 -60 Source: Census 1991and 2001. Three modes of externalreform: communities,contractingout, andPPPs There are multiple modes o freform beingpursuedthat aim to improve service delivery bymoving functions outside o fthe channels o fthe existingline agency structure. As discussions o f these reforms are often hotly charged--particularly around the word "privatization"4t is worth making three points. First, there are multiple types o fnon- state providers-communities, user groups, non-for-profit organizations, for-profit firms-and multiplemodes o fengagement with each (contracting out, regulated entry, market based regulation, demand side transfers). Engagement with non-state providers doesn't mean less government, but it requires different government, inthe sense o f a differentrole, with different skills and capacities to develop, manage, andmonitor service agreements. But a commitment to effective provision and a commitment to aparticular set of providers (e.g. the current employees o f line agencies) are not synonymous. Second, the accountability approach to service delivery tries to avoid ideological positions that somethingmust be good because it i s called "private"-the question i s whether aparticularreform would improve services or not, not what it is called. Third, poor quality public services create a defacto privatization that is more thorough-going, unregulated and inefficient than any proposed planto engage non-state providers: when 87 canal irrigation fails, tubewells result; whenpower i s unreliable private firms and householdsbuybackup systems; when railways are ineffective shippers use trucking. The question i s not whether or not the private sector will be involved in services, but on what terms andhow'to makethat engagement as productive as possible. Engaging communities. One mode o fimproving services i s to pushfor greater community involvement-particularly bythe users o fthe services. Particularly inrural water supply (but also inother rural infrastructure) there has beena shift towards demand-driven schemes inwhich the communities take responsibility for hnctions. A recent study estimated that inUttar Pradesh rural water supply schemes that are demand- driven schemes, such as Swujaldhara, see 90% o f fimds reachingbeneficiaries interms of new capital expenditures, whereas for supply-driven schemes the figure i s as low as 40% (Misra, 2005a). The use o fcommunities is not uniqueto infrastructure. Madhya Pradesh's innovative development o fRogi Kalyan Samiti's (RKS), local health facility committees to manage hnd-raising, has been copied inseveral other states. The national Reproductive and ChildHealthprogram will now support the development o f RKS's for primarylevelhealthfacilities such as Community HealthCenters (CHCs) andPrimary Health Centers (PHCs) as well. Contracting out. A secondmode o f engagement with non-state providers i s through the contracting out of specific services, without transfer o fownership o f assets. This i s commonly usedininfrastructure where the assets are large and competition i s not feasible-but managing the assets i s important. The recent moves towards management of airports are an example (more below). Again, contracting out i s not limited to infrastructure and there are numerous examples from the health sector. A number o f states (Karnataka, Gujarat, Tamil Nadu, among others) contract NGOsto runprimary health care facilities (PHCs, CHCs). The R C Hprogram made extensive use o f contracting out to nongovernmental organizations (NGOs)o f specific maternal and child health services. The RevisedNational Tuberculosis Control Program (RNTCP) contracts NGOs and private physicians and medical laboratories to provide specific services as part of tuberculosis control efforts. Services include case finding, diagnosis, and treatment. TheNational AIDS Control Organization (NACO) which finances state- and local-level HN/AIDS control activities operates largely through service delivery contracts to NGOs for targeted interventions, behavior change communication, and other services. Public-Private Partnerships. PPPs can play an important role inmeetinginfrastructure demand, particularly when investment needs for Few assets are large, but require the public sector to strengthen capacities. Both the central government andthe states are aiming to use public private partnerships (PPPs) more intensivelyto help meet gaps inthe provision o finfrastructure services. India has made progress recently inother sectors, such as ports and roads. A survey undertakenby the World Bank showed almost 90 PPP projects inthe transport and urban sectors that were either operational or at an advanced stage o f implementation, with total project costs o f around Rs 339.5 billion (World Bank, 2005i). 88 Figure5.2: PPPsare makinga growingcontributionto financingof infrastructure in India Number of awarded PPPs by sector Project cost of awarded PPPs by sector (total= 86) (total= Rs 339.5 bn) There i s scope for PPPs to contribute more to meeting India's infrastructure gaps. Over the last decade, private investment ininfrastructurehas averaged around 1% o f GDP in India. Other countries have achieved higher levels. Since 1996private investment in Chile has averaged 3-4% of GDP; inBrazil it has averaged 1.5% and inColombia around 2-3% over the same time period. InChina, private investmenthas accounted for around 10%of total investment inthe roadprogram, compared to around 4% inIndia. Were private investment to increase to 3% o f GDP it would represent $20bn per year. PPPs ininfrastructurerequire at the minimumwilling investors. Presently, private investment ininfrastructure inmiddle and low-income countries i s much lower than the peaks seen inthe late 90's. These were drivenbyprivatizations inLatin America and investmentsinEast Asia butthe increasedperceptions ofrisks involved ininvestingin infrastructurehave reduced transactions volume and the number o f investors participating inthemarket. Indiadidnotparticipateintheboomofprivateparticipation in infrastructure (PPI) inthe `90s. However, with its recent moves to expandthe role of PPPs, it managed to attract a higher level o f investment commitments to private infrastructure projects than either Brazil or China in2004 (see Figure 5.3). 89 Figure5.3: Indiahas overtakenBrazilandChinain attractingPPI Investment Commitments in Private Infrastructure Projects: India, China and Brazil 50 5 45 h 64 v) 2. 40 35 30 =-$ fA 25 C 20 15 - `E=10 F 5 0 Source:World Bank Private Participationin Infrastructure Database Unbundling,delegation,autonomy One keyway that engaging non-state providers ininfrastructure potentially improves performance is that it leads to greater clarity about unbundlingo f fimctions,. more specificity indelegation, and greater day to day autonomy. For instance, inwater and sanitation clarifying and separating policy, regulatory, financing, and service provision roles, which are presently usually combined inone agency, i s an important step. Sector regulatory agencies, now present intelecommunications andports at the national level and inpower at the state and national level, have been an important part o f attempts to increase accountability andperformance. An independent regulatory agency is a crucial step inunbundlinggovernment roles as it clearly separates the "policy maker" function o f government from the "provider" function so that, even ifboth are public sector bodies this creates checks and balances by creating incentives for generating performance information. That said, particularly inthe power sector at the state level, these agencies operate under difficult circumstances. They have been successful inincreasing transparency inthe sector, inparticular consumer participation and better information on the inefficiency of power utilities. However, evidence o fpoor Performance and the pressure o fregulators only transforms into service improvements ifthose responsible for service provision -the owners o fthe utility and the management - act upon them. Without this, even effective regulators will struggle to make much o f a difference. This emphasizes the need for "enforceability" to accompany information-if consumers have no effective choice then even with better information on performance the providers can remain unresponsive. 90 Evenwithin the existing constraints, regulatory agencies inIndia couldbe strengthened by areducedrelianceonretiredgovernment officers orjudges as commissioners, better funding, and clearer mandates stemmingfrom more consistent legislation. The courts have often not proved to be an efficient venue for the resolutiono f appeals against regulatory decisions. The use o f a specialized tribunal intelecommunications canbe extended to other infrastructure sectors. Examples from the health sector suggest that PPPmay be particularly promising insome areas of lagging states, where government infrastructure i s still poorly developed or ina significant state o f disrepair and understaffing. Where governance i s weak, PPP offers an alternative to higherrisk investments in state public sector delivery approaches. A well designed PPP can create greater transparency than may bepossible through the standard treasury expenditure method inwhich all transactions are internal. The extensively documented experience o fthe Janani Program inBihar, which has managedto set up a very large network o fprimary and secondary level service delivery points using a social franchising model i s a case inpoint. Insome cases, private Janani-linked facilities can be found operating successfully inclose proximity to virtually defunct government facilities, which still draw significant budget support. Enforceability: Subsidies anduser fees The needto both expand the nationalhighways network and rampupmaintenance expenditures highlightsthe challenges o f infrastructure finance. Evenifthe GQ and the NS-EW projects are implementedon schedule bythe end o f 2008, Indiawill have reasonably well surfaced, four-lane national highways accounting for just 22% o f the national highways and none of the state highways. India has some 470,000 kmo f `major district roads', and another 2.65 millionkm o f `village and other roads', which are typically unpaved. It i s estimated that the present system of roaduser charges will only generate around Rs 1600 billion total over the next 10 years, leaving a funding gap o f Rs 1050billionper year giveninvestmenttargets and maintenance needs. The same calculations o f "funding gaps" can be produced for sector after sector. Any expansion ininfrastructureservices inIndiaraises the question, who will pay? The answer to that i s clear: the citizens o f Indiawill pay. Whether services are producedby line agency, NGO, for profit private, community or local government inthe endthe citizens o fIndiapay as users or citizens o f Indiapay as taxpayers. There are five reasons why a larger proportion o f the costs o f infrastructure shouldbeborne by citizens of India as users o f infrastructurerather thanby all citizens o f India as taxpayers. 91 Figure 5.4: Comparingthe sources of finance: Either users or taxpayers pay-but sometimes for inefficiency and shortfalls create neglect of assets Finaneiol Who pn,,s Requirements financial Who pays Requirements Interest Reptace- ErnClem n. Australia Source: World Bank 2005e. First, transforming users into paying consumers would help make service providers more responsive to their customers. The more aware individual users are o f the true cost o f services they receive, the more likely they are to take action to exercise their voice. Second,it matches financing automatically to services and helps to improve the financial sustainability o f the sector. For instance, inonly a few mega-cities do tariffs cover even the operational and maintenance costs of water supply. This leaves the water supply agency dependent on general public sector funding, which can increase or decrease for reasons having nothing to do with the provider-and tends to leave maintenance costs squeezed out bywages and investment costs (thejobs and contracts nexus). Figure 5.4 illustrates the "build/neglect/rebuild" mode with an example from irrigation systems. Whennearly all the fundingcomes from taxpayers there is a tendency both for excessive cost and for crucial functions like replacement and operations and maintenanceto be under funded. Third,user charges are, inmany instances, fairer. Very often low prices arejustified on the grounds that theseprices are for essential services andthat higher priceswould hurt the poor. Inwater, as inelectricity, industrial and commercial consumers cross-subsidize domestic consumers, with the rationale beingthat this is pro-poor. The evidence suggests that, by and large, the subsidies that prevail inthe Indian water system do not benefit the poorest domestic consumers (see Box 5.1). Inthe power sector, subsidizedtariffs for agriculture benefit richer farmers more than the poor: a flat rate per pumpbenefits those who have more landto be irrigated by each pump. One study estimated that payments for electricity represented 6% o f gross farm income for larger farmers and 13% o f income for marginal farmers (World Bank, 2001b). The available evidence suggests that subsidies are not at all well targeted. Distortedtariffs for water, as with most other infrastructure services, mean that industrial and commercial consumers cross-subsidize domestic consumers, but these subsidies are not well targeted towards the poor. Moving towards user charges does not mean ignoring the very real concerns o f poorer citizens- 92 buttariffs canbe increasedto reflect sound commercial principles whilst protectingthe poorest. Box 5.1: Do water subsidies inIndia reachthe poor? According to one estimate, subsidies inthe water sector inIndia amounted to around 0.5% o f GDP inthe early part o f this decade. Thoughjustified on the grounds that they help ensure that the poor receive an essential service, the evidence suggests that the non-poor are likely to benefit from these policies more than the poor do. A survey o fwater usage andpricing inBangalore revealed that the richest 10%o fthe populationreceive 2- 3 times the subsidy received by the poorest. Public stand-posts are heavily subsidized and reach a large proportion o fthe poor (77% inBangalore) and are not muchused by the rich. However, the Increasing Block Tariff (IBT) which sees charges increase inline with water usage benefits the rich who have private taps and does not benefit the poor who do not. It was estimated that around 70-80% o f the subsidy associated with private taps through the IBT goes to the non-poor. Source: World Bank 2005e Moreover, low monetary usage charges does not mean that the poor do not pay-- consumers however do pay despite the subsidies-they pay through coping expenditures, they paythrough time. These costs can be borne disproportionately by the poor. Studies inHaryanaandAndhraPradeshestimated that the costs ofrepairingirrigationpump motors, burnt out by erratic voltages, amount to 10%o f gross income for marginal farmers but only 2% for large farmers (World Bank, 2001b). InDelhi, a recent study estimated that households living inJJ colonies spent around Rs 160permonthincoping costs -much o f it time spent collecting water - compared to Rs 220 per monthon average for households living inauthorized colonies (Misra, 2005). Increasingthe quantity o fwater andreliability o f service can bring substantial benefits, particularly to the poor. A study o fwater municipalities inArgentina (Galiani et. al., 2005) showed that those which hadbrought inprivate participation saw the biggest declines (5-9%) inchild mortality attributable to water borne diseases. The impactswere greatest inthe poorest areas, which was where services were worst. The study attributed the declines to improved pressure andhours o f service which resulted from higher investments inthe privatized utilities. A broader review o f experience (World Bank, 2005J) shows that private participation inwater inLatinAmerica lead to increases in coverage inline with historical trends but strong improvements inwater quality and continuity o f service. Where the poor were already consumers, price increases caused negative impacts and in some cases highconnection fees dissuaded the poor from taking upwater connections. The latter canbeavoided, as was done inthe renegotiationo fthe Buenos Aires metro water concession, by levying a fee on all consumers to promote the expansion o f access that can substitute for highconnection costs. All o f this points to the needfor careful design o fwater and sanitationreforms and pricing for these essential services. Fourth,user charges give users the right incentives. Increasing the cess on fuel - as has beenproposed - will help to close some o f financing gap and by linkingroad costs to roadusagehelps inhaving private individuals and firms take into account the true social 93 cost o f vehicle use. Moreover, user charges can be useful to address some o f the existing distortions: for example, lower taxes on diesel mean that heavy trucks do not even cover short runattributable costs and buses are more heavily charged than multi-axle trucks which impose substantially higher wear on roads. The obvious distortion inuse are power tariffs which encourage continuing and increased pressure on ground and surface water. resources as farmers do not have to incorporatethe full social cost o ftheir actions. Fifth, "user pays" as a general principle makes the rationale and design o fpublic sector subsidies when they are adopted clearer and more transparent. Subsidy design also has to take account o f the extent to which they either hinder or benefit the activities o f an incumbent provider incompetitive segments o ftheir market. For example, IndianRailways (IR) i s required to function as a commercial operation, providing critical transport services to the economy. However, it i s also mandatedto perform social functions, inparticular the provision o fpassenger services at highly subsidized fares. Ithas already lost freight traffic to roadhaulage andnow faces increasingcompetition for its profitable premiumpassenger services from the growing budget airline market.Despite the fact that freight services earn IRmost o f its surplus, inadequate investment inrolling stock inrecent years means that failures o f infrastructure and stock on highdensity corridors have lead to congestion and deterioration inservice quality for their most profitable services. Although there has not beenmuchprogress inrealigning andbetter targeting subsidies there has been some progress inthe management and design o f subsidyregimes. The Electricity Act (2003) has set targets for the phase-out o f cross-subsidies. Some states - for example Andhra Pradesh-have been making cash transfers to power utilities for politically mandated subsidies.More generally there i s a greater awareness on the need for better targeting o f subsidies towards poorer consumers. Finally, the Government o f Indiahas developedtwo funds - the Universal Service Obligation Fund(USOF) for telecommunications and the Viability Gap Fund-which will use competitionto select service providers and to determine the cost o fthe subsidy to There is considerable scope for making the auctions for USOF subsidies more competitive (No11 and Wallsten, 2005). Tying the provision o f subsidies to the actual delivery o f service will also help make service providers more accountable for the funds they receive. External accountability and information Where it can be introduced, competition can be a powerful method for introducing accountability and improving performance. This has been most marked in 23The USOFhas beenestablished as one method o f supporting the roll-out o ftelecommunications inless profitable rural areas through the competitive award of subsidies. The Viability Gap Fundwill provide subsidies to PPP projects inthe infrastructure sectors that are felt to be economically justified but not commercially viable, and will also rely on the subsidy being set by a competitive bid for service provision. 94 telecommunications, where it has beena powerful force to reduce prices and expand access inIndia. There has beenparticularly rapid growth incellular phone usage and the rates charged to Indianconsumers are some o fthe lowest inthe world. Competitive pressures generated by the entry o fnew service providers have also improved the efficiency o f incumbents, seennotjust intelecommunications but also inports. Better monitoring o f ex-post outcomes andreturns to investment projects are also needed. Benchmarking the costs o fproviding key services and investments, might be useful in pointing out differences inefficiencies between implementing agencies and approaches. A higher contribution byPPPsinIndiawill need careful selection, structuring and oversight o fprojects. This will require the government to be selective about which projects are best done through the PPP route, and which better done through traditional public procurement. It also requires the capacity to understandwhat i s needed to make PPPs attractive to the private sector andwhen they are likely to perform well. Itis still the case that too often PPPs are viewed primarilyas a source o f financing to undertake investmentsthat the government cannot affordto do. This ignores the claims that PPPs place on public resources. Also, inmany sectors (roads, water and power) revenues from user fees will have to be supplemented bytaxpayers funds for projects to achieve commercial viability. Inparticular, Indiahas the scope to build on its successesto date and see an expanded program o fPPPs inthe transport sector, inparticular highways, ports and airports. PPPswill likelybe more controversialinthe utilities sector, such as power and water, as they have been elsewhere. The starting point has to be an assessment o fhow well the present systemi s performing, and the costs that this imposes on different users. Creating a greater role for PPPs requires addressing the key constraints to their greater use at present. The Viability Gap Fund, which will provide subsidies to PPP projects in the infrastructure sectors that are felt to be economically justified but not commercially viable, is a good step forward. A key principle i s the use o f competitive biddingto minimize the subsidy to be provided bythe government, an approach used by Chile and other Latin American countries for the promotion o f rural infrastructure in telecommunications and power. This will helpbut other measures will be needed to see an increased use o f PPPs. Most countries engaged ina broad-based PPP program have felt the need to develop a cross- sectoral PPP unit although the role that this unitplays i s sometimes restricted to information dissemination and the preparation o f guidance material. The designresponse to two keyissues-the role o f a cross-sectoral unit vis-&vis line ministries and the role o f a national unit insub-national PPPs -will be drivenby the business practiceswithin governments and the fiscal, and other, relations between the center and the states. Nationally, capacities for PPPs must be strengthenedinline agencies but a cross-sectoral pool o f expertise, most likely that runningthe VGF, can help disseminate and introduce good practices more quickly. At the state level, dedicated PPP units can bothbroadenthe PPPprogram and strengthen capacities to prepare and oversee these projects. Finally, 95 funds to help defray the costs of preparing PPPswould lead to better designedprojects, particularly at the state and municipal level. Political economy of external reforms of service provision Systems persist becausethey havepowerful constituencies. This makes the political economy o f any reform a difficult matter. Reforms that engage non-state providers often prove particularly difficult-as theyboththreaten existing providers andnecessitate some "more for more" arrangement whereby users get better services-but are expected to pay more. Naturally botho fthese create resistance and the lessons learned need to be incorporated into future plans. The global reductioninPPPsillustrated infigure 5.3 to some extent also reflectschanges inassessments, not o fthe benefits, buto fthe viability o fthe political economy ofPPPs. Latin America saw the greatest use of PPPs for infrastructure services but inrecent years public perceptions o fthe results have become unfavorable, despitebroadly positive evaluations o f their impacts (see Box 5.2), as the highlyvisible failures are widely touted by opponents andthe less highprofile successesdownplayed. Insectors where competition can readily be introduced, such as telecommunications, deregulation and privatization lead to large and immediate gains andhave beenwidely popular. Insectors with little competition but for which consumers are not the general public, such as ports and airports, PPPs (often contracting out o fmanagement) have also been, by and large, successful as both costs andbenefits o freform are concentrated. Problems and controversies have arisen more often inpower andparticularly water, where revenues from users have rarely covered costs andpolitical economy problems are greater as users are diffuse and coverage issues are important. Although well-designed transactions in these sectors have produced good results, these issues have made them more difficult to sustain. Box 5.2: PPPs in Latin America: an assessment 96 LatinAmerica has attracted around half o f the total o f $786 bnininfrastructure transactions with private participation. A s a result there were major shifts to the private provision o f infrastructure, with around 60% of electricity connections being managedbyprivate companies. Public opinion inthe regionis hostile towards PPI inessential infrastructure services. But incontrast most analyses show broadly positive results althoughthere have been variations inperformance. The poor have not been negatively impacted and have often gained from improvements incoverage. Efficiency has improved, profits have not ingeneral been excessive, andprices have fallen as well as risen following the introduction o f the private sector. Negative public sentiment may reflect a number o f things: a more general rejection o f "Washington consensus" reforms; poorly managedperceptions and expectations; concerns about transparency o fPPI transactions, or perhaps the public placing a different weight on some ofthe outcomes than economists do. Renegotiations and concerns over weak regulatory agencies that cannot adequately protect consumers have fuelled the opposition. Source: World Bank, 2005j. There is similar experience inIndia. Despite some promising results, the Delhi power privatizations remain the subject ofm u c h debate (Box 5.3). The proposed water reforms inDelhi are meetingresistance as well (Box 5.4). Box 5. 3: Power Privatization inDelhi The distributionbusiness o f the DelhiVidyut Board (DVB) was privatized inJuly 1, 2002 with the sale o f 51% equity ineach o f the three distribution companies. BSES Limited, now Reliance Energy Limited, acquired a majority stake intwo o f the companies and Tata Power acquired the majority inthe third, NDPL. The companies were given targets for loss reductions and subsidized bulk power costs for the period through to end 200512006. There has been considerable public debate on whether the reforms thus far have achieved their objectives. Onthe one hand, the companies had, at the last regulatoryreview, met or exceeded their targets for loss reduction, inthe case o fN D P L by as much as 7 percentage points. Limitedinformation o n improvements inthe quality o fsupplysuggests that reliability has improved somewhat. Priceshave however increased considerably. Though less than hadbeen forecast by the Government at the time o f privatization, tariff increases in2003 and 2005 have lead to protests by consumers. Inresponse the Delhigovernmenthas provided subsidies, inthe order o fRs 140crores to the companies to reduce the required price increases for targeted consumer groups. Consumer complaints have perhaps been strongest over delays inmeter replacement, inaccurate billing and what they regard as a slow response by their companies to these complaints. As with utilityprivatizations inother countries, it i s clear that consumers expect rapid service performance improvements once the company i s turned over to private hands. Box 5.4: DelhiJal Board Reforms 97 The quality o f water supplies inmost o f India's major cities remains poor by international standards. The Delhi Jal Board, was recently at the center o f controversy over plans to contract out services on a pilot basis insome areasto the private sector, provides what canonly bedescribedas poor quality service. Despite a highavailability o fwater, at 250 liters per capitaper day, the Board can only distribute water for around 3- 4 hours per day. This levelo fperformance putsit behindmany water utilities that serve major cities in Africa. More affluent consumers can still provide themselves a 24x7 service through overhead tanks and boreholes, the latter adding to the rapid depletion o f the water table. Incontrast, the poor do not have this option, and spend considerable amounts o f time to meet their water needs. As mentioned above, the coping costs for households inslums interms of time wasted are nearly as large as the money costs inmore affluent neighborhoods. Buthowever dire the current performanceor large the potential benefits, reforms needbothto have a transparent process and a strategy for addressing the concerns o f the public. Transparency and consultation inidentifying and implementing reforms are essential, bothinensuringthat all stakeholders have confidence inthe process andcanmaketheir voices heard, butalso to ensure that the right service options are developed and price structures that reflect what consumers are looking for. Also, delay inresponding to concerns about notjust the substance but also the processes around the reform leadto greater controversy and can damage the reputation o f reforms. There are four lessons from the experience to date: 0 Transparent and consultativemethods for involving stakeholders inthe decision- makingleadingto the solution is critical. There is no guarantee that goodprocess will avoid later criticism, but can improve design by anticipating problems and gives a more solid platform for respondingto critiques that emerge. 0 Particularlywhen higher user charges on existing users finance expansion o f coverage, it must be acknowledged that those with coverage (who tend to be powerful) will be worse off if services are not improved. So, either the newly served have to be mobilized as a reform constituency or the service improvements have to be immediate andtangible. 0 A communications strategy is an integral element o f areformpackage-which includes both explanations to users on the process o f decision making, the choices made, their benefits, and for respondingrapidly to criticism that comes (and it will come). This is difficult because the civil servants andtechnocrats incharge of reform may not be empowered to respondto public criticisms. The role of the center inpromoting-or delaying--reform The need for reforms inthe infrastructure sectors havebeen examined and highlightedin a numbero f cross-cutting and sectoral reviews and commissions, o fwhich the India Infiastructure Report (1996) was one o f the earliest and the most comprehensive. The political difficulties encountered inimplementingreforms inthe individual states has been recognized by the Government o f Indiawhich has inrecent years established programs to encourage and incentivize state governments to reform, namely the Accelerated Power Development and Reform Program (APDRP) andthe Urban Reform Incentive Fund (UHF). Bothhave broken ground by targeting support to the 98 implementation o f sector reforms. Key areas for the URIF were the regulation and taxation o f real estate markets, with aims at incentivizing reforms inthe Urban Land Ceiling Act, stamp duties, rent control andproperty tax rates and collection. The APDRP was structured somewhat differently, with one component incentivizing critical physical investments to improve the performance o f the distribution sector, and a second smaller component to provide incentives to state governments that were improving the financial performance o f their power sectors. Bothhave, however, had limitedimpacts. As reported by the Planning Commission in their 1Oth Planmid-termappraisal, the investments indistribution projects underAPDRP are around one third o f the level o f funding allocated. The incentive component paid out underAPDRP has been around one quarter o fbudgeted levels, partly reflectingthe difficulties in improving the financial performance o f the sector. However, this incentive component can bepaid out for merely increasing the average revenuerealization- for example shifting sales from low-tariff users to higher-tariff users -rather than more narrowly by reducing aggregate technical andcommercial losses. The URIF offered only a relatively small amount -Rs 500 crores intotal - for a substantial package o freforms, particularly when compared to funds available that were not conditional on reforms.24 The reforms involved require a cross-sectoral effort at the state level, and state Ministries o fUrban Development and Local Development by and large lack the convening power to bringtogether the mainplayers, including Finance. Finally, the milestones for progress under the URIFwere not clearly specified andhalf of the funds were released on the basis o fthe signature o f Memoranda o fUnderstanding betweenthe state government and the center. The Jawaharlal NehruNational Urban RenewalMission (JNNURM) i s beingdeveloped as a successor to the LJRIF, bringing together a number o freform based initiatives into an integratedprogram linkeddirectly to the financing o finfrastructureprojects at the city level as opposed to budgetary support at the state level. Some o f the lessons learnt from implementation o fthe URIF are being incorporated, including providing the program greater visibility and coordination under an authority to be chaired by the Prime Minister. However, the JNNLrRMmay result in prolonging the time frame within which states must implementkey reforms, as they now have until2010 to implement core reforms rather than 2007 under the URIF. Both initiatives havebeen important inhelpingplace reforms further on the map and in providing models by which central government funding to incentivize reforms can be developed. There are several important lessons from these two efforts. Firstly, the incentive effect will not be large ifa much greater quantumo f funds can be made available unconditionally. Secondly, the measures being supported have to be clearly specified and readily measurable, as well as being important steps inthe overall reform process and ones that can realisticallybe implemented. Giventhe paucity o f data on the cost o f service provision, cost recovery inurban services would prove a difficult target to monitor. 24The total URIFfunds to be released over five years amount to about one sixth of HUDCOreleases and one-twentieth of HUDCO sanctions. 