THE WORLD BANK SERBIA INVESTMENT CLIMATE ASSESSMENT DECEMBER 14, 2004 Finance and Private Sector Development Unit (ECSPF) Europe and Central Asia Region The World Bank CURRENCY EQUIVALENTS (As of January 13, 2004) Currency Unit: Exchange Rate: US$1 = WEIGHTS AND MEASURES The metric system is used throughout this report FISCAL YEAR ABBREVIATIONS AND ACRONYMS BEEPS Business Environment and Enterprise Performance Survey CODB Cost of Doing Business Project CLDS Center for Liberal and Democratic Studies EBRD European Bank for Reconstruction and Development ECA Europe and Central Asia Region EIU Economist Intelligence Unit EU European Union FIAS Foreign Investment Advisory Service FDI Foreign Direct Investments GDP Gross Domestic Product GNI Gross National Income HBS Household Budget Survey ICA Investment Climate Assessment ILO International Labor Organization LFS Labor Force Surveys MAEWS Monitoring, Analysis and Early Warning Signal MCI Ministry for Capital Investments MoE Ministry of Economy MoF Ministry of Finance MoJ Ministry of Justice MoLSA Ministry of Labor and Social Affairs MoT Ministry of Trade OECD Organization for Economic Cooperation and Development PICS Productivity and Investment Climate Survey PRSP Poverty Reduction Strategy Paper SAA Stabilization and Association Agreement SAM Serbia and Montenegro SME Small and Medium Enterprise SOE State Owned Enterprises SOF Socially Owned Firms TFP Total Facto Productivity WDR World Development Report WEF World Economic Forum WTO World Trade Organization Vice President: Shigeo Katsu Country Director: Orsalia Kalantzopoulos Sector Director: Fernando Montes-Negret Sector Manager: Gerardo Corrochano Task Manager: Itzhak Goldberg 2 TABLE OF CONTENTS EXECUTIVE SUMMARY ........................................................................6 CHAPTER 1: INTRODUCTION.............................................................18 CHAPTER 2: PRODUCTIVITY, OWNERSHIP AND THE INVESTMENT CLIMATE IN SERBIA AND WORLDWIDE..............21 CHAPTER 3: MACROECONOMIC CONTEXT AND INVESTMENT CLIMATE PRIORITIES..........................................................................32 CHAPTER 4: LEGAL, REGULATORY AND INSTITUTIONAL FRAMEWORK ........................................................................................39 Ownership, Entry, Competition and Exit .................................................................39 Contract Enforcement ..............................................................................................46 Regulatory Burden ...................................................................................................50 CHAPTER 5: FACTOR MARKETS.......................................................55 Access to Finance.....................................................................................................55 Labor Market............................................................................................................63 Land Issues...............................................................................................................70 CHAPTER 6: KEY POLICY RECOMMENDATIONS .........................73 BIBLIOGRAPHY.....................................................................................78 Appendix 1: Factor Analysis ...................................................................81 Appendix 2: Comparing TFP across countries using firm-level data .....83 Appendix 3: ICA Reference Data.............................................................96 Acknowledgements: The report was prepared by Itzhak Goldberg (Team Leader), Branko Radulovic and Professor Mark Schaffer with the assistance of Katarina Stanic, Alexandra Drees-Gross, and Jasna Vukoje. Comments from Andrew Stone, Ardo Hansson and Roy Pepper, peer reviewers, and Mamta Murthi, Peter Kyle and Ali Mansoor are acknowledged. We would like to thank Deputy Minister Vlatko Sekulovic for very helpful suggestions as well as to participants of the Employment Growth Forum, organized by the Center for Liberal Democratic Studies (CLSD) in Belgrade where ICA was used as a key background paper. Finally, we would like to thank Qimiao Fan and Alberto Cruisciolo for their support through World Bank Investment Climate Capacity Enhancement Program and to participants of the GDLN session on Access to Finance and Contract Enforcement Workshop for very useful discussions. 4 INVESTMENT CLIMATE AT-A-GLANCE Serbia and Montenegro Croatia Poland Bulgaria MACROECONOMIC GDP growth (annual %) 4 5.2 1.4 4.8 GDP per capita, (constant 1995 US$) 1,830 5,440 3,770 1,720 Foreign direct investment, net inflows (% of GDP) 3 4.4 2.2 3.9 Trade (% of GDP) 64.5 100.8 59.5 112.9 Inflation, GDP deflator (annual %) 25.5 2.9 1.3 3.9 Gross capital formation (% of GDP) 16 27 19 20 Gross domestic savings (% of GDP) -7 18 16 13 TRADE AND BALANCE OF PAYMENTS Imports of goods and services (% of GDP) 43.8 54.8 31.4 59.7 Exports of goods and services (% of GDP) 20.7 46.0 28.0 53.1 Total debt service (% of exports of goods and services) 2.8 25.9 22.5 15.9 Current account balance (%of GDP) -8.8 -7.2 -2.6 -4.4 STOCK MARKETS Stocks traded, total value (% of GDP) 0.0 0.6 4.1 0.5 Stocks traded, turnover ratio (%) .. 4.0 26.1 12.9 Market capitalization of listed companies (% of GDP) 0.0 15.8 14.1 3.7 BUSINESS REGULATIONS Number of procedures to start business* 11 12 10 10 Number of days to start business* 51 49 31 30 Cost to start business (in % of GNI)* 9.0 14.4 20.6 9.3 Number of procedures to enforce contracts* 40 20 18 26 Number of days to enforce contracts* 1,028 330 1,000 410 INFRASTRUCTURE Energy use (kg of oil equivalent p.c) 1,507 1,771 2,344 2,428 Internet users (per 1,000 people) 59.7 180.4 230 80.8 Mobile phones (per 1,000 people) 256.6 535 362.6 333 Roads, paved (% of total roads) 62 84 66 92 Telephone mainlines (per 1,000 people) 229 407 295 359 GOVERNANCE AND OVERALL REGULATORY QUALITY Index of economic freedom** 4.28 3.06 2.83 3.26 Government effectiveness*** -0.73 0.19 0.61 -0.06 Regulatory quality*** -0.60 0.19 0.67 0.62 Rule of law*** -0.95 0.11 0.65 0.05 Corruption*** -0.80 0.23 0.39 -0.17 POPULATION, HEALTH, AND LABOR Population, total 8,160,000 4,465,000 38,626,000 7,965,000 Life expectancy at birth, total (years) 72,7 73,8 73,8 71,8 Mortality rate, infant (per 1,000 live births) 16 7 8 14 Age dependency ratio 0.51 0.47 0.45 0.45 Health expenditure, total (% of GDP) 8.2 9.0 6.1 4.8 Unemployment, total (% of total labor force) 22.3 15.8 18.2 19.4 SOURCE: All data are for drawn from the World Development Indicators Database, 2004, World Bank, unless otherwise noted. Data are for 2001- 2002, except for paved roads (1999) *Data are from Doing Business Database, Private Sector Advisory Service, World Bank http://rru.worldbank.org/DoingBusiness/default.aspx. Data for starting business are for 2004 and for contract enforcement for 2003 **Data are from O'Driscoll, Gerald and Edwin J. Feulner, Mary Anastasia O'Grady, "2003 Index of Economic Freedom," Heritage Foundation,http://www.heritage.org/research/features/index/. Measures of economic freedom (or the absence of government coercion or constraint on the production, distribution, or consumption of goods and services beyond the extent necessary for citizens to protect and maintain liberty itself) based on the unweighted average of 10 categories (that are aggregates of 50 economic variables). Each factor is scored according to a grading scale that is unique for that factor. The scales run from 1 to 5 with a score of 1 signifying an institutional or consistent set of policies that are most conducive to economic freedom. ***Data are from Kraay, Aart, Daniel Kaufmann, and M. Mastruzzi, "Governance Matters III: Governance Indicators for 1996-2002," http://www.worldbank.org/wbi/governance/govdata2002.htm. The governance indicators in are measured in units ranging from about -2.5 to 2.5, with higher values corresponding to better governance outcomes (data are for 2002). 5 EXECUTIVE SUMMARY Objectives The objective of this Investment Climate Assessment (ICA) is to (i) provide Serbian experts in academia and in government, as well as Bank staff, with an empirical analysis of the investment climate in Serbia; and (ii) to discuss policy options, based on this analysis, for creating an enabling environment conducive to private sector development, thereby increasing and maintaining enterprise productivity and profitability leading to sustainable growth. The study is part of a Bank-wide effort, managed and funded by the Investment Climate Unit of the Investment Climate Department (CICIC), to analyze the effects of various characteristics of the investment climate on productivity in an international context. One of the main messages of this ICA is the need to increase investor confidence in Serbia; the ICA identifies a number of policies that, if implemented, could improve the perception of Serbian investment climate. In particular, it should be stressed that reversals in policy and lack of respect for property rights on investors that have already invested in Serbia could cause serious harm. Background In the 1990's, the loss of markets, international isolation, armed conflicts, and interruption of long-established production relations left a dire legacy for the Serbian economy. Three years have passed since the October 2000 "velvet revolution" in Serbia and the establishment of the first government in January 2001 that lasted until February 2004. During these three years, the investment climate was heavily affected by political instability. Several major policy steps conducive to a better investment climate were implemented: fiscal stability and trade liberalization were achieved; tax code reforms were carried out; the company law was amended to eliminate pre- registration inspections; a privatization law was adopted in 2001 and amended in 2003; leasing, collateral and concessions laws were introduced in 2003; and tax reforms were implemented in 2001. Accordingly, 4,000 new small enterprises and 20,000 shops have been opened. In addition, around 1,000 mostly medium-sized enterprises were sold in auctions and circa 25 large enterprises were sold by an international tender to major international strategic investors. Yet, the remaining investment climate reform agenda is huge and presents the new government with major challenges. With Bank and other donor support, the authorities began to take important actions in all of the above mentioned policy areas. As will be shown below, the ICA recommends that the new government continue on this path ("stay the course") and support the next generation of reforms to improve the country's investment climate. Moreover, while in the initial reform agenda, program design and legislation played a key role, the short to medium priorities described above require implementation capacity and the maintaining and building of new institutions, such as the Privatization Agency, the Business Registration Agency (a one-stop shop for leasing, collateral and company registration), Foreign Investment Agency (SIEPA), the SME agency, Bankruptcy Administration unit within the 6 Privatization Agency, a supervisory agency for bankruptcy administrators, and a powerful Competition Authority. The ICA reflects the increased emphasis in Serbia on the development agenda. This emphasis is natural for an economy that has largely completed the initial stage of its economic program, by focusing on stabilizing the economy, carrying out urgent reconstruction, implementing "first generation" reforms such as price and foreign exchange liberalization, and laying the foundations for "second generation" reforms. . As in other transition economies in Eastern Europe, the second stage which focuses on ensuring long-term economic stability, sustainable growth, and employment creation, has proven to be more challenging. Initial progress in key areas of reform has been uneven. Significant advances have been made in opening the country to trade and investment, the management of public finances, tax policy, privatization, bank resolution, energy, the labor market, pensions and social assistance. In most areas, however, the reforms are still far from complete, and in several areas (e.g., public administration, health, competition policy), the reform process has hardly begun.1 Moreover, Foreign Direct Investment (FDI)--an important indicator of the investment climate--has yet to play an important role for the private sector in Serbia. FDI per capita in Serbia in 1989-2002 was US$200 in comparison to US$1,400 in Croatia US$560 in Bulgaria, US$460 in FYR Macedonia, and US$419 in Romania. Foreign investors share the views of their domestic peers that Serbia is a risky place in which to invest. The World Economic Forum Global Competitiveness Report ranked Serbia 77 among 102 countries in terms of its overall competitiveness. For attracting FDI and local investment, the new government must, at the very least, ensure consistent treatment of the first group of investors which bought companies in tenders and auctions. With high risk and without abundant natural resources, almost all FDI in Serbia in the first period 2001-2003 was related to privatization. These "brownfield" investments could play a key positive role in moderating the perception of risk for new potential "greenfield" investors. However, this will happen only if the current "brownfield" investors find the investment climate acceptable and are treated fairly. The experience of investors, such as Henkel, LaFarge, Pharmaco, Philip Morris, so far is worrisome.2 Investment Climate Priorities and Policy Implications An empirical analysis of circa 400 Serbian firms in 2003 conducted for the ICA shows that the productivity level of privatized firms is about 90 percent higher than that of socially-owned firms; the productivity level of new private firms is over three times higher. Privatized firms are 15 percent more likely to make profits than socially- owned firms; new private firms are 33 percent more likely3. These strong empirical 1For a more extensive review of reform performance during 2001-03, see World Bank (2003), Recent Progress on Structural Reforms--a document prepared by the World Bank for the Donor Coordination Meeting in November 2003. 2 The experience of Henkel, one of the world leaders in the detergents sector, is particularly illustrative. Henkel acquired the chemical (detergent) company Merima, in an open tender. Recently Henkel had been asked to pay tax on the severance payments for Merima's employees, notwithstanding that Henkel agreed with the Government about the severance pay package as part and parcel of the privatization deal. Other major investors (LaFarge, Pharmaco, Philip Morris) underwent a similar experience. 3 The better performance of some privatized firms may simply reflect the fact that only employees in the better firms decided to take their firms private. The results for new entrants, however, are unambiguous: new private firms in Serbia are more productive, more profitable, and growing more rapidly than socially-owned firms. 