99 A paradigm shift inpolicy directions is requiredto improve the coverage and sustainability o f the rural water supply sector, including a demand-responsive approach, community participation, and decentralization o fpowers for implementing and operating drinkingwater schemes. To addressthese issues, the Go1has launched a national program "Swaj aldhara", with guidelines on policies, institutions, technology options and user contributions. The main challenge i s the implementationand scaling-up o f this ambitious program to deliver 'reliable, sustainable and affordable' services. At present, the traditional supply drivenprograms, including the Accelerated Rural Water Supply Program (ARWSP) continue to fimd most o f the rural water supply activities. Since the States can get financing underthe traditional supply drivenprogram, it does not have any incentive to scale up the sector reforms. Further, the Twelfth Finance Commission (TFC) recommends grants that can be utilized by PRIs for repairs / rejuvenation and O&M cost of rural water schemes, thus reducing any incentive for financing O&M through user charges. The recently announced Bharat Nirman funds follow the ARWSP guidelines and continue to reinforce supply driven programs. Ideally, these funds should be used as incentive funds to scale up reforms. Unless the trendis reversed andthe Swajaldhara funds overtake the traditional funding, there ishardly any motivation for the Statesto adopt reforms (see Figure5.5)-0r communities to support them. Figure 5.5: Most fundingprograms still adopt traditional supply driven approaches. Rural Water Supply & Sanitation: Current Approaches P Source; World Bank 2006 100 Introductionto Part I1 India's success infostering economic growth has created new challenges, as outlined in the overview. This part o f the report focuses on how India can sustain that success and expand the opportunities for all to participateinit and benefit from it. Although Indiahas beengrowing very rapidly, international experience shows that complacency about pushingaheadwith reforms when growth is highis a sure recipe for slower growth. As argued in Chapter 6, infrastructure has now become a major constraint on economic performance. It would nevertheless be unwise to ramp up government infrastructure spending inthe hope that this will lead to a growth spurt and lower the debt-to-GDP ratio without a simultaneous comprehensive fiscal reform. Chief among the fiscal policy priorities are steps to improve the composition and efficiency o f existing expenditure and revenue mobilization. Revenue deficits needto be lowered and capital expenditure raised. The recommendations o f the Twelfth Finance Commission (TFC) aim precisely to do this. Notwithstanding India's impressive growth performance, fiscal adjustment continues to be a live issue as Indiahas shown little sign o fgrowing out o fits debt problems. This could be compounded by interest rate risks looking forward, growing quasi-fiscal deficits, and the persistent problems from food, fertilizer, andimplicit power subsidies. The composition o f government expenditure remains indire need o f improvement. Given the largeneeds ininfrastructureand the many other thingsthe government would like to do that require additional resources, the current fiscal position leaves no "space" to fund them. To undertake new initiatives existing spendinghas to berestructured and taxes have to increase so that governments, both at the central and state level, can continue to meet their fiscal adjustment targets. Chapter 6 thus argues that, unless addressed, infrastructure and the fiscal deficit are the most likely obstacles to overall sustained rapid growth. India's successesso far have not been shared equally, as some states and people have done muchbetterthan others, particularly since the reforms o f the 1990s. Giventhe concerns about the distribution o fthe benefits from economic growth, a special focus needs to be on those areas for action that would accelerate growth by expanding opportunities, which could lead to a more equitable distribution o f the benefits o f growth. The search for "equalizing growth accelerators" inchapter 7 focuses onreforming labor regulations and the financial sector. Current labor regulations seem to be protecting workers injobs by "protecting" other workers from havingjobs. Properly done, labor regulation reform could be an important equalizing accelerator. Financial reforms have made much headway, but have so far primarily focused on serving larger companies, as inthe caseofequitymarketsreform. Improving accessto finance for small andmedium- size enterprises i s a second important growth accelerator. The needto accelerate growth inIndia's agricultural sector prompts special consideration inchapter 8becauseofthe impact ofthe current slowdown on agricultural production, employment, and wages. The traditional, relatively heavy handed approach has been to 101 promote intensification, primarily infood crops, through public investments coupled with inputsubsidies. Although this approachhasbeensuccessful inmany ways, itis largely played out and i s often promoting agricultural practices that are neither economically nor environmentally desirable. A new strategy should rely much more on policies to liberalize agricultural inputs and outputs by reforming land administration and tenancy requirements andby supportingresearch and development that is tailored to specific needs. A workable strategy to accelerate productivity inagriculture has to promote four directions for agriculture inthe overall rural economy: 0 Intenszjication. Increase the productivity o ftraditional crops through greater use o f irrigation, highyield varieties, agrochemical inputs, andmechanization. 0 Diverszjkation. Shift to newmore profitable crops (fruits, vegetables, higher value cereals, medicinal plants) and livestock. 0 Nonfarm Einkuges. Emphasize related activities that add value to agricultural products, such as trading, agroprocessing, and supplyinginputs. 0 Exit. Switch from agriculture to other, more productive activities. The government o f India's justifiable concern with the inclusiveness o f economic growth can be addressed by focusing on expanding the regional scope o f economic growth, expanding access to assets and thriving markets, and expunding equity inthe opportunities for the next generation o f Indian citizens no matter who they are or where they live. These issues are covered inthe final two chapters o f the report. Chapter 9 discusses the issue o f lagging states. Actions are recommendedinfour policy areas. First, it i s important continue to pursue policy reforms that help to integrate India's goods and labor markets by eliminating barriers and distortions to goods moving across states. Second, with integrated labor markets, the large differences inwages across states create incentives either for people to move to jobs (either permanently or temporarily) or for jobs to move to people. But for jobs to move to people inthe lagging states, those states must address the deficiencies intheir investment climate, such as the need for law and order, transport connectivity, and streamlined regulations, so that they can attract industry.Third, the laggingstatesneedto focus on the implementation o freforms inrural areas, where the greatest benefits are likely to come from the expansiono f finance and infrastructure. Finally, fiscal federalism and Finance Commission allocations can assist in makingthe laggingstatesmore effective development agents. Chapter 10focuses on the role o fpublic action inequalizing opportunities by promoting equality inaccess to markets and assets and by scaling up efforts to strengthen livelihood strategies through empowerment. Evenwith the best o f access to markets and opportunities, there are needs for the government to engage insocial protection to assist the poorest and help incoping with the risks and vulnerabilities that citizens face. Social protectionmeasures are important to combat extreme deprivation, but can have equally important dynamic efficiency effects by allowing people to bear risks and undertake profitable investments. 102 Chapter 6. Easingconstraintsto sustainrapid growth India's experience has been very influential inthe continual reshaping o f views about growth. This is not surprising as, besides beinglargeinpopulation and economic size (on the basis ofpurchasingpower parity, Indiais now the fourth largest economy inthe world), India i s also one o f the very few countries that have accelerated growth inthe 1980sand 1990sto rapid levels, avoided a severe macroeconomic crisis, and maintained steady economic growth averaging about 6 percent a year over the past quarter century. A recent World Bankreport compares the experience o f India, China, and Vietnam (the globalizing formerly socialist economies) with that o f Latin America, Africa, the Middle East, and the former Soviet Union. Key conclusions from that report andthe implications for India are discussed inbox 6.1. Box 6.1: Lessons from the 1990s The recent World Bank report Economic Growth in the Nineties compares international growth experiences and highlights the following four lessons. Initiating success does not require getting everything right, but it does require getting the right things right. A growth strategy i s not simply a list o f all o fthe things wrong inthe economy with a reform package for each. Rather, the essence o f a strategy is to identify accurately the bindingconstraints- those that would have the biggest growth payoffifthey could be eliminated-accurately and address them effectively. I Sustaining a boom requires continuous action. Growth accelerations are relatively common-countries have a one-in-four chance o f experiencing an "episode o frapid gr~wth)'~~ inany givendecade-but growth inmost developing countries tends to be highlyunstable, involving booms, busts, and periods o f stagnation alongside periods o f rapid growth. Very few countries have experienced consistently high growth rates over several decades. The good news i s that growth often comes, the badnews i s that growth nearly always goes-almost half of the periods o f rapid growth were followed by slow or negative growth. Complacency with highgrowth is a recipe for slow growth. Sustaining growth requires attacking the bindingconstraints inturn, before the bindbites.26 Commonprinciples, heterogeneous implementation. The experience o fthe 1990s has not demonstrated that the fbndamental principles o f economic growth are wrong-macroeconomic stability, an investment climate conducive to high rates o fprivate investment, and fulfilling core public responsibilities are keys to growth-but it is the case that there are many ways to achieve those principles, and a rigid orthodoxy about how those fundamentals are achieved cannot be supported. Credible steps with institutional continuity: sustained implementation is key. A final lesson, again consistent with the Indian experience, i s that even modest steps, ifthey are credible and are implemented well, can create a positive dynamic o f expectations that sustains growth-while "big bang" changes, even inthe right direction, may have n o impact ifthey are not expected to persist and implementation is lacking. Source: World Bank 2005k 25A growth acceleration i s defined as an increase inper capita growth o f two percentage points or more. T o qualify as an acceleration, the increase ingrowth has to be sustained for at least eight years and the post- accelerationgrowth rate has to be at least 3.5 percent per year (Hausmann, Pritchett, and Rodrik 2005). 26Moreover, international experience suggests that it is very difficult to accelerate from fast to faster growth, andthat only a few countries that have done it (Hausmann, Pritchett, and Rodrik 2005). See also Acharya (2002). 103 India's aggregate growth performance inrecent years has beenparticularly robust as the industrial slowdown observed duringthe NinthPlanperiod (1997/98-2002/03) has been reversed as industryhas recovered, while the services sector has continued its rapid expansion. As discussed inPinto, Zahir and Pang (2006), this robust recent growth performance is closely linked to the still on-going microeconomic restructuring o f firms inresponse to the 1991economic liberalization.Evidenceshows that serious industrial restructuring began only in 1996,towards the end o fthe EightPlanPeriod, when companies felt the heat from risingimport competition and falling profit margins.27After 1996, companies started reinventingthemselves. By 2002, companies had substantially restructured financially and hadraised production efficiency and quality while lowering costs. Startingin2003, companies resumed investinginnew capacity with a noticeable pick-up inmomentum in2005. Notwithstanding India's stellar growth performance, with GDP growth averaging 6.5 percent inthe first three years o fthe Tenth Plan, meeting the Tenth Plantarget o f 8 percent no longer seems withinreach.28Infact, the Tenth PlanMid-Term Appraisal has lowered the expected growth rate to 7 percent and has revised the target for the Eleventh Planfrom 9.3 percent to 8 percent. This chapter aims to identifythe most bindingconstraints on economic activity, where policy changes could yield the biggest growth gains for the least reform effort. We explore two issues as the most likely candidates for important constraints to aggregate growth inIndiainthe medium term: the infrastructure andthe fiscal deficit.29With the process o f industrial restructuring reaching a point where firms are gearing up to make largenew investments, infrastructure i s emerging as an increasingly critical gap. As discussed below, inadequate infrastructure will particularly hurt highemployment- generating, labor-intensive small enterprises; precisely the sector which has the greatest potential to absorb India's fast-growing working-age population. Alleviating the infrastructure constraint is unavoidably going to involve a government role and increased public spending on infrastructure, although the costs o f filling the infrastructure gap and the relative roles o f the government and private sector need to be defined. Other constraints, most notably labor laws and the financial sector, are important equalizing accelerators and are key factors to achieving inclusive growth. They are covered inchapter 7. Chapter 8 focuses on the challenges posedby the agricultural sector. Because the vast majority o f India's poor resides inrural areas and draws their livelihoods from agriculture, rural development is key to inclusive growth inIndia. Chapter 9 addressesthe particular problems o f accelerating growth inthe lagging states. There are many other areas o fpolicy that require attention-such as poor governance, unclear and insecure property rights, and unsustainable use o f natural resources. While reforms inthese areas are desirable, indeed critical, for a long-term development strategy, *'See '*Achievement also Forbes (2002) and Mohan (2004). o f the plan targets is therefore only possible ifGDP growth inthe last two years averages nearly 11percent per year. 29Pinto, Zahir and Pang (2006) highlightthe interdependence between the soundness o f public finances and the microfoundations o f growth. 104 we focus primarilyon the main obstacles to sustained growth inthe medium term. Some o fthese areas are covered inbox 6.2. Box 6.2: Other issues for growth In 1991, faced withabalance of paymentscrisis, Indiaembarked ona historic set o freforms to make the economy more productive by unleashing competition and innovation. The view o f this report is that the two largest policy issues to be addressedto sustain or accelerate growth inthe mediumterm are infrastructure and the fiscal deficit. That said, there are many other issues that could very productively be addressed-it should not be inferred that other reforms are unimportant. The burden of business regulation. Regulatory procedures governing entry, exit, and day-to-day operations of businesses are particularly cumbersome inIndia. Objective indicators on the cost o f doing business are considerably higher inIndia than ina number o f comparable countries, including China. It takes three months to open a business inMumbai: two o f the months are spent inobtaining the Personal Account Number and the Tax Deduction Account Number, and the business i s not allowed to start while waiting to formalize the tax number. There is also room to speed contract enforcement: currently businesses must go through 40 steps, costing well over a year (425 days) and an average 43 percent o f the contract value to enforce a contract through Indian courts. Although secured debt enforcement was recently improved with the Securitization Act, other contract enforcement remains inefficient. The greatest delay comes from enforcing the judgment: an average of 306 days, and this assumes that the debtor does not oppose the seizure. Trade restrictions, tarzfs, and regulatory barriers. At the time o f the trade reforms in 1991 it was feared that lowering barriers would not be possible and that India couldnot export sufficiently to prevent a current account imbalance. Subsequently, there was a boom inexports, but although India i s muchmore open, it remains a relatively protected economy, with tariffs averaging 22 percent (18 percent intrade-weighted terms)-above the average tariffs o f 9.5 percent inemerging Asia and 11.5 percent globally (see Trade Policies in South Asia, World Bank 2004 (Rep. No. 29949)). Moreover, exemptions and partial exemptions are sometimes granted, often inan opaque fashion. Significant nontrade barriers also remain. Further progress indismantling tariff and nontariff bamers will be key to sustaining growth rates. For instance, there i s significant scope to streamline procedures for importing and exporting goods across borders. In India, it takes significantly more time to import and export goods than in,say, China, Malaysia, or Russia. The number o f documents that Indianimporters and exporters have to complete i s also muchhigher than in competitor countries, as i s the number o f signatures needed. Urban land markets. Yet another set o f constraints relates to the use and transfer o f land, which critically affect the performance of firms, particularly larger firms. Obsolete tenancy and rent control laws keep a large part o f urbanreal estate off the market. The freezing o f rents at unrealistically low levels inMumbai, for instance, has raised rents for new properties to phenomenal levels, while keeping rents for old but desirable properties very low. Practices such as this hamper the growth o f domestic retail trade and construction by making it very difficult for new players to start up businesses. A report on "India's Growth Imperative" by the McKinsey Global Institute (2001) argues that land market distortions account for about 1.3 percent o f lost growth per year. The central government has abolished the UrbanLand Ceiling Act, which made changes inlanduse very difficult; however, only a few states have repealedtheir corresponding Urban Land Ceiling Acts, and this needs to be extended to all states. Infrastructureas a bindingconstraint The massive demands now on power networks, transport, urban infrastructure, and ports are the result o f India's success inpromoting economic growth. The danger i s that poor infrastructure now will beput a brake on that growth. Moreover, because infrastructure 105 bottlenecks may be affecting the poor disproportionately, overcoming these bottlenecks can help to equalize as well as accelerate growth. Figure6.1: Infrastructurestocks in India and China, 1998 and2003 300 1 1998 1 2003 1998 1 2003 1998 I 2003 electricity generation, lanof pavedroads per Number of fixed lines watts per person 100,000 hab per 1000people Source: World Development Indicators The comparison o fthe Indiaand China is instructive. In 1980 India had higher infrastructure stocks-power, roads, and telecommunications-but China has invested massively ininfrastructure, has overtakenIndia, and i s wideningthe gap (figure 6.1). From 1990 to 2000, China's installedpower capacity increased by 136percent, compared with an increase o f51percent inIndia.30China has beeninvestingannually around 8-1 0 times India's level inhighways since the mid 1 9 9 0 ~ .The gap ininfrastructure stocks i s ~ l now so large that for Indiato catch up only to China's present levels o f stocks per capita, it would have to invest 12.5 percent o f GDPper year through 2015. The greatest pressures on Indian infrastructure are probably inelectricity, roads, and urban infrastructure. India's power sector problems are legendary. According to a survey of private investors, power outrages occur almost every other day for the average business inIndia, as compared to once every two weeks inChina and once a week in Brazil. The average manufacturer inIndia loses 8.4 percent insales annually on account of power outages, as opposed to less than 2 percent in China andBrazil.The frequency and average duration o foutages are such that generators are standard industrial equipmentinIndia, accounting for as much as 30 percent o f abusiness's power consumption inmany cases. Almost 61 percent of Indianmanufacturing firms own generator sets; the comparable figures are 20 percent inMalaysia, 27 percent inChina, and 17 percent inBrazil. Moreover, India's combined real cost o fpower i s 74 percent higherthan Malaysia's and 39 percent higher than China's. Although power sector 30Generation capacity grew at around 10-12 percent per year before the 1990s. It is estimated that during the 1990s captive generation capacity grew at around 9 percent, compensating for the slow rate o f growth of public generation capacity. 31World Bank, 2004b. 106 reforms have proved demanding, both technically and politically, solutions must be found that meet growingdemands while reducing the fiscal burden. The discussion o fpower sector i s briefinthis report as the issues have beenwell covered elsewhere. Infrastructure constraints inIndia are most bindingon manufacturing, particularly export- orientedmanufacturing. India i s well placedto be a much more significant presence in the world market for exports. The share o fexports and imports inIndia's GDP today is less than halfthat of China's, and at 2.5 percent, India's world market share inthe trade of goods and services i s a small fraction o f China's (10.5 percent). Nonetheless, India exports an impressive mix of goods quality -India's current export composition is similar that that of countries with income almost five times as high(Hausmann and others 2005). So, while Indiai s well placed interms of industrial experience, technical capacity, and available labor, infrastructure hinders India's success inglobal trade. Increasingly manufacturing exports are part of a supply chain inwhich firms source their production inmany different countries, oftenwithmanycountries producing inputs or components o f a final product. This type of manufacturing i s particularly demanding of infrastructure, particularly reliable and efficient transport and production. Because the value o finputs brought inis large relative to the value o f the exports, the margin o f competitiveness is low, and hightransport costs eat significantly into those margins-thus to be competitive firms have to pay lower wages. India i s just beginningto attract this type o f export- orientedmanufacturing, and it i s a potentially important component of fbture growth, one that can help to equalize growth because it employs more semiskilled labor than do the export-oriented services. O f course, this report i s hardlybreaking new ground inpointing out the weaknesses in infrastructure: the problems are widely recognized, not least by the government. It i s clear that Indiahas not invested sufficiently inexpanding infrastructure assets. The 1996 India Infrastructure Report called for an increase ininvestment to around 8 percent of GDPby 2005/6. Recent estimates producedby the Bank suggest that Indiamightneedto invest 8 percent or more of GDP over the period 2006-10 to sustain annual GDP growth at near 8 percent and replace old capital stocks. Over the past decade, however, estimates suggest Indiahas invested only 3-4 percent o f GDPininfrastructure. There are also significant regional andincome-related differences inaccess to much-needed infrastructure, which the government is seeking to address andwhich also require investment. Muchof the funding will have to come from public resources. This raises the issue of "fiscal space" addressed later inthis chapter. Implementationand absorption capacities will have to improve to take advantage o f any increase inthe resources devoted to infrastructure. The Planning Commission, inits recent midtermreview o f the TenthPlan, estimated that only 75 percent o f planned power capacity will be realizedby the end o f the Tenth Planperiod. At the national level reforms to project monitoring andmanagement have reduced cost overruns from 60 107 percent inthe early 1990sto around 20 percent at present (IMF 2004).32Better monitoring o f outcomes and returns to investment projects are also needed (see part I). Policy discussions recognize that a "business as usual" approach will not ease the emerging bindingconstraints to growth posed by infrastructure: even "spending an additional X percent o f GDPwithin the usual institutional structures" will not suffice. Rather, institutional and organizationalreforms are neededto make spendingmore effective and to mobilize more nontax revenues for infrastructure costs. Many o fthese issues were discussed for infrastructure generally inchapter 5, while this chapter focuses on how these apply to the infrastructures key to overall economic performance: transport (ports, rail, national highways, and airports) and power. Many different reforminitiatives are now underway inthe infrastructure sectors, with varying degrees of success. Below we discuss some o fthe ongoing efforts and promising approaches ina number o f sectors. Asset maintenance/asset creation. This issue is particularly acute inroad and rail transport. For many reasons it is more attractive to launch a new project than to finishan existing one, or even less, maintain a completedproject. Actual expenditures for maintenance on national highways are less than a quarter the required level (World Bank 2004b). A Special Railway Safety Fundwas set up to fimd a backlog o f arrears in maintenance. Poor financial performance o fthe power utilities has reduced spendingon maintenance. Competition. Competition inthe market is a powerful force, and technological and regulatory changes intelecommunications, for example, revolutionized the sector. Butin other areas o f economic infrastructure, such as highways, ports and airports competition in the market is less viable andcompetition has to befor the market. This can take the form of contracting out the operation o f assets with no change inownership (e.g. for airport modernization) or public-private partnerships that involve the expansion o f assets (e.g. power, highways). Competition for the market is muchmore technically demanding than merelyallowing competition inthe market (even regulated competition) as it requires the assessment o f bids that are complex and, not matter how stringent the process, will always involve some degree o f contestation. Nevertheless, there are successful experiences around the globe-and inIndia--inthe useofPPPs for infrastructure expansion. Financing economic infrastructure. The enormous costs o f infrastructure expansion raise two related, but distinct, issues: who willpay for infrastructure (to cover the various costs, including operation, maintenance, replacement and financing charges) andwho willfinance infrastructure. As stressed inchapter 5, citizens o f India will pay for the costs o f infrastructure either as users or as taxpayers-and this is true no matter who 32Little evidence i s available o n the extent to which improvements have beenrealized at the state level. India has increased road investment from around $lbillion per year inthe 1990s to around $4 billionper year. But this has not beenmatchedby improvements inimplementation performance, with only a fraction of investments having beendelivered on time. The comptroller and auditor general o f India estimated that NHAIhad completed only 29 percent o f its plannedprogrambythe target completion date o f June 2004. Completed works suffered delays o f up to 28 months, and there were cost overruns o f 22 percent on average. 108 finances the infrastructure. Ifthe govemment borrows money to finance a highway then Indiancitizens haveto pay the interest-either as taxpayers or users. Ifa private contractor builds a highway and finances the construction themselves (perhaps on a Build-Operate-Transfer) basis then Indian citizens have to pay the interest-either as taxpayers or as users. The advantage o fprivate sector engagement ininfrastructure finance ina country like India is not that it puts additional resources on the table innet present value (as was made clear bythe proposalto finance against accumulatingCentral Bank reserves)-it i s that PPPs often offer a transparent commitment to finance the additional infiastructure byusers, not taxpayers. As the needs for economic infrastructure are large andthe fiscal space small this is an important advantage. Again, as emphasized inchapter 5, PPPs do tend to work better inpurely economic infrastructurewith fewer numbers o fusers, who therefore recognize and are willing to pay for service improvements. The fiscal deficit as a binding constraint India's fiscal deficit i s a concern for two reasons.33First, there i s always some risk o f a macroeconomic crisis, which can easily undo extended growth episodes (as inIndonesia). Second, there i s a "fiscal space" problem, given the large needs ininfrastructure andthe many other things the government would like to do that require additional resources, the deficit leaves no "space" to fundthem. To undertake new initiatives, -either existing spending has to fall or taxes have to increase, as the deficit cannot go higher. Below we briefly review recent fiscal developments, andwe elaborate on the two issues identified above. Continuingfiscal adjustment Table 6.1 presents fiscal trends for India's general government revenues and expenditures from 1985/86 to 2005/06. It shows that revenues fell by some 2.5 percentage points duringthe NinthPlanperiod (1997-2002), compared with the second halfo fthe 1980s, largely as a result o f structural reforms. Capital spending fell even more sharply: from 6.6 percent o f GDP inthe second halfo fthe 1980sto 3.1 percent inthe NinthPlanperiod. The fiscal improvement followingthe 1991crisis involveda large compression o fcapital spending. Although it hadbeenhopedthat this compressionwould be offset by rising private investment ininfrastructure, this has not really happened, with the notable exception o fte~ecommunication.~~ Table 6. 1: Fiscaltrends in India, 1985/86-2005/06 (percent of GDP) 33Although some have suggested that the large deficit has crowded out private investment, this argument i s currently not persuasive because interest rates have not gone up and the commercialbanks are holding significantly more government paper than they are required to. Nevertheless, inthe absence of fiscal adjustment now, crowding out could become a bindingconstraint to growth inthe future. j4This section isbasedonPinto andZahir (2004). 109 85/86- 8th 9th IOth Plan 89/90 Plan Plan Avg. Avg. Avg. 2002- 2003- 2004- 2005-06 03 04 05" 2/ Revenues 19.4 17.9 16.9 17.8 18.1 19.1 19.6 Current Expenditure 31 22.0 21.5 24.0 24.6 24.1 23.5 22.9 Social Services 5.4 5.O 5.5 5.3 5.1 5.4 5.3 Economic Services 6.5 5.8 5.7 6.3 6.4 5.8 5.6 General Services 9.5 10.3 12.4 12.6 12.1 12.0 11.7 Capital Expenditure 6.6 3.7 3.1 2.9 3.1 3.7 3.7 GrossFiscal Deficit 9.2 7.2 10.1 9.8 9.1 8.1 7.0 Memo Primary deficit 5.4 2.1 3.8 3.2 2.7 1.9 0.8 Revenue Deficit 2.6 3.6 7.0 6.8 6.0 4.4 3.3 Interest 3.8 5.1 6.3 6.5 6.4 6.2 6.1 EducatiodGDP 2.9 2.8 3.0 3.0 2.8 2.9 2.8 Health and Family Welfare/GDP 1.0 0.8 0.9 0.8 0.8 0.8 0.9 (Irrigation+Power+Transport)/GDP 4.0 3.1 3.1 3.6 4.2 3.8 3.9 (Interest+ Admin.+Pensions)/GDP 6.3 8.1 9.2 10.2 9.9 9.8 9.8 (Interest+ Admin.+ 32.6 45.1 54.2 57.6 54.9 51.5 50.0 PensionsVRevenue 'Revised .~~.~. ~~~ Notes: l'Actuals for the Center and revised estimates for the States; Estimates for the Center and budget estimates for the States; 31Current Expenditure refers to Revenue Expenditure. Net o fpower bonds in03104;4' Refersto capital outlay andnetloans andadvances fromthe Center to States. Recoveryo floans i s net o f power bonds in03104. Table 6.1 highlightsa number o fremarkable achievements. First i s the big rise in revenues so far duringthe TenthPlan period, relative to the NinthPlanperiod. The numbersdepict a dramatic improvement in2004/05: revenues haverisento around 19 percent o f GDP, reminiscent ofthe late 1980s; the primary deficit has fallen to 2 percent o f GDP, comparable to the levels seen duringthe fiscal consolidation that occurred duringthe EighthPlanperiod. Nevertheless, although the general fiscal deficit is estimated to have fallen from more than 10percent of GDP in2001/02 to 7-8 percent currently, the deficit inIndiaremains among the largest inthe world. Over the period 2000-04, India's fiscal deficit was exceeded only by that of Turkey. China's deficit averaged 3 percent o f GDP, while Korea ran a surplus. These fiscal deficits are largely financedby borrowings and explain the large increase ingovernment debt, which has onlyjust begunto stabilize. 110 While significant improvementscharacterize fiscal outcomes inrecent years, Indiai s not growing out o f its debt problem. Notwithstanding the low interest rates over the Tenth Planperiod, interest payments have continuedto increase. Moreover, capital expenditures remain low. 35 Government indebtedness has risen even as the spending composition has shiftedaway from capital and development expenditures. Few people think India is in any imminent danger o f a macroeconomic crisis, nor does the World Bank, but the fiscal situation is serious. Figure6.2 presents an average over 2000-04 o fthe ratio o fpublic debt (external plus domestic) to GDP for 12 countries, every one o fwhich, except China, had a macroeconomic crisis. InIndia's 1991 crisis, the main issues were low reserves and liquidity, and, unlike inArgentina, Brazil,Russia, and Turkey, where default risks were judged highby the market, solvency was not an issue. However, India's ratio o fpublic debt to GDP, inexcess of 80 percent,36is more than three times higher than China's. India's interest payments averaged 6 percent o f GDP (and35 percent o frevenue), exceeded only by interest payments incountries with chronic fiscal and debt problems such as Turkey, Argentina, Brazil, andPhilippines (indescending order of interest payments) over 2000-04. Incontrast, the fast-growing economies have had interest payments averaging around 2 percent o f GDP (Korea and Thailand) or less (1 percent inChina). Nevertheless, the growth rate of India far exceeded that of other highly-indebtedcountries, with only China andRussia growing faster inthe 2000-04 period. Figure6.2: Totalpublicdebt andinterestpaymentsinIndia and selectedcountries, 2000-04 Total Public Debt (% of GDP, 2000-04Average) Interest payment (% of revenue, 2000-04Average) 120, ahzEi4 -- - Rn 100 80 60 60 40 40 20 20 0 0 35As discussed inPinto, Zahir andPang (2006), when outcomes are compared with the benchmark period, 1985186-1989190,we see that capital expenditure cuts have served to exactly offset the rise ininterest payments and fall inrevenues, a remarkable persistence of the response that began during the EighthPlan period. 36Brazil and Turkey conventionally report net debt, whereby certain items on the balance sheet o f the central bank are subtracted from the government's debt. The figure reports gross debt for comparability, but also shows net debt. 111 Source: Pinto, Zahir and Pang (2006). IMFWEO, IMF country and Article IV reports, and World Bank staff estimates. For India, the sources are Government Budget Documents, Handbook of Statistics on Indian Economy, RBIBulletins (for Turkey, GNP not GDP). As discussed hrtherinPinto, Zahir and Pang (2006), the apparent disconnect between growth outcomes and public finance indicators for India can be rationalized by appealing to "country-specificity". India i s not likelyto have a major crisis for well-known reasons: limited capital account convertibility, a pliant banking sector, highforeign exchange reserves and a flexible exchange rate.37 India also has relatively low levels o f contingent liabilities inrelation to its financial system. Currency mismatches on the balance sheets o fbanks, enterprises and the government are virtually unknown-unlike inBrazil, Turkey andespecially Argentina. Therefore, vulnerability is low andin particular, the government is not at the mercy o fthe external debt market. With India's increased integration into the world economy, its external position has strengthened. Foreign exchange reserves have grown rapidly, and they currently exceed India's gross external debt exposure. The balance o fpayments also remains comfortable: after three years of current account surpluses, the current account has now begunto show a deficit. Significant deficits on the trade account (some 5 percent o f GDP in2004/05) are balanced by large inflows inthe invisibles account, including remittances from abroad (net invisibles grew at 21 percent in2004/05). Figures for 2003/04 put private transfers at $18.9 billion, making India the largest recipient inthe world. But with large debt stocks and already large fractions of revenue devoted to interest payments, increases ininterest rates (which have been at historic lows) can cause the financing costs to rise very rapidly. The key question going forward i s how to get back on track to meeting the Fiscal Responsibility and Budget Management Act (FRBMA) targets, while creatingthe needed space for developmental spending.Fiscal policy i s set to turn expansionary inthe 2005/06 budget, with a number o fimportant new schemes proposed. Increasingpriority developmentalspending The UnitedProgressive Alliance government elected in2004 is launchingmany potentially desirable initiatives that require additional expenditures for rural infrastructure (Bharat Nirman), employment (NREGA), education (SSA), rural health (NRHM),and urban infrastructure (JNNLTRM). But those new initiatives must be financed with some combination o fhigher taxes (or user charges) or lower expenditures (cutting other existing funding). The existing deficit leaves the government no fiscal space even for highlyproductive new spending that can sustain the current rapid growth. This situation calls for a focus on raisingrevenues inways that do not choke off growth and cutting out unproductive spendingto make space to support better programs. 37Ahluwahlia (2002a). Shah and Patnalk (2005) argue that the rupee is defacto pegged to the dollar. 112 A fundamental challenge facing Indiais that greater capitalinvestmentand developmental spending are needed, but the fiscal position i s tight. Some have argued that the government should alleviate the infrastructure constraint by increasing public investments even ifthis means a further rise inthe fiscal deficit. The argument infavor would say that with strong micro-foundations, growth and revenue collections would accelerate, bolstering government solvency and leadingto a steady fall inthe debt-to- GDPratio. However, a sustained and largejump ingrowth rates wouldberequired for Indiato grow out o f its debt problem and such growth rates are highly unlikely to materialize.38 A preferable alternative wouldbe to addressthe weaknesses inthe public finances and the infrastructure gap simultaneously through an overhaul o f the public finances. While increased tax revenue i s part o fthe answer, as is increasedprivate sector financing of infrastructure, expenditure restructuring will also be particularly important. Higher taxes? Significant progress has beenmade instrengtheningrevenue performance since the 1990s.Gross central tax revenues have increased from 8.2 percent in2001/02 to 10.2 percent in2004/05, inline with the goal o fthe authorities to raise taxes to 10.3 percent by the end o f the TenthPlan. General government revenues have also been growing and are now at around 19percent o f GDP. Inspite o fthese recent improvements, further growth inrevenues is essential to ensure that developmental spendingis sufficient to achieve the desired outcome andthat fiscal indicators are sustainable. Moreover, India's combination o fhighmarginal effective tax rates and numerous tax exemptions i s distorting and constraining investment and growth.39Stamp duties on property transactions are among the highest inthe world (sometimes above 10 percent, compared with 1-2 percent inmany countries), as are combined central and state indirect taxes (often 30 percent, comparedwith halfthat inmany Asian countries). These highrates are combinedwith anarrowbase, reflective inparticular o fthe inabilityof India's states to tax agriculture and services. Thus the great bulk o f taxes are raised from industry,which only constitutes 25 percent ofthe economy. One o fthe key challenges facing India's states i s thus to broadenthe tax base. Another is to simplify India's tax system and reduce corruption. India's indirect tax systemis probably the most complex in the world, and surveys have shown state tax offices to be among the more corrupt government agencies inthe country.40 38An illustrative example inPinto, Zahir andPang(2006) shows that ifcapital expenditure hadbeen maintained at 6.4 percent o f GDP, the benchmark level duringthe EighthPlan Period, growth would have hadto be substantially higher to keep debt-to-GDP on its observed path: an average o f 3.6 percentage points higher over the Eighth Planperiod; 5.4 percentage points over the NinthPlanperiod; and 3.1 percentage points over the Tenth Plan period. Further, this rise inthe growth rate i s a lower-bound as all other macroeconomic variables are kept at their observed levels: at a minimum,we would expect a rise in real interest rates and an increase innoninterest current spending (such as O&M) to complement the increase incapital expenditure. These would then require an even faster counterfactual rate o f growth to keep debt-to-GDP on its observed path. 39IMF(2005b), Carey andRabesona (2002). 40A survey o f industries inKarnataka found that 31percent o frespondents paidbribes to the Commercial Tax Department, a higher number than for any o f the other 13 agencies mentioned(Public Affairs Centre 2002). 113 Revenue performance inrecent years has been excellent, on the back o f a buoyant economy and reforms intax policy and administration. Nevertheless, there is further scope to increase government revenues. Removing exemptions and broadening the service tax base is critical; a tax reform allowing for lower rates andbroader bases would enhance India's low tax productivity compared with that inother countries. Efforts are underway to broadenthe tax base, for instance with the inclusion o fnew taxable services. The introduction of the value added tax on goods inmany states inApril 2005 has been an important achievement. Nonetheless, the preparations for a national goods and services tax should be continued, as this tax would bebroader based and more efficient, thus helpingto achieve FRBMArevenue targets with relatively lower rates. Tax administration reforms and improved interstate and center-state coordination are arguably evenmore important than tax policyreforms. Strengthened enforcement technology and procedures, coupled with better staff incentives, management flexibility, and effective anticomption institutions might well have the greatest potential to lead to a significant and sustainedrevenue increase. Getting rid of unproductive expenditures. There is some scope to increase revenues, but the emphasis shouldbe on fiscal "rightening" not only to ensure fiscal sustainability, but also to make governments more effective agents o f development. Any significant increase inpublic investment and developmental spending needs to be largely compensated by accelerated reduction o fthe revenue deficit andmovement toward generating revenue surpluses. Not only the quantity, but also the quality o f the deficit remain a concern: boththe center andthe states continue to borrow to finance current spending. The share o f interest, administration, andpensions remains high, having actually risen from the already highlevels duringthe NinthPlanperiod. Spendingon education and on health and family welfare has stagnated. Moreover, as summarized in box 6.3, international experience suggests that the more successful fiscal adjustments rest at least partly on the restructuring o f recurrent expenditures. Fiscal consolidations achieved through cutting the wage and subsidybills tend to be more lasting and trigger higher growth rates than adjustments based solely onrevenue increases and cuts inmore productive spending. Moreover, protecting capital expenditures during a fiscal adjustment leads to higher growth, as does an increase inthe share o f current spending on nonwage goods and services. While increasingrevenues i s critical for India's central and state governments, the revenue side cannot bear the entire brunt o f adjustment. Box 6.3: International experience with fiscal adjustment Over the past decade a number o f studies have evaluated the success o f fiscal adjustments. This work i s based largely on experience o f countries inthe Organization for Economic Co-operation and Development (OECD), although one study has found the same results for low-income countries (Gupta and others 2002). Following Alesina and Perotti (1999, fiscal adjustment strategies are sometimes broadly divided into two categories: Type 1, which relies primarily o n cuts inrecurrent spending, and Type 2, which relies primarily on tax increases, with spending cuts mostly limited to public investment. Ina study o f 20 OECD countries for the period 1960-94,60 episodes o f fiscal consolidation were identified. Of these episodes, only 16 were lasting, and among these successful cases 73 percent were based at least inpart on recurrent spending cuts. Although most fiscal adjustment efforts rely o n tax increases to lower the deficit and the debt burden, those successful inaddressing fiscal imbalances rely more heavily o n cuts incurrent expenditures than tax increases. McDermott and Wescott (1996) find similar results for 74 episodes o f fiscal adjustment in20 countries during 1970-95. Whereas a little less than half o f the Type 1 adjustment cases were successful, 114 only one out o f the six Type 2 adjustment cases was successful. They also found that Type 1adjustments are more likely to reduce the public debtratio. Complementary to this, researchby Alesina and Ardagna (1998) finds that the success o f fiscal stabilization is notjust determined by the size o f the adjustment, but also by the composition. Adjustments including cuts ingovernment transfers and wages are more likely to succeed inreducing the primary structuralbalance. Inaddition, such adjustments not only last longer, but can also be expansionary. Adjustments relyingmore heavily on tax increases and cuts inpublic investment tend to be contractionary and unsustainable (von Hagen, Hallett, and Strauch 2001). There are problems with unproductive spending at boththe center and the state levels, as also discussed inWorld Bank (2003b). One key issue is that o f salaries andpensions, which are estimated to account for roughly 25 percent o f general government spending. Withthe states havingresponsibility for the major services o fhealth, education, andlaw andorder, mucho fthis spendingis at the state level.41Itis estimated that at the state level salary spending alone accounts for about a third o f all public spending, so the success o f any expenditure reform strategy will depend heavily on strategies adopted in this area. As discussedinHowes andMurgai (2004) there are significant savings that can feasibly be extracted from the salary bill, through both wage and hiringrestraint, without sacrificing expenditure quality. Ifreal wage increases are avoided, the combined state- central government salary bill could fall bytwo percentage points o f GDP over the coming decade. Such wage restraint does not seem uncalled for, since survey data show that public sector employees inIndia are greatly overpaid relative to their private sector counterparts (see Chapter 3). At the senior level, civil servants are widely regarded to be underpaid, but this i s evidently swamped by overpayment at all other levels. A large public-private wage differential exists inall states, with average premiums in 1999/2000 rangingfrom 89 percent inMaharashtra to 170percent inRaja~than.~~ Pensions are also a rapidly mounting liability, andthe pensionbill cannot be expected to decline as a share o f GDP.Although the data are uncertain, estimates suggest that central and state pension liabilities could amount to 25 percent o f GDP, with a significantly higherrelative figurefor some states. The issue here is containment o fcosts through reforms ofpension characteristics and structure. A mandatory defined contribution scheme for new central civil servants was introduced in2004, and 13 states havejoined so far. The passage o fthe pendingPensionBill, which would among other things set up a new regulator, will also help to improve long-term fiscal sustainability. Nonetheless, these reforms will not have an immediate fiscal impact and thus need to be complementedbyreforms to the existing system, which can have a large fiscal impact now. Attemptedandpossible reforms include: (a) use o f longer averaging periods for the 41 In2000J01, 57 percent o fIndia's total government expenditure was financed bythe states, as was 97 percent o f irrigation maintenance, 90 percent o f public health spending, and 86 percent o fpublic education spending. Infact, India's states are responsible for a higher proportion o f general government spending than inany other developing country, except China. 42 Inpart, the premiums reflect differences inhumancapital as the public sector tends to employ workers with greater education and experience. Adjustingfor these differences incharacteristics brings down premiums, but they are still large, ranging from between 62 percent to 102 percent with the private formal sector and between 164percent and 259 percent over the private-informal sector, depending on the type o f methodology used for adjustment (Glinskaya and Lokshin 2004). The premiums are highest instates like Punjab, Rajasthan, Tamil Nadu, and Uttar Pradeshthat have been less restrained than others intheir pay awards. 115 calculation o f benefits; (b) reduction o fpension abuse; and (c) use o f a lower limit for the maximum amount that can be commuted at retirement. Subsidies, another area o funproductive spendinginIndia, have remained stubbornly high.As discussed inGovernment ofIndia (2004a), there hasbeenvery little change in the share o f subsidies to GDP since 1987/88. At the central level, the largest subsidies are for food, fertilizer, andthe below-cost provision o fpropane and kerosene. At the state level, explicit subsidies are relatively small, and the largest subsidies are implicit, most notably power sector losses. Most o f these subsidies are bothinefficient and regressive. For instance, the propane subsidy benefits largely the urbanmiddle class (World Bank2003b). A largeportion o f the subsidy on food goes to cover food storage costs. About 73 percent o f the rice and 84 percent o f the wheat purchased by the Food Corporation o f India(FCI) i s from Haryana, Punjab, and Andhra Pradesh, even though these states produce only 26 percent o f India's rice and 35 percent o f its wheat. Farmers inthese states enjoy guaranteed sales at prices that are much higher than instateswhere the FCI is not active. Highpurchase prices for food benefit net producers, but harmnet consumers inrural areas, who tend to be poorer. The Indian government (2004a) estimates that nonmerit subsidies accounted for 58 percent o f total subsidies in2003/04. The National Common MinimumProgram pledged "all subsidies will be sharply targeted at the poor andthe truly needy," but no concrete actionhas been taken yet, althoughthere are numerous proposals to address the food and fertilizer subsidies without adversely affecting the rural and urbanpoor. Subsidies have proved difficult to cut, largely for political reasons. To make further progress inpower sector reform, tackling the lack o f commercial discipline inthe sector has to be a top priority. Interest savings will largely follow from reduced borrowing, but states can also take advantage o f a low-interest regime by aggressive debt restructuring. The quality o f spending at the state level can and must also beimproved, as discussed inpart Io fthis report. All inall, the next section shows that a sustained cross-sectoral program o f state and central reforms could enable an elimination o f the revenue deficit by 2008/09, while increasingcapital spendingas a share o f GDPby some three percentage points. For this scenario to be achieved, it is particularly important that institutional and organizational reforms discussed inpart Io f this report are pursued. Prospects and risks India's hture success insustaining growth and expanding opportunities i s closely linked to progress inthe policy areas discussed throughout this report. Below i s a demonstration of the net effect of the proposed policy measures by comparing two ofmany possible macroeconomic scenarios.43 43Actual outcomes could fall within or even outside the range; the scenarios are primarily designed to show the impact o f different broad development strategies. 116 Prospects under scenario 1.The first illustrative scenario shows the results o f a haphazard implementation o f structural reforms and the failure to attain a high-quality fiscal adjustment-all o f which would be associatedwith a disappointing (but inthis scenario not dismal) total factor productivity performance (table 6.2). Even though growth i s sustained at 5 percent under this scenario, India's past successes are not taken forward further. Under this scenario the composition o fpublic spending would not support shared economic growth. Interest payments as a share o f GDP would increase from 6 percent currently to 8 percent by the endo f the period, and as a result the consolidated deficit would reach nearly 10percent o f GDP. Inspite o f these increasing deficits, there would be no room for increased capital and developmental spending. Moreover, not only is the composition o f spending suboptimal inthis case, but so too is the quality o f spending: the sense o f"bang for the buck" would leavemuchto be desired. Itis obvious that such apolicyenvironment would not beconduciveto advancing India's social agenda. Prospects under scenario 2. Under a different illustrative scenario, Indiahas much greater success (table 6.3). It reaches much higher growth rates than under the first scenario, reflecting high-quality fiscal adjustment andkey policy reforms. Real GDP growth is projected to reach up to 9 percent inthe outer years. Further fiscal reforms at the central and state levels would increase revenues and improve the composition o f public spending, with a lower share spent on civil service wages and interest and a higher share spent on operations, management, andkeypublic investments, which would also stimulate private investment.The strong growth performance would arise from actions to ease the keybindingconstraints, including the reduction o f the deficit and the removal o f infiastructuralbottlenecks, especially inIndia's rural areas. A more effective delivery o f key services would accelerate progress ineducation, health, and other measures o f social welfare. Manymore states, regions, andpeople would be able to share inIndia's success under scenario 2 than under scenario 1. Risks. As highlightedthroughout this report, shared growth and improved human development outcomes inIndia cannot be achieved without addressing the special needto expand opportunities to people inlagging states andregions. Failure to address these challenges can pose risks to the sustainability o f India's overall growth process and can thwartits development prospects. It also increases the important andreal risk that the reform effort i s delayed or derailed bypolitical concerns. Otherrisks could threaten development prospects. Interestrates have already started to increase from their historic lows. Evena small increase inthe assumed interest rate inthe scenarios would lead to a significant diversion o fresources away from productive spendingand to debt service purposes. Another keydownward risk facing Indiacurrently i s that o f sustained highoil prices. Nevertheless, neither increasing interest rates nor high oil prices are likely to derail the overall trajectory o f the economy. 117 z?--""" OI G O r - w r-r- 2 3 c Chapter 7. Accelerators to equalize the benefits of growth Although there is a strong link between overall economic growth and the speed o fpoverty reduction, easing growth constraints can have either an equalizing or a disequalizing impact across states, regions, andpeople-even when growth is particularly robust, as it has beenin India. This chapter focuses on reforms intwo constraining areas that are most likely to accelerate efforts to equalize the benefits o f growth: labor laws and the financial sector.44 Labor laws Suggestingthat labor laws havebecome abindingconstraint to growthis controversial. The National Common MinimumProgram "recognizes that some changes inlabor laws may be required," but also statesthat "the UPArejects the idea o f automatic `hire and fire"' and that "such changes must fully protect the interests o fworkers and families." This report agrees: labor regulations to protect the interests o fbothworkers and employers are important, and India should not simply adopt some simple "solution" to labor regulation. But,by the same token, ill- designed regulation can have large economic costs and, evenworse, can be disequalizing among workers. The existing labor regulations, infact, do not protect "workers and families," but only protect those workers with protectedjobs. If,as it increasingly appears, those regulations prevent goodjobs from being created, they are working against the general interests o f workers and their families. Investment climate surveys and recent research suggest that Indianlabor laws are among the most restrictive and complex inthe world.. Firmswith more than 100workers consider labor regulations to be as important a constraint to the investment climate as power shortage problems (World Bank, 2004e). Labor laws have been "criticized as being ad hoc, complicated, mutually inconsistent, ifnot contradictory, lacking inuniformity o f definitions andriddled with clauses that become outdated and anachronistic" (ILO, 2004). There are four interrelated issues. First, clause 5(B) o f the Industrial Disputes Act (IDA) places onerous administrative andjudicial constraints on retrenching workers. Second, some parts o f IDA and court decisions that have stressedthe bargainingpower o fworkers inindustrial disputes also work to raise labor costs and discourage employment. One example o fthis is seen inthe share o f workers involved inmajor work stoppages (strikes and lockouts); at about8-10 percent between 1995 and 2001, India's share i s highby international standards. Third, too many laws inplace govern a host o fpractices inthe formal sector, relatingto service conditions, industrialrelations, wages andbenefits, social security, andinsurance. A typical firm inMaharashtra thus has to deal with 28 different acts pertainingto labor. Fourth, the labor dispute settlement mechanismhas grown more unwieldy andcomplex, with the state government and courts increasingly playing a more prominent and costly role insettling enterprise-level disputes.Numerous, not always consistent, court decisions create uncertainty regarding the interpretation o f these laws. While protecting firms from arbitrary and capricious actions by other firms i s desirable, the current situation has not done enough to allow firms to protect themselves from capricious actions byworkers: labor courts 44Cross-country evidence suggests, for instance, that growth i s more powerhl inreducing poverty in some countries than inothers: greater poverty reduction i s observed where policies are inplace to enhance the capacity of poor people to participate ingrowth. (World Bank, 2005k) 120 have prevented firms from sheddingworkers who have stolen from the company or carried out criminal assaults on company employees. Labor law reform should not be caricatured as taking the side o f workers against firms or vice versa-rather the real interest is providing all workers with the best possible opportunities. The current protections do not benefit"workers" as a whole. Although they mightbenefit the roughly 7 percent o f workers inthe organized sector, this protection comes at a highcosts to other workers. Inan economy with four times as many unemployed as organized private sector workers, the primary concern should be vibrant creation o f employment. Recent research has demonstrated that these restrictive labor laws and practices impose large costs-in lost output and productivity growth, inlostjob growth, indistorting the size o f firms, and incasualization of labor. La11and Mengistae (2005) use data from manufacturing plants locatedin40 cities to show that both inadequate power supply and the rigidity o f labor laws result inlower productivity. Besley and Burgess (2004) show that states that amended the Industrial Disputes Act to increase restrictiveness in shedding workers and to increase worker bargainingpower indisputes had substantially slower manufacturing value added growth and slower expansion o fmanufacturing employment. Ahsan andPages (2005) find that more restrictive labor laws or even more restrictiveness inthe application o fthe existing laws adversely affects manufacturingvalue added, employment, and productivity growth. The formal manufacturing sector, where these laws have the widest application, has had slower output growth and less robust employment growth than the less-regulated service sector. 121 Figure7.1: ManufacturingemploymentinIndia and selected Southeast Asian countries, 1980s India Korea 1989.90 1986 500+ 500+ 200-499 200-499 100-199 100-199 50-99 50-99 10-49 10-49 5-9 5-9 0 10 20 30 40 50 0 10 20 30 40 50 % distribution of employment YOdistribution of employment Hong Kong Malaysia 1982 1981 500+ 200-499 500+ i 200-499 100-199 1 100-199 50-99 - 50-99 10-49 - 10-49 5-9 1 5-9 0 10 20 30 40 50 0 10 20 30 40 50 % distribution of employment % distribution of employment Source: Mazumdar (2005); World Bank, forthcoming(a). The costs that a firmmust bear for joining the formal sector or reaching the size o f enterprise subject to labor law enforcement can be inferred from the size distribution of firms. InIndia, muchmore than inother countries, there are many small firms, a few very large firms, and much less inbetween-a highlybimodal size structure with a missingmiddle(figure 7.1). In1989-90, more than 40 percent of the employment inmanufacturing took place infirms with five to nine workers, compared with only 4 percent insuch firms inKorea. Recent data suggest that dualistic trendshave persisted, as the share o f firms belonging to the formal organized sector (employing more than 10 workers) i s less than 2 percent, and its share of employees has declined to 25 percent. These distortions to avoid labor laws have severe consequences: productivity inthese small firms was only 20 percent or less than that o f large firms (Mazumdar 2005, World Bank forthcoming(a)). Specifically, restrictive labor laws appear to create diseconomies o f scale, particularly for middle-size firms whose scale is not large enough to overcome the higher labor adjustment costs imposedby laws. Interviews suggest it is the medium-and small-scale firms that suffer the most from these diseconomies inlabor adjustments and are the most threatened by permanent loss o fbusiness and clientele inprolonged disputes.A recent McKinsey report on India's textile industryhas pointed out that Indian manufacturers often set up multiple small plants instead of a single bigone to take advantage o f easier labor laws. As a result, Indian 122 clothing plants typically have 10to 20 percent o fthe capacity o f Chinese plants andwork at lower levels of efficiency (Business Standard February 19, 2005). Restrictive labor laws thus end up creating a bias to protect already employed formal workers at the expense of creating more and better jobs for workers outside the formal manufacturing sector or encouraging firms to enter the formal sector. These laws create massive inequality. They divide a tiny enclave o frelatively better-paid salaried formal sector workers, who have goodjob security andbenefits, from the vast majority o finformal or unorganized sector workers, who work for much lower wages and with little or no social protection. Salaried workers make about 28 percent more than casual workers-even whenworkers are similar inage, education, etc. Moreover, as discussed inchapter 9, the dualism created between "good" jobs and casual work or self-employment allows discrimination against women and facilitates caste bias inhiring. Itis worth emphasizing that the dualism inIndianlabor markets isnotbetweenthe rural and urban sectors, but between the "organized" and "unorganized." After adjusting for humancapital characteristics, such as age, gender, and education, there is little evidence o f any premium between real wages for rural and urbancasual labor, which together account for close to 40 percent of the total labor force and 75 percent o f the wage labor market. The convergence betweenrural andurban casual wages, the fast growth o f even casual agricultural wages, which are at the bottom o f wage distribution, and the relative slowdown inrural to urbanmigration (compared to migration within rural markets), all tend to support the notiono f some convergence inthe urban-rural casual labor market. Of course, it will be difficult to addressthe restrictive practices that make the labor laws a bindingconstraint on growth andjob creation, while still maintaining adequate and appropriate protection of worker rights. The lessons learned from previous reforms will be valuable in tackling this hardissue. Those lessons point to the need for solutions specific to India (not imported from abroad or assumed to be a single solution), for institutional continuity with credibility for future progress (which will require consultation among all parties, including trade unions), and for attention to implementation. The financial sector After more than a decade o ffinancial reforms and deregulation, India's financial sector has changed, insome respects, beyond recognition (see Basu 2005). Beginning inthe early 1990s, interest rates were largely liberalized, and the burdens o f priority sector requirements eased. Capital markets were deregulated, restrictions on capital inflows were gradually eased, and private entry was allowed. As a result, large firms can now access equity markets as a source of finance. Although the bank's required holdings o f government debt have been reduced, a high proportion o fbank assets continues to be invested ingovernment debt. As o f March2005, about 38 percent o fbank assets were invested ingovernment debt, far inexcess o f the prescribed statutory minimumratio o f 25 percent. Indianbanks have one o f the highest ratios of government debt to deposits inthe world. Reforms so far have maybe beenmost effective inliberalizing equity markets-mostly benefiting India's largest firms. Banking is still tightly regulated andmostly government owned. 123 Although India's financial sector has been successful inmobilizing resources following reforms of 1 9 9 0 small and medium-size enterprises are not able to receive credit at the levels one would ~ ~ expect. The ratio ofprivate credit (from deposit money banks and other financial institutions) to GDP inIndia remains low at under40 percent, compared with over 100percent for countries such as China, Korea, andMalaysia. Financial sector risks andreform issues have been covered elsewhere, includinginBasu (2005) and also inthe previous Development Policy Review (World Bank 2003). Onone level is appears that the demandfor investment, not the supplyo ffinancing, i s the key issue: interest rates are low, banks are holding substantiallymore government debt that they are requiredto, and, inaggregate terms, Indiahas been (at least untilquite recently) exporting capital as its national savings exceeded investment. However, what the aggregate picture masks is that the financial sector reforms have moved very far inequity markets, but have stalled indirect lendingto the private sector. Withbooming stock markets drivenbyportfolio investment, large firms have been able to finance expansion easily-while massively reducing their leverage. But the sector has been slow to expand its lending, especially to small and medium-size firms, because o f a combination o f factors: the banking sector is closely regulated andpublicly owned; the large deficit prompts fears o fmakingnew loans; and few incentives encourage innovation. Access to finance for small and medium-size enterprises (which are an important engine o f growth and productivity) i s further stymiedbecause they are too small to interest equity markets or other foreign direct investment. Improving the efficiency o f financial intermediation and ensuringbroader access to financial services is a critical accelerator for equalizing growth. Problems inaccessing finance are often cited as a major impediment to the performance o f small and medium-size businesses inIndia. The ratio of private credit (from deposit money banks and other financial institutions) to GDP in Indiaremains low, at under40 percent, compared with over 100percent for countries such as China, Korea, andMalaysia. Only 54 percent o f small businesses inIndia have active bank credit lines, against Brazil's 75 percent. Problems incredit access are rooted in: (a) weaknesses inthe legal framework for loan recovery, bankruptcy, and contract enforcement, together with inefficiencies inthe court system, with the latter largely accounting for interstate variations inthe time and cost o floanrecoveryandbankruptcy; (b) institutional weaknesses, such as the absence of good credit appraisal andrisk management and monitoring tools inbanks, which increase transaction costs indealing with small andmedium-size enterprises; (c) the absence o freliable credit information on such enterprises; and (c) the lack o f sufficient market credibility among small andmedium-size enterprises. It is difficult for lenders to assess risk premiums properly for small and medium-size enterprises, creating differences inthe perceived versus real risk profiles and resulting inuntappedlendingopportunities. A recent World Bank (2004b) study indicates that over 60 percent o f India's ruralpoor do not have a bank account, and 87 percent have no access to credit from a formal source. Informal financiers, who charge exorbitant rates o f interest, remain a strong presence inrural India. Ifthe financial sector i s to contribute more fully to inclusive economic growth, it must reach out to more people. Indeed, there i s a growing consensus, based on theoretical and empirical work, that better access to finance canbe an extraordinarily effective tool for promoting economic growth and poverty reduction. Financehelps the poor catch up with the rest o f the economy as it grows, 124 and it helps to extendthe range o findividuals, households, and firms that can get a foothold in themodem economy. Policiesto expandcompetitioninthe finance markets for rural and small andmedium-size enterprises and to improve the governance and management o ffinancial institutions can go a longway inhelping the underserved access finance on better terms. These policies need to be accompanied bybetter laws andregulations governing financial transactions; ajudiciary that can enforce contracts, however small; the demarcation o fproperty; improvements in landtitling; andbetter credit information. 125 Chapter 8. Reforming agriculture to combatruralpoverty Home to nearly three-quarters o f India's poor, rural areas are the ground on which the major battles against poverty are waged. Evenwithin relatively prosperous states, some cities have thrivedwhile rural areas have stagnated. Gettingthe rural economy movingwill require measures to facilitate rural (nonfarm) entrepreneurship, as discussed inchapter 9.But, with over half o fthe labor force still employed bythe agricultural sector, raising rural incomes will also require sustained efforts to improve agricultural productivity and enhance agricultural growth (figure 8.1).45 Furthermore, because agriculturehas such a profound impact on the rural economy and particularly on employment andwages (as rural and urban markets for casual labor are increasingly linked), sustained agricultural growth can be an important accelerator for equalizing growth. However, agriculturalperformancehas beenpoor inrecent years. Such weak agricultural performance, especially among the poorest states, makes it more difficult to narrow India's widening income disparities andreduce poverty. Figure 8.1. Laborproductivityandshare of the labor force in agricultureinIndian states, 2001 Pun Pun Ker Ker Har Har WB WB Uttn Uttn Ass ASS All India All India AP AP Kar Kar GUJ Guj UP UP Mah Mah TN TN Raj Non Ag Raj Oris Ag Oris MP MP Bih Bih Jha Jha Chh Chh 0.0 50.0 100.0 150.0 0% 25% 50% 75% 100% Labor Productivity, Rs 000/worker % of labor force Source: Census 2001, Ministryo f Agriculture. Theproductivity of agricultural labor is low, particularly inIndia's poorer states. This is linked to the large number o fworkers tied to agriculture in almost all states, the slowing o f agricultural growth, and limited opportunities for rural nonfarm employment. The challenge of improving 45 According to the 2001 census, 58 percent ofthe total labor force, numbering approximately 235 millionpeople, is employedinthe agricultural sector inIndia. These include 127.6 million cultivators and 107.4million agricultural laborers (Government of India, Department of Census and Statistics, "Provisional Population Totals: India, Census o f India 2001"). 126 agricultural productivity over the longer term i s complicated by increasing environmental degradation inmany areas. The signs o f degradation include soil nutrient imbalances due to unbalanced applications o f fertilizer, declining groundwater tables due to overpumping o f groundwater, and waterlogging and salinity o f soil due to overapplicationo f surface irrigation. Table 8.1: Averageyields of major cropsinIndia andother major producing countries, 1998-2000 (inkg/ha) Crop India Brazil China Indonesia Pakistan Thailand Vietnam Rice 1938 2,875 6317 4283 3000 2501 4101 Wheat 2619 1713 3790 2299 639 Sorghum 801 1612 3484 610 1624 Maize 1768 2767 4938 2693 1730 3523 Groundnut 1007 1819 2956 1715 1087 1583 1389 Rapeseed/mustard 833 1551 1420 961 Soybean 1106 2375 1743 1209 1240 1445 1159 Sunflower 522 1508 1599 1222 Sugarcane 71514 68340 68902 64783 47981 54831 50094 Potatoes 17053 16375 14212 14480 15690 12505 10970 Cotton 640 1995 3130 1281 1776 1396 994 Tea 1906 1906 772 1442 296 951 Notes: The period 1998-2000 was selected to eliminate impact o f successive droughts inthe early 2000s. Source; FA0 Statistical database (http://as~s.fao.ora/ua!elcollections'~subset=apriculture), Center for Monitoring the India Economy, 2002. The pace o f agricultural growth decelerated inthe second halfo f the 1990s, from 3.4 percent during 1985/86-1994/95 to 1.8 percent in 1995/96-2002/03, with the largest decline inthe Green Revolution states, even though yields of many crops inIndia are still halfthose o f comparator countries (table 8.1). Weather shocks, particularly extensive droughts inmany states due to poor monsoons46and flooding insome northern states, contributed to the recent slowdown in agricultural growth, but do not hlly explain it. Even under better circumstances, it is unclear whether Indiacould regainits past agricultural growth performance-or attain an even higher growth trajectory-without fundamental changes inagricultural policies and strategy. The government o f India's agricultural strategy, which was largely founded on achieving food self-sufficiency through highprice supports, large input subsidies, and highlyregulatedmarkets, has been successful inmany ways. Now, however, that approach i s largely played out and i s often promoting agricultural practices that are neither economically nor environmentally desirable. The government's price supports for rice and wheat and its input price subsidies (including those on fertilizer, sugar, electric power, and canal irrigation) encourage the perpetuation o funsustainable and environmentally harmful cropping practices and discourage agricultural diversification that could also enhance agricultural growth. The unbalanced composition o f expenditures infavor o f subsidies crowds out productivity-enhancing investmentsinrural infrastructure and services and limits available resources for operations and 46Relative to the long-term average, taken as 100percent, 1999 had 96 percent o f average rainfall, 2000 and 2001 had 92 percent and 2002 had 81 percent (MinA~y f Finance Economic Survey 2002-03). o 127 maintenance to sustain past investments. Notably, these price subsidies are often highly regressive and tend to benefit only some farmers ina few states, with the larger share of the benefits generally captured by large farmers (World Bank 2005k) (figure 8.2). Markets are overregulated (such as through the Essential Commodities Act, Agricultural Produce Marketing Regulations Act, and trade controls), which raises transaction costs, reduces competitiveness, and discourages private investments. Figure 8.2: Food grain price subsidies, by state, 2001/02 25,000 20,000 c ._ 15,000 E0 2E 10,000 5,000 0 5 Rice Subsidy Wheat Subsidy Note: Pnce subsidy= (mnimumsupportprice-total cost of production) x procurementvolume. Source: World Bank 20051. Box 8. 