7 results imply that a continued commitment to private ownership and a supportive business environment, facilitating new private entry or via privatization, supporting the rights of owners and creditors, is an engine to increased profitability and productivity of enterprises and therefore should continue to be a key policy priority. Sustainable economic growth can only come from investment by the private sector. Thus, issues related to the investment climate are key to putting the country on the road to long term prosperity. While the government in Serbia has made progress in implementing business enabling reforms such as liberalization and deregulation of foreign trade and investment, privatization, simplification of the tax regime, and modernization of the labor legislation, it has yet to focus on the related institution building that is necessary. The implementation of recently drafted legislation aimed at improving business entry, corporate governance, access to finance, and business exit has yet to start and requires extensive institutional building to take effect. In view of the limited capacity of the government and the decreased foreign assistance, the prioritization of investment climate reforms and implementation efforts is critical. Based on the surveys, experiences of foreign and local investors, the ICA proposes the following priority areas. Goal #1: Strengthening contract enforcement. The court system in Serbia is markedly pro-debtor. Weak judicial capacity also increases the backlog of cases, and encourages firms to use courts only as a last resort. Accordingly, firms rely more on personal relationships and are less willing to deal with new customers and suppliers, which in turn, limits the number of transactions and also limits market dynamism. Weak contract enforcement contributes to uncertainty and non-transparency that create fertile ground for corruption. Goal #2: Increasing access to credit. Lack of access to credit limits businesses to using only their own funds, retained earnings, or money borrowed from friends and family. This constraint limits market dynamism in a post-conflict and post-socialist environment in which retained earning have not yet been accumulated. The lack of access to credit through banks and other formal financial institutions removes an important motive for businesses to formalize--thus contributing to the persistence of the informal economy. Goal #3: Reducing the regulatory burden. The "regulatory burden" is the consequence of a series of government actions and polices such as a legacy of inefficient and anti-market laws and ministerial regulations, complex company registration, frequent multiple and overlapping inspections or lengthy certification. This regulatory structure distorts and hinders market development, and, for those businesses who succeed in becoming viable, create opportunities for corruption (the bribe tax) and wasting of businessmen's time (time tax). Administrative harassment and corruption as well as the unpredictability of business rules, and the unfairness in their application represent a serious problem. Improvement of the quality of regulation by streamlining, improving transparency and market-friendliness, and consistent application should also reduce the so-called "time and bribe taxes" mentioned above. In addition, increasing the tax administration capacity to support rule-based tax collection rather than punishing successful businesses is an important goal Goal #4: Strengthening property rights in land and reducing labor rigidities. An ongoing unresolved issue is the failure to adopt a legal framework for dealing with the 8 nationalized property and the claims for restitution. The failure to resolve the restitution issue regarding both land and other properties, although easy to understand in view of the recent political history (the Law was ready just before the assassination of PM Zoran Djindjic) continues to create uncertainty and deters many real estate and other investors from investment. The restrictions on use and state ownership and control of urban land (particularly construction land) are major obstacles for the construction and services sectors. The Serbian land administration system is in a very poor state, with most of the apartments in Belgrade and other cities not registered. Analysis and removal of the remaining burdensome construction permits procedures deserve the immediate attention of the authorities. The proposed amendments to the labor law will introduce very detrimental labor rigidities. They include the obligation to conclude a collective agreement for employers who have more than 10 employees and administratively setting the level of minimum wages to 50 percent of the average wage, making dismissal for economic reasons more difficult or costly. The introduction of these rules would bring Serbia back on the list of the countries with the most rigid labor market; please see specific recommendations. Goal #5: Facilitating entry, change of ownership, competition and bankruptcy.Schumpeterian "creative destruction", facilitating entry and exit, is especially important in the post socialist industrial structure, in which restructuring and bankruptcy are channels for a dynamic entry of small new enterprises. The process, whereby bad firms exit and new ones replace them is an important condition for the sustainability of growth in Serbia. - Sub-goal #5.1 Stop backsliding in privatization. Retroactive actions to undo legal privatizations could undermine the confidence of investors. Major changes in the Privatization Law, such as those now discussed by the Parliamentary Privatization Committee also will undermine investors' confidence. - Sub-goal #5.2 Build bankruptcy institutions. Reallocating the capital and human resources of un-restructured and failing firms requires a framework for bankruptcy and liquidation that frees the tied-up capital for alternative, higher productivity uses. - Sub-goal #5.3 Start building a powerful competition authority. Trade liberalization by opening the markets to foreign competition, together with the privatization and establishment of SMEs, may not be sufficient to bring about competitively structured markets and the resulting competitive conduct of business operations. Specific policy recommendations based on the analysis are presented below. Specific Policy Recommendations Based on a discussion of the priorities, ICA proposes that the following steps be taken by the government to achieve its goals: Goal #1: Contract Enforcement 9 · The legal framework simplifying contract enforcement procedures should be further developed and applied. The new Law on Execution Procedure4 and the the new Law on Civil Procedures have been adopted in November 2004. The introduction of new concepts--the establishment of appropriate treatment of debtors and creditors, coupled with non-repressive acceleration of enforcement procedures is expected to help clear the backlog in the courts and facilitate doing business in Serbia. · The performance of bailiffs should be strengthened. By-laws on bailiffs should be adopted followed by training. This should strengthen their skills and lower the possibility of delaying the execution and corruption. · Court resources should be improved.5 Education of judges should be a priority. As the new legislation will broaden the competences of the commercial courts their enforcement benches should undergo additional training. Besides the state of the Serbian court system is such that urgent resources for training of judges and court personnel in the area of case filing and tracking systems, etc. are needed. · Information-sharing systems should be improved. Although Serbia has established public credit registry, it ranks among the 10 countries worldwide in which the legal and regulatory environment is least conducive to information sharing. Therefore it is important to better coordinate and facilitate operations of enforcement benches, and strengthen the link between commercial court enforcement benches and the competent registries (registry of securities, pledge registry and organization for coercive compensation). · Effective usage of court statistics should be developed and used. Empirical data are necessary to identify bottlenecks in court performance and give policymakers information where reforms are needed most. This would improve court administration and case management systems. · Removal of non-dispute cases from the courts. The new registration system in Serbia will transfer business registration away from the courts to the Business Registration Agency freeing up judges' time to deal with proper disputes. Goal #2: Access To Credit · Pass a new Mortgage Law and establish the Mortgages Registry. The development of sustainable mortgage finance is still deterred by problematic enforcement of contracts unsuitable legal framework and underdeveloped practice. 4The Law on Enforcement Procedure is being prepared by a Working Group appointed by the Minister of Justice in 2004. 5It is clear that successful implementation of the law will depend upon the other reforms in this area. USAID will provide US$15 million for reform of commercial courts, and putting the right legal framework in place is a precondition for court reforms. 10 · Establish and streamline a publicly accessible leasing registry. Amendments that will reflect changes with respect to the introduction of VAT also should be prepared. · Increase the volume of the credit lines at unsubsidized interest rates and only through financial intermediaries that meet minimum eligibility criteria. However, such credit lines only makes sense if Serbia maintains macroeconomic stability and further develops sound financial sector and prudential regulation with improved supervision. · Establish a sound legal framework for the emergence and operations of investment funds to encourage savings while protecting investors, especially non-institutional non-sophisticated investors. · Introduce improved disclosure and financial reporting of state-supported funds (the Development Fund, the Guarantee Fund and the Transition Fund). Linking the disbursements of severance pay from the Transition Fund and other budget subsidies should be based solely on the fulfillment of the Restructuring Plan prepared by the Privatization Agency. Goal #3: Regulatory Burden · Introduce public consultation on new laws as a rule, not as an exception. There is a need to introduce public hearings and public discussion of laws and decrees. Recently, the Regulatory Reform Council initiated reform of the Rules of Operations of the Government to establish mandatory public discussion. This could significantly improve the transparency of policy making by bringing in a wider range of affected interests. The Government has already conducted several successful public consultation processes, such as a public-private sector dialogue on the draft Leasing Law, the Company Law, the Business Registration Law and the Bankruptcy Law. · Enhance the capacity of the Regulatory Council to implement Regulatory Impact Analysis (RIA). The objective of RIA is to reduce the cost of regulation on the economy and to improve the performance of new laws, by checking the need for and impacts of proposed actions before they are adopted. Regulatory reform should not be regarded as a one-time measure, but as a continuous process. In order to improve the quality of regulations and to anticipate their impact on the economy, the Government has introduced as a mandatory step an expanded justification statement for all laws and other regulations that explains the expected benefits and costs of the actions, including the results of public consultations, which will have to precede each law or regulation. A step-by-step approach based on an OECD model and expertise is recommended to build the capacity for RIA into its policy processes. · The effectiveness of the inspections system should be improved. The amendments to existing legal regulations should clearly define and separate the currently overlapping responsibility areas of market inspection vis-à-vis other inspection authorities 11 · The national system for accreditation and certification should be adjusted to meet EU standards and technical norms. An incomplete system of accreditation, certification, and conformity assessment pose a serious barrier to the unhindered flow of goods and services. Since the country is aiming for EU accession, a solution to the problem of technical obstacles lies in adopting EU norms relating to the standards and technical regulations. The adoption of European solutions in this area would enable the development of a recognized system of accreditation, standardization and conformity assessment. Goal #4: Strengthening Property Rights In Land And Reducing Labor Rigidities · The Denationalization Law should be based on the following general principles: (i) compensation will be done, as a matter of general policy, through bonds, and will be substantially in accordance with principles of 2001 Privatization Law; (ii) compensation will be done in a fiscally prudent manner; and (iii) to avoid the creation of new injustices, the legally owned non-state property shall not be restituted (in kind) to original owners; other forms of compensation will be used. The resolution of the restitution process should be based upon the best practice from the other transition countries, (not jeopardizing privatization process and taking into account the fiscal impact. · Accelerate the legalization of buildings that are already recognized by most authorities. · Accelerate the ongoing work on the land administration system, supported by the World Bank Cadastre Project. · The newly proposed amendment to the Labor Law requiring collective labor agreements in businesses with as few as 10 employees and the clauses regarding employee rights in bankruptcy should be abandoned. Goal #5: Facilitating Entry, Ownership, Competition Change And Bankruptcy. · Continue the rapid implementation of the Laws on Business Registration and on the Business Registration Services Agency (BRSA). The Government should also continue to build the capacity the agency, which will house the company, leasing and collateral registries. · Establish the institutions, raise the resources and train the personnel required to build the Supervisory Agency and the unit which will serve as administrator of socially-owned companies. Although the new Law on Bankruptcy was adopted in July 2004, the Law alone is not expected to improve creditor rights unless the institutions are established and unless the Government shows the political will to initiate bankruptcy proceedings in its role as creditor of SOEs through the tax administration, the electric power company (EPS), and state-owned banks. In view of the experience of other post socialist economies where bankruptcy cases were fraught with collusion among bankruptcy administrators, judges and creditors, the creation of a specialized institution that will take over bankruptcy administration of a socially- or state-owned enterprise is promising if properly implemented. 12 However, the Government did not decide in which agency the specialized institution would be located. Given its weakening capacity, if the Privatization Agency is selected, it will need significant help to implement this new task. Yet, the agency's new management is reluctant to use foreign consultants while it keeps losing its best local professionals. · Adopt the new Competition Law. Adoption of a new Competition Law, expected in 2005, is necessary but far from sufficient for its implementation and impact. The latter requires a powerful and independent authority that will have the power to deny firm mergers and set maximum prices for monopoly products (sometimes at the behest of the monopoly or the cartel). Most difficult for the competition authority will be to order and oversee the break up of companies that abuse their dominant position, in particular of those companies controlled by the most powerful vested interests in the economy. In essence, the establishment of such an authority will not be an easy task, and it is unlikely that the authority will be fully operational for several years. The Government should agree with the European Union (EU) about the draft of the Law and start building the capacity of the authority by organizing training for the future officials. 13 ) yrs3 ver Long-term (o -term yrs)3ot X X X X Timing (1 Medium ) -termt yr1naht X X X X Shor ess(l rityoirP h h Hig Hig umdie h h h Hig Hig Hig umdie umdie M M M Plan ytilibisn no y e e cil CRR E,O Action unoC atorlu (CRR), M CRR E T T T slativi CRR, slativi J,O M MO MO MO MO Reg Leg Secreteriat GoS, Leg Secreteriat Respo ent GoS, Reform GoS, SBRA, MOF, 14 m ht ots ysisal wi d ent s to sse eht ld thats on ital the yel An for local Develop actp onitatul an ss ofnoi cap n ns wsaltf endm lation acce am build Im coci .ntemn shgni creat busines Regu eth in sinuBethfo ou dra of ildub awL(dte mumi each op etitiopm effectivsn min and ry w with internet acity ng;itad ad Coa budgeteht tioc lato Regutcu ublprof neno duceortnI veroG rysti for s). in rity free acity erb fun lish ng upso ns.oiat Reg leb (cap secula updnar edent num weringol- Law.nio tions programno nd esrud ofnoi capeth ena cy ild enm etitpm estab itstuo gr en bu ern tities) Co to funditn tly. Ac and co carry en d actina oce ed gulre lege Agn ansferta enebsahtcAyna En to to fectaf dat ide mp ents. Lawae sufficie ndepe arep ). acity prhsibla raltneCeht nte and tifyingn lish sitivop .sn gove rnm lish Co icmo weneth Pre cap (RIA Est joram joramr ovr reu heot eratpOfoseluReht latio stratioi Estab with Gove regu Estab Reg udeclni que serv uni will weN act ority.ht orityht indd Econ requirem En App Au Ens Au an Recommende · · · · · · · · n Area andy rofk or teg ewm tionlau etitio Stra fra reged mp Business registration Co X X X X X X X X h h Hig Hig umdie umdie umdie umdie umdie umdie M M M M M M cali PAd rehto rehto n oni PA an E T T lan naloi iatio cyn mmer ts E, esirtsi esirtsi MO MOT, MOT, MO MO Nat, zatidr ur Co MO min min Natio Age Co Accred CRR, CRR, odyB anda St 15 d ni in la to c tcy SOEs. an fo et up EU nkrab tcy Es.OS se specifi eduresc and bankrupt of ts licena otdedne itiesr rmits.ep yna ope tho and andn au rkameth es.ussi ngi Law. zed ofe nkrupab educlnidluo esristniimh ses reduce ortid sceht shs of nimimehtred tain ob ancrb worr tono cus nsio rform pe proethgni itatio s. eciali truste itiatein cre to of extebd localfo nkruptcy spa licengni foeth onitceotrp for fined accredr rmon as ase easurem umencod of nte ns.oi eetmotdestujda s. ities er fo cal Ba w ofgni issu spectini ift les act to neeht ildub willtnem vem to sh m bed activeht nte nte oulhs encyga rnev rolstinisgndi onital erbm nadluohs um recdluohs ents gure de nueht ecessaryns volnieht proces oulhssrtoffe ruehtd spectnieht an systela izationr ater related rnm rketameht andsn of onscnonoi rnm quiremre oni echnitdnas rd acity Goe her cing licensing catifi ocee Enforce Cap trustee Th pr Furt duer tho au ngitimil onitalug byd the Dere corpoin areaseth oveG The work licatiopud oveG ities spectni use tionane The technical activ be Th cert