1: Priorities in the Tenth Five-Year Plan for agriculture Utilization o f wastelands andunutilized or underutilized lands. Reclamation and development o fproblemsoils and lands. Rainwater harvesting and conservation for the development o f rainfed areas. Development o f irrigation, especially minor irrigation. Conservation and utilization o fbiological resources. Diversification to high-value crops and activities. Increasing cropping intensity. Timely and adequate availability o f inputs. Strengthening of marketing, processing, and value addition infrastructure. Revamping and modernizing the extension systems and encouraging private sector to take up extension services. Bridging the gap between research and farmers yields. Cost-effectiveness while increasing productivity. Promotion o f farming systems approach. Promotion o f organic farming andutilization o f organic waste. Development of eastern and northeasternregions and hill and coastal areas. Reforms to introduce proactive policies for the farm sector. Source: GoI, Planning Commission, 2005b. The TenthFive-YearPlan(2002-07) identifies a number o fpriority areas for raisingthe agricultural growth rate to 4 percent a year (box 8.1). Inmaximizing the impact of these measures on agricultural growth andpoverty, some strategic prioritization will be critical. Indian 128 agriculture is not homogenous, andthe potential for achieving a higher agricultural growth trajectory inany area is significantly influenced by area-specific natural, infrastructural, and human capital endowments and by access to input and output markets. A workable strategy for acceleration o fproductivity inagriculture has to tailor the right mix o f actions with the specific potentials o fregions; a "one size fits all" agricultural strategy will not be effective. Pursuinga differentiated strategy, across states and within states, will require a shift inthe government's stance from "command and control" to a more market-orientedapproach-in agricultural marketing, landpolicy and administration, agricultural research and extension, andwatershed management. This calls for a changingrole for the government, a shift from direct market participation to catalyst and facilitator through policies and investments. As discussed later, implementing a new agricultural strategy is, however, complicated bypolitical dynamics. The tradeoff is between retainingpresent productivity levels and protecting farm incomes today versus sustaining natural resources andcreating a more equitable system for tomorrow. The strategy will have to be one o fnegotiated rationalization o fthe present system. Agriculturalproductivityandgrowth Agricultural performance and potential varies widely across Indian states, and even within states, presenting challenges and opportunities for achieving more rapid growth (table 8.2). One approach to assess a state's or area's potentialto achieve higher agricultural growth i s to differentiate agricultural systems based on their endowment^.^^ For simplicity, three broad categories are considered here: 0 Highpotentia1irrigated systems with good accessto markets 0 Moderate to highpotential areaswith limited access to markets 0 Marginal drylands. These categories could cover broad regions (including several states) or subregions within a state. Table 8.2: Variationin agriculturalperformance,inputuse, and infrastructureacross Indianstates Regiodstate Agric. labor Avg. annual agric. Average yield, TE 1999/2000 Average Households Share of productivity growth rate (%) (kgha) fertilizer using villages 2000/01 1985/86 1995/96- Rice Wheat Maize Sugar cane Cotton use, 2001/02 improved connected by (RsOOO/ 94/95 2002103 (kg/ha) seeds, 2003 roads, worker) (%I 1997 (%) a States with low agricultural Iabor productivity Jharkhand . 7.9 5.1 729 1,070 804 19,024 22 Bihar 8.4 4.6 1,538 2,071 2,092 45,978 88 34 48 Chhattisgarh 10 -3.1 596 547 899 46 12 100 Madhya Pradesh 13.9 -2.1 853 1,733 1,435 41,512 105 40 29 28 Orissa 14 0.6 -0.7 1,241 1,273 1,336 59,838 287 40 19 49 States with moderate agricultural labor productivity Rajasthan 16.3 4.5 -2.2 1,213 2,507 1,130 45,919 250 41 54 52 Tamil Nadu 16.5 5.5 0.3 3,370 1,604 111,085 302 144 56 51 Maharashtra 16.6 5.8 1.2 1,655 1,185 1,538 87,364 132 76 70 71 Uttar Pradesh 17.9 1.8 2,089 2,625 1,343 59,010 67 130 48 50 47These typologies andthe discussionof developmentpathwaysthat follows draw extensivelyfromWorld Bank 2005111. 129 Gujarat 18.9 4.0 -3.9 1,553 2,307 1,547 70,274 329 90 62 94 Kamataka 19.8 4.0 3.2 2,499 706 2,865 99,083 226 103 68 100 Andhra Pradesh 20.3 4.3 3.0 2,611 572 3,154 75,608 237 150 33 86 Assam 22.9 0.4 1,384 1,196 730 40,266 37 64 75 Uttaranchal 23.3 1.2 1,976 1,873 1,386 61,559 153 98 States with high agricultural labor productivity West Bengal 34.4 5.4 3.0 2,243 2,221 2,709 74,028 123 63 49 Haryana 39.5 4.6 1.6 2,472 3,912 2,086 54,657 320 163 47 99 Kerala 71.7 4.9 -3.2 2,079 80,445 278 59 16 99 Punjab 78.3 4.0 2.2 3,323 4,294 2,319 59,649 247 171 44 99 Average for 22.2 3.4 1.8 1,937 2,618 1,766 70,941 219 91 57 India Notes: a: Values for Bihar, Uttar Pradesh, andMadhya Pradeshare for unsplit states; b: While more recent crop yield data are available, triennium ending (TE) 199912000figures are used to eliminate the impact of widespread droughts during the early 2000. Source: CMIE 2004. Punjab and Haryana qualify as having high-potential irrigated systems (figure 8.3). Irrigated areas often tend to be relatively better provisioned interms o f access to roads and transport systems, markets, electricity, communications, and other public services. Other states or regions, such as Bihar, EasternUttar Pradesh, Madhya Pradesh, Andhra Pradesh, and Northeast India, have moderate to highagricultural production potential because o f their good biophysical endowments (e.g., good climate and soils andaccess to surface or ground water), but maximizing their potential ishampered byweak linkagesto input andoutput markets. Limited communications and road connectivity increase transaction costs, which leads to lower farmgate prices and limits diversificationto higher value crops, tying farmers to the production of nonperishable products or food for home consumption. irrigation good rainfall or irrigation "Green Revolution areas" Bihar, EasternUttar Pradesh, Western Rajasthan, Northern (Punjab, Haryana, Western Madhya Pradesh, Andhra Pradesh, Karnataka Uttar Pradesh, Tamil Nadu) Orissa, Northeast states Marginal arid areas are generally sparsely populated and often remote, with very limited agroecological potential andpoor access to markets. Agricultural activities there are subject to highproductionrisks becauseo fdroughts. Livestock production, dominated byhardier animals such as goats and sheep, and off-farm employment are often important sources of livelihood. Still, people living inthese marginal areas are often very poor. Two areas with these characteristics are Western Rajasthan andNorthern Karnataka. Regions or districts within a state can also be classified according to these three broad agricultural production systems (table 8.3 applies this fi-amework at the district level in 130 Rajasthan). The characteristics of the production systems will favor particular crops or livestock. Productionof more water-sensitive crops, such as wheat, rice, and maize, is concentrated inthe areas with moderate to highagricultural potential and the high-potential irrigated areas, while less water-intensive crops, such asjowar and groundnuts, are concentrated inthe marginal drylandareas. Hardier livestock is concentrated inthe marginal, drought-prone areasinthe state. Table 8.3: Agriculturalsystems inRajasthan,by soil conditions Production Soil quality Availability of water Market access: Average system/district Avg. % ofarea avg. % of agriculture rainfall irrigated villages National (mm) connected by Domestic road, 2001 Product per cultivated area (Rs/ ha) Marginal drylands Jodhpur, Ajmer, Churu, Desert soils and sand Bikaner, Jaisalmer, dunes, seirozones 100-500 13.2 25.8 7,008 Banner, Tonk Moderate to high agricultural potential Sikar, Jhunjhunu, Desert soils, seirozones, Jaipur, Chittor, yellowish brown, Udaipur, Ganganagar mediumblack, red loam, 100-700 41.2 37.1 15,271 s i l t y clay High-potential irrigated Jalore, Sirohi, Alwar, Sandy and hilly soils, Bharatpur, Kota, alluvial black and Banswara, Bundi, yellowish soils, 300-750 67.4 40.8 20,011 Sawaimadhopur, yellowish brown, Jhalwar medium black, red loam Notes: Marginal drylands are those with marginal soils and limitedrainfall and irrigation. Moderate- to high- potential areas have moderate to good soils and moderate rainfall or moderate access to irrigation. High-potential irrigated areas have good soil and highrainfall or good irrigation (greater than 60 percent). Source: Government o f Rajasthan. Developmentpathwaysfor growth Four possible development pathways could be exploredto maximize agricultural growth potential and achieve the government of India's agricultural growth goals. These are intensification, diversification, nonfarm linkages, and exit. I ntensification Intensification involves increasing the output of existing activities. This could be achieved through a number o fmeans, including (a) increasing cropping intensityby expanding access to irrigation, (b) increasing yields andor reducing losses through the adoption of improved farm practices, seedvarieties, and livestock breeds and through the use of other inputs, and (c) relaxing some bindingconstraints, such as labor shortages through the use o f mechanization, water shortages through the use of drip irrigation, or access to credit through the use of warehouse receipts. 131 The development onirrigation systemsimproves agricultural potential, especially inhighly fertile areas, ina number of ways. It improves agricultural productivity and performance directly by increasingyields andby enablingmultiplecropping, but also indirectly by increasing the returns to other factors o fproduction, such as improved seeds, fertilizer and other agrochemicals, andlabor. Itreduces rainfall-related output volatility andrisks, which also makes agricultural diversification to higher-value crops more economically viable. Irrigation expansion, however, is becoming difficult and costly inmany states, becausethe best areas have beendeveloped and expansion i s beingpushedto more difficult areas. Moreover, increasing water scarcity and competition for water from other sectors (industryand drinkingwater) firther heightens the need for better allocation and management o fwater resources inmany states. Diversification Diversification involves the shift inproductionto higher-value crops or products (fruits, vegetables, higher-value cereals, medicinal plants, and livestock) to take advantage o f new, more profitable market opportunities. The potential gains from diversification come from improving crop rotations, spreading labor demand, raisingincomes, and reducing risks. Agricultural diversification i s often viewed as a possible panacea for raising agricultural performance everywhere inIndia. Agricultural diversification for sustained increases inhouseholdincome, however, is most successful ifit i s market driven. Nonfarm linkages Connections between farms and related economic linkages can be strengthened through activities that foster greater value addition, such as trading and agroprocessing, as well as activities that provide inputsto the farm sector. Exit Exit i s the shift away from farming to nonagricultural occupations. Exit can implymigration of farmworkers to anew location to take advantage of higher-paying employment opportunities in the industryor service sectors, usually ina town or city. Or it can also simply meanthe reallocation o f labor to the rural nonfarm economy withinthe same local area. Matching agricultural strategieswith production potential Just as the development pathways for achieving more rapid agricultural growth will vary across the different productions systems, so too must the agricultural strategies vary from region to region (table 8.4). Inhigh-potential irrigated areas, increasing land and water scarcity heightens the importance o f diversification to higher-value products andvalue addition as a means to maximize the returns from existing resource endowments. Changing consumer demand, as incomes grow, cities expand, andmarkets and trade are increasingly liberalized, will provide hugeopportunities for producers to diversifyinto higher-value products, including fmits, vegetables, livestock, aquaculture, and associated livestock feeds andforages. Demandfor 132 processed products also increases with risingincomes, which presents new opportunities for movingup the value chain into agroprocessing andrelated services. Giventhe pressures on the natural resource base inhigh-potential irrigated areas, another development path is also the shift inemployment from agriculture to the industryand service sectorsthrough measures to strengthen rural nonfarm entrepreneurship and develop humancapital. Table 8.4: Tailoring developmentpathwaysto agricultural production systems High-potential irrigated areas Moderate- to high-potential areas Marginal drylands Diversification ofproductionto Increased intensification through Short- and medium-term: higher value products to cope increased adoption of high-yield Sustainingproduction. with increasing landand water varieties, irrigation, and modem Long term: Exit through off-farm scarcity. inputs. employment (agroprocessing, Value addition through Diversification by W h e r services, industry). agroprocessing to cater to improving linkages to markets. increasedperishability o f most highvalue products. Exit to serviceslindustry sectors. Inmoderate- to high-potential areas, goodbiophysical endowments offer the opportunity to maximize their untappedpotential through further intensification o fproduction and agricultural diversification. But this necessarily involves improvinglinkages to markets. In contrast, the strategy for improving livelihoods inmarginal drylands, because o fthe limited production potential and significant associated production risks inthose areas, needs to be differentiated betweenthe short to mediumterm and the long term. Inthe short to medium term, the focus will needto be on maximizing or at a minimumsustaining agricultural (crop and livestock) and nonfarm incomes from existingnatural endowments. Over the longer term, an important pathway to reducingpoverty i s to exit agriculture by encouraging the shift to other higher-paying employment locally or inother areas, as well as more profitable nonfarm uses for the land, such as manufacturing and services to conserve naturalresources andpromote off-farm employment. Because exit i s a longer-term objective, especially inpoorer areas, where opportunities to participate innonagricultural labor markets are fairly limited, an intermediate step is to reduce poverty by developing those agricultural systems that can be viable. Improvements at the farm level should helprural households enhance food security, conserve natural resources, and ifpossible enter markets for selected products inwhich they have a comparative advantage (e.g., arid zone h i t s andnuts, spices, honey, herbs, agroforestry, and small livestock). The development process inmarginal areas needs to be managed carefully to ensure a smooth transition for rural households over time. They will need a combination o frisk management at the farm level, safety nets at the household level, and other forms o f transitional support. These differentiated strategies inturnhave important implications on the policies and investments for promoting agricultural and rural growth. These are elaborated inthe following sections. Strategiesfor high-potential irrigated areas 133 Fostering agricultural diversification and value addition inhigh-potential areas requires reorienting policies to ensure that incentives favor diversification and sustainable agricultural practices and fostering new institutions to meet the modem marketing needs o f higher-value products (table 8.5). Key public sector institutional changes, reforms, and investmentsrequired to provide such an environment are: Continuing liberalization of agricultural market andtrade policy (e.g., removal of movement and storage restrictions, except duringemergencies; amendment o fthe Agricultural Produce MarketAct to allow direct salesbyfarmers; andremoval o fFDIrestrictions onretailing, etc.). Expandingagricultural risk management mechanisms, such as negotiable warehouse receipts, forward and futures contracts, and crop insurance, and developing financial systems for savings, capital redistribution, and risk management. Encouraging investment ininfrastructure for irrigation (development and rehabilitation), markets (e.g., market yards, cold chains, port infrastructure), and agricultural services (e.g., agricultural research, extension, market intelligence and information systems). Supporting capacity buildingfor businesses to conform to market grades and standards and for public agencies to regulate and certify food quality and safety. Strengthening agricultural research and technology transfer systems. With landand water becoming scarce inmany states, diversification and productivity growth will become highly knowledge-intensiveprocesses. Indianfarmers will requiremore market-oriented agricultural research and extension systems to supplynewtechnologies and information. The private sector canplay animportant role inthese innovation systems, but to do this, appropriate incentivesmust be available, such as a business- friendlyinvestmentclimate. To improve the effectiveness o fthe public research system the roles andresponsibilitieso fthe IndianCouncil o fAgricultural Research institutes and state agricultural universities needto be redefined to minimize overlaps and duplication and to focus on the key constraints facing farmers. For strengtheningstrategic focus andbuildingthe critical mass o f scientific human capital inpriority areas, there is needfor consolidation of research programs, amalgamation o f some institutes and departments, need-based redeployment o fhuman resources, and establishment o f a long-term human resources development plan. A rigorous priority-setting exercise is necessary to ensure that resources are allocated to drive the future agricultural growth and diversification agenda. The growing consensusto reform landpolicy, particularly land tenancy policy, and the land administration system at the state level can help to drive action and should be encouraged. While landdistribution hasbecome less skewed, landpolicy andregulations to increase the security of tenure (including restrictions or ban on landrentals or conversionto other uses) have had the unintendedeffect o freducing access by the landless and discouraging rural investments.In considering land tenancy reform, the experience o f states that do not have tenancy restrictions canprovide usefullessons inthis regard. State government initiatives to improve land administration through computerization o fland records have reduced transaction costs and increasedtransparency. For example, Lobo and Balakrishnan (2002) found that the computerization of land records inKarnataka reducedprocessing delays by 1.3 million person days per year and eliminated petty corruption estimated at Rs 800 millionper year. But it has 134 also brought to light operational and institutionalweaknesses, such as institutional fragmentation4* and problems o f inconclusive proofo f ownership.49Over the longer term, a more holistic approach to land administrationpolicies, regulations, and institutions i s necessary to ensure tenure security, reduce costs, and ensure fairness and sustainability o fthe system. Priority actions include the expansion o fthe computerization o f land records inall states, the integration o fthe systems for records o f rights and registration, andprogress toward a single integrated landadministration agency. Table 8.5: Policy and investment implications of a differentiated agricultural strategy High Potential Irrigated Areas Moderate to High Potential Areas Marginal Drylands Policy Policy Policy Market (input/output) 0 Market (inputloutput) 0 Natural resource deregulation deregulation management .. Land policy (tenancy) and Land policy (tenancy) and Drought risk management admhstration administration Water resource allocation Price risk management and management Price risk management Public Investments Public Investments Public Investments Research and extension to Irrigation development and Watershed management . support diversification, SPS management Minor irrigation requirements Research and extension to R&D on drought tolerant Rehabilitation and support intensification/ crops, small ruminants, maintenance of irrigation diversification NRM - systems Rural infrastructure (roads, Disaster management-safety electricity, markets) nets during droughts Rural infrastructure to attract nonag investments e------- Improved human capital and governance With increasing competition for water expected, especially inhigh-potential imgated areas, better management o fwater resources will be essential. Many states lack both the incentives and the policy, regulatory, and institutional framework for efficient, sustainable, and equitable allocation anduse o f water, or for internalizing the environmental costs o f inefficient use. Public expenditures on irrigation have tendedto put lower priority on operations andmaintenance, leading to the rapid deterioration o f existing infrastructure. Improvingthe planning, management, and allocation o fwater resources requires placing greater priority on (a) the establishment of institutions for state-level water resource management and allocation; (b) adopting new instruments to govern incentives for water use (e.g., water entitlements, management contracts, water pricing); (c) reforming and modernizing Imgation and Drainage 48The Revenue Department i s responsible for handlingthe record o f rights, while the Department of Stamp Duties handle the registration of deeds. There is generally no link between the databases o f these departments. 49Existingland documentation does not provide adequate proofof land ownership. Landregistration does not require proofof ownership, and the government is not responsible for any errors inregistration. 135 Departments to integrate the participation o f farmers and other agencies inirrigation management; (d) improving the efficiency o f irrigation through appropriate water pricing and cost recovery andthrough conversion to higher-value crops combined with the provision o f agricultural technical assistance; (e) rationalizing public expenditures, with priority to completing schemes with the highest returns; and (f) allocating sufficient resources for operations and maintenance for the sustainability o f investments. Off-farm employment i s vital to the strategy for high-potential irrigated areas. Upgrading rural infrastructure and facilitating access to finance, as discussed inchapter 7,will help to ease two key constraints to growth inthe nonfarm sector. Growth inthe farm sector can also spur growth inthe nonfarm sector byincreasingthe demand for laborto trade, process, andpackage food and other products. This requires well-functioning product markets,which can be developedby carehlly reorienting the government's role away fi-om direct intervention and overregulation toward creation o f an enabling environment for greater private sector participation. Strategiesfor moderate- to high-potential areas Creating the enabling environment for growth inmoderate- to high-potential areas will require manyof the same elements outlinedinthe strategy for high-potential areas. The main difference i s that in contrast to high-potential areas, moderate-potentialregions still have scope for improving productivity through agricultural intensification. For bothintensification and diversification inthese areas, substantial public andprivate investments to increase labor and landproductivity are needed. A keypriority i s the expansion of irrigation where economically feasible, as it has the potential to generate higher returns and agricultural employment (table 8.6). As discussed earlier, however, anumber o freforms are neededto achieve the fullpotential o f irrigation. Table 8. 6: Impactof irrigationon agriculturalproductionand employment Variable All farms Irrigated Rainfed b/ Total crop value (Rs) 44327 50224 29655 Crop area (acres) 6 7 6 Total labor (mandays) 173 186 140 Family labor (mandays) 60 61 58 Hired labor (man days) 112 125 81 Total labor expenditures (Rs.) 7595 8509 5322 Value o f agricultural assets (Rs) 43620 51267 24596 Observations 4161 2968 1193 Notes: These states include Assam Bihar, Madhya Pradesh, and Uttar Pradesh (unsplii states), Gujarat, Haryana, Himachal Pradesh, Karnataka, Himachal Pradesh, Kerala, Maharashtra, Orissa, Punjab, Rajasthan, Tamil Nadu, and West Bengal. a: Irrigated: Share of irrigated area > 0.25; b: Rainfed : Share o f irrigated area c0.25. Source: Jin and others forthcoming 2005. Another critical priority will be to raise farm profitability by improving access to markets for inputs and outputs. Market development requires support for key market institutions. Support 136 will beneeded for institutional innovations that facilitate coordination along the supply chain, including contract farming, out-grower schemes (inwhich an agribusiness complements its supplies through arrangements with surroundingsmallholders), and farmers' and traders' association. Capacity buildingi s required for regulation and monitoring, which are crucial institutional elements for agricultural markets to function well. It may be necessary to develop official systems o f grades and standards andwell-functioning market information systems. Access to markets must be improved, but so must access to new technologies and advisory services to improve profitability and expand smallholders' awareness o f market opportunities. To meet these needs, agricultural research and extension priorities for these areas will increasingly have to focus on issues relating to marketing policy, post-harvest technologies andpractices, livestock and high-value commodities with strong market demand, and cost-saving technologies. As the productivity andprofitability o f food crop production increases, priority should also be givento diversification. By strengthening links to market outlets (for example, through investments inroads, markets, electricity, and information systems), farmers canproduce more crop and livestock products for the market. Diversificationto products that require more inputs, processing, andhandling will also offer additional nonfarm employment opportunities. To help farmers cope with changingmarkets and production systems, investments mustbe made in financial systems for managingrisk (e.g., futures markets, crop insurance schemes, warehouse receipt systems, commodity exchanges, and social safety nets). Givinggreater priority to agricultural diversification, however, does not implyphasing out support for further productivity improvements inthe traditional crops. Although Indiahas reached food self sufficiency, and growth indemand for traditional foodgrain crops is slowing down, the bulk o f demandwill probably still have to bemet through domestic production. It will not be feasible for India to import largevolumes o f foodgrains from the world market without substantially increasingworld prices. Thus, the needwill remain to support public investments in agricultural research andtechnology transfer to increase the productivity and profitability o f staple crop production. The returns to greater input use are declining inmany o f the Green Revolution (high-potential) states, and further intensification inthese areas will exact a heavy toll on the environment. There i s growing consensus inIndia over the need to foster the shift o f the breadbasket to the moderate- to high-potential eastern and northeastern regions o fthe country, which can support rice-wheat systems more sustainably over the long term. However, achieving such a transition will require a radical shift inthe government's price policies, as well as significant investmentsto strengthenagricultural support services and infrastructure. Strategiesfor marginal areas While the long-term strategy inmarginal areas with limited agricultural potential is exit to off- farm opportunities, through promotion o fnonfarm entrepreneurshipand development o f human capital, as a transitional strategy, policy should focus on those aspects o f agriculture that are viable inthe harsher agroecological conditions o f marginal dryland areas (see table 8.5). This necessitates increasedpriority to the research, development, and diffusion o f drought-tolerant crop andlivestock varieties, suitable livestock feed and fodder, appropriate soil andwater 137 conservation measures, and technical advice on alternative cropping systems that can mitigate agriculture production risks associated with low rainfall. Because the returns to research and extension associated with drylandcrops and environmental resource conservation are limited,the government will have to play a major role indeliveringthese services. Box 8.2: Issues for effective watershed management at the national, state, and local lev1 Is I Recent experience with watershed programs inIndia points to a number o f institutional, design, and implementation issues that are hindering the programs' full effectiveness. Institutional constraints include (a) a multiplicity ofprograms financed and delivered through different central agencies; (b) weaknesses in state government capacity for watershed management planning, monitoring, and evaluation; (c) weak coordination among different government agencies charged with planning and delivery; (d) lack of coordination between decentralization policy and local authority capacity for delivering watershed programs; (e) weak community institutions inmany projects; and (f)uncertainties relating to the legal framework for common lands. Design and implementation concerns include (a) unequal distribution o f benefits across households; (b) lack o f sustainability o f some assets created; (c) insufficient attention to hydrology inprojects; (d) inadequate consideration to externalities; and (e) weak systems for monitoring and evaluation and for management information. At the national level, some options for improving the effectiveness o f watershed programs include (a) consolidating different watershed programs at the national level; (b) strengthening national knowledge- sharing and learningnetworks to help buildgovernment capacities; (c) developing state and community capacities and participatorymicroplanningbefore undertaking major field investments. K e y state and local institutional reforms include: (a) increasing local exposure visits to help buildcommunity capacities; (b) establishing district-level technical coordination units within the lead implementation agency for line agencies to work with; (c) increasing support for nongovernmental organizations incommunity institution building,monitoring, etc.; and (d) seeking agreement betweenthe local agency and community over rights to use and share benefits from common lands. The strategy for marginal dryland areas will require making the best use o f scarce water resources, reducing soil erosion andwater runoff, and conserving natural pastures and vegetation. The success o fwatershed management programs inmany states shows that well- designedand targeted initiatives can reverse resource degradationand/or improve productivity. These successes generally combine elements o f water conservation or small-scale irrigation, management o fnatural resources heldincommon, and grants to communities for small-scale infrastructure and income-generating activities, both on and o f f the farm. But there are areas for improvement, requiringconcerted efforts at the national, state, and local levels (box 8.2). Livestock are often the key to improving food security and reducingpoverty inmarginal drylands. Thus, the availability o f feed i s a critical variable, and efforts are often neededto reduce the effects of annual and seasonal variations infeed availability through strategic fodder production (e.g., inhigher potential sites indryland areas), range management, and livestock marketing. For livestock producers to gain access to regional markets, they require access to veterinary services, animal health products, and more developed market channels, including provisions for compliance with sanitary standards (to minimize the risks o f spreading contagious diseases) and food safety standards (to ensure safe processing). Safety nets, discussed inchapter 10, will be appropriate incertain circumstances to ensure that people attain a minimum standard o f living (especially when there i s a drought). Investments in producers' capacity to prepare for andrespond to drought should provide for preparedness 138 training, early warning systems, better drought management capabilities, and assistance to recover from drought. The politicaldynamicsof ruralreform Implementingthe government o fIndia's agricultural development strategy is complicated by the political dynamics o f reform. Increasing public expenditures inproductivity-enhancing investments, such as agricultural research and extension, irrigation, rural infrastructure, and market support services, will be critical to reinvigorating agricultural performance and meeting the goal o f4 percent average annual growth per year. Providing the budgetaryresources for these investments, however, is made difficult bythe current large fiscal burdeno f agricultural subsidies, These include the food and fertilizer subsidies o fthe national government andthe power (for groundwater pumping) andimgation subsidies o f state governments. Recent studies have shown that the present agriculture subsidy regimei s inequitable and inefficient, directly benefiting mainly owners o fbigger farms inbetter off districts. But rationalizing subsidies i s a politically complex and sensitive subject, and a strategy o fnegotiated rationalization o f the present regime will be required. Thepolitical rationalefor agricultural subsidies Thegovernment is aware of theproblem of misdirected and unsustainable subsidies. Inits approach paper to the Mid-Term Appraisal o f the TenthPlan (2002-07), the Planning Commission stated that the present price support andprocurement systems combined with input subsidies on fertilizer, electric power, and canal water "have ledto a sharp increase insubsidy based support while public investment inagriculture has suffered. The outcome i s distributionally inequitable since the subsidies typically go to the richer farmers inareas o f assured irrigation, while the lack o fpublic investment hurts poorer farmers and those inarid region^."^' The subsequentMid-Term Appraisal has identified the adverse effects o f several current pricing and subsidy mechanisms and has recommended "focus on reducing those subsidies that leadto distortions and have deleterious effects on natural resources and cropping patterns." Inhislast budget speech, the finance minister spoke ofthe needto "now take upthe task of restructuring the subsidy regime."51 And the primeministerhas stated that "the hture direction o fpolicy clearly has to recognize that we are no longer inan era o f chronic shortage, and that our emphasis now has to be on providing rapid growth inagriculture-based livelihoods. For this we would need to correct the various distortions that have crept into our olicy framework both in terms o f geographical focus as well as incentives to specific crop^."^ Despite this recognition o f P the problem, however, there is no visible strategy o f how the government proposes to address the j0 Explicit central subsidies amounted to 4.2 percent o f GDP in2003-04; implicit subsidies and state subsidies are similarly significant. The main agriculture-related central subsidies are on food and fertilizers, while states subsidize irrigation and power. j` Chidambaram, P., Budget speech, February 28, 2005, available at http:llindiabudget.nic.in/ub2005- 06ibslspeecha.htm. 52 Singh, Manmohan, interview, IFPRI Forum, newsletter of International FoodPolicy Research Institute, Washington, March2005. 139 issue inpractice. There is a proposal to make grain procurement more cost effective through decentralized mechanisms, especially innontraditional states, and a working group inthe department o f fertilizers is examining next steps ina new fertilizer pricing scheme proposed to come into effect from April 1,2006. But a politically feasible strategy to restructurethe agriculture subsidy regime has yet to beunveiledor discussed. The complexity and sensitivity of reforms Subsidies, especially those directed at agriculture, are apolitically complex and sensitive subject. Since 2004, several states, such as Punjab, Tamil Nadu, andAndhra Pradesh, that had made a beginning inreforming their power sectors reversed track to once again provide free power to their farmers. Most states have yet to begin seriously addressing the issue o f sustainable water management, andthe national government too i s cautious about touching the Minimum Support Price (MSP) regime.53The logic that the present regime benefits better-off farmers disproportionatelyhas failed to carry muchpolitical weight with either the rural elite or the rural poor. Richer farmers believe that they are better off precisely because o f these subsidies andthat they still needthe subsidies for sustenance, and to avoid falling into poverty. Besides, being better off does not necessarily free one from vulnerability, giventhat even so-called big landlords inGreen Revolution areas often operate smaller than optimal farms or own scattered parcels o f land.For the landless poor and marginal farmers, there i s a big stake inthe present system, despite its obvious inequity, as their employment opportunities stem from the subsidized viability o fbigger farmers. And small and medium farmers (especially those inaridareas) depend crucially on subsidized inputs, such as seeds, fertilizer, and groundwater (often purchased cheaply from a richer neighbor using free electricity to operate a pump to mine it). Attacking subsidies, therefore, translates politically into a direct assault on all farmers' interests. Farmers understand the benefit o f movingto a more sustainable andnondiscriminatory system based on agricultural technology and infrastructure investments instead o f subsidies, but neither rich nor poor farmer i s confident o fwithstanding the transition. These insecurities stem from the small-farm structure o f the rural economy andthe vast numbers dependent on it. Among policymakers, there are many who argue that ina sector with low returns, highrisk, uneconomic average farm size, and excess labor, the government will have to subsidize production inone way or another. In a set-up of numerous small holdings, goes this argument, subsidized incentives are a more workable way o f improvingproductivity than capital-intensive technology, and removal o f subsidies now will cause both decline inproduction andpauperization o fmany.54 Several other factors contribute to political resistance to reform in agriculture. The first is the broadperceptioninlarger Indian society that the present regime has led to increased farm j3Inhis 2005 budget speech, Finance Minister Chidambaram made clear that food grainprocurement wouldbe made cost effective "without impairing the present MSP-based procurement." This is not to say that sometimes the government o f India does not "touch" the MSP ina backdoor fashion. The minimal adjustments to the MSP since 2002 have resultedinthe MSP declining inreal terms-which has incentive and fiscal effects. s4For example, the Expenditure Reforms Commission (2000) estimated that ifthe farmgate price ofurea were raised to its import-parity price without a corresponding increase inthe procurement price o f food grain, production would fall by 13.5 milliontons. 140 productivity, helped the country achieve self-sufficiency incereals, expanded economic access to food, and i s thus an investment innational food security. The second, which also finds support in broader society, is the view that it is unfair to dismantle the protections available to vulnerable Indianfarmers when advanced economies such as the United States, Europe, and Japan provide hugesubsidies to their farmers. The third, a sentiment emanating from the farm community, is that it i s not right to target rural support mechanisms when urban populations continue to receive heavily subsidized services, such as electricity, water supply, and urbantransportation, when the industrial sector too receives various benefits, and when government employees are cushioned against all risk. Linkedto this is a perceptioninrural areas that the urban middle class has benefited from liberalization and globalization in a manner that the rural community has not. These sentiments together translate into political pressure from all rural classes to notjust retain but even expand the present regime. For political decision makers, the tradeoff i s betweenretaining present productivity levels and protecting farm incomes today versus sustaining natural resources and creating a more equitable system for tomorrow. The political compulsion universally i s to lean infavor o f today. The fiscally logical argument that saving wasteful subsidies will make greater resources available for investment i s not one that works politically, for this assurance i s not demonstrable inthe short term and its potential beneficiaries are too diffused a political constituency. Incontrast, those who standto lose are already politically influential, more visible to each other, therefore better able to organize and so politically more potent. More directly, the highlevel o frisk and vulnerability inIndianagriculture produces a political response, which leads to the breakdown o f commercial discipline inrural subsidy delivery mechanisms. Announcing subsidies, reducingtariffs, refusingto collect dues from farmers- these are important signaling devices that enable politicians to demonstrate their responsiveness to the difficulties of the rural community. For all these reasons, there appears to be, perhaps more than the apparent unwillingness to pay, a political reluctance to charge. Inthe wake o f these political constraints, it i s clearly overoptimistic to expect subsidies to be eliminated or even substantially restructured. Nonetheless, it i s entirely possible, as India's economy grows, agriculture's share init falls, and rural investments enable more farmers to access electricity and irrigation connections, that governments might be tempted (or pressured) to find the fiscal space to continue indulgingthe present regime. Negotiated solutions Thestrategy will have to be one of negotiated rationalization of thepresent system. Drastic reforms are neither politically feasible nor even desirable given the vulnerabilities o f India's farmers, both big and small. The challenge before India's political andbureaucratic leadership, then, isto devise an acceptable strategy that leads to more rational management ofthe present system while correcting its imbalances and inequities over time. A concertedfocus on addressing the fundamental issues o f landproductivity, water management andrural off-farm livelihoods with specific strategies for different regions is more likely to yield long-term solutions than the present tactic o f fiscal arithmetic drivingminor advances on the subsidy front only to be followed by dramatic reversals underpolitical compulsion. 141 There are forms o fnegotiated rationalization already happeninginsurface irrigation. A strategy of linking increased cost recovery o firrigation operations and maintenance with the (a) transfer of irrigation systems management to water user's association (participatory irrigation management) and (b) improved quality o f service, have helpedto increase buy-ino fthe pricing reform from both politicians and farmers. O f course the process has not always been smooth (e.g., Andhra Pradesh), but the move i s inthe right direction. To encourage states even more, the Planning Commission has anAccelerated IrrigationBenefits Programwherein, ifstates increase the recovery o f operations and maintenance costs, they are rewardedwith additional funds to complete some irrigation systems. Several states have taken this up-Maharashtra, Rajasthan, Madhya Pradesh, Uttar Pradesh. The approach must be to address farmers' problems rather than attack their perceived interests. Making appropriate infiastmcture investments, creating a climate to encourage off-farm job creation, and utilizingparticipatorymechanisms to introduce new cropping incentives, improving targeting, recoveringuser costs, and managingcommon assets and resources would together result inmore change andrationalization inthe agriculture sector thanany top-down attempt to enforce reform. 142 Chapter 9. Bringingup lagging states Seven poor states-Bihar, Chhattisgarh, Jharkhand, Madhya Pradesh, Orissa, Rajasthan, and Uttar Pradesh-are together home to over 40 percent o f India's population and nearly 50 percent of India's poor. These states, as shown inchapter 1, have failed to produce sustained accelerated growth inthe 1990s.Although they are growing at modest rates, they are increasingly lagging the rest o fthe country ininvestment,manufacturing, and income. These lagging states accounted for just 23 percent o ftotal investment in2003, while six richer states-Andhra Pradesh, Gujarat, Karnataka, Maharashtra, Punjab, and Tamil Nadu-attracted over 66 percent o f the total. Just four states (Maharashtra, Tamil Nadu, Karnataka, and Gujarat) along with Delhi, accounted for 72 percent o f foreign direct investment, while the sevenpoor states capturedjust 13 percent, The share of the lagging states inmanufacturing output declined from 30 percent in 1990/91 to only 17percent by 2002/03, and their share inAll-India GDP fell from 32 percent in 1990/91 to 25 percent by 2002/03. As a result, the average per capitaincome o f these seven lagging states relative to ten large richer states55decreased from 71percent in 1980/81 to only 54 percent in 1999/00. The gaps across states are widening inIndia, not because growth inthe lagging states has decelerated, but because growth inthe middle-income states has accelerated. And this is largely because the middle-income states were muchbetter placed to capture the gains o fthe economic liberalizationprogram launched in 1991,andthey have managed to attract muchmore private investment over the past decade than the lagging states. But there i s no reasonwhy the lagging states cannot, accelerate their growth as well, through improving their investment climate. While it is the statesthemselves that ultimately need to take advantage o fthe growth opportunities presented, compensating transfers instate fiscal flows, especially against the background o fthe challenges created by increasingincome disparities across states o f India, can also play a role. This i s discussed inthe final section o f this chapter. IntegratingIndia: goods and labor The most straightforwardway to address disparities across regions within a country is to ensure that there is a nationally integratedmarket inassets, goods, and labor. Policies that fragment markets-such as border taxes-can make it difficult for lagging states to attract job-creating investments. A recent study (Virmaniand Mittal, 2005) commissioned for this report examined whether Indiawas becoming more or less integratedas a national market by examining the differences inprices across regions o f India. It appears that commodity markets are muchmore integrated in2004 than a decade ago: the coefficient o fvariation o fprices (the ratio o f the standard deviation o fprices to the mean), which is a measure o f dispersion o fprices across markets, fell for nearly every commodity, and the median fell from 0.14 to only 0.08 (figure 9.1). Continuing to eliminatebarriers to movement o f goods across state borders and across regions is important to make the lagging states more attractive as a location for production for the domestic (or international) market. j5These states are Andhra Pradesh, Gujarat, Haryana, Karnataka, Kerala, Maharashtra, Punjab, Tamil Nadu, Uttaranchal, and West Bengal. 143 Figure9.1: Dispersionof pricesacross marketsin India,by commodity, 1994 and2004 0.40 .-m 58 0.35 1994 2004 0.30 .- 0.25 -.-g 0 0' 0.20 c 2C 0.15 .-0 8 O . I 0 0.05 0.00 Note: CV o fprices indicates the coefficient o fvariation inprices (the ratio o fthe standard deviation o fprices to the mean). It is a measure o f dispersion of prices across markets. Source: Virmani andMittal2005. It has longbeen argued that, for a variety o f social and economic reasons, labor marketsare not well integrated across Indianregions and states. Wage differentials across regions create pressures for people to move, either temporarily or permanently, to regions with higher wages. There are signs o f convergence inwages across regions, albeit at a slow rate. Figure 9.2 plots the growth of real rural (nonfarm) andurban casual wages from 1983 to 2000 against initial wages in 1983. Figure9.2: Growthin casualwages inIndianstates, 1983-2000 Rural (non-farm) Urban (casual wages) -0.5 &-, m 2 -0.5 ~ I 2.0 2.5 3.0 3.5 4.0 4.5 2.0 2.5 3.0 3.5 4.0 log wage in 1983 log wage in 1983 Source: Ahsan andPages (2005) There are other signs o f important movements o fpopulation that have economic impacts. Figure 9.3 shows the growth o fpopulation and Gross State Domestic Product (GSDP) per capita on 144 equally spaced axes as a box with dimensions set at the 10thand 90th percentiles o fthe states on each o fthose axes. If population movements across regions are large and differences inper capita growth are small, consistent with labor markets that are very tightly integrated, then the box is a long and thinrectangle (which i s true o f economically integrated, highlymobile countries, such as the UnitedStates or Canada). Incontrast, ifmovements inlabor are small and differences ingrowthinoutput per person are large, thenthe box i s a tall thinrectangle, consistent with poorly integrated labor markets (and which is true o f data across countries). Over time, it appears that differences inpopulation growth across stateshave grown-the 1OtW90th box is wider inthe 1990s-while the dispersion inoutput growth rates across states has not diminished. So, although the evidence on wage convergence suggests some increasing integration o f labor markets, so far labor mobility has not beensufficient to cause convergence across states inIndia. Figure 9.3: Decadal growth of per capita gross state domestic product and population in India, 1961-99 2a 5.0% - .-m 14 4.0% - 0 a& 3.0% - (I) c3 'c 0 c 55 2.0% 0.9% 0 5% 1.0% 1.5% 2.0% 2.5% 3.0% -1.O% J Decade wise Population Growth, PPA 1961-1970 1971-1980 1981-1990 - 1991-1999 Source:Author's Calculation Unleashing the industrial potential of lagging states 145 A fully integrated national market isthe first key step to addressing the issue o flagging states. With an integrated market, wage differences will cause some mix o fpeople movingtojobs and jobs moving to people. While it is important to allow all Indians the freedom to pursuejob opportunities no matter where they are, it i s also important to encourage opportunities close to home-jobs movingto people-to encourage growth especially inthe lagging states. But for that to happen, the lagging states have to be able to attract the investment sufficient to createjobs. Andthe investment will likelyhaveto be inindustry,particularly manufacturing, which still has the greatest potentialto provide high-wage employment for the largeproportion o fthe labor force still working insubsistence agriculture or as casual laborers and to integrate these workers into the modem, formal economy. Hence, the improvement inthe climate for investment in industryinthe lagging statesis apriority, but a difficult one. Table 9.1: Investment performance of India's lagging states The lagging states are: Thebetter performers are: Bihar Andhra Pradesh Jharkhand Gujarat Madhya Pradesh Kamataka Chhattisgarh Maharashtra Orissa Punjab Rajasthan Tamil Nadu Uttar Pradesh Account for 23 percent o f investment Attracted over 66 percent o f the in2003; 13percent offoreigndirect total investment in2003; 72 percent investment during the 1990s. o f foreign direct investment during Share inAll-India GDP: 25 percent. the 1990s. Average per capita income: 54 percent o f the average per capita income o f other major states. Industrial output and employment inIndia remain concentrated ina few, better-performing states (and within them, ina few major metropolitan areas) (table 9.1). This reflects large gaps in manufacturing labor productivity across states (and cities). According to the World Bank's India Investment Climate Assessment 2004, labor productivity is almost 20 percent higher inthe six states that have attracted the most foreign direct investment to the country, comparedwith the rest o f India.56Inpart, this i s because states with highforeign direct investment have managed to attract more investment inplant and equipment than other parts o f India. The rate o fplant-level net fixed capital formation inthose states i s 6.3 percent (against 1.6 percent inother states).57 The cumulative outcome o fthese differences incapital formation is that the average employee is better equippedwith machines andtools instates with highforeign direct investment (and in cities with low costs) than inother states and cities. Furthermore, because wage rates are higher inthe high-investmentstatesandthe low-cost cities, the average employee there ismore skilled andmotivated. 56 Iti s also 85 percent higher in about half a dozen metropolitan areas inthe better performing states than in all other major cities. 57 The rate o fplant-level net fixed capital formation inlow-cost cities i s 10.3 percent (against under 2 percent in high-cost cities). 146 However, gaps inworkforce skills and equipment are only part o f the explanation o f the spatial disparities inmanufacturing labor productivity. Evenifthere were no skill gaps between states andbetween cities-if every employee was paid the same, exerted the same effort, and used the same technology-labor productivitywould still be considerably higher inhigh-investment states (andlow-cost cities) than inother parts o f Indiabecause o fthe differences inexternal economies stemming from the geography o f locations andthe policy environment inwhich businesses operate-that is, inthe investmentclimate. While external economies arising from the (economic) geography o f any location are given inthe short to mediumterm, unfavorable productivity gaps arising from deficiencies inthe investment climate can, by definition, be remediedthrough appropriate policy changes.'* Thebindingconstraints on manufacturingproductivity inthe current investment climate in India's lagging states are (a) inadequate infrastructure-particularly unreliablepower supply, which leads to excessively highpower costs for firms, andpoor road networks, which add to firms' costs; (b) poor economic governance-particularly the redtape involved instarting and closingbusinesses, and the continuingburden o fthe "Inspector Raj," characterized bythe interference o f government inspectors, such as tax and labor inspectors, inthe daily operations of firms-which places an excessive regulatory burdenon firms and raises the costs o f doing business; (c) serious law and order problems; (d) access to land; and (e) access to finance. Reducing infrastructure bottlenecks Table 9.2: InfrastructureindexrankingofIndianstates Rank State High Goa, Maharashtra, Punjab Highmiddle Gujarat, Haryana, Kerala, TN Middle AP, Karnataka Lower Middle HP, MP, Orissa, UP, Uttaranchal, WB L o w Arunachal Pradesh, Manipur, Meghalaya, Jharkhand, Mizoram, Nagaland, Assam, Chhattisgarh, Sikkim, Tripura, Bihar, Rajasthan Source: Report o f the Twelfth Finance Commission, Government o f India, 2005. Inthe InfrastructureIndex presentedinthe report ofthe TwelfthFinance Commission (Government o f India 2004), among the major Indian states, Jharkhand, Chhattisgarh, Bihar, and Rajasthan are ranked inthe lowest category for access to physical infrastructure, while Madhya Pradesh, Orissa, andUttar Pradesh are inthe second lowest category (table 9.2). 58 See India Investment Climate Assessment, World Bank 2004e (op. cit.). 147 Figure 9.4: Access to power in India's lagging states 70 Lv) 60 BL E 50 2000 m2003 E% 40 .-P, 5 30 0 g 20 v= Fl 10 s 0 1 Source: World Bank 2004a According to the India InvestmentClimate Assessment 2004, the problem that firms inIndia's lagging states face in accessing reliable power at reasonable cost is among the most significant factors constraining business profitability, andthus it is a key factor inreducing the attractiveness o f these states as investment destinations. With a few notable exceptions, firms in India's lagging states generally have inferior access to power from the grid than firms inthe richer states. This leads to heavy reliance by firms on their own generators (figure 9.4). By far the worst situation appears to be inOrissa, where 60 percent o f the firms surveyed reported relying on their own generators; inUttar Pradesh, over 30 percent o f firms report owning generator sets. Contrast this with the better states, such as Gujarat, wherejust 17 percent o f the firms surveyed rely on their own generators, or Andhra Pradesh, where the figure is 17.6 percent (similar to Brazil and lower than inChina). Inturn, this translates into very highcosts o fpower for businesses located inthe lagging states, reducing profitability. Indeed, firms inUttar Pradesh's capital city, Lucknow, currently report paying $0.1O h hfor electricity, compared with the Indian average o f $O.OS/kwh for industrial use, and lower costs insome Indian states. Our calculations presentedinthe World Bank's India Investment Climate Survey 2004 indicate that, ifthe cost o f electricity for firms inLucknow were to be reduced to the nationwide average, firm-level profitability would increase by almost 8 percent. Furthermore, ifprices were around those o f Thailand at $0.05/kwhYthen profits would increase by almost 30 percent relative to the baseline scenario. Ingeneral, the laggingstatesalso havepoorer accessto other infrastructure services. For instance, among the major Indianstates, Madhya Pradesh has among the highest proportion of habitations unconnected by roads, with Orissa, Rajasthan, andUttar Pradesh not that far behind (figure 9.5). As shown infigure 9.6, access to sewerage facilities also differs greatly across states. The challenge o f reforming infrastructure has been discussed earlier inchapter 6. 148 Figure9.5: Percentageof habitationsnot connectedby roads, by Indianstate 82% Andhra Pradesh Source: All data from Ministry o f RuralDevelopment,Government o f India. Figure9.6: Percentageofthe populationwith access to sewerage facilities,by Indianstate Rajasthan Orissa Chhattisgarh MP AP WB TN Karnataka UP Uttaranchal Maharashtra Gujarat 0 10 20 30 40 50 60 70 Percent access Source: CentralPublicHealth& EnvironmentalEngineeringOrganization, 2000. Figure9.7: Cost andtime requiredto start a businessinIndia's laggingstates 149 52 - RJ, 51 5 49 ~ UP, 46 8 46 - OR, 44 9 w , 4 3 4 43 . Cast to Start Business (percent Time to Start a Business GNP per capita) (Days) Source: World Bank 2004h Improving economicgovernance Poor economic governance-characterized byredtape andregulatoryhassles-places aburden on businesses across India, but the problem i s particularly severe inthe lagging states. Regulatory procedures governing entry, exit, and day-to-day operations are cumbersome and require private investors to deal extensively with the state bureaucracy. Delays and harassment in these interactions are common complaints, and such governance failures impose severe costs on firms. In2004, starting a business cost 51.5 percent o fper capita income inRajasthan, 45.8 percent inUttar Pradesh, and44.9 percent inOrissa (figure 9.7). This was considerably higher than the cost o f startup inWest Bengal (the best state interms o f the cost ofbusiness startup). Similarly, while it took as many as 79 days to start a business inOrissa (most countries inwhich ittakes as many days or more to start abusiness are concentrated inSub-Saharan Africa), it took 57 days inKarnataka or Punjab (the best states on this indicator). As these better states introduce measures to further streamline business startup, they can look to other emerging market economies (China: 48 days, Russia: 33 days, or Malaysia: 30 days). Another entry indicator is property registration. Registering a property takes longer in India (on average, 67 days), than inChina (32 days), Brazil (47 days), or Russia (52 days). But the gap between India's better states and the lagging states is muchwider (figure 9.8). Ittook, for example, 123 days to register a property inOrissa; this number is comparable to many Sub- Saharan African countries. Incontrast, registering a property inKarnataka (the best Indian state on this indicator) took 35 days, not much longer than what it took inChina (32 days). Similarly, the cost to register aproperty was muchhigherinlagging states like Orissa, Rajasthan, or Uttar Pradesh, compared with Punjab (the best Indian state on this indicator). 150 Figure 9. 8: Cost and time required to register a property inIndia's lagging states Cost to Register Property (% of Time to Register Property propertyvalue) (Days) Source: World Bank2004h A keychallenge for the lagging statesis to introducereforms to streamline clearances and approvals. They can learn fiom the experience o f the better states. Some states (e.g., Kamataka) have increasingly moved away from "in-principle" to "actual" clearances and from sequential to concurrent approvals. Andhra Pradeshhas introduced "deemed clearance." The better states have also introduced improved legal frameworks for business entry (e.g., Karnataka Industries Facilitation Bill).Another set o f reforms i s the introduction o fbusiness facilitation agencies (e.g., Gujarat's Index B, Maharashtra's Udyog Sarathi, Tamil Nadu's Industrial Guidance Bureau, and Andhra Pradesh's APFIRST for information technology firms). Simplified application forms for starting a business or registering property have also been introduced in manyplaces, as have improvementsto the functioning o fthe court system. Cumbersome bankruptcyprocedures make exit as difficult as entry. In2005, it took 10years to close a business inIndia, compared with 2.3 years inMalaysia, 2.4 years inChina, and 3.8 years inRussia. Recoveryrates inIndia, while higherthaninChina, were significantly lower than in Malaysia or Russia. The ease o f closing a business also vanes significantly across India. In2004, closing a business took more than 15 years inUttarPradesh, 11.34 years in Orissa, and 10.42 years inRajasthan. Contrast this with Kamataka (the best Indian state on this indicator), where it took 8 years to close a business; while considerablyworse thaninmost emerging East Asian countries, this was still better than Brazil, where it took 10years to close a business in2004. Recovery rates are also much lower inUttar Pradesh, Orissa, or Rajasthan, compared with Karnataka. While the License Raj has been substantially reduced at the center, the pervasive Inspector Raj survives at the state level, as previously noted. Indian manufacturers face, on average, 7.4 visits a year from government officials who visit firms to inspect compliance with various laws and regulations (e.g., on tax, labor, and environmental standards) and to renewbusiness permits. This is an improvement over the reported 11.7 visits a year that firms faced from government inspectors in2000, and the number i s lower than inBrazil and much lower than inChina. But government inspectors still appear to take up a higherproportion o fthe time of senior 151 management inIndianfirms each year (14.2 percent) than inChina (8.1 percent) or Brazil (7.8 percent) (table 9.3). The burdeno fthe Inspector Raj also varies across the Indian states. While firms inMaharashtra facedjust five inspections a year, and those inDelhi faced less than four inspections a year, their counterparts inOrissa faced manymore inspections a year. Table 9.3: Number of inspectionsa year andrequiredmanagementtime inIndia and comparator countries No.of Senior management time inspections a spent dealing with year regulations (percent) India 7.4 14.2 Brazil 9.6 7.8 China 36.0 8.1 Source: World Bank, 2004e. A key challenge for the lagging states that face aparticularly heavyburden o fthe InspectorRaj is to streamline business procedures to reduce delays and opportunities for rent seeking. This may require reengineeringthe entire gamut o fprocesses governing business operations, on the basis o f clear principles o ftransparency, absence o f discretion, and accountability. The experience o f the better-performing states points to some important lessons. For example, in Karnataka, only a random sample o f firms i s inspected. InAndhra Pradesh, statutory inspections havebeenreduced to once a year. InGujarat, the inspectionprocess has been outsourced to accredited private agencies. Some states have introduced "self-certification": inAndhra Pradesh, barringmajor hazardous industries, all industries can self-certify compliance with labor laws. Annual returns andregisters for labor regulationshavebeen streamlinedinAndhra Pradesh, Tamil Nadu, Gujarat, andKarnataka. Some o f India's lagging states-including Bihar, Jharkhand, Orissa, andRajasthan-have a rich mineral resource base that could be developed by miningand quarrying enterprises. But many o f the bottlenecks identified above have preventedthese states from achieving their potential inthis sector (box 9. Some states have also questioned the current royalty regime which governs the distribution o fbenefits between the center and states. Box 9.1: NaturalResourcesin India's poorer states Figure 1presents a simple plot o f per capita state income and the share o f mineralproduction instate gross domestic product. India's lagging states generally have a higher dependence o n mineral resources, but lower per capita incomes. Other measures of development produce a similar pattern: the mineral dependent states have consistently poorer human development scores, lower growth rates and higher levels o f infant mortality. Global evidence indicates that inmanycountries, a highlevel o f mineral dependence i s associated with weaker economic performance, lower growth rates and inferior development outcomes. Inparticular countries that are dependent on point resources - i.e. resources extracted from a narrow geographic base (such as minerals) - perfoimpoorly across a range o f development indicators. This result i s termed the re,sourcecurse. 59 See "Turning the Minerals and Metals Potential o f Eastern India into a Goldmine," CII and McKinsey, April 2005; "Rajasthan Economic Report," World Bank 2005; "Orissa-Investment Climate Assessment," World Bank 20050. 152 I Figure 1:Mineral dependence correlated withpoor per$ormance t 1 I Rajasthan 0.16 7 0.14 v) $ 0.12 4 `Oriss E 0.06 cv) $ 0.04 2 Gujarat Q) 0.02 0.00 0 2500 5000 7500 10000 12500 15000 17500 20000 Income per Capita There are numerous explanations as to why the resource curse might occur, but there is a broad consensus that weak governance, coupled with the technology o f mining, are a key explanation. Extractive industries tend to be highly capital intensive, rely o n a small number o f skilled workers and are geographically concentrated. T o establish linkages with the mining sector and catalyze growth across the economy requires good institutions and an investment climate that enhances the opportunities for firms to invest and create jobs. But the evidence shows that mineral dependent economies tend to have weak institutions that impede investment and development. Many o f the reforms that are necessary to create a growth friendly investment climate place few demands o n the budget. Hence the problem is not merely due to a shortage o f finance. Instead, the prospect of rapid gain from mining reorients institutional and administrative priorities from concerns about the size of the pie (growth) to concerns about the share of the pie (distribution). Minerals and other point resources are.an enticing target for rent seeking. However, this outcome i s not inevitable. Many countries havejudiciously harnessed their resource wealth to create prosperous economies. Reviewing the diversity o f outcomes indicates that there i s no single recipe for success. Norway and Australia's strong accountable institutions ensure that minerals are exploitedjudiciously, environmental impacts are minimizedand the windfalls are used to deliver public goods that sustainbroad based development. In the developing world, Botswana's unique governance structure and fiscal rectitude promotes benefit sharing through theprovisiono finfrastructure andpublic goods that promote investment, while Chile's deep reform agenda has generated cross-sectoral growth. Inso far as there are common threads, they include the following three elements o f success: (i) a strategy that generates broad based growth, (ii) strong institutions and (iii) sequencing o f reform. Broad Based Growth Strategy: Mininginisolation i s unlikely to generate broadbased growth and development. The success stories are ineconomies where mineral extractionhas spurred or complemented development inother sectors o f the economy. But this can only occur ifthere i s an investment friendly climate that promotes linkages with upstream or downstream industries, or creates investment opportunities inother sectors. These generate the jobs and growth that spread the benefits o f mining.Broadbased development is therefore a necessary adjunct to a mineral dependent growth strategy. Strong Institutions: Strong institutions are needed to prevent the capture o fbenefits and promote investment Institutions are also necessary to address the many and varied environmental and social costs o f mineral extraction that impede growth. Sequencing of Development: Mineral developments incountries with strong and accountable institutions have 1 produced virtuous outcomes (e.g. Norway and Australia). Where institutions are weak, similar developments have had negative outcomes (e.g. Nigeria and Venezuela). Hence the timing o freform and development matters. Institutionalstrengthening and reform should be the center piece o f a mineral intensive growth strateev. ~ 153 Introducing law and order A pressingconcern facedbyinvestors, particularly inIndia's laggingstates, isthe morebasic problem o f law and order. Investors will not invest unless their investmentscan be preserved and protected. A recent surveybyIndia Today,6owhich provided a rankingo f states accordingto their ability to maintainlaw andorder, found that Bihar andUttar Pradeshhadthe worst security situation among the major India states in2005. The situation inJharkhand, which has the lowest number o fpolicemen (32) per 100,000 people, is not that muchbetter. A fbndamental component o fthe rule o flaw is the ability to enforce acontract. The inefficiency of the court system inthe lagging statesmeans that it takes much longer to enforce a contract (1,165 days inUttarPradesh, 875 days inRajasthan, and 765 days inOrissa) than inthe best Indianstate onthis indicator (Maharashtra, 425 days) (figure 9.9). The cost o fenforcing a contract was also muchhigher inUttarPradesh, Rajasthan, and Orissa, compared with the best Indianstate onthis indicator (Karnataka). Figure 9. 9: Time required to enforce a contract in India's lagging states 10004 I RJ ,a75 800 OR, 765 700 600 500 400 Cost to Enforce a Time to Enforce a Contract Contract (%of Source: World Bank 2004h Improving the efficiency of land markets Some 90 percent o f landparcels inIndia are reportedly subject to disputes over ownership, which take decades to settle incourt. Furthermore, obsolete tenancy andrent control laws keep a largepart o furbanreal estate offthe market. The central government has abolished the Urban Land Ceiling Act, which made changes inland use very difficult; however, only a few states have repealed their corresponding acts, and the lagging states are typically not among these 6oSee "State o f the States Survey," India Today,August 2005. 