andast · · · · · · · · itxe d an n ng tio noitazdi dar censi d an Business Li Inspections Inspections Certifica an St X X X X X X X h Hig umdie h Hig umdie umdie h h - Hig Hig M M M cialer cialer d J J EO MO mmoC urts Co mmoC urts anJ F M Co MO ciationso MO F, NBS MO Business Ass MO 16 MOJ, MOJ, fo ec ral . thedna eenbevah be cas urt evedhi suran gn . . edar onlydna be lduohs ngniairT. e eb d. rof ge proved ld se .s of im ou garto eth sh devi ac judicial M to ep IFIseht .d collate be al Leasila rates eetm .g edvo area bed coeht (ADR)" tionlu hodstem wena eth Inegagrt to sedvirey hedsibla ect preb ofe atth Law ouldhs liffs est statistics resp ld procedurest accessible nanciiF volum terestni es)r bai prmi the form in needesis shoulds oulhs omrf beosald resoetu ofyteri thee Mola adopt usloiut ly bed on shou iaries en educ on bailiffs of bedluo gye related ld ca oul w the ed onnelrs emst stem ourtc of ratst tsid in izedd ou sals sh withseg also criteria. sy syg casesde coulemti spide vaagn ou public T forcemnE Pro -laws aininrtybdew pe an ch ance By. shs follo ngik arin VA usage sputid es' propo Nationethfot bed Laeht and n- tos increase unsubsi ilitybi on w (Civiln edn udgj ofn at elig and urceo .d. ation-sh onitpurroc-i no sputesid up "alternativ udiclnisms rale lined ent an . shtnemnrev enm rystigregn reflect ctio terminlaicnanfi La atio ted. perform theg ted op op The Litig ad The stren ad restruoC urtocdna acrtdna lines rm effective ope Sev ugh lish erd judges ngilfi ant fo In and devel An ovemeR ingvl easil will ugh mumi eeing endm Fr roth echanim ocess Goe reso pr Th Law. estab oulhsnoiatroproC stream A and Am atth odutrin Consi credit roth min · · · · · · · · ct ement to ntra Co Enforc Access finance X X X X X X X X X h h h h h Hig Hig Hig Hig Hig umdie umdie umdie M M M F SBA LSA National MO A, fficeOro F F RGZ, MCI MO MO MCI MOF, MO LSO Lab MCI 17 M eb to ngillornt d an atht m Office .de nges syste seeh(t . tax e tion). dnagn fis eedsns co na llufht ysaw fonoi ors dnaln sfer tiva lish in titus nnia entm undftr forsms auerubit rtop wiu rcept rketam ba lizationan tran estab edcr eta gnri Pln end supot rea but otstpmetat Laborlan chadna sharing rty tioa est strinim Law ba Natio ation entsm peeht ope vestning ureht Cont estate pr rei of alre ad quicafo ecessaryn est Urno am ppo su echanim bedlu adopt rm end ld der sho ou edircd stsired fo Denno zei am real gen ealreht ngivorp sugg oul ocess Law endmmo state- ousr orni 0012ehte onali base currenethyb sht ang ativel imno andd of cenam orf goridna,det itiesv lan acti'sd shtnemnrev oniatslgiel d- .snoitrostdie ility,bomro chaeht oa ation. ratp br,evi enn ch tly labe infoethevorp provemiotdre ided Lawethtp ngsiaer legis quire ado dec .s prehtni recdnanoi inform ovrp the,reuntefo actrtatotdnagn ovrpt held and erd acyv esrud tionc entrrucehte uctr ry. Per audi funeth Goe veroG canifin ducortnidul im imdl are nci ou udeclnidlu oulw saction studyaear uat Th pri effect credit .The sig wo To sh Introduce atht security nafi epcn sho co Draft onsiC atht ep oce nstru tran Pr pr co measures. Eval onstC necessa · · · · · · · · · aleR d an d te Labour Lan Esta CHAPTER 1: INTRODUCTION Background and Objective Three years have passed since the October 2000 "velvet revolution" in Serbia and the establishment of the first government in January 2001 that lasted until February 2004. During these three years, the investment climate was heavily affected by political instability. In particular, the country experience uncertainty related to the future of the union of Serbia and Montenegro and the future status of Kosovo; the ICTY Hague criminal tribunal; and doubt about the country's creditworthiness. After the assassination of Prime Minister Djindjic in March 2003, the situation deteriorated into a stalemate in Parliament, disintegration of the ruling coalition, a collapse of the government, and new parliamentary elections which took place in December 2003. These events have increased the difficulties of moving ahead with the necessary reforms. In the 1990's, the loss of markets, international isolation, armed conflicts, and interruption of long-established production relations have left a dire legacy for the Serbian economy. In the economic sphere, several major policy steps conducive to a better investment climate were implemented: fiscal stability and trade liberalization was achieved; tax code reforms were carried out; the company law was amended to eliminate pre-registration inspections; a privatization law was adopted in 2001 and amended in 2003; leasing, collateral and concessions laws were introduced in 2003; and tax reforms were implemented in 2001. Accordingly, 4,000 new small enterprises and 20,000 shops have been opened. In addition, around 1,000 mostly medium-sized enterprises were privatized. Yet, the remaining investment climate reform agenda is huge and presents the new Government with major challenges. The objective of this Investment Climate Assessment (ICA) is to (i) provide Serbian experts in academia and in government, policy makers, as well as Bank staff conducting policy dialogue with the Government, with an empirical analysis of the investment climate in Serbia; and (ii) provide policy options for creating an enabling an environment conducive to private sector development, thereby increasing and maintaining enterprise productivity and leading to sustainable growth. With Bank and other donor support, the former Government began to take important actions in all of the above mentioned policy areas. As will be shown below, the ICA recommends that the new government continue on this path ("stay the course") and support the next generation of reforms to improve the country's investment climate. What is the Investment Climate? Work to improve the investment climate is recognized as a key pillar of the World Bank Group to promote economic growth and poverty alleviation in developing countries. The World Development Report 2005 defines the investment climate as "the factors in a particular location that shape the opportunities and incentives for firms to invest productively, create jobs, and expand." Essentially, it is the economic, institutional, legal and cultural environment in which firms operate, and as such it has many dimensions: geography, competition, institutions, security of property rights, taxation, the operation of the markets for capital and labor, and so on. Some factors are difficult or impossible for governments to influence. But in most dimensions of the investment climate, governments have a powerful or decisive influence (Figure 1). Figure 1: Policy uncertainty dominates microeconomic constraints Crime Courts 3% 3% Policy Skills uncertainty 3% 33% Finance 5% Infrastructure 8% Corruption 11% Regulation 11% Tax 23% Source: World Development Report 2005, World Bank. (Investment Climate Surveys in 49 countries, based on rankings by country) Structure and Methodology of the Serbia ICA The Serbia ICA systematically analyzes the conditions for private investment and enterprise growth, drawing on the experience of local firms to identify the areas where reform is most needed to improve the private sector's productivity and competitiveness. Much of the analysis takes place at two levels. First, the ICA assesses the country climate, drawing on the many datasets now available (Box 1) to compare Serbia across countries or groups of countries. Second, the ICA analyzes in detail various aspects of the investment climate in a country-specific context. This level of analysis draws on a wide range of evidence and sources, including but not limited to evidence gathered at the firm level. The report reviews all aspects of the investment climate, but, in order to be useful to experts and decision makers, it focuses on the subset of those barriers that can be influenced or changed through policy measures. 19 Box 1: Background on Surveys used in the ICA6 Productivity and Investment Climate Survey (World Bank): PICS utilizes a standard core questionnaire intended to calculate firm productivity. The PICS sample conducted in Serbia is different from the standard one as it does not cover exclusively manufacturing. The survey was administered to managers of firms and consisted of a core set of questions as well as several modules that can be used to explore specific aspects of the country's investment climate and links to firm- level productivity. The core survey had 11 sections: General information about the firm: ownership, activities, and location; Sales and supplies: imports and exports, supply and demand conditions, and competition; Investment climate constraints: evaluation of general obstacles; Infrastructure and services: power, water, transport, computers, and business services; Finance: sources of finance, terms of finance, financial services, auditing, and land ownership; Labor relations: worker skills, status and training, skill availability, over-employment, unionization, and strikes; Business-government relations: quality of public services, consistency of policy and administration, customs processing, regulatory compliance costs (management time, delays, bribes), informality, and capture; Conflict resolution/legal environment: confidence in legal system and resolution of credit disputes; Crime: security costs, cost of crimes, and use and performance of police services; Capacity, innovation, and learning: utilization, new products, planning horizon, sources of technology, worker and management education, and experience; and Productivity information: employment level, and balance sheet information (including income, main costs and assets). The sample size was relatively large with 408 firms and includes a representative sample of the composition of the Serbian economy. The PICS survey was undertaken in the aftermath of the assassination of the Prime Minister at a time when the country was experiencing a great deal of uncertainty, so the survey results must be viewed in this light. Business Environment and Enterprise Performance Survey (EBRD and World Bank): The BEEPS utilizes a standard survey instrument applied to nearly all countries in Eastern Europe and Central Asia, thus ensuring comparability. BEEPS II is a follow-up of an earlier BEEPS effort. In Serbia the BEEPS II survey had a sample size of approximately 230 firms in 2002 (BEEPS I did not cover Serbia). Generally, the sampling strategy in BEEPS II differs from that of the PICS, as the sample design of the BEEPS is highly skewed toward smaller firms. Cost of Doing Business Project (World Bank): The Cost of Doing Business Project investigates regulations that enhance business activity and those that constrain it. Quantitative indicators on business regulations and enforcement are being gathered to allow comparisons across more than 130 countries over time. The indicators are based on assessments of laws and regulations, with input from and verification by local experts. Information is gathered from surveys of experts rather than firms. The questionnaire is administered to local professionals experienced in their fields, such as corporate lawyers and consultants for business entry, or litigation lawyers and judges for contract enforcement. 6 PICS 2 was initiated before the Memo of Understanding (MOU) arrangement between the World Bank and EBRD on the BEEPS. In Serbia, PICS was programmed by the Bank in advance of the MOU with the EBRD, hence had to be grandfathered. However, it should be acknowledged that BEEPS 2 was a collaborative effort under the MOU, and that Serbia is exceptional in having both a detailed PICS 2 and the BEEPS. 20 CHAPTER 2: PRODUCTIVITY, OWNERSHIP AND THE INVESTMENT CLIMATE IN SERBIA AND WORLDWIDE Introduction This chapter provides an overview of Serbia's investment climate in an international context. First, we examine how Serbia compares to other countries, in particular to other transition economies, in terms of its output and productivity. Second, we review which characteristics of firms worldwide are associated with high productivity and strong performance, show how investment climate variables affect productivity, and show how the characteristics of firms worldwide affect complaints about the investment climate. Third, we show which Serbian firms have high productivity and are growing quickly, and present the investment climate constraints that they face, using the PICS sample of over 400 Serbian firms. TFP, Productivity, and Serbia in the International Context Total factor productivity (TFP) is a multi-factor productivity measure that represents the efficiency of the firm in transforming factor inputs into outputs. A firm that has a high level of TFP is one that can produce a high level of output for given quantities of capital and labor. TFP can be affected by wide range of factors: technology, managerial quality and incentives, corporate governance, government policies, and, of course, various dimensions of the investment climate. TFP is usually analyzed in a production function or growth accounting framework. One approach is to account for the contribution of measured inputs, and label the "residual" ­ how much output is produced taking into account the volume of inputs ­ as an estimate of TFP. Another approach is to estimate the production function to try to measure the additional factors that affect productivity and output: when a characteristic of firm such as private ownership is associated with higher output, we say that after accounting for the contribution of the factors of production (capital and labor) to output, private ownership has a positive impact on TFP. Increases in the level of TFP of firms are an essential feature of economic growth: rich countries are countries with firms that are highly productive. This section demonstrates how TFP at the firm level is related to GDP per capita at the country level, and compares the level of TFP in Serbian firms with the levels in comparator countries. These comparisons use the large PICS-BEEPS dataset of about 27,000 firms from 50 countries.7 To obtain estimates of the average TFP level by country, the production function equation is first estimated on country-averages: averages were calculated for firm 7Estimating TFP requires measures of capital and output. For both PICS and BEEPS we could use the book value of fixed capital as a measure of the capital input; value added, however, is available in PICS but not BEEPS. We therefore used sales as the measure of output. The production function estimations were generally acceptable, however, and the strong correlation between estimated TFP and GDP per capita reported below is a good validation check of the results. See the appendix for further discussion. 21 capital, labor and output were calculated for each of the 55 country surveys, and a regression estimated using these 55 observations. Appendix 2 contains further details on the methodology and results of the analysis. The basic estimation results are reported in column 1 of Table 1; the coefficients in the parentheses show the productivity gaps between Serbia and selected countries. Croatia, Hungary, Poland and Slovenia all had higher TFP than Serbia in 2002 or 2003.8 The TFP "residuals" as defined above are plotted in Figure 2 versus GDP per capita. The figure shows the strong positive correlation between country-average TFP and GDP per capita: richer countries have more productive firms.9 Table 1: Basic TFP regressions using country-survey-averages and firm-level data (manufacturing only) Regression on country- Regression on firms with survey-averages country-survey dummies Fixed capital 0.574** 0.387** Labor 0.445** 0.663** Country survey dummy variables included No Yes Selected estimates of relative TFP Benchmark: Serbia 2002=0 Croatia 2002 [0.612] 0.798** Hungary 2002 [0.817] 0.922** Poland 2002 [0.543] 0.717** Poland 2003 [0.938] 1.087** Slovenia 2002 [1.146] 1.347** Number of observations 55 country-survey-averages 14,687 firms based on 14,687 firms ** significant at the 1 percent ; heteroskedastic-robust standard errors; [ ] indicates estimate based on a residual The figure and the table also show that Serbia is well below its East European peers in productivity. Croatia is a natural comparator country. The country-survey-averages TFP regression implies that the productivity of Serbian manufacturing firms in the BEEPS 2002 survey is 61 percent10 below that of Croatian firms; the gap based on the PICS 2003 survey is even greater, 93 percent behind Croatia.11 The TFP gaps with other leading East European countries are as big or bigger. In column 2 of Table 1 we show an estimation of the production function which uses the same PICS-BEEPS dataset as the country-averages estimation but with dummy 8 Table 3.1 reports TFP gaps relative to Serbia in 2002. The TFP estimates for firms in the Serbian PICS 2003 survey showed lower productivity levels than in the Serbian BEEPS 2002 survey: the estimated gaps were -0.321 for the regression using country-survey averages (column 1) and -0.286 for the regression using country-survey dummies (column 2). The latter figure was not statistically significantly different from the Serbia 2002 benchmark. 