154 states. The inefficiency o f land marketsi s a constraint to business development outside o f the cities, too. Improving access tofinance for business Problems inaccessing finance are often cited as a major impedimentto the performance o f small and medium-size businesses inIndia, particularly inthe lagging states. Per capita credit-to- deposit ratios are considerably lower inlagging states such as Bihar, Orissa, andUttarPradesh thaninbetter-performingstates such as Andhra Pradesh, Karnataka, Maharashtra, andTamil Nadu (figure 9.10). Figure9.10: Access to finance inIndian states, 2003 Per capita credit to deposit ratio, March 2003 (%) 100 80 60 40 20 0 Figure 9.11: Growth incredit and industrial output in Indian states, 1991/92-1999/2000 All SCB Public sector +Industrial growth rate 155 Another indicator of the difficulty that firms face inaccessing finance is the flow o f credit to small-scale industry. The data cover credit to small-scale industryunits in 14major states by all scheduled commercial banks and public sector banks for the period 1991/92 to 1999/2000. On this indicator, too, Bihar and Orissa perform the worst (figure 9.11). Problems incredit access are attributable to a combination o f factors that are rooted in(a) weaknesses inthe legal framework for loan recovery, bankruptcy, andcontract enforcement, together with inefficiencies inthe court system, with the latter largelyaccounting for interstate variations inthe time and cost o f loan recovery andbankruptcy; (b) institutional weaknesses, such as the absence of good credit appraisal and risk management and monitoringtools inbanks, which increase transaction costs indealingwith small andmedium-size enterprises; (c) the absence of reliable credit information on small and medium-size enterprises; and (d) lack of sufficient market credibility for such enterprises. It i s difficult for lenders to assess risk premiumsproperly, creatingdifferences inthe perceivedversus realriskprofiles o fsmall and medium-size enterprises. As a result, these enterprises are often unable to tap lending opportunities. All o fthesebusiness constraints-infrastructure bottlenecks, government red tape, legal enforcement problems, and obstacles to securing landand finance-impose a heavytoll on productivity. Calculations presented inthe India Investment Climate Survey 2004 attempt to quantify some o fthe cost. If,for example, power supplyproblems couldberesolved so that the typical business neednot rely on its own generators, manufacturing labor productivity inlagging stateswould increase bymore than 80 percent. Reforms that would improve business regulation would increase productivity bymore than 60 percent. Ifboth reforms took place at the same time, alongwithreforms increasingaccessto land andto formal finance, average manufacturing labor productivity would rise bymore than 160percent. These counterfactuals for the level of labor productivity translate to large gains inbusiness growth and business investmentrates from the same hypotheticalreforms inthe investment climate. For example, a 10percentage point reduction inthe indicators o f deficiency inpower supply, tax and customs administration, access to land, access to finance, andlabor regulation that we have usedinthis report would raise the average firm-level sales growth rate from the current 11.3 percent to 15.9 percent a year. Transforming the rural economy of lagging states Evenifall o fthese bindingconstraints to industrial growth and employment inthe lagging states could be put inplace, there would still be a need to get the rural economies o fthese states moving, given that the overwhelming majority o fpeople inthe lagging states still live inrural areas. Transforming the rural economy o f lagging states will require not only improving agricultural productivity-addressed inchapter 8-but also facilitating rural (nonfarm) entrepreneurship. The latter calls for efforts on a number o f fionts-but withparticular attention to the bindingconstraints of rural infrastructure and rural finance. 156 Upgrading rural infrastructure A better rural infrastructure (including access to power, roads, and telecommunications) is critical to rural economic transformation. Better access to roads can play a dramatic role in transforming the rural landscapeby integrating the rural population into the formal economy. But more than 50 percent o frural habitations are not connected byroadinthe states o f Bihar, Jharkhand, Madhya Pradesh, Orissa, Rajasthan, and West Bengal, and inChhattisgarh, for instance, some 82 percent of rural areas remain unconnected by road (figure 9.5). Investing in road infrastructure i s a critical priority inthese states. Access to power can also have a transformative impact on the rural economy. Among the major Indian states, the challenge o f rural electrification i s the greatest inUttar Pradesh, which reports the lowest proportion o f electrified villages (under 60 percent). Access to a phone line can help link rural producers with markets, raise incomes, and change lives. Again, the lagging states, particularly Bihar, Chhattisgarh, Jharkhand, and Uttar Pradesh, have much catching-up to do on this front. To addressthe issue o fupgrading rural infrastructure, the United Progressive Alliance government recently (May 2005) announced the Bharat Nirmanprogram. Under this program, the government is set to earmark a major investment for rural infrastructure insix areas: irrigation, drinkingwater and sanitation, roads, electrification, telecommunication, and housing:61 We are committing over Rsl,74,000 crores Bharat Nirman should unleash the growth potential of our villages...in the nextfour years we need to ensure that every habitation haspotable water. Every village of over 1,000 population or over 500 in hilly and tribal areas, must have an asphalted road...we must also ensure that over one crore hectares of land is irrigated and that at least 60 lakh houses are built."62 While Bharat Nirman signals the government o f India's commitment to the transformation o f the rural economy and i s a potentially useful program, the key to success will lie inthe implementation. Bharat Nirman can be a vehicle for changing the way business is done by addressing the issues highlighted inpart Iof this report on service delivery-or it can be a fiscally costly continuation o fbusiness as usual. Ruralflnance Beyondrural infrastructure, another critical ingredient o f rural transformation i s access to finance. A growing body o fresearch from around the world-some o f it by the World Bank- shows that well-developed and inclusive financial systems are associated with faster growth and better income distribution. Financehelps the poor catch upwith the rest o f the economy as it grows. Finance also helps extend the range of individuals, households, and firms that can get a foothold inthe modern economy, and it reduces damaging concentrations o f economic power. Largely thanks to microfinance, there i s now a growing appreciation o fthe "empowement" http:Ilwww.hindu.cod2005/08/2l/stories/2005082112900800.htm. 62Prime Minister Speechon October 2005. See http:linews.indiainfo.comi2005110108108l0nirman-flagship- upa.html. 157 dimension o f finance, o f the extent to which it can give ordinary people and the poor access to opportunity andthe ability to escape ossified social structures. The financial sector strategies o fsuccessive post-Independence governments inIndiahavebeen shaped by the goal of "serving better the needs o fthe development o fthe economy inconformity with the national policy and~bjectives."~~To this end, the 1950s and 1960s saw the growth o f a large number ofrural credit cooperatives across India. The late 1960switnessed the nationalization o f commercial banks, which were charged with loosening the grip o ftraditional informal sector moneylenders through the use o f targeted, low-priced loans. Between 1969 and 1980, thousands o fnew bank branches were established across rural India. Rural credit planning, involvingquantitative credit targets and subsidized credit, became the order o f the day. Mandatory requirements were placed onbanks to direct largeproportions o f their credit to priority sectors, including agriculture and small-scale industries, identified as critical for bringing about economic and social change inrural areas. Competition was limitedby the "service area approach," which limitedentry. Severe constraints were placed on the operational and financial autonomy o fbanks. While the 1990s saw increased competition and liberalizationinthe Indian financial sector, some of the key features of rural credit planninghave persisted. Over the past decade, interest rates have been largely deregulated, although lending rates on small loans (under Rs 200,000) are capped at the prime lendingrate, which banks are free to set; there is also a floor on short-term deposit rates. Priority sector credit requirementshave beeneased, but remain highat 40 percent. Debtor interest waiver schemes are usedfrequently. Competition inthe banking sector has increased, but the public sector banks (including the nationalizedbanks, the State Bank o f India, and regional rural banks) continue to dominate the banking system, accounting for 73 percent o f commercial banking assets and 52.4 percent o fthe assets o f all financial institutions inthe system. Competition is particularly weak inrural areas, although the ReserveBank o f India's recent decision to dispense with some o f the restrictive provisions o fthe service area approach may help stimulate the entry o fnew branches inrural areas. Butthe evidence suggests that the provision o fdirected credit to farmers andsmall-scale rural industryon subsidized terms through the public sector banks has not met with much success (Basu 2005b). Recent data from a World Bank-NCAER Rural Finance Access Survey (World Bank 2003d) indicate that access to formal finance i s a problem across rural India and is a particularly severeproblem insome o f the lagging states, such as Uttar Pradesh, where 80 percent o f all rural households have no access to formal credit, and 87 percent o fpoor rural households have no access to credit from a formal source (Basu and Srivastava 2005). According to the Rural Finance Access Survey 2003, for those rural households that do have access to formal finance, commercial banks are by far the most dominant source. They account for 51 percent o fhousehold deposits and are also the most important source o f credit. Regional rural banks account for 34 percent o f household deposits and 31percent o f credit. Other formal sources, such as credit cooperatives and post office branches, appear to play a modest role in providing savings and credit services to rural households. 63Reserve Bankof India:Functions andWorking, Murnbai, 1983. 158 For the minority with access to finance from a formal source, transaction costs are high.The 2003 survey indicates that all types o f formal institutions demand bribesbefore approving loans. The bribe amounts are hefty, and higherinUttar Pradesh than inAndhra Pradesh. Procedures for opening an account or seeking a loan are cumbersome and costly, with loan processing times tending to be much longer inUttar Pradesh thaninAndhra Pradesh. Longer processing times for loans, together with bribes, result inhigheffective costs for small borrowers (table 9.4). Table9.4: Aspects offormalborrowinginAndhraPradeshand Uttar Pradesh State Banks of India Regional h r a l banks Coops Others Andhra Uttar Andhra Uttar Andhra Uttar Andhra Uttar Pradesh Pradesh Pradesh Pradesh Pradesh Pradesh Pradesh Pradesh Interest rate (median) as 15 12 12 14 12 14 21 14 percentper year Loanamountreceivedas 87 93 81 95 88 79 72 95 percento famount applied Householdsreporting 1.9 38.8 4.4 47.4 2.6 19.2 0 19 bribesas percentof all households Bribesas percento f 6.5 10.2 9.8 23.8 4.5 16.7 0 13.9 amount approved Time taken to process a 25.7 36.5 25.9 32.3 25.4 22.0 7.5 22.0 loanapplication(weeks) Source: World Bank 200d. Fiscal federalismandthe TwelfthFinanceCommissionrecommendations Just as the overall performance o f India's states has beenincreasingly divergent, so has the fiscal performance o f its states, with fiscal indicators deteriorating much more rapidly inpoorer states. Not only have the poor states suffered greater variability inrevenues, with higher debt stocks and salary bills, they have also sufferedmore from recent expenditure shocks. Interest payments as a share of own revenues are nearly twice as highinpoor states as inother states: inBihar and Orissa debt service preempts more than 90 percent o f own revenues. The tight fiscal situation in many o f India's states is severely hamperingthe role o f the state as an effective developmental agent. Against this background, the recommendations o f the Twelfth Finance Commission are welcome, and their implementation should assist the states inbecoming more effective agents o f development. As a result o fthe commission's proposals, there is a one percentage point increase intotal taxes transferred to all states as a shareofthe government ofIndia's tax take (from 29.5 percent to 30.5 percent). An even bigger increase i s projected ingrants: the average annual grants for all states are 143 percent higher duringthe Twelfth Finance Commission's tenure (2005-10)' compared with the previous five-year period (2000-05). Loans to states from the government o f India, however, will be much lower, consisting only o f on-lending o f external loans and credits. The commission has thus correctly sought to reverse the damagingtrend established over the 1990s o f fewer transfers and more loans. 159 Nine o f the ten grants that the Finance Commission has rewarded are tied. The Twelfth Finance Commission has set up a Debt ReliefFacility to replace the FiscalResponsibility Facility. The total incentive offered by the Debt ReliefFacility is bigger that the previous arrangements and i s made up o ftwo schemes. Enacting fiscal responsibility legislationqualifies states for restructuring o f old central debt at significantly lower interest; and the second scheme offers a waiver on debt servicing as a reward for meeting prescribed annual targets for reducing the revenue deficit. Tied grants for maintenance (separate grants for roads, buildings, and forests) are conditional on states' increasing their own spending inthese areas, as are the health and education grants, which are only for the poorer states. The obvious challenge for the states is to ensure that these funds are productively spent (box 9.2).64 Box 9.2: State financial accountability issues The World Bank has recently undertaken a series o f State Financial Accountability Assessments, from which a number o f lessons have emerged onpublic expenditure reforms: Budget realistically and implement the budget aspassed. A good budgetary set-up is one inwhich it is difficult to get some project into the budget, but then, after budget approval, implementation is automatic. Inmost Indian states, the opposite i s the case. States are endemically overstretched, and their reach far exceeds their grasp. Any number o f initiatives are introduced, and then underfunded. This shows up ina number o f ways. Budget revenue estimates are systematically overly optimistic: for the five years ending 2002/03, budget revenue estimates exceeded actuals by an average o f 8 percent. The systematic nature o f this bias suggests that the problem is a political one and that revenue forecasts are inflated to allow artificially highexpenditure levels to be projected. And then during the year, new projects are added through policy pronouncements and supplementary budgets, which often add another 5-10 percent to total spending. The result is that not only do deficits exceed targets, but cash and administrative rationing has to be usedto prevent too many budgetedprojects from actually proceeding. Projects thus lie incomplete, huge arrears o f unpaidbills pile up, and an enormous amount o f administrative time i s consumed inpersuading Finance to release funds. The most important budgeting reform that state governments could undertake would be to base the budget o n realistic revenue forecasts, to restrict new policy initiatives to the budget period, and then to relax post- budget central controls on spending. This i s harder than it sounds. It requires very tough decisions to be made o n what governments can and cannot afford to d, and strong political leadership and ownership o f the budget. Enhance departmental accountability andflexibility in the budgetary process. InIndia's states, budgets are typically puttogetherby scheme, rather thanbydepartment, andthere is very little discussionofwhat money isbeing spent to achieve. Departments needto be given muchmore flexibility to spend money as they best see fit to achieve agreed targets within an agreed envelope of resources. Ths reform itselfneeds to be seen within the larger context o f focusing departments on targets and results. Most departments do not have mission or vision statements; transparent performance monitoring is often absent, as is systematic citizens' feedback on services provided and individual accountability; management information systems are rudimentary; and anticorruption institutions are often ineffective. Tighten budgetary controls over open-ended obligations and capital projects. While many o f the micro controls typically exercised by finance departments canbe relaxed, there are some areas where controls are too weak. Most subsidy obligations are open-ended and need to be redefined on the basis of "purchaser-provider" agreements, under which Finance commits to a certain subsidy level inreturn for an agreed delivery o f services. Similarly, control over 64 The TFC has set up a Debt Relief Facility to replace the Fiscal Reforms Facility. The principles o f the two funds are the same: states are rewarded for revenue deficit reduction, albeit now inthe form o f debt relief or restructuring rather than the cash grants given earlier. However, the corpus o f the Debt Relief Facility is muchbigger than that o f the Fiscal Reforms Facility. States can get significant restructuring simply for passing a Fiscal Responsibility Act, which has to meet certain minimumelements including the elimination o fthe revenue deficit by 2008109 and a reduction o f the fiscal deficit to 3 percent o f GDP. Ifstates not only pass such an act, but also reduce their current deficits on track to reach zero by 2008109 and contain their fiscal deficit, they will inaddition have their interest and principal repayments waived on outstanding debt to the government of India for the period o f the TFC. 160 capital projects (over boththeir entry into the budget and their implementation) is weak, with far too many capital projects receiving minuscule amounts o f funding and thus never being completed. Tighten accounting and audit arrangements. India's accounting and audit arrangements are fine onpaper, but neglected inpractice. State governments have little information on their accounts inthe course of the year; audit observations are not responded to; and many local governments do not even produce accounts, let alone audits. The challenge is to reactivate the system to reduce the gap between theory and practice. Source: World Bank 2004a. Not only does the Finance Commission aim to reduce the quantity o f loans, but it also has tried to reform the borrowing regime, which has been characterized by soft budget constraints inthe past. This i s particularly important because India's states seem to be the most highly leveragedin the While it is welcome that the central government has agreed to the commission's recommendationto stop direct lendingfkom the central government to the states, it is unfortunate that the central government continues to compel statesto borrow all the proceeds o fthe rapidly growing "small savings" hnd(largely savings mobilized through India's post offices). This is a form o f "forced savings," which is expensive andwhich teaches the states that they should borrow whatever is available, not what they can afford. The commission is silent on reform o f small savings, but this i s infact the next important area o f center-state fiscal reform. Unfortunately, despite an expressed intentionto maketransfers more progressive, India's Twelfth Finance Commission seems to havemade themmarginally less progressive. The four biggestgainers inpercentage terms are the richest states: Punjab, Gujarat, Haryana, and Maharashtra; while the average gainers are the poor states.66It i s important to note though that these changes are made at the margin, and that overall the Finance Commission's transfers to the states remain progressive. Infact, although central transfers do not come close to achieving anything like horizontal equalization, Finance Commission transfers are still the most progressive o f all the various channels o ftransfers fkom the central government to the states (negatively correlated with the level o f income o f states).67 For all that the center can do, through implementation o f Twelfth Finance Commission's recommendations andother, follow-up reforms, India's states will ultimately be the masters o f their own fiscal destiny, as showninthe case studies o fRajasthan and Bihar inthe following annexes. The combination o f a large increase ingrants andtighter controls over borrowing will reverse the adverse resource trends the states have experienced since the early 1990s andwill provide an ideal framework for fiscal adjustment by the states. But, while compensating transfers instate fiscal flows areimportant, inthe endit is the statesthemselves that needto take advantage o f the growth opportunities presented. 65In2000, for all ofIndia's states combined, the ratio o fdebt to revenues stood at 203 percent. Canada was next with 189 percent, followed byBrazil with 170percent andPakistanwith 100percent. 66Howes and Prasad (2005). 67Beyond these formal transfers there are hidden or implicit revenue transfers among states. For instance, a large hidden transfer is associated with the purchases o f farmproduce by the Food Corporation o f India, which procures at above-market prices from some states. Attempts to quantify the benefits o f this subsidy (World Bank 2004) suggest that Haryana andPunjab together account for 67 percent o f the subsidy and receive more inFood Corporation of India subsidies than they do through the formal transfer system. These informal transfers vitiate the modest progressivity of the formal transfer system. 161 Annex 9.1: Rajasthan: closingthe development gap Long considered one o fIndia's lagging states, Rajasthan, infact, made significant progress duringthe 1980s and 1990s: growth rates accelerated to become one o f the fastest among Indianstates; literacy and school enrollment increased rapidly, and poverty rates declined sharply. Indeedtoward the end o f the 1990s Rajasthanpresented an example o f how a lagging state could make progress inclosing the development gap with India's more developed states. Now, though, Rajasthanpresents an example o f the deep challenges to sustaining development even after two decades o fprogress.68 Development prospects are now under threat from several directions. While India, on average, i s growing rapidly, Rajasthan's growth has faltered markedly. A growing crisis inwater supply and erratic rainfall calls for fundamental changes inagricultural and water managementpolicies. Falling private investment rates-at a time when investment flows to other faster growing states are rapidly increasing-increase the risks that Rajasthanwill be caught ina low- level development trap. Inadequate public investment inthe late 199Os,poor maintenance, and management have led to a large shortfall ininfrastructure. Weak public service delivery mechanisms impede further progress inhuman development, a key requirement for long-term growth. This report recommends a medium-termreform strategy built around four goals: Reviving growth. IfRajasthanis to close its development gap withthe national averages, its economy has to grow at a sustained rate o f 7-8 percent annually, well above the trend o f 4-5 percent inrecent years. Agricultural diversification, supported by liberalizing agricultural marketing, modernizing researchand extension, and a regionally differentiated strategy, will be critical to reviving agricultural growth. Since nonagricultural sectors account for three-quarters o f gross state domestic product, restoring and accelerating growth inthese sectors to at least their trend rate o f the 1990s will be crucial. Rajasthan can expect to increase private investment rates significantly only ifit offers a better investment climate than competing states. To achieve this, the government has to address still extant significant barriers to entry in key sectors, labor market rigidities ingeneral, and poor infrastructure. The public-private partnership framework needs to be strengthened to attract private investment, and road connectivity needs to be improved through public investment innew roads and better maintenance. And while much has been achieved inincreasing power supply inthe state, large power sector losses threaten future supply and the state's fiscal stability. Creatingjscal spacefor development and using it efectively. After several years o f acute fiscal stress, fiscal adjustment inthe past two years has created fiscal space for much neededpublic investment. But because Rajasthan's fiscal position continues to be fragile, the mainpriority now will be to sustain the fiscal adjustment trend by effectively implementing the Fiscal Responsibility and Budget Management Act recently passedby the RajasthanAssembly. With larger fiscal space and additional resources to spend, ensuring the quality o fpublic expenditure has assumed greater importance. It will be particularly important to ensure that the rapidly increasing capital expenditures have highreturns and are not thinly spread over too many programs. Containing expenditures on salaries will also be crucial for fiscal adjustment. Strengthening civil service, local governments, andpublic accountability. Despite strengths insome traditional areas o f government, service delivery mechanisms in Rajasthan are weak, contributing to poor human development outcomes. The government o f Rajasthan needs to take steps to improve the skills and composition o f the civil service, reduce transfers and increase tenure, and make service providers more accountable for better public service delivery. Fiscal decentralization needs to be strengthened and an effective local govemment civil service cadre introduced to realize the full potential o f local governments. Sector specific interventions for human development and social protection. While Rajasthan has made substantial progress inexpanding education, challenges remain inenhancing its quality, closing the gender gap insecondary education, and extending services to disadvantaged groups. The health sector remains a cause for serious concern, even though there have been some improvements inrecent years. Nonetheless, This annex draws on the Rajasthan; Closing the DevelopmentGap (Economic Report) WorldBank 2006 162 immunization rates have stalled, and child and maternal mortality rates remain high.Given the recurrence of droughts, social protection has assumed greater significance inRajasthan. Rajasthan's weak monitoring and evaluation systems need considerable improvement to help achieve the state's human development and social protection goals. 163 Annex 9.2: Bihar: towarda developmentstrategy A recent World Bankreport (2004) identifies the key challenges that face Bihar: reducing poverty; increasing and sustaining the growth rate; improving the delivery o f services, inparticular inhealth and education; strengthening the accountability, perfonnance, and transparency o f institutions and government systems; and improving the law and order situation. The report suggests that making economic and social development objectives central to government policy i s vital for transforming Bihar's human and economic landscape. A development strategy for Bihar will require a multidimensional approach across sectors and institutions. For improving educational and health outcomes, it means workmg with different communities and organizations-private, public, and nongovernment-to leverage scarce resources. To strengthen institutions and update systems, it also means developing a vision and strategy for reforming the administrative system. It will also require improved fiscal performance to manage the mounting debt and carefully shepherd the state's scarce resources inthe right direction. The report suggests that the basic development strategy presented for Bihar could rest on two pillars: Enhancing Bihar's growth performance by establishing a better investment climate to encourage entrepreneurship, investment, and the spread o f improved technologies, particularly inthe rural sector. Supporting human resource development through improved quality and access to social services, particularly for the poor and socially disadvantaged communities. The most critical feature the state needs to strengthen to improve growth i s the delivery o f core public services. Chief among these are the improvement o fbasic infrastructure, particularly inm a l areas; the support o f agricultural researchand extension services; and the provision ofbasic law and order. K e y infrastructure services inwhich the private sector cannot substitute public delivery include road and water management, especially for boosting investment and output inagriculture, agroindustry, and related services, such as transport, storage, and marketing. The public sector has an importantrole inbasic capital investment and maintenance o f large and medium-scale water management systems. Private and community management o f small-scale and feeder systems has been effective in some pilot cases where communities have organized themselves, and this now needs to be scaled up. T o underpin the development strategy, efforts to reformpublic finance andpublic administration will be essential. Bihar is more dependent than other states on the government o f India's support to the State Plan and Centrally Sponsored Schemes to meet its development expenditure. One o f the challenges confrontingthe government o f Bihar i s to increase its utilization o f the fairly substantial amount o f resources allotted to the state under Centrally Sponsored Schemes. Bihar's relatively low utilization under these schemes i s a concern also shared by the central government. Inaddition, the government o f Bihar must buildcapacity for absorbing additional assistance from external donors in due course. A central constraint that has beenflagged inthe report and i s recognized by government o f Bihar, i s the weak public expenditure management environment inthe state. Addressing Bihar's large administrative reform agenda is fundamental both to efficient use o f public resources and for improving government performance. 164 Chapter 10. Achievingequitablegrowth India's rapid progress inthe 1990s has not beenuniformly shared among its people. As documented inchapter 1,income inequality has risen, although India still has a relatively balanced distribution of income by global standards. Byother measures o f well-being, it remains a deeply unequal society. There are large differences ineducation and health status, for example, across groups defined along axes of wealth, gender, caste, ethnicity, and location o f residence. Inprinciple, beyondsome absolute thresholdofdeprivation, inequalitiesinstatus maynotbe a cause o fworry, as long as they reflect differences inpreferences, effort, talent, or luck. Indeed, income differences play an important role inproviding incentives to invest ineducation and physical capital, to work, and to take risks. However, inequality inopportunities across people i s of concern for intrinsic reasons and also because it may have an instrumental impact on the development process.69The 2006 World Development Report on Equity andDevelopment focuses on two broad sets o f channels through which inequality inopportunities can constrain long-run development (World Ban&2005a). First, there are manymarket failures (notably in markets for labor, land, credit, and insurance) that distort resource allocation, so that resources do not flow to those who have the highestreturns. For example, highly capable children may fail to complete primary school, while others who are less able may finishuniversity. Second, high levels of inequality tend to lead to economic institutions and social arrangements that systematically favor those with more influence. Society as a whole is likely to be more inefficient and miss out on opportunities for investment and growth. Policies that promote equality of opportunities can thus not only enable the poor and excluded to participate inthe development process, but can also leadto more rapid long-run growth. In general, correcting the market failures that leadto structural differences inopportunities i s the ideal response. Where this is not feasible, or too costly, some forms o f redistribution (e.g., assets, access to services) can increase economic efficiency. It i s important to note that pursuit o f equity as definedhere interms of equality of opportunities does not implya focus on redistribution of income, which at least rhetorically drove many o f the regulatorypolicies pursued from the 1950s to the 1970s, without achieving either growth or redistribution. Liberalization o fthe last two decades has unlocked enormous potential and shown India's capacity for sustained growth and poverty reduction. To make the most o fthis potential, the challenge i s to continue rapid growth, but also to makethe growth and development process more inclusive. This chapter discusses the role o fpublic action inequalizing opportunities bypromoting equality inaccessto markets and assetsandby scalingup efforts to strengthen livelihood strategies through empowerment. Evenwith the best o f access to markets and opportunities, there are needs for the government to engage in social protection to assist the poorest and help in coping with the risks and vulnerabilities that citizens face. Socialprotectionmeasures are important to 69Inequality o f opportunities implies that individuals born indifferent groups (e.g., caste, gender, or class) have unequal chances o f acquiring assets, earn unequal returns to equivalent assets (for similar effort), have unequal access to adequate quality basic services and are treated unequally inthe processes that govern economic, social and political life. 165 combat extreme deprivation, but can have equally important dynamic efficiency effects by allowing people to bear risks and undertake profitable investments. Equality of access to markets and assets Therehas been a seismic shift from a view that the poor needto beprotectedfrom markets to the view that a better route out o fpoverty is to helpthemto strengthen their own livelihood strategies through empowerment and fair access to markets for labor, credit, land, andproducts. This section focuses onpolicies to expand access to markets for labor, a key asset o fthe poor, and for credit, which impartsboth economic and empowerment benefit^.^' Despite a variety o f schemes and policies, tribals remain one o fthe most marginalized groups inthe country. For tribals, the critical issue is not access to markets; rather, a central factor affecting them, as described below, is secure access to and control over natural resources. Finally, this section considers government andprivate sector initiatives to strengthen rural livelihoods by improving people's access to marketsand services, focusing at the same time on increasingpeople's sense o f empowerment. Addressing caste and gender bias in labor markets Caste has historically beenthe key axis o f stratification inIndia, responsible for major inequalities inaccess-in areas as diverse asjobs, technology, education, and health. Caste i s especially important for labor markets because it has at once a ritual and an occupational logic. Thus, strict rules o fdiningandmarriage, basedonritual purityandpollution, have historically governed the relations between castes, and an equally strict division o f labor has meant that only certain castes or subcastes undertook certain occupations. Inspite o f far-reaching changes, some occupations continue to be castebased (figure 10.1). Even withinthe public sector, memberso f the Scheduled Castes (SCs) dominate the manualjobs o f sweeping and cleaning-hstorically assigned to them inthe caste hierarchy. Other occupations-notably the nonagricultural semiskilledjobs-also tend to be caste-based.Members o fthe Scheduled Tribes (STs) are less beset bythis demarcation, since they were traditionally assigned a role outside the pale o f the caste system and since they for the most part own some land for subsistence agriculture. 'O Measures to improve the functioning of land and product markets have been discussed inchapters 8 and 9. 