9The results also confirm that TFP estimates are feasible using the input and output measures available. 10In log percentages, e0.61=1.84, so this means Croatian TFP is 84 percent higher than Serbian TFP using standard percentages. 110.61+0.32=0.93 22 variables included for each country. The coefficients of the country dummies capture country-average TFP.12 The estimated difference in TFP levels between Serbian manufacturing firms in the BEEPS 2002 survey and Croatian firms is 80 percent13, and 108 percent using the 2003 Serbia PICS survey. These differences are statistically highly significant.14 These cross-country differences in TFP partly reflect the differences in GDP per capita between the countries, as is apparent in the figure. But lower TFP cannot be blamed on having too little capital: holding capital and labor constant, Serbia is generating much lower output than Croatia, Slovenia, or the other East European countries that have joined the EU. Serbia will need to invest in new fixed capital. However, to catch up with these countries, it will also need to invest in improving the investment climate, and in changing incentives for firms--including privatizing those that are still socially- or state-owned.15 Figure 2: GDP per capita and TFP, country-survey-averages regression 1 Bulgaria2002 Slovenia2002 Latvia2002 Peru2002 Lithuania2002 Estonia2002 Poland2003 .5 Brazil2003 Slovakia2002 Hungary2002 Moldova2002 Indonesia2003 Morocco2000Turkey2002 Ecuador2003 Croatia2002 Azerbaijan2002 India2002 Poland2002 Uzbekistan2002 Kyrgyzstan2002 India2000 Algeria2002 0 Armenia2002Bolivia2000 Georgia2002 Czech2002 Zambia2002 Philippines2003Romania2002 Bangladesh2002 Pakistan2002 Belarus2002 Kazakhstan2002 Guatemala2003 Honduras2003 China2002 Serbia2China2003BiH2002 Russia2002 002 Cambodia2003 Tanzania2003 Moldova2003 -.5 Ukraine2002 Albania2002 Montenegro2003 Nicaragua2003 Serbia2003 Ethiopia2002 FYROM2002 -1 Tajikistan2002Kyrgyzstan2003 Uzbekistan2003 Tajikistan2003 -1.5 5 6 7 8 9 Log GDP per capita in US dollars TFP (between regression residual) Fitted values 12See Appendix 2 for further details. 13 In log terms. 14The coefficients on the country dummy variables are the estimated TFP differences reported in column 2 of Table 1, and as estimated coefficients, can be the subject of standard statistical significance tests. The differences reported in column 1 are residuals and hence these standard tests do not apply. See Appendix 2 for further discussion. 15In addition Serbia needs to invest in schooling as well as in R&D to better link its universities and science institutes with the private business sector. 23 What types of firms perform better than others? Evidence from the PICS- BEEPS surveys In this section we account for the variation in productivity (TFP) across firms that can be attributed to various characteristics of firms. The key characteristics of interest are ownership and export orientation; it is particularly important in the Serbian context to establish that, based on international experience, private ownership enhances productivity. The previous section also established that the productivity of firms is highly correlated with the country in which they operate. The analysis here augments the estimation equation used in column 2 in Table 1 with measures of ownership, export orientation, and also location (whether or not the firm is operating in a capital city). In the regression results, the estimated impacts on TFP of these firm characteristics are impacts that raise or lower the TFP of a firm compared to other firms in the same country. International evidence. The estimation used a sample of firms from 51 PICS-BEEPS surveys of firms in 44 countries. The estimations were done separately for manufacturing and services firms (Table 2). The omitted or benchmark category is state-owned, domestic, non-exporters and not in a capital city.16 The results show that privately-owned firms, whether privatized or new private, have levels of TFP that are 30-38 percent higher than state-owned firms in the same country. Foreign ownership adds another 23-43 percent to TFP levels. Export orientation adds 22 percent to TFP (for manufacturing firms only).17 Table 2: TFP regressions using firm-level data Manufacturing Services Fixed capital 0.374** 0.268** Labor 0.686** 0.698** Privatized 0.382** 0.295** New private 0.382** 0.289** Foreign-owned 0.433** 0.226** Exporter 0.216** 0.089 Capital city 0.092 0.178** Country-survey dummy variables included Yes Yes Number of observations 11,364 firms from 51 country 3,314 firms from 35 country surveys surveys ** significant at the 1% level; heteroskedastic- and cluster-robust standard errors This analysis can be extended to other measures of firm performance. Appendix 2 reports the results using whether or not the firm reinvested any profits, invested in new plant in the preceding three years, and the log growth rate in its level of employment. The results reinforce those for TFP: private ownership and export orientation are good for firm performance. 16 The results were similar when the data were pooled and sector dummies included (see Table 3.3 below). In alternative regressions, an SME dummy was included but was insignificant, probably because size is already captured by the capital and employment variables. 17All in log percentages. 24 Serbian evidence. What are the characteristics of productive, growing firms in Serbia, and how do these compare to those in the other PICS-BEEPS countries? We answer the first question by estimating the same regressions using only the sample of Serbian firms in the PICS-BEEPS data. The second question is addressed by testing whether the estimated effects for Serbian firms are different from those for firms in other countries. The 2003 Serbia PICS survey covered 408 firms; the 2002 Serbia BEEPS survey covered an additional 230 firms. About half of these firms were in services, about 10 percent were in construction, and the rest were manufacturing firms of varying sorts. Two thirds of the sample were small and medium-sized firms (employing 100 or more persons). Over half (57 percent) of the sample was composed of new private firms, 29 percent were socially- or state-owned, and the remainder had been privatized. Almost one-quarter of the sample were exporting, and about one in ten had significant foreign ownership. Because of the relatively small Serbian sample, the estimations of pool manufacturing, construction and services firms and employ sector dummy variables; the results are similar if the regressions are done separately by sector. Table 3 shows the results of the TFP estimations. Table 3: TFP regressions using firm-level data ­ Serbia vs. other PICS-BEEPS countries 49 country Serbia different surveys Serbia from 49 others? Fixed capital 0.374** 0.394** No Labor 0.693** 0.590** No Privatized 0.359** 0.599** No New private 0.359** 0.846** Yes* Foreign-owned 0.205** 0.282 No Exporter 0.216** 0.089 No Capital city 0.089** 0.398* No Country-survey dummy variables included Yes Sector dummy variables included Yes Number of observations 15,050 firms from 51 country surveys * significant at the 5 percent level; ** significant at the 1percent level; heteroskedastic-robust standard errors The estimations for Serbian firms show that new private firms in Serbia have productivity levels that are 85 percent higher18 than the benchmark socially-owned firms, and the productivity of privatized firms is 60 percent higher. The estimations based only on the Serbian firms in the sample do not show that foreign ownership and export activity are associated with high TFP levels, but this result is probably due to the relatively small sample size of the Serbian surveys. The last column in Table 3 compares the effect of the firm characteristics in Serbia to those in other countries: there is no evidence that Serbia is any different from other countries, with one exception--new private firms are even more strongly associated in Serbia with higher TFP than they are in other countries. Thus Serbia provides positive evidence in support of the general pattern found internationally, and no evidence against it. 18In log terms. 25 These findings extend to the other three measures of performance used above (see Appendix 2 for details). New private firms in Serbia are more likely to have higher employment growth, to have reinvested in profits, and to have invested in new plant, than state/socially-owned firms; Serbian exporters are also more likely to have invested in new plant. Moreover, there is no evidence that Serbian firms are any different from their counterparts in other countries in any of these dimensions: there are no significant differences between the estimates for Serbia and the other countries for any firm characteristic. These results need to be interpreted with some caution because of the possibility of reverse causality. For example, privatization may boost TFP, but higher productivity firms may also be more likely to be privatized or sold to foreign investors. Serbia provides an example: the 1997 privatization law allowed employees to take firms private if they so chose, and many of the best firms were privatized in this way; the better performance of some privatized firms in Serbia may simply reflect the fact that only employees in the better firms decided to take their firms private. The results for new entrants, however, are unambiguous: new private firms in Serbia, like those in the rest of the world, are more productive, more profitable, and growing more rapidly than state/socially-owned firms. The problem is simply that there are not enough of them: although the new private firms are better performers, their scope for growth, and the possibility of entry by new entrepreneurs, are both severely limited by the investment climate. Below the aspects of the investment climate that are particular constraints for these and other firms are analyzed. Later chapters of this ICA discusses what needs to be done to create an investment climate that supports entrepreneurship and facilitates entry, both greenfield and brownfield. What aspects of the investment climate affect TFP? Evidence from the PICS- BEEPS surveys The investment climate is an important component of the determinants of TFP. This section extends the analysis in the previous section to examine the impact of the investment climate on TFP, using the reported perceptions of firms of the major obstacles to conducting business that they face. An important qualification must be made: these are the reported perceptions of firms about the business obstacles they face. Firm characteristics and firm performance will be a major determinant these perceived constraints. This creates a major reverse causality problem: the reported constraints on doing business are determined partly by the country-average investment climate, but also by the particular situation the firm faces. Consider, for example, a rapidly growing firm in a country with a good average investment climate including a skilled labor force. The managers of this rapidly growing firm may still perceive difficulties in finding skilled people to hire. They will report that compared to the typical firm in this country, it faces skills shortages, but this perceived skills shortage is the result of its rapid growth. Yet the country as a whole may not have a skills shortage. The correlation between growth and skill shortages is better interpreted as evidence that rapidly growing firms are more likely to face skills 26 shortages than stagnating firms. This line of analysis is picked up in the next section, which asks which constraints are better performing firms likely to face. An approach to estimating the impact of the country-average investment climate on TFP that possibly mitigates the reverse-causality problem is to use the country- averages regression framework. The average responses reported by firms in a country about the business obstacles they face are estimates of that country's average investment; the individual variations in investment micro-climate are "averaged away", leaving an estimated country-level investment climate. The BEEPS and PICS datasets have a very rich set of investment indicators, and the question is how to make use of them. Two sets of investment climate indicators were used. The first set was the simple country-survey-average of each of the raw indicators reported by firms, scaled from 1 (no obstacle) to 4 (major obstacle). These variables are those in the block of questions used in the factor analysis reported in the previous chapter. The second set used sets of indicators corresponding to aspects of the investment climate, combined an analysis described in Appendix 2, (e.g., a single "finance" indicator was created using the responses to questions about financial access and cost of finance). The other combined indicators were for infrastructure, taxation, regulation, the macroeconomic environment, and property rights (corruption and crime). These were scaled to be comparable to the raw data on perceived business obstacles (the minimum was 1 and the maximum was 4). In addition to augmenting the country-survey-averages regression equation (2) with the average investment climate variable, country-survey-averages of ownership variables (privatized, new private, foreign owned) and export orientation were also included. The analysis was conducted on 50-51 country-survey-averages based on about 14,000 firms from manufacturing, services and construction; the analysis separating manufacturing and services firms yielded similar results. In these TFP estimations most of the investment climate measures in the regressions were insignificant, with one very clear and consistent exception: infrastructure constraints generate lower TFP, with an estimated coefficient of about ­0.7. Countries in which firms complain a lot about infrastructure are, ceteris paribus, countries with less productive firms. To put the estimated impact of infrastructure constraints in the Serbian context, the Serbian surveys average for the combined infrastructure investment climate constraint measure is 1.9; in Croatia and Slovenia it was 1.3. If the infrastructure constraint perceived by Serbian firms fell to Croatian or Slovenian levels, the impact on TFP would be (­0.6)*(­0.7) = +0.42. This represents an increase in average TFP of about 40 percent in log terms. What types of firms report investment climate constraints? Evidence from the PICS-BEEPS surveys The previous section pointed out that what firms report about the investment climate constraints they face is determined in part by the characteristics of the firm. This section, on the other hand, explores what types of constraints various types of firms face. 27 A simple and straightforward approach adopted, for example, in the Malaysia ICA, consists of splitting the sample of firms into groups according to the characteristics that affect performance in the regression analysis above or according to their actual economic performance, and identifying the differences in various groups of firms' perceptions of the investment climate. Thus, for example, it is possible to group firms according to their ownership, and ask whether privatized firms, new private firms, or state-owned firms, perceive infrastructure as a major constraint; or they can be grouped according to their estimated TFP or employment growth rates, and the same question asked. The main idea of the approach is to identify the differences in the investment climate faced by good performers versus that faced by bad performers. This section takes the analysis a step further and uses regression analysis to answer the question: compared to other, similar, firms in the same country, what types of firms report significant investment climate constraints? The advantages of using a multivariate regression framework are that it allows for formal statistical testing, and it controls for multiple characteristics of firms simultaneously. Reverse causality is also an issue here. It is possible that better performing firms will report that they are more likely to face certain constraints, but it is also possible that the causality can run the other direction: firms that are particularly constrained may suffer in their performance. Nevertheless, in the results reported below it is usually possible to identify which channel dominates. The reason is that when causality runs from the investment climate to TFP and other performance measures, we expect bigger constraints to generate lower performance (i.e., there will be a negative coefficient on the investment climate constraint measures). When, however, causality runs from performance measures to the perceived investment climate constraints, we expect better performance to generate higher reported IC constraints (i.e., there will be a positive coefficient, because rapidly growing, high productivity firms are more likely to be constrained). The exception to this is financial constraints, where poorly performing firms are if anything more likely to report that they are short of money than growing, profitable ones. Thus for most constraints, the two effects work in the opposite direction of eachother, and if we find coefficients that are significantly different from zero, this will tell use which effect dominates in the data. In fact, the second channel usually dominates: the coefficients are usually positive, meaning that better performing firms face bigger constraints. The exception also fits with the priors expressed above--poorly performing firms report bigger financial constraints. International evidence. The main results from the 51 PICS and BEEFS surveys are summarized as follows and discussed in more detail in Appendix 2: · Privatized and especially new private firms perceive greater business constraints practically across the board compared to similar state-owned firms in the same country. · Foreign-owned firms also generally perceive greater business constraints, except for constraints related to finance and crime, about which they complain less than domestically-owned firms. 