166 Figure10.1: Distributionof occupationsinIndia,by caste status, 1983-2000 Professional, Clerical Sales and service Agricultural Other manual Out of the labor technical, workers force/unemployed administrative Source: Das 2005. The differences inoccupationby caste are inpart related to differences ineducation levels, but this is only apart o fthe story. Differences inoccupation betweenSC/STs andother groups persist evenwhen "observationally equivalent" persons (Le., people with the same level o f education, inthe same region, etc.) are compared. Even controlling for such characteristics, SC/STs are muchmore likely to be incasual labor and less likely to be engaged innonfarm self- employment (Das 2005). Recent evidence on the low mobility inthe Indianlabor market also finds significant effects o f caste-basedoccupations (Munshi andRosenzweig 2005). Gender is another important axis o f stratification that has a bearing on labor market participation. InIndia, despiterobustgrowthrates, female laborforceparticipationrateshaveremained stubbornly low compared with other developing countries. This has happened despite rising education levels among women; in fact, ina patternpeculiar to India and Pakistan, education lowers the likelihood o f labor force participation by women. Two kinds o f explanations have beenarticulated for women's low labor force participationinIndia. First, family honor inmany parts o f Indiarests on women's restrictionto the home, thus affecting their ability to work outside the house (Chen 1995). These norms, however, tend to be lax among the poorest, who cannot afford the cultural trappings o f status. Hence, the majority o fwomen who do enter the labor force are those with weak options-uneducated women who take to casual labor. Second, there i s evidence to suggest that low participation rates o f educated women are caused not merely by cultural norms o f status and seclusion, but inlarge part by lack o flabor market opportunities for educated women (Kingdon and Unni 1997). A recent study shows that nearly a third of the women who do only domestic work state that they would like to be employed- primarily inregular part-timejobs-and the response does not vary by education level, suggesting that lack o f appropriate employment opportunities is likely a constraint (Das 2005). Apart from scarcity o f goodjobs, there i s also evidence o fwage "discrimination," which i s probably a further deterrent to female labor force participation. One study finds that among casual laborers, women get about halfthe wages o f men, and observed differences in 167 endowments, education levels, and demographic characteristics explain only 28 percent o f this differential inwages. Unobservedexplanations, such as type o fwork, skill not reflected in education levels, and outright discrimination against women inthe casual labor market, accounts for the rest o fthe wage gap (Das 2005). Other studies have also found similarly highlevels o f wage discrimination inthe urban labor market (Kingdon and Unni 1997). Inthe current context, job reservations are atool for promotinglabor market participationby those who are excluded. India has long had caste-basedreservations injobs that untilrecently have beenthe most coveted-regular salaried work inthe public sector. Preferential treatment for SC/STs (and more recently for the group identifiedas Other Backward Castes) inother areas (such as age relaxation, waiver o f application fees, etc.) and quotas inpublic employment works also aim to make labor markets more inclusive. While it is difficult to evaluate precisely the impact o f these affirmative action policies, there is some evidence to suggest that these policies are helpingSC/STs to overcome occupational barriers. Inregular salaried work-which i s still predominantly inthe public sector and where the reservation policy operates-there i s actually an advantage to SC/ST status inurban areas (Das 2005). However, while there are positive impacts o f some types o freservations, the mere extension o freservationto larger and larger spheres without accompanying reforms creates real risks o f freezing up rather than freeing opportunity. The same analysis finds that SC men suffer a disadvantage inregular salaried jobs, iftheyhavemorethanprimaryeducation. This appearsto indicatethat educational attainment of SC menhas risen over time, and the reservation policy, which seems to work well, also creates a system o f rationing ofjobs for SCs, exacerbating the existing shortage o f formal sector jobs. To make the playing field more evenacross castes and gender, the most significant reformthat Indiacan undertake i s to reform labor laws. As discussed inchapter 7, current labor market protections benefit the small share o fworkers inthe organized sector at the expense o f creating more andbetterjobs for workers outside this sector. The dualism created between formal sector jobs and the casual and self-employed workers facilitates discrimination, such as gender and castebias, inhiring. Facilitating access tofinancefor thepoor As with labor markets, credit markets show considerable evidence o fmarket failure. The vast majority o f India's rural poor still do not have access to formal finance. According to a recent Rural FinanceAccess survey (WAS 2003), 87 percent o f the poorest households surveyed (marginal farmers) do not have access to credit, and 71 percent do not have access to savings from a formal source. This has resultedin a heavy reliance among poor ruralhouseholds on informal finance, mostly moneylenders and shopkeepers, who charge exorbitant interest rates (World Bank 2004). Inthe rural sector, interest rates are high, but they are also variable. A survey o f six villages inKerala and Tamil Nadu found that, while the richpay a relatively low rate (33 percent), the poor pay rates o f 104percent and get only 8 percent o f the credit. There may bemany economic reasons for interest rates to vary across borrowers, including informational asymmetries, lack of collateral, etc. However, these market failures affect the poor disproportionately, inways unrelated to their investmentopportunities, thus leading to both greater inefficiency and the perpetuationo f inequalities. 168 There have been several initiatives to improve accessto finance for the rural poor, through centrally sponsored credit-orientedrural livelihood schemes-starting with the Antyodaya Program inthe early 1970s, followed by the Integrated Rural Development Program:* which was later revamped into the Swamjayanti Gram Swarozgar Yojana (SGSY) in 1999.These programs have typically not yieldedthe intended outcomes.72The most notable effort is the Self- help Groups BankLinkage model, the growtho fwhich-from just 500 self-help groups linked to banksinthe early 1990s,to over 700,000 by 2003-has been remarkable. Data from the WAS 2003 show that the bank linkage program appears to have targeted the rural poor effectively, reducing the vulnerability o f clients. One particularly successful scaling-up o fthe banklinkagemodelisthe IndiraKrantiPrathamproject inAndhra Pradesh (box 10.1). Box 10.1: The Indira KrantiPrathamProjectinAndhra Pradesh After a long gestation period, beginning with the Development o f Women and Children inRuralAreas groups in 1982-83, followed by the UNDevelopment Programme's South Asia Poverty Alleviation Project in 1995, w h c h was then scaled up across the state with the formation o f Velugu (recently renamedIndira KrantiPratham) in2000, this movement has experienced exponential growth. T o date approximately 6 million rural poor women have been organized into self-help groups and village organizations at the village level, Mandal Samakhyas at the block level, and Zilla Samakhyas at the district level. Buildingonthe growing capacity o fthese self-help groups andthe mandatory lending that National Bank for Agriculture andRural Developmentprovides, Velugu has made significant strides inlinking the self-help groups and their federations with the financial sector by increasing the access to commercial credit. Total annual credit to poor households and their groups has increased twelvefold from Rs 2.3 crore in2000 to Rs 27.6 crore in2005. The total credit flow from commercial banks to these groups since 2000 has grown to Rs 63.1 crore and i s expectedto reach 100crore by the time the project closes. As ths movement has matured, it has expandedinscope. Itnowprovides access to social safety nets andrisk management instruments (insurance products) and greater food security (rice credit lines). The project is also supporting livelihoods promotion and expansion inseveral key areas o f agriculture, horticulture, services, and agribusiness. Inrecent years, other models ofmicrofinance-modeled on the Grameen Bankmodel pioneered byMuhammadYunus inBangladesh-have also emerged inIndia, although the outreach o f these specialized microfinance institutions i s modest incomparison to the Self-help Group Bank Linkage and microfinance institutions elsewhere inthe world. The collective outreach (concentrated mainly inthe south) of Indianmicrofinance institutions is limited: inMarch2004, the microfinance sector hadloans outstanding o f about Rs 5 billion ($116million), reachingless than2 millionpeople, a tiny fraction ofIndia's poor.73Furthermore, Indianmicrofinance institutions tend to have a narrow scope, offering a limitedrange o f financial services beyond credit. Only a handful of institutions, such as VSSU (West Bengal), offer savings as a service. 71 The programwas rife with elite capture, poor loanrecovery, as well as nonproductive investments. 72 A large number o f self-help groups were formed under this scheme, but as o f 2003-04, only one inevery 50 self- help groups hadtaken on economic activities. 73 Incontrast, MFIsinBangladeshare estimatedto reachmore than 60 percent ofthe poor, withthe larger programs such as Grameen, BRAC, and A S A all reaching over two million clients, each. Grameen Bank's loanportfolio alone exceeds that o f the entire microfmance sector inIndiaby a factor o f over two whereas bothBRAC's and ASA's portfolio is more than one and a half times that of all microfinance institutions inIndia. 169 Few provide insurance, and only the top three inIndia offer a composite set o f services to their customers. The limited scale and scope o fIndianmicrofinance institutions, relative to the giants in Indonesia and Bangladesh, reflects, at least inpart, the absence o f an enablingpolicy and a legal andregulatory framework, which limits the ability o fmicrofinance institutionsto mobilize member deposits, trade inequity, andraise debt from external sources. Microfinance institutions are also constrainedbythe lack o f adequate capacity and skills infinancial control and management, information systems, new product design, etc. Furthermore, since most microfinance institutions inIndia lendto self-help groups, this means that microfinance institutions inIndia are constrained by many o fthe same factors that have heldback the outreach andscale o fthe Self-help Group BankLinkage. Inparticular, capacity, time, and cost issues related to group formation have posed constraints. Strengthening the microfinance and self-help sector can, inthe short term, serve as a quick way to deliver finance. But the medium-term strategy to scale up access to finance for the poor should be to "graduate" microfinance clients to formal finance institutions where they can access standard individual loans, possibly on a fully commercial basis. Efforts to promote microfinance should go handinhandwith efforts to make the formal sector better at "banking the poor," and boththe government and the private sector canplay a critical role inthis context. A better dealfor tribals. Notwithstanding the provisions o freservation, affirmative action, and a multitude o f schemes designedto enhance their economic and social status, tribals are among the most marginalized groups inthe country. Tribal populations, constituting 8 percent o f the population, suffer from geographic and sociocultural exclusion, as they inhabit relatively underdeveloped, remote, sparsely settled areas with inadequate access to basic amenities (roads, communication, education facilities, and drinkingwater). Their lack o frepresentationand powerlessness result inan inability to negotiate with the state. The central factors affecting tribal livelihoods are secure access to and control over natural resources. Large numbers of tribals, who live inand around forest areas, have depended upon forests for their livelihoods for generations. However, their customary rights over land and forest produce have not been adequately recognized and recorded at the time o freservationo f these areas. As a result, they continue to be treated as encroachers inforest records, and their vulnerability is hrther exacerbatedbyperceived conflict with conservation priorities and displacement by development projects (mining, dams, etc.). One important policyto redress inequities i s the Panchayats Extension to Schedule Areas (PESA) Act, 1996.The act formally recognizes the tribal traditional system as the basic unit o f self-governance, by empowering tribals to redefine their own administrative boundaries, and theirvillage council (Gram Sabha) becomingtheir core institution. Inaddition, PESA gives rights to the ownership o fnontimber forest produce to the respective Gram Sabhas. However, the envisaged transfer of decision-making power to the traditional Gram Sabhas has not been reflected inthe state laws. Similarly, tribals have not yet gained rights to nontimber forest produce and corresponding benefits because o f uneven implementation o fthe act, inconsistencies betweenPESA and state government provisions, and continued state control over collection and trade o f key nationalized high-value nontimber forest produce. 170 A more recent effort, based onthe assurances giveninthe Common MinimumProgramme, i s legislation to recognize and vest the unrecorded forest rights and occupation inforest land in forest-dwelling Scheduled Tribes and to provide for a framework for recording the forest rights. The draft billhas attracted broad support and also intense criticism (particularly from some conservation activists). The central issues o fthe debate relate to reconciling socialjustice concerns with ecological concerns (particularly relatedto parks, sanctuaries, andbiodiversity hot spots) and to implementingthe bill, given the doubts over the institutional capacity and commitment o f the implementingagencies. One set o f critics also arguethat, while resolving socialjustice for tribals, the rights o f forest-dwelling nontribal poor communities are not addressedbythe proposedbill. While the bill represents a progressive step inrestoring social justice, at least partly, its success will depend largely on resolving implementationconstraints. Unless the governance system (including institutional issues and accountability mechanisms) i s addressed along the lines envisaged by PESA, affirmative action or other policies alone will not produce desired outcomes on the ground. What needs priority attention i s developing a tribal development framework and strategy that addresses empowerment, service delivery, and livelihood dimensions ina holistic manner, with a clear focus on implementation. Promoting rural livelihood and empowerment. The government o f Indiaand state governments are simultaneouslyworking from the bottom up by implementing a number o f rural livelihood initiatives to help ensure inclusive agricultural and rural growth. These initiatives focus on three key priorities: (a) economic support o f livelihood activities that could increase incomes and reduce vulnerability; (b) strengthening the delivery o fbasic services, such as drinkingwater supply, education andhealth services, and access to rural infrastructure (rural roads, minor irrigation); and (c) empowerment, to enable the poorest to manage their resources well and benefit from economic opportunities. For example, the Madhya Pradesh District Poverty Initiative Project provides large grants to common interest groups o f menand women federated byeconomic activity. The groups undertake avariety o f activities, ranging from income generation (animal husbandry, service sector and village industries) and development of village infrastructure (e.g., water storage facilities) to land-based activities (e.g., land leveling, investment intubewells, etc.). The private sector has also launched a number o f initiatives aimed at expanding credit and linkingrural people more effectively to input and output markets (box 10.2). These initiatives show that there is no intrinsic contradiction betweencommercial viability and fair access for the poor. With concerted efforts at mobilization, some types o f livelihood interventions, such as the credit-oriented self-help groups, havebeen found to have significant impact on the confidence, communication skills, and decision-making abilities o f women (table 10.1). 171 E x 10.2: Examples of private sector initiatives to improve rural livelihood ICICI bankingservices for the poor T o address the ruralpoor's unequal access to finance, ICICI Bank has ledmultiple initiatives to provide banking services at an affordable cost to the poor. ICICI has partnered with others to colocate automatic teller machines with rural Internet kiosks and explore Smartcard technology to provide secure, low-cost transactions and loan management. More important, ICICIBank has created a network o f 8,000 self-help groups, covering about 160,000 women, to serve as the vebcle for creating successful, microfinanced businesses. Inthe process, ICICI Bank has given these women the means to transform their social and economic lives, their families, and their villages. ITC e-Choupal I T C e-Choupal illustrates how improvements intechnology and communications infrastructure canbe good for both equity and efficiency inproduct markets. ITC e-Choupal today reaches out to and empowers over 3.5 million farmers in31,000 villages by enabling them to readily access crop-specific, customized, and comprehensive information intheir local language. Vernacular websites provide real-time information to even the smallest marginal farmers on the prevailing Indianand international prices and price trends for their crop, expert knowledge on best fanning practices, and microlevel weather forecast. This significantly improves the farmers' decision-making ability, thereby helping thembetter align their agricultural produce to market demand and ensure better quality, productivity, and price discovery. The ITC e-Choupal model has been specifically designed to tackle the challenges posedby the unique features o f Indian agriculture, which is characterized by fragmented farms, weak infrastructure, information asymmetry, and numerous intermediaries. Over the next decade, the ITC e-Choupal network aims to cover more than 100,000 Indian villages, representing a sixth o f rural India, and to create more than 10million e- farmers. Table 10.1: Impact of participation in self-help groups on women's attitudes and behavior 172 Indicator of change Before Afler participation participation in selfhelp in selfhelp group group (percent) (percent) Self confidence and self-worth Exudes confidence 21 78 Can confidently meet financial crisis 33 85 Gets respectful treatment from family 40 89 Comes out to help neighborslothers 51 95 Decision-making Makesjoint decisions on purchase o f 39 74 household assets Makesjoint decisions on social 42 69 matters, such as education of children and marriage Communication skills Speaks out freely 23 65 Talks only ifasked 40 9 Behavior changes Protests drinkingand gambling 37 81 Protests wife-beating by husbands 52 78 Suffers domestic violence 67 49 Has increased mobility 45 75 Source: Prahlad 2005. These efforts are an important tactic within an overall growth strategy. Policies to create a thrivingrural economy also requirecomplementary approaches to ensure inclusion o f all into that thriving economy-neither can fully succeed without the other. For rural livelihood interventions to be successful, their design andmode o f empowerment needs to be tailored to local conditions. Depending on the key constraints inan area, livelihood interventions can vary infocus, with some credit-oriented (e.g., self-help group, SGSY), asset-oriented (e.g., Madhya Pradesh District Poverty Initiative Project), and market-oriented (ITE e-Choupal) programs. Successes inone state may not necessarily beperfectly transferable to others, because local conditions (natural, institutional, human, and cultural endowments) matter. Ultimately, the goals o fthe program will determine the optimal approach taken, interms o fwho finances and who delivers the service (figure 10.2). 173 Figure 10.2: Alternative paths for scaling up rural livelihood initiatives, depending on the objective and mode of empowerment r h RuralLivelihoods RWSS 1 NRFGI Bharat Nirman Line ( Stronger -\ Ministrv Local Government Singlesector Program Community organization Small enterprise e.g. Cwperative formation e g Formationof SHGr Market Agriculture NABARD Strategy IClCI Providing social protection The social protection system canplay a role not only inmitigating poverty andinequality through redistribution, but also inhelping address distortions that limit the opportunities for the poor to participate more fully ingrowth. Incontrast to earlier thinkingthat assumed a tradeoff between equity and growth, recent research indicates that well-designed andwell-implemented social protection systems can enhance current and future opportunities by mitigating failures in credit, insurance, andother markets that affect the poor most strongly.74 They can also play a role infacilitating efficiency-enhancing reforms. Thus, social protection can have equity and dynamic efficiency functions. Indiahas a long tradition o fprograms that have attempted to mitigate chronic poverty, most notably the Public Distribution System, but also housing programs for the poor; categorical cash transfers to the destitute elderly, widows, and people with disabilities; and stipends and other programs targeted to SC/STS.~~There have also been manypromotional programs that attempted to smooth or raise permanent income, includingpublic works schemes, subsidized 14See World Bank (2005a).See Townsend (1994); Morduch (1995,2003);Ravallionand Chaudhuri (1997), and MunshiandRosenzweig (2005) for evidence oninformal support networks inpromoting consumption smoothing. ''Programsto support the elderly are likelyto become increasingly important over time as the population ages and, as already evident, traditional co-residence patterns become less common. 174 credit for the rural poor, and school feedingprograms. This section reviews the extent to which performance has achievedobjectives o fredistribution, protection, or opportunity enhancement. India appears to have entered aphase o f expansion insocial protection spending, driveninlarge measure by a shift toward greater central financing. Spending on core antipoverty programs has expanded inmost recent years inreal terms (andas a share o f total spending inGDP). The expansion has been drivenprimarilyby an increase inpublic works programs (which reversed a decade-long real decline), as well as the scheme to provide midday meals. While it i s unlikely that the rate o f expansionwill continue, legal commitments (such as the rural employment guarantee) and initiatives (such as the expansion o f social security to the unorganized sector) suggest that reversal i s unlikely. Institutional arrangements for social protectionare unusually complex inIndia. This is partly driven by the multiplicity o fprograms and partlyby the multitiered structure o f governance. Despite the commitment o fthe government o f Indiato rationalize programs,76 and some progress inthat respect (e.g., consolidationofmultipleprogramsinto SGSY), institutionalresponsibility for different social protection efforts is fragmented. The political economy o fprogram reform may also militate against consolidation. The situation is made more complex bythe increased role for PRIs inimplementation, which brings inthe new challenges o fbuildingcapacity, delineatingresponsibilities, and increasinglocal flexibility inprogram design. The impact o fthe major social protectionprograms inachieving redistribution andprotection objectives appears relatively limited. With the exception o f the Public Distribution System, coverage o f the main social protection programs i s very low, suggestingrather low redistribution (figure 10.3). Evenwithinthe Public Distribution System, coverage is still surprisingly low on core food items.At the household level nationally, targetingo f public works is mildly progressive, and targeted credit is neutral. However, the variation across states i s significant, with no clear pattern by income levels. Targeting performance is not particularly impressiveby international standards, and overall the mitigation o fpoverty i s likely to be Figure 10.3: Coverage and targeting of major social protection programs Share of total population, IRDP, and public works beneficiaries Coverageof mainwelfare programs -- 100.0% in HHs below Rs 400 PC expenditure, 1999.00 I 25 80 0% 2E I 20 60 0% 15 2m 40 0% I 10 s 5 20 0% 0 00% Publcworks Targeted TPDS rice TPDS -food . credit or wheat 76 See the TenthFive-Year Plan. 77 Coady, Grosh, and Hoddinott (2004) for targeting evidence from developing country transfer programs. 175 Notes: TPDS refers to TargetedPublicDistribution System. IRDP refers to IntegratedRuralDevelopmentProject. Source: Public Distribution System, 5gthNSS; credit andworks, 55thNSS. Bank staff estimates. Administrative data indicatethat differentials inimplementation across states have lessened the redistributiveimpact of social protectionprograms (figure 10.4). For the Public Distribution System, most poor states are notably below national averages inboth total offtake and offtake o f Below Poverty Line/Antyodaya Anna Yojana grains. The ratio between the grains offtake and total offtake is also generally lowest inthe poorest states. The picture for public works is more mixed.Inall states, employment generationhas beenwell below policy targets. Across states, some o fthe poorest states (e.g., Bihar andJharkhand) have achieved evenmore marginal employment generation, while others such as Orissa have performedbetter than average. Figure 10.4: Differencesinimplementationof social protectionprograms,by state PDS: total and BPUAAY offtake, 2002-05 SGRY workdays peragncultural worker and per BPL HH, 2003- 100 04 7 Another factor driving targeting outcomes is the Below Poverty Line (BPL) system, which inits most recent version i s a form o fproxy means test and i s required for access to various schemes. Despite its importance, the BPL current andprevious methodology has beenpersuasively criticizedY7*and claims o f abuse, including inflation o fBPLrolls and exclusion o fvery poor households, have been widespread. The claims appear to be supported by evidence on estimated BPLhouseholds andissued BPLrationcards. While the two figuresmatch at the national level, they often do not match inindividual states, where both over- and undercoverage are pronounced (table 10.2). Table 10.2: Implementationof the BPL targetingsystem,by state, 2000 State BPL BPL Ratio households ration B/A (A) cards (B) All-India 652.0 655.5 1.o AP 40.6 128.1 3.2 Karnataka 31.3 63.6 2.0 Orissa 33.0 43.4 1.3 UP 106.8 74.4 0.7 Rajasthan 24.3 18.6 0.8 M P 41.3 42.6 1.o Source: Indiastat.comfrom Department o f Food andPublic Distribution. 78See Sundaram 2003. 176 While social protection spending i s significant, the government o f India's evaluations indicate that spending has not beencost-effective inpromoting social protection for the poor, inpart because o f rigid program designs and inpart because o fmajor implementation problems driven byineffective accountability mechanism^.^^ Major issues inimplementationinclude (a) poor absorptive capacity, particularly inpoorer states; (b) significant leakage o f funds and food grains (e.g., 58 percent o f Public Distribution System grains do not reach BPL households, accordingto the PlanningCommission); (c) lack of public information for the target population on scheme eligibility and fundusage; and (d) weak monitoring and evaluation systems. Increasedsocial protection spending therefore risks promoting greater inefficiency inresource use and exacerbating governance problems, unless a range o fwell-known design andimplementation issues canbe addressed. A further factor inassessingthe structure o fsocialprotection spending is the major role of food intransfer programs. The PublicDistribution Systemremains the largest program, with workfare programs relyingsignificantly on food (typically, 75 percent o f wages). The midday meal program by its nature is food-based. Such reliance on food-based transfers is unusually strong in South Asia, while other regions andother lower-income developing countries have relied more on cash-based systems." Untiedcash transfers to the poor remain remarkably limitedinIndia. TheNational Rural Employment Guarantee Act (NREGA) may improve the balancebetween food andcash, ifproposals eventuate to make the center-to-state transfer incash. There is also a proposal within the Ministryo fFinance to experiment with greater use o f cash inthe Public Distribution System. However, to date the heavy reliance on food transfers limits options for poor households inthe range of opportunities that transfers provide. Interventions that enhance opportunities for the poor have operated inIndia for some time, but have been insignificant inspending terms andor beset by implementation problems that have dilutedtheir potential impact. Promotional programs include social insurance, targetedcredit, andto some extent public works. Social insurance spendinginIndiais concentrated almost entirelyinthe organized sector (andwithinthat, the public sector). Spending is dominated by civil service pensions, which accounted for 2.1 percent o f GDP in2003-04, with one of the fastest rates o f spendinggrowth inrecent years.'* As a result, coverage o f social insurance remains very low, with the exception o f life insurance (table 10.3), and its contribution to risk management for the population remains minimal. There have been several attempts at the national level to promote coverage of social insurance inthe unorganized sector, but they have failed to achieve significant penetration.82A number o f states have also introduced welfare funds, and insurance schemes through nongovemmentalorganizations are proliferating (e.g., SEWA; Yeshasvini). Welfare funds have insome states achieved significant coverage, but there are concerns about weak financial protection, financial viability, and high administrative costs. Community-based schemes incontrast are more focused, but face challenges o f scale. The 79 See Governmentof India evaluations of SGSY (CAG, 2003), TPDS (Planning Commission, 2005a), EAS (Planning Commission, 2002). See also Radhakrishnaet a1(2004); Dev et al. (2001); Nayak, Saxena, andFarrington (2002). 8o See Coady, Grosh, andHoddinott (2004), which found that simple cashtransfers accountedfor 27 percentof programsinthe poorestcountries, but 54 percentinhigher-income developing countries. 81A DCpensionscheme has startedinseveralstates andnational legislation is proposedfor early 2006: This could contribute to a declining civil servicepensiodGDP ratio over time. 82 See ILO (2003); Dev et al. (2001); O'Keefe (2005). 177 government o f India i s again proposing a major social insurance program for the unorganized sector, covering health, life, maternity, disability, and old age insurance. While the draft proposals address issues of intermediation and transactions costs ignoredby previous efforts, they seem very ambitious inlight of Indian and developing country experience. The benefit o fpublic works programs inenhancing opportunities for the poor has been largely disappointing, with notable exceptions, such as the Maharashtra Employment Guarantee System inthe 1970sand 1980s.Public works programs were hopedto provideimplicit insurance fbnctions for the poor, to create assets that could boost economic performance, and to encourage gender equality bypromoting opportunities for women. Most o fthese benefits either have not beenrealized (e.g., women's share inpublic works is less than their overall rural labor force share) or have not beenmeasured because o fpoor evaluationmethods (e.g., the economic impact of assets created; insurance functions). The National Rural Employment Guarantee Act offers an opportunity to achieve better outcomes (andmeasure them), butprevious experience also suggests major risks (box 10.3). Subsidized credit for the poor, primarily through the Integrated Rural Development Program inthe 1990sand subsequently through SGSY,has also failed to achieve significant penetrationbecause o f a range o f supply and demand problems. Table 10.3: Coverage rates of social insurance in India's organized and unorganized sectors, 2004 Social insuranceprogram Organizedsector Unorganizedsector coverage coverage (percent) (percent) Public schemes Employees' Provident Fund 25.1 0.18 Employees Pension Scheme 12.2 0.02 Government Pension Scheme 48.7 0.24 Government Provident Fund 54.0 0.21 Contributory Provident Fund 4.0 0.02 Commercial schemes Life insurance (endowment) 54 23 Personal accident insurance 3.6 1.2 Private health insurance 2.0 0.5 Nonlife general insurance 2.8 1.4. Source:ADBiMOF survey 2004. Bank staff estimates. Box 10.3:The National RuralEmployment Guarantee Act The National Rural Employment Guarantee Act (NREGA) i s the government o f India's most ambitious public works initiative, guaranteeing every rural household at least 100days employment per year at the agricultural minimumwage, subject to a nationally mandated minimumo f 60 Rsiday. Coverage would initially be confined to 200 backward districts, with nationwide rural coverage within five years. Analysis to date suggests a mixedpicture on costs and benefits. First, there i s significant potential for a lean season Employment Guarantee System (EGS). Simulations o f a nationwide 100- day EGS suggest (Murgai and Ravallion, 2005): The lean seasonrural poverty rate could be reduced from37 percent to around 23 percent, or to around 30 percent o n a yearly basis. 0 The fiscal cost would be around 1.7 percent of GDP annually, if implementednationally. The gains should be progressive, with the poorest quintiles accounting for 29 percent (and the 178 richest quintiles for 10percent) o fparticipants, and gains from EGS coming to 51 percent of pre- EGS consumption levels for the poorest quintiles (7 percent for the richest). The bulk o f expected participants would be casual laborers. However, there are design issues that raise concerns o f effectiveness, efficiency, andpotentially equity. They include: Usingstate agriculturalminimumwage rates as the scheme wage rate is likely to be problematic. Scheme rates above market rates typically result inrationing, which experience from Maharashtra EGS confirms applies also to "guarantees." The direct transfer impact of EGS is significantly less than an untargeted transfer, because of the opportunity costs o fparticipation. An untargeted transfer usingthe wage portion o f EGS only would reduce poverty to around 15 percent-as against 23 percent from EGS. While the comparison is imperfect, because it measures only the transfer impact o f EGS and not other economic impacts, nontransfer gains from EGS would need to be substantial to dominate an untargeted transfer. The point highlights the importance o f acheving significant economic returns to EGS assets and for the poor to capture a reasonable share o f the gains. A shortcoming o f previous public works programs inIndia (and internationally) has been the absence o f evaluations o f the economic impact o f assets. A significant improvement o fNREGA over previous works schemes is the strengthenedrole for panchayats in design, implementation, and monitoring. However, it will be important to develop accountability mechanisms w h c h avoid the bundlingo f functions by specific actors (inparticular DRDAs),which has contributed to implementationproblems inprevious schemes. Itremains to be seen ifthe incentive and accountability structure ensures that GPs are at the heart o fNREGA implementation. Inaddition, the act offers limited guidance on funds flow mechanisms. IfGPs are to be empowered, it would be important for them to have direct control over a greater portion o f scheme hnds than currently stipulated. The scheme benefits from a stronger monitoring and evaluation system than previous works schemes, including earmarked funds. To expand the system o f concurrent evaluation to include robust impact evaluation, it will be critical to collect good baseline data before beginning the scheme. Looking ahead to a more coherent social protection system What i s the desirable trajectory o f social protection systems inIndia over the coming 10-20 years? Three lessons are clear from India's experience. First, the program mix between equity and opportunity-enhancinginterventions for the poor should become more balancedto maximize both redistributional and dynamic efficiency objectives. There should also be flexibility for states to tailor programs to their conditions (e.g., food-based schemes are less neededinsome states, where cash-based transfer options could be considered). Such an approach could intime evolve into a broader, conditional block grant system. Second, even inprograms with a strong equity focus-which will remaincrucial-policy reforms are needed and accountability mechanisms need major improvement, ifthey are to realize their redistributional objectives more effectively, particularly inseveral poorer states. Third, the mass expansion o f contributory social insurance requires a sequenced strategy that takes account o f country and state-level conditions and avoids the past experience o fpublic schemes with high"announcement value," but limitedimpact. Differences across states inthe pace of demographic transition, presence o f intermediary organizations for unorganized workers, and other factors suggest that the states have very different potential for broad-scale insurance programs inthe unorganizedsector. 179 Improving safety nets Because amajor expansion o f social assistance spending inIndiai s unlikely, given India's current commitments and trends inother countries,83it i s thus all the more important for Indiato bemore efficient inthe way it spends safety net funds. The improvements neededto increase accountability inprogram delivery have beenreviewed innumerous assessmentsby the government o f India and others. Specific, high-priority actions include: Greater accountability. There i s a need to consider where different elements o f responsibility for implementationshould lie within the system. This will ina number o f cases require "unbundling" implementationfunctions to exploit the comparative advantage o f different levels o f administration andlocal bodiesp4 e More innovation andflexibility at the local level. Accountability improvements should be accompanied bypolicy adjustments to broaden the scope for innovation inthe Public Distribution System (e.g., vouchers, increased community involvement indistribution, decentralized procurement, andpossibly cash-based transfer options). e Better targeting. There is a need to refine targeting mechanisms inprograms that do not rely on the BPL system (e.g., community-basedtargeting, well-grounded proxymeans testing, and categorical targeting where appropriate). e More cooperation with local groups to benefit thepoor. Inimproving access to credit for the poor, the government could collaborate with local organizations by, for example, permittingstronger roles for nongovernmentalorganizations and self-help groups in preparingpoor people and groups to access credit programs. e Better implementation of public worksprograms. Workfare can play a potentially important role as an insurance-type intervention, ifit i s well designed and well implemented.8sHowever, the NREGA and SGRY will need to overcome a range o f implementation problems that have plaguedworks schemes. The most critical i s enforcing a credible accountability structure at all stages o f implementation. There are also a number o f design issues inNREGAthat raise concerns o f effectiveness, efficiency, andpotentially equity. e More links with human capital development. It may be worth linking existing social assistancetransfers to human capital development inthe form o f conditional-or partly conditional-transfers. While the supply-side issues are substantial, experience in 83Even in OECD and transition countries, social assistance spending averages only around 2 percent of GDP, a level that India will approach as the NREGA is rolled out. 84See World Bank (forthcoming(b)) for recommendations on the appropriate functions by administration and PRIs at different levels inthe context o f public works. *'Walker and Ryan (1990); Cain and Lieberman (1983) o n South Asia; Pritchett and others (2002) on Indonesia. 180 education inBangladesh and more recently Punjab inPakistan suggeststhat there may be a role for demand-side interventions, even ina difficult supply environment. Greater reliance on cash transfers. More proactive efforts to expandthe relative role o f cash rather than food insocial protection programs seem overdue. Suchreforms are linkedto reforms inthe public procurement system, but seemwarrantedbothbythe evidence o f inefficiencies and governance problems inthe use o f food transfers and by the likely empowerment benefits o f allowing poor households greater choice inhow they use transfer income. More attention to monitoring and evaluation. Monitoring and evaluation o f social protection programs needs to move beyond the practice of concurrent evaluations to include several additional tools: (a) impact evaluations based on timely surveybaselines that address a broader range o fimpacts thanthe input/output focus o f most concurrent evaluations; (b) periodic use ofprocess evaluation tools, such as public expenditure tracking surveys, for programswhere leakage is substantial; (c) effective implementation of social audits, which are allowed for insome schemes but rarelyused inpractice (though effective when used, as Rajasthan's experience demonstrates); and (d) experimentationwith tools such as citizens' report cards, which have proved valuable in states such as Karnataka. The new Right to Information Act should inprinciple facilitate social audits. Expanding social security coverage Expansiono f social insurance coverage is closely linkedto the level o f formality inthe labor market. Thus, reforms are needed outside the social protection system inorder to create incentives for employers and workers to participate inthe formal sector. Specific, high-priority actions include: Wider coverage of thepension system. Roll-out o fthe proposed DC pension system seems desirable on several fronts. Fiscally, it will gradually create space for social protection interventions covering a wider population. Administratively, it may provide a platform into which existing unorganized sector initiatives may "plug in" through group insurance. In equity terms, it should reduce the current public subsidy to a relatively privileged group. More evaluation of existing initiatives. Efforts to expand social security to the unorganized sector shouldbe preceded by a comprehensive evaluation o f existing initiatives, including public, private, and nongovernmentalefforts. This should focus on features that are common to successful initiatives (such as effective intermediaries betweenworkers and insurers) and those common to failing or unsustainable schemes (such as hightransactions costs, and benefits with no actuarial linkage to contributions). This would inform a realistic strategy for gradual expansion o f social insurance coverage. This is likely to focus first on the "organizable" sector (Le., workers for whom intermediary organizations exist) and on forms of insurance such as life and disability, where moral hazard is lower, administration less complex, and premiums less costly. 181 More consideration of noncontributory pension schemes. Giventhe challenges of expanding contributory social insurance to the unorganizedsector, rigorous evaluation o f social pension schemes should be undertakenwith an eye to a possibly expanded role for suchbenefits as a noncontributory pensionsystem. A number of developing countries have implemented such schemes as a steppingstone inbroadening social security coverage, where labor market and institutional conditions have made contributory systems challenging.86 86See Palacios (forthcoming). Brazil, SouthAfrica, Bolivia, andNepalprovide examples o f this approach. 182 References Acharya, S. 2002. "India: Crisis, Reforms and Growth inthe Nineties." Working Paper 139, Center for Researchon Economic Development andPolicy Reform, Stanford University. Ahluwahlia, M. S. 2000. "Economic Performanceof States inPost-Reforms Period." Economic and Political Weekly,May 6, pp. 1637-48. Ahluwahlia, M.S. 2002. "India's Vulnerabilityto External Crisis: An Assessment." inM.S. Ahluwahlia, S.S. Tarapore andY.V. Reddy(eds.) Macroeconomics and Monetary Policy: Issuesfor a Reforming Economy. Essays in Honor of C.Rangarajan, Oxford University Press. Ahsan, A. and C. Pages. 2005. "Helping or HurtingWorkers? Assessing the Effects of dejure anddefacto Labor Regulation inIndia." BackgroundPaper for forthcoming World Bank Report, IndiaLabor andEmploymentStudy. Alesina, A andR.Perotti. 1995. "Fiscal Expansion andFiscal Adjustments inOECD Countries." Economic Policy, Vol. 21, pp. 205-48. Alesina, A. and S Ardagna. 1998. "Tales of Fiscal Adjustment." Economical Policy: A European Forum, No. 27, pp. 487-546. Alsop, R.and B.Kurey. 2005. "Local Organizations andDecentralization." inDecentralized Development: Their Functions and Performance in India, Directions inDevelopment Series, World Bank. Aten, B.,A. Heston, andR.Summers. 2001. PennWorld Tables 6.0. Center for International andInterareaComparisons, UniversityofPennsylvania. Banerjee, A., E. Duflo, and A. Deaton. 2004a. "Health Care Delivery inRuralRajasthan." Economic and Political Weekly,February 28, pp. 944-50. Banerjee, A, S. Cole, E. Duflo, and L.Linden. 2004b. "Remedying Education: Evidence from Two Randomized Experiments inIndia." Mimeo. Banerjee, A. andT. Piketty. 2005. "Are the Rich GrowingRicher? Evidence from IndianTax Data." InA. Deaton and V. Kozel (eds.) The Great Indian Poverty Debate. Macmillan India:New Delhi. Basu, P. 2005a. India 's Financial Sector: Recent Reforms Future Challenges. Macmillan India: New Delhi. Basu, P. 2005b. "A Financial System for the Poor." Economic and Political Weekly, September 10, pp.4008-4012. 183 Basu, P. andP. Snvastava. 2005. "Scaling-up Microfinance for India's RuralPoor." Policy Research Working Paper 3646, World Bank. Besley, T. and R. Burgess. 2004. "Can Labor Regulation Hinder Economic Performance? Evidence from India" Quarterly Journal of Economics, 119(1) pp. 91-134. Bourguignon, F. 2004. "The Poverty Growth InequalityTriangle." Paperpresentedat the Indian Council for Research on InternationalEconomic Relations, New Delhi, February 4. Cain, M.and S.S. Liebeman. 1983."Development Projects and Projects for Fertility Decline in Bangladesh." Bangladesh Development Studies 11(3), pp.1-38 Carey, D and J. Rabesona.2002. "Tax Ratios on Labor and Capital Income and on Consumption." OECDEconomic Studies 35(2), pp. 129-174. Centre for Monitoringthe IndianEconomy.2002. IndianHarvest. Mumbai:CMIE. Centre for Monitoringthe IndianEconomy. 2004. IndianHarvest. Mumbai: CMIE. Center for Monitoring the IndianEconomy. 2005. Studyon vertical share inWest Bengal. Background paper for forthcoming World Bank report "India: Rural Governments and Service Delivery". Chattopadhyay, R. andE. Duflo. 2004. "The Impact ofReservation inthe PanchayatiRaj: Evidence from aNationwide Randomized Experiment." Economic and Political Weekly, February 28, pp. 979-986. Chaudhury, N.andJ. Hammer. 2005. "Oh No! Not Another Paper on IndianPoverty inthe 1990s?An Examinationwith Data fromthe National FamilyHealth Survey." Mimeo. Chen, M. 1995. "A Matter of Survival: Women's Right to Employment inIndiaand Bangladesh." inM.Nussbaumand J. Glover (eds) Women, Culture and Development. Oxford: Clarendon Press. Chen, S. and M.Ravallion. 2004. "How have the World's PoorestFaredsince the Early 1980s?" Policy ResearchWorking Paper 3341, World Bank. Coady, D., M.Grosh, and J. Hoddinott. 2004. "Targeting o f Transfers inDeveloping Countries: ReviewofLessons and Experience." World Bank, Mimeo. Central Public Health & Environmental Engineering Organization. 2000. Solid Waste ManagementManual. Das, J. and J. Hammer. 2004a. "Strained Mercy: Quality o fMedical Care inDelhi." Economic and Political Weekly, February 28, pp. 951-64. 184 Das, J. and J. Hammer. 2004b. "Money for Nothing: The Dire Straits o f MedicalPractice in India." PolicyResearch Working Paper 3269, World Bank. Das, J., P. Gertler andK.Leonard. 2005. Presentationat CEPR basedon data from ISERDD. October 2005. Das, M.B.2005. "Do Traditional Axes o fExclusion Affect Labor Market Outcomes inIndia?" Background Paper for forthcoming World Bank Report, IndiaLabor and Employment Study. Deaton, A. andV. Kozel. 2005. The Great Indian Poverty Debate .Macmillan India: New Delhi. Deaton, A., J. Freidman, andV. Alatas. 2005. "Purchasing Power ParityExchange Rates from Household SurveyData: India and Indonesia." Mimeo. Dev, M.,P. Antony, V. GayathriandR.P. Mamgain. 2001. "Social and Economic Security in India." Institute for HumanDevelopment, Delhi. Duflo, E. andR. Hanna. 2005. "Monitoring Works: Getting Teachersto come to School." Department of Economics and Poverty Action Lab, Massachusetts Institute of Technology, Mimeo. Filmer,D., E.King, andL.Pritchett. 1998. "Gender disparity inSouthAsia: Comparisons between andwithin Countries." Policy ResearchWorking Paper 1867, World Bank. Forbes, N.2002. "Doing Business inIndia:What has Liberalization Changed?" InA.O. Krueger (ed) Economic Policy Reforms and theIndian Economy, University of Chicago Press. Galiani, S., P. Gertler, andE. Schargrodsky. 2005. "Water for Life: Impact of Privatization of Water Services on Child Mortality." Journal of Political economy, 113(1), pp. 83-120. Glinskaya, E. andM.Lokshin. 2005. "Wage Differentials betweenthe Public and the Private Sectors inIndia." Policy Research Working Paper 3574, World Bank. Government of India. 2000. Report ofthe Expenditure Reforms Commission. Government of India. 2005. Report of the TwelfthFinance Commission. Government o f India, Comptroller-Auditor General o f India. 2003. Evaluation of Swamjayanti GramSwarozgarYojana Government o f India, Department o f Census and Statistics. 2001. Provisional Population Totals: India, Census of India2001. Government of India, Ministryof Finance. 2003. Economic Survey 2002-03. 185 Government of India,Ministryof Finance. 2004. Central Government Subsidies inIndia: A Report. Government of India, Ministry of Finance. 2005. Outcome Budget 2005-06: Outlays and Outcomes: Towards Achieving Better Results. Government of India, Ministryof PanchayatiRaj.2004. Task Force Report o f Officials in Charge of PanchayatiRaj inStates to Examine the Centrally Sponsored Schemes. Government o f India, Planning Commission. 10thFiveYear Plan. Government of India,Planning Commission. 1992. Vaidyanathan Committee: Report of the Committee on the Pricingof IrrigationWater. Government of India, Planning Commission. 2002. Studyof Employment Assurance Scheme. Government of India, Planning Commission. 2005a. Study o f Targeted Public Distribution System. Government of India, Planning Commission. 2005b. MidtermAppraisal of 10thFiveYear Plan (2002-2007). http://planningcommission.nic.in/midterm/midtermapp.html Government of Nagaland. 2004. Communitisation and Resurgenceof Naga Social Capital. Hagen, Von, J., H.Hallett and R. Strauch. 2001. "Budgetary Consolidation inEMU."Journal of Japanese and International Economics Vol. 16, pp. 512-535 Hausmann, R., L.Pritchett andD.Rodrik. 2005. "Growth accelerations." Working Paper 10566, National Bureau of Economic Research. Hausmann, R., J. Hwang, and D.Rodrik. 2005. "It is not how muchbut what you export that matters." Mimeo. Hausmann,R., D.Rodrik andA. Velasco. 2005. "Growth Diagnostics." Harvard University. Mimeo, Postedat http:://ks~home.harvard.edu/-drodrik/barcelonafinalmarch2005 .pdf. Hoff, K andP. Pandey. 2004. "Belief Systems andDurableInequalities: An Empirical Investigation of Indian Caste." Policy Research Working Paper Series 2875, World Bank. Howes, S. andM.Prasad.2005. "Reforms to The India's FederalTransfer and GrantRegime Proposedby the Twelfth Finance Commission: What Will They Mean for the States?" Mimeo, World Bank. 186 Howes, S. and R. Murgai. 2004. "Subsidies and Salaries: Issues inthe Restructuringo f Government Expenditure inIndia, "NIPFP-IMF Conference on Fiscal Policy inIndia, New Delhi. ILO. 2004. Report of the SecondNational Commission for Labor, Macroeconomic Reforms and a Labor Policy Frameworkfor India. IMF. 1977. The Mysteryofthe IndianGrowthTransition. IMFWorkingPaper 04/77. IMF.2004. Public Investment andFiscalPolicy. IMF.2005a. ForeignDirect InvestmentinIndia: How canitbeIncreased? India: Selected Issues. Country Report 05/87. IMF.2005b. The Tax System inIndia: CouldReform Spur Growth? India: Selected Issues. Country Report 05/87. IndiaInfrastructureReport. 1996. Policy Imperatives for Growth & Welfare. Report prepared bythe Expert Group on the Commercialization ofInfrastructure Projects. Jalan, B. 2004. TheFuture of India: Politics, Economics, and Governance. PenguinBooks: New Delhi. Jayal, N.2004. "Federal Anxieties, Democratic Desires: The Politics o f Governance Reformin Two IndianStates." www.cnsisstate.com. Jha, S., V. Rao, and M.Woolcock. 2005. "Governance inthe Gullies: Democratic Responsiveness and Leadership inDelhi's Slums." Policy Research Working Paper 3694, World Bank. Jin, Songqin, Klaus Deininger and DinaUmali-Deininger (forthcoming) "Does the farm-size productivity relationship change over time? Evidence from India." PolicyResearch Working Paper, World Bank, forthcoming. Kapur, D.andP.B.Mehta. (eds.) 2005. Public Institutions in India: Performance and Design. Oxford UniversityPress: New Delhi. Kingdon, G. andJ. Unni. 1997. "How muchdoes Education affect Women's Labor Market Outcomes inIndia?An Analysis usingNSS Household Data." Working PaperNo. 92, GujaratInstitute o fDevelopmentResearch. Komives, K.,V. Foster and J. Halpern. 2005. "Water, Electricity, and the Poor: Who Benefits from Utility Subsidies?" Mimeo, World Bank. Kremer, M.,K.Muralidharan, N.Chaudhury, J. Hammer andF.H. Rogers. 2004. "Teacher absence inIndia: A Snapshot." Mimeo, HarvardUniversity andWorld Bank. 187 Krishna, A. 2005. "Are Villagers Readyfor Decentralization." inL.C. Jain (ed) Decentralization and Local Governance,Orient Longman. Lall, S. andT. Mengistae. 2005. "The impact ofbusiness environment and economic geography on plant levelproductivity: an analysisof Indianindustry .',PolicyResearchWorking Paper 3664, World Bank. Lobo, A. and S. Balakrishnan. 2002. "Report Card on Service of Bhoomi Kiosks: As Assessment ofBenefitsbyUsers of the Computerized LandRecord System inKarnataka". Public Affairs Center, Bangalore, Mimeo. Mazumdar, D.2005. "Employment andEarnings inthe Formal Sector o f India." Background Paper for forthcoming World Bank Report, IndiaLabor andEmployment Study. McDermott, C. John and R.F.Wescott. 1996. "An EmpiricalAnalysis ofFiscal Adjustment." IMFStaff PapersNo. 43. . MehtaP. B. 2003. TheBurden of Democracy. Penguin: New Delhi. Milanovic, B. 2005. "Half aWorld: Regional InequalityinFive Great Federations." Policy ResearchWorking Paper, World Bank, forthcoming. Misra, S. 2005. "Delhi Water Supply & Sewerage Services: Coping Costs, Willingness to Pay and Affordability." Mimeo, South Asia Environment andInfi-astructureUnit,World Bank. Mohan, R. 2004. Finance for IndustrialGrowth.RBIBulletin, Reserve Bank of India, Mumbai. March. Morduch, J. 1995. Income Smoothing and Consumption Smoothing" Journal of Economic Perspectives 9(3), pp. 103 - 114. Morduch, J., 2003. "Consumption Smoothing Across Space: Tests for Village-Level Responses to Risk." In Stefan Dercon, ed., Insurance Against Poverty, Oxford UniversityPress. Munshi,K.andM.Rosenzweig. 2005. "Why is MobilityinIndiaso Low? SocialInsurance, Inequality and Growth." CID Working Paper No. 121, July. Murgai, R. and M.Ravallion. 2005. "Employment Guarantee inRural India: What Would it Cost andHow MuchWould it Reduce Poverty?" Economic and Political Weekly,July 30, pp. 3450-3455. Nagaraj, R. 2005. "Industrial GrowthinChina andIndia: A Preliminary Comparison." Economic and Political Weekly,May 21, pp.2163-2179. 188 Narain, A. 2005. "Participation Rates, Wage Trends, and Determinants." BackgroundPaper for forthcoming World Bank Report, IndiaLabor and EmploymentStudy. Narayana, D.2005. "Local Governancewithout Capacity Building: Ten years o fPanchayati Raj." Economic and Political Weekly,June, pp.2822-2832. Nayak, R., N.C. Saxena, and J. Farrington. 2002. "Reaching the Poor: The Influence ofPolicy andAdministrative Processeson the Implementation of Government Poverty Schemes in India." OverseasDevelopment Institute Working Paper 172. Noll, R. and S. Wallsten. 2005. Universal Telecommunications Service inIndia. AEI-Brookings Joint Center for Regulatory Studies. O'Keefe, P. 2005. "Strengthening Employment and Social Security for Unorganized Sector Workers inIndia." World Bank, Mimeo. Oommen, M.A. 2005. "Twelfth Finance Commission and Local Bodies." Economic and Political Weekly,40(20), May, pp. 2022-2026. Oxford Analytica, November 142005 Palacios, R.Forthcoming. "Expanding Social Security Coverage inthe Informal Sector for the Government of India." BriefingNote for the Government o f India, World Bank. Panagariya, A. 2004. "India inthe 1980sand 1990s:A Triumph o fReforms." ConferencePaper. Paul, S. 2006. "Public Spending, Outcomes, and Accountability: CitizenReport Card as a Catalyst for Public Action." Economic and Political Weekly,January 28, pp. 333-340. Paul, S., S. Balakrishna, K.Gopakumar, S. Sekhar, and M.Vivekananda. 2004. "State of India's Public Services: Benchmarks for the States." Economic and Political Weekly, February 28, pp. 920-34. Pinto, B., and FZahir. 2004. "Why Fiscal Adjustment Now." Economic and PoliticalWeekly, March 6, pp.1039-1048. Pinto, B., F.Zahir, andG. Pang. 2006. "Rising debt to rising growthinIndia: Microeconomic Dominance?" Powis, B. 2003. "Grassroots Politics and SecondWave ofDecentralizationinAndhra Pradesh." Economic andPoliticalWeekly, June 28. Prahlad, G.K. 2005. TheFortune at the Bottom of the Pyramid: Eradicating Poverty through Profits. Wharton School Publishing. Pratham, 2005. Annual Status of Education Report 2005. postedat www.pratham.org. 189 Pratichi Trust. 2002. ThePrutichi Education Report, Kolkata. Pritchett, L.,S. Sumarto, andA. Suryahadi. 2002. "Targeted Programsinan Economic Crisis: EmpiricalFindings from Indonesia's Experience." Harvard University. Public Affairs Center. 2002. Benchmark Surveyon Service Standards and Constraints Encountered by Entrepreneurswith IndustrialRegulations inKarnataka. Radhakrishna, R., K.H.Rao, andK.Subbarao. 2004. "Reducing Vulnerability andEnhancing OpportunityTowards aMoreEffective UseofIndia's SafetyNet Programs." Draft Report, SouthAsia HumanDevelopment, World Bank. Ram,F., B.Paswan, L.Ladusingh,N.Khan, andR.Ranjan. 2004. "District Level Household Survey2002: EstimationofInfant Mortality, Major States.'' Mimeo. Rao. V, T. Besley, andR.Pande. 2005. "The Political Economy of Gram PanchayatsinSouth India: DrawingPolicy Implications from a ResearchProject onDemocratically Elected Village Governments." http://www.cultureandpublicaction.org/bijupdf;/panchayatreport.pdf Rao, V. andM.Walton. 2005. "Culture andPublic Action: An Introduction." inV. Rao andM. Walton (eds). Culture and Public Action, Stanford UniversityPress. Ravallion, M.and S. Chaudhuri. 1997. "Risk and InsuranceinVillage India: A Comment." Econometricu 65(1): 171-84. Ravallion, M.and S. Chen. 2004. "China's (uneven) Progress against Reducing Poverty." Policy ResearchWorking Paper 3408, World Bank. Rodrik,D.andA. Subramanian.2004. "From "Hindu Growth" To Productivity Surge: The Mysteryofthe IndianGrowthTransition." IMF Working Paper04/77. Rodrik, D., A. Subramanian, andF.Trebbi. 2004. "Institutions Rule: The Primacy o f Institutions Over Geography and IntegrationinEconomic Development." Journal of Economic Growth, 9(2), pp.131-165. Saxena, N.C. 2005. ImprovingDelivery o fProgrammesthrough Administrative Reforms in India. Paperwritten for the National Advisory Council. Postedat nac.nic.in/concept.htm. Sen, A. and Himanshu. 2005. "Poverty and Inequality inIndia." InA. Deaton and V. Kozel (eds.) TheGreat Indian Poverty Debate. Macmillan India: New Delhi. Shah, A. and I. 2005 ."India's Experience with Capital Flows: The Elusive Quest for a Patnaik. Sustainable Current Account Deficit." Working Paper 1187, National Bureau of Economic Research. 190 Shah, T. 2005. "Accountable Institutions." Backgroundpaper preparedfor India's Turbulent Water Economy: Bracing for a Turbulent Future, South Asia Region, World Bank. Sundaram, IS. 2003. "On Identification ofHousehold Below Poverty Line inBPL Census 2002: Some Commentson the Proposed Methodology." Economic and Political Weekly, March 1,2003. Tarozzi, A and A. Mahajan. 2005. "Child NutritioninIndia inthe Nineties: A Story o f Increased Gender Inequality?" Mi'meo. Townsend, R. 1994. "Risk and Insurance inVillage India." Econometrica 62(3), pp. 539-91. Transparency International. 2002. Corruption inIndia: An Empirical Study. Transparency InternationalIndia and ORG-MARG Research Private Ltd. Virmani A. 2004a. "Sources ofIndia's Economic Growth: Trends inTotal Factor Productivity." Working Paper 131, Indian Council for Research on InternationalEconomic Relations, New Delhi. Virmani, A. 2004b. "India's Economic Growth: From Socialist Rate o f Growth to Bharatiya Rate o f Growth." Working Paper 122, Indian Council for Research on International Economic Relations, New Delhi. Virmani, A. 2005. "Policy Regimes, Growth andPoverty InIndia:Lessons Of Government Failure And Entrepreneurial Success!" Working Paper 170, Indian Council for Research on International Economic Relations, New Delhi. Virmani, A. andMittal, S. 2005. Domestic Market Integration.WorkingPaper, IndianCouncil for Research on International Economic Relations, forthcoming. Walker, T. and J. Ryan. 1990. Village and Household Economies in India 's Semi-Arid Tropics. Johns Hopkins University Press. World Bank. Forthcoming(a). India: Labor and Employment study. World Bank. Forthcoming(b). India: RuralGovernments and Service Delivery. World Bank. Forthcoming(c). Cost Effectiveness ofRuralWater Supply Schemes inIndia. World Bank.2001a. Why Some Village Water and Sanitation Committees are better than others: A study ofKarnataka andUttarPradesh(India). World Bank. 2001b. India: Power Supplyto Agriculture. 191 World Bank. 2003a. World Development Report 2004: Making Service Work for Poor People. OxfordUniversityPress:New York. World Bank. 2003b. India: SustainingReform, ReducingPoverty. OxfordUniversityPress: New Delhi. World Bank. 2003c. Costs o fDeferred Maintenance inIndia. World Bank. 2003d. RuralFinanceAccess Survey. World Bank. 2004a. Services Revolution: Access to ForeignMarkets, Domestic Reform, and InternationalNegotiations. World Bank. 2004b. India: Financing Highways. World Bank. 2004c. India-FiscalDecentralization to Rural Governments. World Bank. 2004d. State FiscalReforms inIndia: ProgressandProspects.MacmillanPress. World Bank 2004e. India: InvestmentClimate Assessment 2004 -ImprovingManufacturing Competitiveness. World Bank 2004f. India: Scaling-up Access to Finance for India's RuralPoor. World Bank. 2004g. Bihar:Towards aDevelopment Strategy. World Bank. 2004h, DoingBusiness -IndiaRegional Profile. World Bank. 2005a. World Development Report 2006: EquityandDevelopment. Oxford UniversityPress: NewYork, World Bank. 2005b. India's UndernourishedChildren: A Call for Reform and Action World Bank. 2005c. Economic growthinthe 1990s:Learning from a Decade ofReform. World Bank. 2005d. Will IndiaMeet the MillenniumDevelopment goal o f "Halving by 2015 the Proportion o f the Peoplewithout Sustainable Access to SafeDrinking Water and Basic Sanitation?' Draft report. World Bank. 2005e. India's Water Economy: Bracing for a Turbulent Future. World Bank. 2005f Amplifying Voice and Client Power. World Bank. 2005g. Reforming Services inIndia: Drawing from Lessons o f Success. World Bank.2005h. India- UrbanFinance and GovernanceReview. 192 World Bank. 2005. BuildingCapacities for Public-Private Partnerships. World Bank. 2005J. Infrastructure inLatinAmerica: Recent Developments and Key Challenges. World Bank. 2005k. Pro-poor Growth inthe 1990s:Lessons and Insights from Fourteen Countries. World Bank. 20051 India: Re-energizingthe Agricultural Sector to Sustain Growth andReduce Poverty. OxfordUniversityPress, New Delhi. World Bank. 2005m. Agricultural Growth for the Poor: An Agenda for Development. Directions inDevelopment. World Bank.200%. Jharkhand Policy Note. World Bank. 20050. Orissa Investment Climate Assessment: Towards a High-Performing State. World Bank. 2006. India Water Supply and Sanitation: Bridging the Gap between Infrastructure andService. 193 StatisticalAnnexes 194 Statistical Annex Table of Contents Table Al. Gross Domestic Expenditure and Product (shares based on current price data) Table A2. Gross Domestic Expenditure and Product (Rs. billioncurrent prices) Table A3. Annual Growth Rates of National Income andProduct at Constant Prices (annual growth rates) Table A4. Gross Domestic Product by Expenditure, National Income and Savings (Rs. billion at 1993-94 prices) Table A5. Exchange Rates and Prices Table A6. Central Government Finances Summary Table A7. Budgetary Classification o f Central Government Finances Table A8. Budgetary Classification of State Government Finances TableA9. Budgetary Classification o f General Government Finances Table A10. Transfers Between Center and States Table A11. Outstanding Debt (Center and States) Table A12. Banking Survey and Interest Rates Table A13. Balance of Payments Table A14. Exports and Imports Table A15. External Debt and Debt Service Table A16. Financial Sector Indicators Table A17. Investment Climate Table A18. Vulnerability Indicators Table A19. MillenniumDevelopment Goals Indicators Table A20. Development Indicatores - India and Comparator Countries Table A 21.Unemployment Rates: Alternative Measures D ? 5 Y s 5 a $a 3 - g - r ~ n e i ~ ~ m9o9g r - o '4 k N '9 - --? 9 N cf Z v - - - ? t ? ? -. -NmNL n N N g r - O - w O 09 9 9 9 a - r m = Y $a ~ ; a 9 n ; 1 0 ? n 1 - 9 5 0 ? t '4Lc? - L ~ ~ c? t 9Nc? k 9 W O O ~ O N N -~ mN m Z~ 9 D ? 51 s $ 9 z VI 0 hl Q w 4 0 0 N m 4 8 0 N N 4 z N 0 2 0 D N z D r m m m 2m T m 2mT 2n h T w !B n n r n 3 5) r .-8 v) h 9 9 9 9 z z z z 0 0 0 0 /a I m m m '3 N 7 a m m c 2 7 2 m m vi N " v1 N d r- m w W K - W m d N w N vi - m x d w m '3 m e w z a. N 0 3 x ..- N N vi N x c m N 0. h 9 N N N m M 7 % -m 0 x N 2 8 N v1 7 7 N N m N 0 8 % v) v) m J - 1I r) 2. f Y v) J rn 1 ss 9 E2 ._ c 1 +. P E P E - & g o s f: hl c Table A16. Financial Sector Indicators 1996-97 1997.98 1998.99 1999.00 2000.01 2001.02 2002.03 2003-04 2004.05 A. Banking System 1. Depth and Structure Total domestic credit (%of GDP) 42.4 41.9 40.2 40.4 41.6 42.1 45.0 44.4 47.4 of which private credit (% of GDP) 27.5 28.5 28.5 29.9 31.9 33.2 36.7 37.0 40.8 Number of banks Public Sector 27 27 27 27 27 27 27 27 28 Private Banks 29 29 29 29 Foreign Banks 31 31 31 31 2. Efficiency and Strength Spread over LIBOR 10.4 8.1 8.0 7.1 5.8 8.3 10.1 10.2 9.2 Non-performing loans as % of total Public sector banks(27) 17.8 16.0 15.9 14.0 12.4 11.1 9.4 7.8 5.4 Private banks (34) 8.5 8.7 10.8 8.5 8.5 9.7 9.2 5.8 3.9 Foreign banks (45) 4.3 6.4 7.6 7.0 6.9 5.5 5.4 4.8 3.0 B. Stock Market Market Capitalization (%GDP) 31.9 31.2 25.4 41.4 32.4 23.0 25.7 46.5 56.1 Value Traded (%GDP) 24.9 38.6 35.8 62.4 111.5 2.0 38.7 47.4 54.8 No. of listed companies 5999 5843 5860 5863 5937 5795 5650 5644 4730 S&P/IFC investable index (annual % change) -2.0 5.8 -23.0 81.o -31.1 -19.9 6.8 76.5 20.1 Sources : RBI and World Development Indicators. 16 Table A17. Investment Climate Year India Low Income Regional Middle-Income Average Average Average Private Investment Environment Domestic Credit to Private Sector (stock, % GDP) 2003 32.0 27.0 31.0 64.2 Starting a business Number of start-up procedures Jan-04 11 11 9 10 Time required days Jan-04 89 63 47 51 Registering Property Number of procedures Jan-04 6 7 6 Time required days Jan-04 67 100 56 ao7 Rigitidy of employment index (0 less rigid to 100 more rigid) Jan-04 48 52 42 38 Enforcing contracts Number of procedures Jan-04 40 35 30 31 Time required days Jan-04 425 418 349 401 Disclosure index (0 less disclosure to 10 more disclosure) Jan-04 4 2 3 3 Time to resolve insolvency (years) Jan-04 10 4 5 3 Governance ICRG Corruption Rating (1-6, bad to good) Oct-05 2.5 ICRG Bureaucratic Quality Rating (1 - 6) Oct-05 3.0 ICRG Law and Order (1 - 6) Oct-05 4.0 Openness Trade (imports+exports)lGDP (%) 2004 32.6 34.9 24.1 58.3 FDI inflows (% GDP) 2003 0.7 1.3 0.7 2.4 W i O member? Y Weighted Mean Tariff (%) 2002 8.9 Infrastructure Paved Roads, % of total 1999 57.3 14.6 30.8 54.0 Motor vehicles (per 1000 persons) 2000 9.3 9.6 56.1 Cost of Calls to US (US$per 3 min) 2001 3.20 4.5 2.70 2.3 Internet Users (per 1000 people) 2003 18 16 11 117 Electricity consumption (kwh per capita) 2002 380 313 344 1,388 GDP per unit energy use (PPP $ per Kg oil equivalent) 2002 5.2 4.3 5.3 4.3 Wages and Productivity Minimum Wage (US$ per year) 1995-99 408 Labor Cost Per Worker in Manufacturing (US$ per year) 1995-99 1,192 Value Added Per Worker in Manufacturing (US$ per year). - . . 1995-99 3,118 R&D Expenditure (% of GNI) 1989-2000 0.62 0.62 Source: World Bank, World Development Indicators Database, 2005. 17 Table A18. Vulnerability Indicators 1996-97 1997-98 1998-99 1999-00 2000.01 2001.02 2002.03 2003-04 2004-05 A. Market Indicators Annual percent change in average exchange rate (%) 6.1 4.7 13.2 3.0 5.4 4.4 1.5 -5.1 -2.2 Annual change in stock market index (%) -2.0 5.8 -23.0 81.o -31.I -19.9 6.8 76.5 20.1 B. Risk Ratings ICRGcomposite (1-100, bad to good) ai 68.5 66.3 64.3 64.3 64 65.8 66.8 72.0 71.8 Euromoney(1-100, bad to good) Institutional Investor (1-100, bad to good) C. Financial Annual growth in real domestic credit (%) -3.3 2.6 -2.9 9.2 7.0 5.0 10.3 7.0 16.1 Foreign currency to total deposits (%) b/ 15.8 17.0 18.0 18.0 18.5 20.0 21.4 23.7 25.4 Non-perfm. loans of commercial banks (YOof total) Pubiic Sector 17.8 16.0 15.9 14.0 12.4 11.1 9.4 7.8 5.4 Private Banks 8.5 8.7 10.8 8.5 8.5 9.7 9.2 5.8 3.9 Foreign Banks 4.3 6.4 7.6 7.0 6.9 5.5 5.4 4.8 3.0 D. Reserve Cover Indicators Reservecover of imports (months of imports) 6.5 6.88 8.2 8.2 8.8 11.5 14.0 16.7 14.1 Reservesto short term debt 3.9 5.8 7.5 9.7 12.2 19.7 16.5 23.6 29.6 ReservesiM2 (%) 12.2 13.2 13.7 14.8 15.3 18.0 E. Prices Annual change in terms of trade (%) -8.6 15.6 3.0 -10.4 -4.9 -2.2 -9.5 8.6 Annual Depreciation REER (%) 0.3 5.0 -5.3 -0.2 5.1 2.9 6.3 1.9 3.8 F. External Current account balance (%of GDP) -1.3 -1.4 -1.o -1.1 -0.6 0.6 1.2 1.7 -1.o External Debt (% of GDP) 24.3 23.0 23.6 21.7 21.5 20.4 21.o 18.9 16.0 G. Fiscal sustainability indicators (Genral Govt.) Totai Public Debt (% of GDP) 68.4 70.2 70.2 71.8 75.6 80.7 85.7 85.8 83.6 Fiscal Deficit (% of GDP) 6.4 7.3 9.4 9.5 9.6 10.1 9.8 9.1 8.1 Primary Deficit (% of GDP) 1.2 2.1 4.0 3.9 3.6 3.8 3.2 2.7 1.9 a/ Endof the fiscal year ratings (for the month of March). b/ Share of Net Foreign Exchange Assets to Totai Deposits including Net Foreign Exchange Assets. Sources: IFS, WDi, Reserver Bank of india and World Bank Staff Estimates. 18 Table A19. Millennium Development Goals Indicators 1990 1995 1998 2001 2004 Goal 1: Eradicate extreme poverty and hunger Incomeshare held by lowest20% 9.0 Malnutritionprevalence,weight for age (% of childrenunder 5) 64.0 47.0 Povertygap at $1 a day (PPP) (%) 8.0 Povertyheadcountratioat $1 a day (PPP) (% of population) 35.0 Povertyheadcountratioat nationalpoverty line (%of population) 36.0 29.0 Prevalenceof undernourishment(% of population) 21.o 20.0 Goal 2: Achieve universalprimary education Literacyrate,youth total (%of peopleages 15-24) 64.0 76.0 Persistenceto grade 5, total (% of cohort) 62.0 61.O Primary completionrate,total (% of relevantage group) 77.1 72.6 75.8 88.5 Schoolenrollment,primary(% net) 83.0 87.0 Goal 3: Promote gender equality and empower women Proportionof seats held by women in nationalparliament(%) 5.0 7.0 9.0 9.0 Ratioof girlsto boys in primaryand secondaryeducation(%) 69.8 77.2 78.0 87.7 Ratioof young literatefemalesto males (% ages 15-24) 73.9 80.5 employment) 13.0 14.0 16.0 17.0 18.0 Goal 4: Reduce child mortality Immunization,measles(% of childrenages 12-23 months) 56.0 72.0 51.O 56.0 56.0 Mortalityrate, infant(per 1,000 live births) 80.0 74.0 72.0 66.0 62.0 Mortalityrate, under-5(per 1,000) 123.0 104.0 94.0 85.0 Goal 5: Improvematernal health Birthsattendedby skilled healthstaff (% of total) 42.3 42.5 Maternalmortality ratio (modeledestimate, per 100,000 live births) 540.0 Goal 6: Combat HIVIAIDS, malaria, and other diseases Children orphanedby HIV/AIDS Contracep~veprevalence(% of women ages 15-49) 52.0 47.0 Incidenceof tuberculosis(per 100,000 people) 167.8 167.8 Prevalenceof HIV, female (% ages 15-24) 1 Prevalenceof HIV,total (%of populationages 15-49) 1.o.o 1.o Tuberculosiscases detectedunder DOTS (%) 0.3 1.7 23.8 57.1 Goal 7: Ensure environmentalsustainability C02 emissions(metrictons per capita) 0.8 1.o 1.1 1.I Forestarea (%of land area) 22.0 23.0 23.0 GDP per unitof energy use (constant2000 PPP$ per kg of oil equivalent) 4.0 4.0 5.0 5.0 5.0 Improvedsanitationfacilities(% of populationwith access) 12.0 30.0 Improvedwater source (% of populationwith access) 68.0 86.0 Nationallyprotectedareas (% of total land area) 5.2 Goal 8: Developa global partnership for development Aid per capita(current US$) 1.7 1.9 1.6 1.7 0.6 Debtservice (PPGand IMF only, % of exportsof G&S, excl.workers' remittances) 29.0 32.0 23.0 13.0 20.0 Fixed line and mobile phonesubscribers(per 1,000 people) 6.0 12.9 23.2 43.6 84.5 Internetusers (per 1,000 people) 0.0 0.3 1.4 6.8 32.4 Personalcomputers (per 1,000 people) 0.3 1.3 2.7 5.8 12.1 Total debt service(% of exportsof goods,servicesand income) 32.0 30.0 21.o 12.0 19.0 Unemployment,youth female (% of female laborforce ages 15-24) 8.0 10.2 Unemployment,youth male (%of malelaborforce ages 15-24) 8.4 10.1 Unemployment,youth total (% of total labor force ages 15-24) 8.3 10.1 Other Fertilityrate, total (birthsper woman) 3.8 3.4 3.3 2.9 2.9 GNI per capita,Atlas method(currentUS$) 390.0 380.0 420.0 460.0 620.0 GNI, Atlas method(currentUS$) (billions) 330.6 349.6 412.9 479.3 673.2 Grosscapitalformation(% of GDP) 24.1 26.5 21.4 22.4 24.1 Life expectancyat birth,total (years) 59.1 61.4 62.2 63.4 63.5 Literacyrate,adulttotal (%of peopleages 15 and above) 49.3 61.O Population,total (millions) 849.5 932.2 982.2 1032.5 1079.7 Trade (% of GDP) 15.7 23.2 24.1 27.7 41.6 Note:Where data is not availablefor the year mentioned,the closest year was taken. Source:World DevelopmentIndicatorsdatabase,April 2006 19 Table A20: Development Indicators India and Comparator Countries - Indicator Unit Year India China Brazil Indonesia Pakistan Per Capita Income US$ 2004 620 1290 3090 1140 600 PPP Per capita income US$ 2004 3100 5530 8020 3460 2160 Population Million 2004 1080 1296 179 218 152 Male literacy rate Percent 2000-2004 73.4 95.1 88.3 92.5 61.7 Female literacy rate Percent 2000-2004 47.8 86.5 88.6 83.4 35.2 Population below the poverty line Survey year in 26.1 (2000) 4 6 (1998) 22.0 (1998) 27.1(1999) 32 6(1999) Percent brackets Population below US$l/day Survey year in 35.0 (2000) 16.6 8.2 7.5 17.0 Percent brackets Infant mortality rate Per 1000 2003 63.0 30.0 33.0 31.O 74.0 live births Life expectancy at birth Years 2003 63.4 70.8 68.7 66.9 64.1 Access to improved sanitation facilities Percent 2002 30.0 44.0 75.0 52.0 54.0 [I] alternatemethodologyusedbyDeatonandDrezeestimatedthepovertyheadcounttobe28.6%in1999/2000 an Source: World Development Indicators, 2005, The World Bank; Poverty data on India (199912000)sourced from Planning Commission, Go1 20 Table A21: Unemployment Rates: Alternative Measures Unemployment rates Usual Usual Current Current Daily Principal Principal and Weekly Status Status Subsidiary Status Status (UPS) (UPSS) (CWS) CDS) All India 1987-88 3.8 2.6 4.8 6.1 1993-94 2.6 1.9 3.6 6.0 1999-00 2.8 2.2 4.4 7.3 Urban 1987-88 6.6 5.3 7.1 9.4 1993-94 5.2 4.5 5.8 7.4 1999-00 5.2 4.6 5.9 7.7 Rural 1987-88 3.1 2.0 4.2 5.3 1993-94 1.8 1.2 3.0 5.6 1999-00 2.0 1.4 3.9 7.2 21