28 · Exporters state they face greater constraints from infrastructure, skills shortages and regulation (especially, and not surprisingly, customs regulations), and complain less about crime, than do comparable non-exporting firms. · Large firms complain more about infrastructure, regulation, and skills shortages, and less about financial constraints, than do smaller firms. · Firms in capital cities, not surprisingly, complain less about electricity cuts and more about land access. They are also more likely to face perceived skills shortages, and complain more about the macroeconomic environment and about regulation. There are also are clear relationships between performance and perceived investment climate constraints: · High TFP firms perceive higher constraints deriving from regulation, and complain less about finance. · High employment growth firms face greater constraints from land access, regulation, and complain less about finance. · Profit-reinvesting firms, and firms that have opened plants in the preceding three years, complain more about infrastructure, more about regulation, more about skills shortages, and more about property rights. The interpretation of these results is straightforward: firms that have high productivity, are growing and investing, and that have characteristics that are associated with good performance and growth (private/foreign ownership and export orientation) are more likely to report that they face significant investment climate constraints of all sorts. The exception is finance, where poorer performers are more likely to complain they are constrained. The usefulness of these findings from the international PICS-BEEPS surveys is primarily to set a benchmark against which to compare the investment climate constraints perceived by Serbian firms. Serbian evidence. Do firms in Serbia face the same kinds of investment climate constraints as those types of firms in the rest of the world? The same methodology used above was applied to just the firms in the 2003 Serbian PICS survey. The investment climate variables are those described earlier, plus composite measures for infrastructure, taxation, regulation, macroeconomic conditions, finance, and property rights. We estimated the effects on these investment climate measures of ownership (privatized, new private, foreign-owned), export orientation, size, sector (manufacturing, agriculture, services), and location (operating in Belgrade, Vojvodina, or central Serbia). The main results from the analysis are summarized below and presented in more detail in Appendix 2: · Privatized firms face greater skills shortages and bigger constraints from business licensing (Figure 3). · New private firms also perceive licensing as an important constraint (Figure 3).19 19Note that skills shortages faced by new private firms appear lower than those faced by privatized and socially- owned firms in the figure, but this is only because the bar chart presents a simple correlation and does not control for other factors. In particular, SMEs tend to complain less about skills shortages, and new private firms are 29 · Foreign-owned firms perceive greater constraints than domestically-owned firms across almost the entire investment climate: infrastructure, regulation, macro conditions, and property rights problems. The only areas in which they do not perceive greater constraints are finance and taxation (Figure 4). · Exporters complain about excessive customs and trade regulations (Figure 5). Figure 3: Ownership and Business Figure 4: Domestic vs. Foreign Constraints Ownership and Business Constraints 60% 60.0 50% 50.0 40% 40.0 30% 30.0 20% 20.0 10% 10.0 0% 0.0 Cost of Financing Business licensing and S kills and education of Tax admin. Cost of Elect ricit y Transport Access t o operating permit s available workers Finance Land DENOVO P RIVATIZED S TATE/S OCIAL Domestic Foreign Figure 5: Exporters vs. Non-Exporters and Business Constraints 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 Customs and Transportat ion Access to Land Skills and trade regulations educat ion of available workers Export er Non-Exporter The characteristics of firms associated with greater perceived investment constraints are also characteristics that the earlier analysis showed was typical of productive, growing firms in the other PICS-BEEPS countries: private ownership, foreign ownership, and export activity. The exception, as before, is finance; these firms smaller than average. Once size is controlled for, the fewer complaints about skills shortages by new private firms disappears. 30 complain less about their finances, presumably because of the revenue their better performance generates.20 Conclusions The main lessons of the analysis are: · After accounting for labor and capital inputs, the productivity of Serbian firms is still well below that of its East and Southeast European competitors. · Cross-country evidence points to infrastructure as an important element of the investment climate that raises the productivity of firms. The same surveys show that complaints by firms about infrastructure constraints are substantially higher in Serbia than in comparator countries with higher productivity such as Croatia and Slovenia. · Evidence based on surveys of firms in PICS-BEEPS countries and in Serbia itself shows that private ownership, FDI and export-orientation are all strongly associated with firm productivity, investment and growth. · The same international and Serbian surveys also show that firms with the characteristics associated with strong firm performance are the firms that complain the most about investment climate constraints on the operation of their businesses. Reducing investment climate constraints will benefit primarily the firms that can generate growth. · Closing the productivity gap between Serbia and its near neighbors will require, as well as investment in fixed capital, improvements in the investment climate for firms, privatizing firms that are still socially- or state-owned, and investing in human capital, schooling and R&D. 20Performance measures were also included in the Serbia-only analysis, but no strong correlation between these measures and reported investment constraints was found. The likely explanation is the relatively small sample size for the Serbia survey, and multicollinearity between the performance measures and the other regressors (shown above to strongly correlated with each other). 31 CHAPTER 3: MACROECONOMIC CONTEXT AND INVESTMENT CLIMATE PRIORITIES Investment climate conditions: 1989-2000 Even though Serbia began its economic transition relatively well integrated into the world economy, the loss of markets, international isolation, armed conflicts, and interruption of long-established production relations have left a dire legacy. The period of 1989 to 2000 was marked by macroeconomic instability, aggravated by badly managed economic policies. Rigid real government spending together with eroding real revenues stimulated hyperinflation in 1992-93 that destroyed the national currency and all dinar-denominated assets. In addition to macroeconomic instability and international isolation, the real sector became increasingly inefficient due to the continued dominance of socially-owned firms and ill defined property rights. As the rule of law deteriorated, property rights became very weak. In addition, the regulatory burden was large due to frequent and discriminatory administrative control and taxation of companies. Price controls were also widely used to substitute for effective social policies and to suppress inflation, while the tax system was highly distortionary. Rigid labor regulation was used to protect employees in the social and state sector, preventing labor shedding. The commercial banking and payments systems were used for quasi-fiscal purposes and to administer state support in priority areas, limiting access to finance for the private sector. While a large share of the formal sector of the economy came to depend on explicit or implicit state subsidies, the growing informal sector developed sophisticated means of evading taxes and circumventing laws to finance its activities. This resulted in further expansion of grey economy, accumulation of losses in the real sector, and numerous microeconomic distortions. The use of the commercial banks and state-owned utilities to subsidize households and enterprises left these sectors devastated by the end of the 1990s. In addition to causing pervasive disruption to economic activity, the 1999 war resulted in significant damage to infrastructure. By 2000, recorded per capita GDP had fallen to below one half of its 1989 level. The Economic Reform Program In 2000 the new Government launched an ambitious reform program with the objective of removing large distortions in the economy and placing Serbia on sustainable growth path. The implementation of this program has been successful in establishing macroeconomic stability. By employing the exchange rate as an anchor, the central bank was able to curb inflation. Inflation declined from 111.9 percent in 2000 to 7.8 percent in 2003. Following the liberalization of the foreign exchange market and unification of the exchange rate in October 2000, the national currency has remained relatively stable, while official foreign reserves grew from US$516 million at end-2000 to US$3.55 billion at end-2003. International support, 32 immediately following upon political changes, made macroeconomic stabilization possible. Serbia's external debt declined significantly from 132 percent of GDP in 2000 to 71 percent in 2003. The successful implementation of stabilization and reform program has played an important role in promoting economic recovery. Serbia's GDP has recovered from a sharp contraction of about 18 percent in 1999 owing mainly to the war. After growing by 5 percent in 2000, real GDP continued to grow at an annual average rate of 4.2 percent during 2001-03. Real GDP growth in 2003 is estimated to be around 3 percent. Fiscal performance has also been positive. The government has implemented a fiscal reform program to address major macroeconomic imbalances and reduced the consolidated budget deficit. Tax reform initiated in 2001 aimed to remove tax distortions and simplify the tax system. In 2002 the government introduced tax incentives for investment and employment, while preparations for the introduction of value added tax are underway. In parallel, the government explicitly recognized the size of quasi-fiscal deficit (losses in state-owned utilities) and initiated relative price adjustment. As a result of acceleration of reforms in early-2001, inflows of foreign aid and remittances, recovery in real wages and incomes, and the real appreciation of the dinar, the current account has been recording large deficits. About two-fifths of this increase can be attributed to rising investments following a decade of under- investment, with the remaining recorded decline in savings primarily being driven by the government sector as it began to unwind the unsustainable cash rationing and arrears accumulation of the 1990s and to incur costs of transition. The increased deficit was funded via donor support, remonetization (the repatriation of funds held abroad and the return of cash foreign exchange into the banking system), and rising FDI from privatization. The government has made significant progress in the implementation of other structural reforms, but has not done so evenly across problem areas. There has been considerable progress in opening the economy, restructuring the banking sector, and adopting legislation in many areas aimed at harmonization with the European Union. Improving the institutional framework for investment will help Serbia integrate into EU structures. However, Serbia is behind in the process, and negotiations on a SAA have not yet started. In many areas, the reforms are still incomplete, badly designed or badly implemented, and in several priority areas the reform process has barely begun. During the 1990s, investment was persistently too low to replenish the existing capital stock, and this had an enormously detrimental effect on capacity over the decade. In addition to the low volume and questionable efficiency of fixed capital investment during the 1990s, the war caused damage to the capital infrastructure of the country. Following a decade of underinvestment, the ratio of investment to GDP did grow modestly from 14.2 percent in 2000 to a projected 15.7 percent in 2003. However, the latter level remains extremely low by regional standards and remains insufficient to support rapid sustained growth. The direct negative consequences of this lost decade of investment have been a stock of fixed capital that is largely obsolete. The low levels of capacity utilization in state-owned enterprises (SOEs) cannot be increased by 33 providing working capital alone; sustainable growth will require new investment and restructuring. While the government in Serbia has made progress in implementing business enabling reforms such as liberalization and deregulation of foreign trade and investment, privatization, simplification of the tax regime, and modernization of the labor legislation, it has yet to focus on the related institution building. The implementation of recently drafted legislation aimed at improving business entry, corporate governance, access to finance, and business exit has yet to start and requires extensive institution building to take effect. In view of the limited capacity of the government and decreased foreign assistance, the prioritization of investment climate reforms and implementation efforts is critical. What are the barriers that discourage the private sector the most and what policy steps could encourage investors? Perceptions of the domestic business community In the 2003 PICS survey, Serbian managers were invited to rate 18 dimensions of the investment climate in terms of whether they were important obstacles to doing business. These dimensions were rated from 0 (factor is no problem at all) to 4 (factor is a major obstacle). Economic policy uncertainty was considered the biggest obstacle (Figure 6), which is understandable in light of the assassination of Prime Minister Djindjic in March 2003, not long before the survey took place. The factor that managers believed presented the least problems was access to land, a finding that is not as surprising at it first appears, for reasons discussed below. Figure 6: Perceived Investment Climate Constraints in Serbia and other transition countries 3.5 3 2.5 2 1.5 1 0.5 0 Accesstoland easingofland Electricity s afia der Corruption judiciary lations ycu... Taxrates Titleorl Transportation TelecommunicationOrganisedcrime/Meetcrime/theft/disorlsandeducationofava...Laborregulationsicensingandper... AccesstofinancingTaxadministrationeconomicinstabilityCostoffinancing Str Skil Businessl FunctioningoftheCustomsandtradereguContractviolationsbAnti-competitivepractice... acro M Economicpolicyuncertainty Serbia Other transition countries Source: BEEPS II, PICS 34 What can we learn for these results? Although the survey should not serve as the only basis for defining policy priorities, it should be taken into account together with other inputs in making decisions about the business environment21. The survey results need to be interpreted with certain precautions.22 Land is not rated as a major obstacle in large part because the respondents are managers of going concerns. The entrepreneurs who want to set up firms but cannot because they cannot get access to land are not covered in the survey. Taxes are rated by managers as a major business obstacles, but their responses would be different if the question also identified what public services would have to be cut (or what the level of inflation would rise to) as a consequence of tax cuts. Indeed, in surveys such as PICS and BEEPS, taxes are usually identified by managers as a major obstacle to doing business, whatever the country. Perceptions of Foreign Investors Foreign Direct Investment (FDI) has yet to play an important role for the private sector in Serbia. FDI per capita in Serbia in 1989-2002 was US$200 in comparison to US$1,400 in Croatia US$560 in Bulgaria, US$460 in FYR Macedonia, and US$419 in Romania.23 Foreign investors24 share the views of their domestic peers that Serbia is a risky place in which to invest. The World Economic Forum (WEF) Global Competitiveness Report ranked Serbia 77 among 102 countries in terms of its overall competitiveness (Table 4). Serbia also leads SEE region in terms of overall riskiness to investors, according to the Economist Intelligence Unit. 21A recent OED report21 discusses at some length. 22See Doing Business in 2004­Understanding Regulation, pp. 11-12 on shortcomings related to enterprise surveys. 23EBRD Transition Report 2003. 24The Foreign Investment Council (FIC), which represents most foreign investors in Serbia, has just published its 2004 White Book which elaborates the perceptions of FIC members and priority list of essential policy measures. The FIC list of Laws to be adopted is fully consistent with the priorities in the IMF and the WB's policy adjustment credits. A similar list was adopted by the ad hoc committee of donors on Legal Reform (chaired by the WB) before the new Government was formed. 35 Table 4: Perceptions of Foreign Investors: Global Competitiveness and Country Riskiness World Economic Forum Rankings* EIU** Growth Macroeconomic Public Technology Overall country competitiveness environment institution risk Slovenia 31 37 35 24 25 Hungary 33 38 33 32 38 Poland 43 46 53 33 38 Croatia 53 55 67 41 43 Romania 75 81 86 55 47 SAM 77 87 77 66 69 FYROM 81 80 93 70 61 Bulgaria 64 73 62 63 41 Source: * World Economic Forum Global Competitiveness Report 2003-2004. The table shows the rank of the countries based on the indexes. The growth competitiveness index is composed of three component indexes (the technology index, the public institutions index and the macroeconomic environment index). ** EIU Country Risk Service (December 2003). Overall risk scores are in the range 0 ­ 100 (higher score equals higher risk). The scores are obtained on the basis of the county political, economic policy and liquidity risk. Almost all FDI in Serbia in the first period 2001-2003 was related to privatization. Such brownfield investments potentially could play a positive role in encouraging greenfield investors. However, this requires that current investors find the investment climate acceptable and they are treated fairly. While experience so far is limited, the treatment of investment at times has been uneven.25 Investors have been mistreated in other ways: after the transactions are signed, the Government cannot ensure registration of land ownership or ensure that state creditors of the enterprise write off debts or overrule a reluctant management which refuses to cooperate with then new owners. In view of the new Government's reluctance to deal with the legacies of excessive debts or excessive employment, combined with their high price expectations, investors are turning away from Serbia and are pursuing investments in neighboring countries with which Serbia and Montenegro has signed free trade agreements. Priority areas for improving the Serbian investment climate The views of Serbian managers can, however, help answer the question of whether Serbia differs from other countries, in the sense of whether Serbian managers perceive the investment climate in the same basic ways as managers and policy makers in other countries. This question can been addressed using the technique of factor analysis (detailed in Appendix 1). The analysis shows shows what Serbian managers or businessmen interviewed perceive to be the largest consraints to their businesses in order of importance: (i) policy instability; (ii) high taxation and bad enforcement; (iii) lack of access to finance; (iv) insecure property rights; (v) high burden of regulations; 25For example, the experience of Henkel--one of the world leading firms in the detergent sector--investing in the privatization of Meirma has not positive. Without prior notice, Henkel was asked to pay tax on the severance payments for privatized firm's employees, although it had already agreed on the terms of the severance pay package with the government. Other major investors, such as LaFarge, and Philip Morris, have had similar experiences after buying companies that were being privatized. 36 (vi) bad infrastructure. Serbian perceptions compared to other transition countries is presented below (Figure 7). Figure 7: Combined investment climate measures 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 n cro latio Infrastructure Regu Property rights Finance Ma Taxation All 26 transition countries Serbia Source: BEEPS II, PICS 2003 Serbia 37 Box 2. Reducing the Informal Economy in Serbia According to the 2003 Poverty Reduction Strategy Paper for Serbia, an estimated one million people were involved in the "grey" or informal economy in Serbia in 2002. This represents almost one third of those who actively participated in the labor market. In addition, 10.8 percent of the persons that had a primary job had an additional job in the grey economy. Among poor employees, those involved in the grey economy represented 42 percent in 2002. Serbia's large informal economy is a main consequence of a poor investment climate. Given the depth and persistence of the informal economy, a main government priority is to identify policy reforms and other measures that will create incentives for firms to formalize. A number of works have analyzed the root causes of informal economies and ways to increase formalization. In his well-known work on Peru, Hernando De Soto argues that formalization of business could be encouraged by secure property rights which in turn require contract enforcement (as discussed in Chapter 4). Based on these findings, it is recommended that Serbia focus on the following three issues to attract entrepreneurs to the formal sector: · Contract enforcement: If it is hard to enforce contracts, an informal firm gains little from formalization; · Access to credit: If formal companies cannot get credit from banks, the incentive to formalize is small. · Land and property rights: If formal entrepreneurs cannot get clear title to land or real property, it will not be a factor pushing a businessman to formalize. According to the EBRD's Transition Report 2000, incentives for participation in informal sector differ depending on the country's stage of reform. In economies at a mature stage of reform (e.g., central and eastern European countries), the motives tend to be more "market- related" and guided by a desire to evade taxes and avoid other bureaucratic constraints. In economies at an initial stage of reform, informal activity is interpreted as being driven by poor opportunities in the formal sector and is seen as providing a "coping strategy" for survival. Both motives exist in Serbia at this stage, and as PICS indicates, formal enterprises try to keep their activities "blended" in order to avoid the administrative and financial burden of regulation. 38 CHAPTER 4: LEGAL, REGULATORY AND INSTITUTIONAL FRAMEWORK Ownership, Entry, Competition and Exit Growth during the early phase of transition requires movement of factors of production across enterprises. The ease of such movement via entry and exit of firms in the private sector is an important determinant of productivity, investment, and entrepreneurship. Relatively easy entry and exit allow poorly performing firms to leave the market and dynamic new ones to enter. This process, whereby bad firms exit and new ones replace them was dubbed "creative destruction" by Schumpeter over half a century ago. Evidence from the past 15 years across transition countries shows that the new private sector plays a leading role in generating growth. Countries that provide new entrants with friendly investment climates see rapid entry and growth of these firms; countries with environments that are hostile to new entrants (i.e., with excessive regulation or corruption) do not.26 Reallocating the capital of un-restructured and failing firms does, however, require a framework for bankruptcy and liquidation that frees the tied-up capital for alternative, higher productivity uses. Although the machinery, equipment and inventory of failing firms may have little in the way of market value, the buildings and land they occupy do. Many firms are in good locations of which new entrants and well-managed privatized firms could make better use. Dominance of the social and state ownership. As mentioned above, the performance of new private firms and even insider-dominated firms privatized before 2001 is clearly better than that of the socially-owned firms. Apart from the accumulation of losses, SOEs are characterized by low management turnover, overstaffing and limited flow of new human capital. Management Employee Buy-outs (MEBOs) created by the previous Law on Privatization, with few exceptions have resulted in weak corporate governance practices due to a distorted incentive structure.27 In the socially- owned and insider dominated firms the time horizon is short and the investment decisions are inefficient, and this can partly explain recent rapid growth of real wages. By cutting enterprise profitability, real wage growth reduced the resources available to enterprises to fund new investments from retained earnings. The development of the private sector still has a long way to go. The share of employment of private enterprises is consistent with the EBRD estimate that 50 percent of Serbian GDP is produced in the private sector28 (Figure 8). In this respect, Serbia compares unfavorably to Hungary and the Czech Republic with 80 percent, Bulgaria and Poland with 75 percent, Romania with 70 percent, Slovenia and FYR Macedonia with 65 percent, and Croatia with 60 percent. These facts show that Serbia 26World Bank, Transition: The First Ten Years, 2002. 27The Private Sector Note (PSN) examines these issues in greater detail. 28EBRD Transition Report, 2004. 39 has a long way to go in creating a private sector comparable to that in other East European countries. As the private sector is the engine of growth and employment generation, further privatization and new business development are essential to sustain growth. Figure 8: Employment by ownership sector at end 2003 Shops, Arts and Privatized Crafts 18% 20% Private enterprises (De novo) Socially-Owned 11% 18% State-owned Enterprises Financed from the 6% Enterprises not Budget (education, subject to health, etc.) privatization 23% 4% Source: Privatization Agency and authors calculations. Supporting Entry and Effective Registration Procedures Government role in supporting entry. Heavier regulation of entry is generally associated with greater corruption and a larger informal economy.29 Entry decisions, as any investment decision, depend on the prospective profitability of the investment, which in turn depends inter alia on the investment climate. Therefore, all the parameters discussed below--financing, regulatory burden, contract enforcement, infrastructure, land--affect the cost and benefit decision of the entrepreneurs.30 Serbia is a transition country, and like most transition countries, entry is driven primarily by the new private sector, (i.e., small domestic entrepreneurs) and by domestic and foreign investors that acquire SOEs through privatization and move them into new markets as part of restructuring. Both these sectors are major sources of growth for transition countries and Serbia is no exception, as the Serbia PICS data confirm. The Government has already begun to implement some important policy reforms that would promote either foreign investment or new private sector entry and 29See Djankov S., R. La Porta, F. Lopez de Silanes and A. Schleifer. "The Regulation of Entry", Quarterly Journal of Economics, 117, 1-37, (2002) 30A number of other institutional and administrative factors that play a key role in constraining the entry and growth of new firms will be discussed in this and other sections. 40 growth. The cumbersome system of business registration has become a particular a target for policy measures that are relatively easy to implement. Priority measures underway are discussed below. · Removing unnecessary legal barriers to entry. In 2002 the Government amended two laws regulating the status of enterprises and shops--the Law on Enterprises and the Law on Entrepreneurs. Amendments to the Law on Enterprises simplified the procedure to start a business by removing inspection approvals as a condition for beginning the business. Changes in the Law on Entrepreneurs simplified the procedure for opening shops by abolishing a number of redundant procedures. Table 5 shows data on changes in the number of registered firms and of full-time employees in the SME sector between 2000 and 2002. These data suggest a 10 percent increase in the number of active small firms as well as sizable increase in the number of shops. Table 5: The SME sector: Number of Firms and Workers, 2000-2003 2000 2001 2002 2003 Number Workers Number Workers Number Workers Number Workers Registered SMEs 216780 223796 234027 239270 Active SMEs 61722 610619 63985 581193 66219 654768 67703 669442 Number of shops 176724 353448 188812 377624 195186 390372 207596 415192 Source: SME Agency and Solvency Center · Establishing a "one-stop shop". The Serbian Parliament took a major step when it adopted in May 2004 a new business registration system that, once operational in 2005, will be one of the most modern and streamlined in Europe.31 The new system is supposed to reduce the average time needed for company registration from over 40 days to less than 10. The Business Registration Services Agency (BRSA), an independent and user-friendly agency, needs to be established by the Ministry of the Economy to facilitate this process of entry and will provide a "one-stop shop" for company, leasing and collateral registration. By supporting reforms these areas, it is designed to lower the cost of the financing for new investment--a crucial element of any entry decision. The figure below shows how Serbia compares to other countries in terms of business entry and how it improves over time. The reduction in the number of days from 2002 to 2003 is due to the changes in the Company Law, which made pre-registration inspections unnecessary in all but hazardous types of business. As seen in the Figure 9, a further decrease is projected in 2005 assuming a successful introduction of the new system. 31Serbia's present system of business registration is highly decentralized and legalistic. Currently, 16 commercial courts register companies and 161 municipalities register entrepreneurs, while a central database is maintained by the Republican Statistics Office. Moreover, registration drains resources away from adjudication of commercial law because it is located in the commercial courts and requires judges to review and approve all corporate filings. Serbia took important steps in 2002 to streamline business registration for companies and entrepreneurs. 41 · Lowering minimum capital requirements. The most commonly used legal form of enterprises in Serbia is the limited liability company. The contributions to the share capital may be made in the form of cash or contributions in kind. However, the minimum cash share capital was set unnecessarily high at an equivalent of US$5,000. Such a rationale has much less relevance or application in a modern market economy as creditors should be primarily protected by appropriate debt covenants and by laws prohibiting fraudulent conveyance. The new Company Law has lowered minimum capital requirement to EUR 500, and there is no requirement of capital maintenance, as is required in open joint stock companies. In summary, the capacity to implement the reform of business registration, in particular the shift from the courts and municipalities to the Business Registration Services Agency is a critical challenge. The objective is not only to provide a better infrastructure for opening a business but get out of the way as far as judicial prerogatives for opening a business is concerned. Therefore, the move of this function from the courts to the agency will not only make registration easier but also remove a psychological barriers to entry. More practically, it will decrease the load of the overburdened commercial courts. Figure 9: Starting a Business Cost (% of income per capita) 80 70 Greece Albania Hungary 60 50 40 Turkey 30 Moldova Italy Poland 20 Croatia Serbia 2002 Romania Serbia 2003 Czech Republic Ireland Slovenia Slovak Republic 10 FYROM Serbia 2005 Bulgaria Germany (projected) Austria France 0 0 20 40 60 80 100 120 Time (days) Source: CODB, Note: Serbia 2005 projected 42 Supporting Exit and Effective Bankruptcy Procedures Un-restructured companies are tying up capital and labor. A country's bankruptcy regime promotes efficiency and productivity in two ways. First, it should support creditors' rights and facilitate lending to firms that are going concerns. Ideally, the bankruptcy regime should provide creditors with a set of contingent property rights. When the borrower defaults, these rights become real. In a regime in which creditor property rights are strongly protected, the creditors in effect become owners of the defaulting firm's property. Second, the bankruptcy/liquidation regime promotes efficiency by enabling bankrupt firms to be financially restructured (if the firm is viable as a going concern and has a pure debt problem), physically or organizationally restructured (if the firm is viable as a going concern but only after, say, managerial change or restructuring or sell-offs of some production facilities), or by shutting down the firm and disposing of the assets (if the firm is not viable as a going concern). Persistent and large excess capacity. Although the capacity utilization rate depends on many factors, including business sector, market conditions or business cycle, the high rates of excess capacity reported by the PICS data indicate that the exit mechanism is not working (Table 6). Persistent and large scale excess capacity indicates that there are plants with obsolete machinery that is unlikely ever to be used again on a large scale, and many of these plants are occupying buildings and land and located next to infrastructure (electricity, water and sewage connections) that can be better used in other productive activities. Almost 8 percent of surveyed private companies were inactive, while 7 percent of privatized companies and SOEs stopped their production. Moreover, "Exit" in the sense of stopping production activity (plant closures) is not enough. What is also needed is the reallocation of the assets that are freed up, and this is what a good bankruptcy framework facilitates. Table 6: Excess capacity, plant closure and plant opening (percent) Average Excess Share of Companies that Share of Companies that Capacity closed at least one opened at least one plant/outlet plant/outlet New private 30.5 4.0 15.5 Privatized 44.2 11.1 7.9 State/Social 44.0 23.3 2.3 Manufacturing 37.3 16.6 14.5 Services 35.0 3.6 6.0 Commerce 44.1 11.8 8.6 Agriculture 38.9 15.4 2.6 Construction 31.8 6.4 8.5 Source: PICS 2003 Cumbersome and ineffective implementation of exit rules. The overall number of enterprises that have entered bankruptcy or liquidation has been low due to cumbersome and ineffective implementation of the exit rules. As shown in Table 7, 43 bankruptcy procedures in Serbia are lengthy, though better than in some other countries in the region. However, some bankruptcies last several years, and creditors often appeal the final distribution of proceeds so that the process takes an additional six months to one year. Thus, creditors are not willing to start the procedure due to the uncertainty about when they will be able to collect proceeds. The existing law contains serious deficiencies. The trustee/manager had no personal accountability, while incentive (compensation) was not merit based and effectively encouraged prolongation of the procedure. The rules for selling the assets were unclear and rigid, while forced settlement provided inflexible and limited opportunities for reorganization. In addition, in practice the role and protection of the creditors was minor with unclear deadlines, causing the procedure to last for years. Table 7: Bankruptcy Indicators Actual Time Recovery Rate (in years) Cost (% of estate) (cents on the dollar) Bulgaria 3.3 8 34.2 Croatia 3.1 18 26.1 Macedonia FYR 3.7 38 7.9 Poland 1.4 18 68.2 Germany 1.2 8 50.3 Romania 4.6 8 6.9 Serbia 2.6 23 20.8 Source: The World Bank "Doing Business in 2005 ­ Removing Obstacles to Growth" The new Law on Bankruptcy was adopted in July 2004.32 However, the Law is not expected to improve creditor rights by simplifying and accelerating bankruptcy unless the necessary institutions are established and the Government shows the political will to initiate bankruptcy proceedings in its role as creditor of SOEs. Under the law, the court will make more judicial rather than operative decisions, while a bankruptcy administrator will have a more active role, yet still be subject to oversight and supervision by the court and the creditors committee. The administrator will have much more authority to manage the procedure without prior judicial approval or direction. The centralization of the administration function for unviable social and state owned companies in an the Privatization Agency (PA) is as much a challenge as an opportunity. The PA will need significant technical assistance to implement this task. 32The new Bankruptcy Law follows the general form of other modern bankruptcy laws in the region, notably that of Croatia, which itself is based on the German Insolvency Act of 1994. The Bank, with others, has developed Principles and Guidelines for Effective Insolvency and Creditor Rights Systems which it uses to assess whether a draft insolvency law meets international design standards. The EBRD has also recently completed an assessment of the structure and content of insolvency laws of the 27 transition countries to which it lends. This assessment ranked the new Serbian Bankruptcy Law, together with the Romanian law, as being superior to those of the other 25 countries whose laws were assessed. For more detailed discussion see "Serbian Privatization and Restructuring - Regaining the Momentum Private Sector Note". 44 Competition Building institutions that engender inter-enterprise competition is essential to improving the investment climate. The inherited industrial structure of the Serbian economy is characterized by large enterprises relative to market demand, thus resulting in diseconomies of scale and scope. An apparent trend, however, over the last three years suggests that competitive pressures are moderately increasing, because of firms' greater exposure to the international market and due to implementation of structural reforms. Opening the markets for foreign competition in 2000 and 2001 was a major step towards increasing competition in tradeables. Although the role of trade is somewhat limited by non-tariff-barriers, which are used to protect local producers, for example, by requiring certification.33 Trade liberalization is not sufficient. The implementation of reforms, opening the markets for foreign competition, together with the privatization and establishment of SMEs, may not be sufficient to bring about competitively structured markets and the resulting competitive conduct of business operations. It was estimated that the lowest level of competitive market structure even after the introduction of imports as a source of competition for Serbia are recorded in industries which produce ceramics, glass, paper and iron products, construction materials, as well as products of the basic chemical industry.34 From the experience of other transition countries, privatized firms, through mergers, acquisitions or investment, have acquired dominant positions in their sectors. In the absence of competitive pressure exerted by rivals or threat of entry of new rivals, inefficient or commercially unviable businesses survive, with little restructuring, avoiding bankruptcy or liquidation35. Private firms face a greater competition. Both BEEPS 2 and PICS 2 show that the surveyed de novo private firms face a greater number of competitors than privatized businesses or SOEs. According to PICS, in the last two years, while 29 percent of the surveyed firms in 2001 stated they have no competition in the domestic market, only 12 percent of the surveyed enterprises perceived no competition in 2003. Strengthening the regulatory framework. A new Competition Law has been prepared but its adoption has been delayed due to the EU requirement that it be compatible with EU standards. The adoption of the Law is a necessary but far from sufficient for its implementation and impact. The latter requires a powerful and independent authority that will have the power to deny request firm mergers and set maximum prices for monopoly products (sometimes at the behest of the monopoly or the cartel). 33 Imports are subject to examiniation to establish whether they meet conditions for the Serbian market. A particular problem is that imports are often tested in authorised laboratories owned by domestic producers that are their direct competitors. In those conditions, it is in the interest of the domestic producers to at least postpone attest issuing. Therefore an absurd situation is created in which products of the companies with world reputation, which already have all necessary attests in European Union, are waiting for domestic attest issuing. 34See Begovic, B. et al. "Competition Policy in FR Yugoslavia" (2002) pp. 98 -114. 35Carlin et al. found that competition has an important and non-monotonic effect on the growth of sales and of labor productivity. (see Carlin, W., S. Fries, M. Schaffer and P. Seabright "Competition and Enterprise Performance in Transition Economies: Evidence from a Cross-Country Survey", EBRD Working Paper No. 63 (2001) 45 Most difficultly, it will order and oversee the break up of companies that abuse their dominant position. Powerful interests may oppose the break up of their groups and its unrealistic to expect the Authority to cope with those forces. Therefore, the establishment of such authority will not be an easy task and it is unlikely that the authority will be fully operational for several years. Contract Enforcement Contractual compliance and enforcement mechanisms--both formal and informal-- represent key features of the investment climate. In Serbia, as in most other transition countries, there is a widely-held view that the courts do not function well and that the administration of justice is slow and inefficient. Among the range of institutions that facilitate contractual compliance and enforcement, courts represent the key instrument. Without a well functioning dispute resolution mechanism and the expectations that courts will uphold their contractual rights, firms are less willing to deal with new customers and suppliers, and fewer transactions take place. As a result, firms rely on other institutions and informal practices to enforce contracts and reduce uncertainty in transactions by relying heavily on pre-payments and by building long-lasting relationships with suppliers and customers. However, relational contracting (i.e., relying on personal relationships) limits market dynamism as enforcement rests on the ability of the firm to punish its trading partner if it does not cooperate.36 Such a situation impedes development of more efficient impersonal transactions. Thus, a more workable legal system should emerge to stimulate impersonal transactions and the role of the courts is expected to increase overtime and to substitute relational contracting.37 Weak contract enforcement in courts. According to the Cost of Doing Business Survey, it would take almost three years (1,028 calendar days), and 36 steps to bring a relatively straightforward lawsuit to an end.38 In this respect, it is interesting to compare Serbia with other former Yugoslav republics as well as with Germany and Austria, countries with legal systems that provide the foundation for many Serbian laws. Table 8 shows that the average time to collect on a debt through the court system in Germany, or in some neighboring countries is several times lower, although some of the more economically advanced transition countries with similar legal traditions, such as Poland and Slovenia, do not fare much better in terms of duration. Serbia has far more procedures than any other country in the region, and is among the 10 countries with the highest number of procedural steps worldwide. Thus, commercial court reform should be put on the government and donor agenda as soon as possible. 36 More efficient solution requires that the society develop and use institutions that will permit anonymous, impersonal exchange across time and space (see North, Douglas "The Evolution of Efficient Markets in History") 37Johnson, McMillan and Woodruff, 1999 38The methodology was developed in Djankov, Simeon, Rafael La Porta, Florencio Lopez-de-Silanes, and Andrei Shleifer, (2003) "Courts," Quarterly Journal of Economics, 453-517. It assumes a hypothetical transaction, where a client reneges on the payment of a fulfilled contract, amounting to 50 percent of country's GNI per capita. 46 The number and complexity of procedures result in high costs to firms in terms of legal and administrative fees and corruption. Serbia needs to simplify or improve its procedures and reduce the time required by the first-instance courts for trials. On average, a contested trial involves about 10 to 12 separate court appearances and takes at least 12 months. Table 8: Contract enforcement indicators Number of procedures Duration (days) Cost (% GNI per capita) Bulgaria 34 440 14.0 Croatia 22 415 10.0 Macedonia FYR 27 509 32.8 Poland 41 1,000 8.7 Germany 26 184 10.5 Romania 43 335 12.4 Serbia 36 1,028 23.0 Source: The World Bank "Doing Business in 2005 ­ Removing Obstacles to Growth" Another measure of the cost of going to court is that total attorney and court enforcement fees costs can amount to 40 percent of the claim value. However, apart from having problematic way of resolving disputes, procedural complexity index is close to the regional average and to countries with similar legal tradition, pointing to the fact that the Serbian judiciary system has a number of other deficiencies. Evident pro-debtor bias in the courts. Apart from the heritage of the Serbian judicial system and of the social ownership, much of pro-debtor bias is the result of debtor favorable provisions in the old laws. However, a set of laws relating to the contract enforcement was passed in November 2004. The new Law on Civil Procedure (Litigation Law) has been severely improved by laying the burden of proof from courts to parties in dispute, setting the deadlines for submitting evidence, introducing the obligation of the second instance court to reach the verdict and not to give the case back to the first instance court, etc.). The second problem was related to the Law on Enforcement Procedure. The old law's provisions were not fully transparent or adapted to every day's life, and included problems such as the requirement for the debtor to disclose assets to facilitate execution. In addition this was compounded by inactive bailiffs and a lack of support from the authorities, as well as substantial resistance of the debtor. Thus, a security instrument such as a mortgage do not serve its purpose. Weak judicial capacity. Serbia's judiciary has been under pressure for an extensive period of time. The system is characterized by passivity and corruption, and there is a strong need for court personnel to be trained, evaluated and compensated adequately. Information sharing also remains weak. Specifically, judges do not have adequate information on how the High Commercial Court interprets and applies the law, leading to an increase in avoidable court cases and unnecessary appeals. At the moment, expert witnesses represent one of the main generators of corruption. When it comes to enforcement, corruption of court experts takes place both prior to the enforcement procedure and during the enforcement stage. 47 Heavy case backlog. Reducing the current case backlog in the Commercial Courts is another area requiring immediate attention.39 Commercial Courts are overloaded with inactive or abandoned case files that create additional administrative burdens. In addition the arbitration division of the Serbian Chamber of Commerce System is not functioning in the practice, thus leaving the court as the only place to resolve a dispute. Figure 10: Case back log in Serbian Courts Cases completed and new cases, by case type Case backlog in 2002- - 03 in Commercial Courts (2002) 250000 30000 January 2002 Other cases 25000 January 2003 200000 Litigation Economic violations Decision Execution July 2003 Business registration Bankruptcy 20000 150000 15000 100000 10000 50000 5000 Cases solved in 2002 New cases 0 Commercial Council of District Courts Magistrates Court Source: DFID and Commercial Courts Reputation is central to ensuring contract performance. In deciding whether to contract with a new partner, firms are guided by what they know about the potential partner's history of complying with contractual obligations. 39The Commercial Courts are "specialized" courts in the Serbian court system as opposed to the "regular" courts. There are 16 regional Commercial Courts of the first instance and the High Commercial Court. The third instance court in the system is the Serbian Supreme Court that has a division to hear appeals from the High Commercial Court. The Commercial Courts are responsible for commercial disputes between enterprises and other legal entities, trademark and copyright disputes, cases involving status changes, enterprises commercial offenses, conducting bankruptcy and liquidation proceedings, the enforcing judgments, and the validity of foreign court and arbitration decisions. 48 Table 9: Pre-payments Average share of sales paid in Share of the firms that use pre- advance by clients (%) payment Bulgaria 6.47 30.8 Croatia 6.63 49.5 FYROM 9.27 41.5 Poland 4.99 38.7 Romania 5.77 42.7 Serbia PICS 1 23.7 72.2 Serbia BEEPS 2 22.9 69.8 Serbia PICS 2 20.0 66.1 Source: PICS 2001 and 2003 for Serbia, BEEPS 2002 for other countries Table 9 shows that in Serbia the proportion of sales paid in cash as pre-payment (20- 23 percent) is significantly higher than in any other Eastern European country. The percentage of firms using pre-payment (65-70 percent) is also higher than in any other country in the region. Pre-paid transacting limits the dynamism of the market and slows down the expansion of business. Development of credit rating agencies coupled with changes in the banking sector should make trade credit easier. Policy implications. Following are the policy implications for the government in order to improve the investment climate in the areas outlined above: · The legal framework simplifying the procedures should be further developed. This includes the adoption of a new Law on Execution Procedure40 and amendments to the current (or the adoption of the new) Law on Civil Procedures. The introduction of new concepts--the establishment of appropriate treatment of debtors and creditors, coupled with non-repressive acceleration of enforcement procedures--is expected to help clear the backlog in the courts and facilitate doing business in Serbia. · The performance of bailiffs should be strengthened. By-laws on bailiffs should be adopted followed by training. This should strengthen their skills and lower the possibility of delaying the execution and corruption. · Court resources should be improved.41 Education of judges should be a priority. As the new legislation will broaden the competences of the commercial courts their enforcement benches should undergo additional training. Besides the state of the Serbian court system is such that urgent resources for training of judges and court personnel in the area of case filing and tracking systems, etc. are needed. · An anti-corruption strategy should be devised. In addition to improving the financial status of the judges, a new Law on Court Experts (expert witnesses) should be developed. 40The Law on Enforcement Procedure is being prepared by a Working Group appointed by the Minister of Justice in 2004. 41It is clear that successful implementation of the law will depend upon the other reforms in this area. USAID will provide US$15 million for reform of commercial courts, and putting the right legal framework in place is a precondition for court reforms. 49 · Information-sharing systems should be improved. Although Serbia has established a public credit registry, it ranks among the ten countries worldwide in which the legal and regulatory environment is least conducive to information sharing. Therefore it is important to better coordinate and facilitate operations of enforcement benches, and strengthen the link between commercial court enforcement benches and the competent registries (registry of securities, pledge registry and organization for coercive compensation). · Effective usage of court statistics should be developed and used. Empirical data are necessary to identify bottlenecks in court performance and give policymakers information where reforms are needed most. This would improve court administration and case management systems. · Removal of non-dispute cases from the courts. The new registration system in Serbia will transfer business registration away from the courts to the Business Registration Agency freeing up judges' time to deal with proper disputes. Freeing up judges' time could also be achieved through "alternative dispute resolution" (ADR) mechanisms including a variety of methods for resolving disputes outside the formal judicial process. Countries in Central and Eastern Europe are introducing ADR as a complement to judicial reform, and it is suggested that Serbia explore the suitability of ADR mechanisms to local conditions. Regulatory Burden A country's general governance structure and business-government interactions are an important component of the investment climate. In general, these include the burden firms face in complying with regulations, the quality of the services provided by these regulations and the extent to which corruption is associated with the procurement of such services. The network of government bodies with control functions in Serbia is complex, and their responsibilities are often vaguely defined and overlapping. In addition, pervasive and confusing business regulations impose high costs on formal businesses. The information provided by the surveys shows a comprehensive picture of the relative position of Serbia to other countries. The PICS survey explores the issue of administrative harassment and corruption (i.e. the overall cost of compliance with regulatory norms), as well as the predictability of business rules and fairness in their application. The FIAS Administrative Barriers Study examines the issues of administrative burden in greater detail. High cost of administrative barriers. Administrative barriers are still a great impediment to successful business. The surveys indicate that the regulatory burden is one of the key constraints to long-run growth and competitiveness in Serbia. Very often enterprises simply do not know which documents should be submitted or which procedures should be followed. The process commonly involves corruption and a high degree of inconsistency in the application of the rules, while public servants have wide discretionary powers in making decisions on minor issues that affect the issuance of permits, licenses and certificates. 50 The time tax. The time that senior managers have to spend dealing with government officials and regulatory activities provides a measure of the burden (so called "time tax") that regulation places on firms. Very often the responsibilities of the various bodies are not clearly defined and or not co-coordinated thus causing overlap. According to PICS 2, senior management spends an average of 17.2 percent of its time on dealing with requirements imposed by the government regulations, such as taxes, customs, labor regulations, licensing and registration. As Table 10 shows this is far more than in any other comparable transition economy.42 Table 10: Cost and time tax of administrative barriers (percentage) Senior management's time spent in dealing with public Total annual sales paid in Unreported sales officials unofficial payments Bulgaria 3.7 1.95 17.2 Croatia 5.9 0.64 12.7 FYROM 9.0 0.79 36.2 Poland 9.5 1.21 9.8 Romania 9.3 2.57 13.4 Serbia PICS 1 16.8 2.58 23.7 Serbia BEEPS 2 12.3 1.70 23.7 Serbia PICS 2 17.6 3.39 23.1 Source: PICS 2001 and 2003 for Serbia, BEEPS 2002 for other countries The bribe tax. The PICs data also shows the direct costs to firms from one aspect of corruption - the so-called "bribe tax". The procedures for obtaining licenses, permits and certificates required for business operation take too long and are not sufficiently transparent, resulting in corruption. The PICS data show clearly that new private firms are burdened most heavily by the bribe tax, as they are paying on average 5 percent of their total sales while socially- owned enterprises are paying only 1 percent. Firms paid more of their annual revenues in bribes compared to 2002, but payments were not as often, suggesting not a decline in corruption but a non-competitive bribe "market" among bureaucrats (Serbia appears higher on the bribe tax but more moderate on frequency suggesting more concentrated bribery). In general, Serbian firms perceive corruption to be a less severe obstacle to their business as they experience perhaps a more important impediments in which the weight of corruption relative to other problems has declined. According to PICS, the presence of irregular payments to get connected to public services or to obtain land or get construction permit is still very high. After the adoption of the new Law on Public Procurement, the ability to obtain government contracts has been significantly improved, though the current law has number of 42To some extent the higher percentage of time tax recorded in PICS can be explained by the sampling weight given to socially-owned enterprises, and the fact that managers of such enterprises contact public officials more frequently. 51 deficiencies. In addition, administrative barriers, such as frequent inspections or excessive licensing regulations, increase the operational costs of firms. The amendments to the Enterprise Law, which were adopted between two surveys, stipulate that an inspection of newly established enterprises (except for some specific activities) should be carried out their start­up and not before, as prescribed by the previous legal solutions, which prevented corruption in this segment to a considerable degree. However, entrepreneurs very often do not have adequate information on what is needed to fulfill some technical requirements, thus having problems once the inspection comes to check the facility. Inspections and certification. The average number of inspections in surveyed enterprises last year remained at around nine times per annum as in 2001. In the future, the challenge will be to find an optimum between using inspection as a tool in reduction of the informal economy and yet not exposing firms to a frequent and unnecessary inspections of their business activities. In addition, according to PICS, enterprises have experienced a high increase in certification costs from EUR 2,224 annually in 2001 to EUR 6,200 in 2003. Compared to the previous survey, corruption in this area has been reduced, but bribery continues to be widespread. In addition there is a need to bring national standards and technical norms in conformity with EU and international standards, as well as to reform the system of accreditation and certification to remove barriers to the unhindered flow of goods and services. High unpredictability of laws. The high incidence of the time and bribe tax is caused to some extent by the constantly changing regulatory environment. Such an environment generates ambiguity and uncertainty, which in turn creates scope for rent-seeking activities. With the beginning of the reform process, the new authorities had to change many laws and adopt new ones. In the past, due primarily to the absence of a political consensus, the adoption of some system-related laws has slowed down. Regulatory reform should not be regarded as a one-time measure, but as a continuous process. Over the past three years, many laws have been replaced with new, modern ones. In order to improve the quality of regulations and to anticipate their impact on the economy, the amendments to the Rules of operations of the Government that require a justification statement for all laws and other regulations were adopted in November 2004. Firms are particularly concerned that some firms are in a more favorable position because of connections with certain political parties or their size (i.e., dominant firms or conglomerates in key sectors of the economy). One half of surveyed enterprises perceived these firms as ones with the highest influence on laws and regulations, followed by international financial institutions or foreign governments. Firms believe, for example, that such companies benefit from postponement of tax payment, easier access to government procurement and better treatment by various state agencies. 52 Box 3: Market Inspections The market inspections system was inherited from the former socialist era when it served as an instrument of control of private businesses. The inspection service is organized at the republican level, within the Market Inspection Section of the Ministry of Trade. The Market Inspectorate is responsible for supervising the application of laws and secondary legislation in the areas of consumer protection, trade, the catering industry, small business and crafts, prices, and the safety of non-food products. The Ministry of Trade also issues a certificate ("license") of fulfillment of the minimal technical requirements for performing specific activities. Recent evidence from FIAS, the G17 Institute and Serbia MAEWS Report, reveals that different types of inpections are often overlapping and result in frequent and duplicate inspections. According to the 2004 MAEWS Report, 60 percent of companies were subject to market inspection, spending from 5.5 days per annum on it. The Ministry's figures are consistent with this finding as there was 17 percent increase in the number of market inspection performed (over 180 000) compared to 2002. Although the main difference between market and tax inspections is that tax inspection controls only papers and cash flow, while market inspection controls the goods, there is a clear overlap of activities. Consequently, there is an urgent need to: · narrow the scope of work of the market inspection to reduce any duplications and shift the focus of the market inspection on consumer protection issues; and · change the minimal technical requirements for performing specific activities and the rules defining the procedures to be used by the inspections. In addition there is anecdotal evidence that points to the selective usage of market inspections to discouridge competitiors or that inspectors are punishing every firm or shop visited but most of their decisions are overrulled by the second instance authority. In order to address these issues, the Government has recently established a Commission for Monitoring the Inspection Activities of Government Agencies. The basic tasks of the commission are to propose measures to improve coordination of inspection activities carried out by government agencies, as well as the organization of joint inspection activities carried out by several government agencies. However, much more is needed to precisly define the scope of work of the market inspection, including amending current legal framework and significant capacity building. Figure 11: Percent of enterprises visited by various inspection authorities 70% 60% 50% 40% 30% 20% 10% 0% Labor Financial Sanitary Tax Market protection police inspection authorities inspection inspection 53 Policy implications. Following are recommended policy changes to reduce the regulatory burden: · Public discussion should be introduced as a rule, not as an exception. While the recent efforts were focused on transparency in party financing, disclosure of parliamentary votes, and on preparing new conflict-of-interest rules through the Law on Conflict of Interest in order to introduce political and public accountability, the inclusion of various stakeholders that would contribute to improved government decision-making and oversight through voice and participation is missing. There is a need to introduce public hearings and public discussion of laws and decrees as a rule and not as an exception. Recently, the Regulatory Reform Council has initiated reform of the Rules of Operations of the Government that are needed in order to establish the mandatory public discussion. This could significantly improve the transparency of policy making by bringing in a wider range of affected interests. Until now, the Government introduced several successful public consultation processes, such as public-private sector dialogue on the draft Leasing Law, the Company Law, and more recently the Business Registration Law and the Bankruptcy Law. · Build the capacity for Regulatory Impact Analysis. In November 2004, the Government empowered the Regulatory Reform Council so that it could also provide the capacity for regulatory impact analysis (RIA). The objective of RIA is to reduce the cost of doing business and improve regulatory reform by enhancing the authority and capacity to assess need for and quality of regulations and market impacts of proposed rules on business.43 · The effectiveness of the inspections system should be insured. Preparation of the amendments of existing legal regulations (the Law on Requirements for Turnover of Goods, Provision of Services in the Framework of Turnover of Goods and Inspection Control, Law on Trade and Law on Consumer Protection) should clearly define the responsibilities, procedures, authorities and duties of market inspection with regard to other inspection authorities. · Establish Central Registry of Regulations with positive legal security within the Government and enable free internet access to regulations. · The national system for accreditation and certification should be adjusted to meet EU standards and technical norms. An incomplete system of accreditation and certification can seriously hinder the flow of goods and services. One way to address technical obstacles in this area is to adopt the EU norms relating to standards and technical regulations. The adoption of European solutions in this area would enable the development of a recognized system of accreditation, standardization, and certification. 43Although only two or three OECD countries were using RIA in 1980, by 1996, more than half of OECD countries had adopted RIA programs. By October 2000, 14 out of 28 OECD countries had adopted universal RIA programs, and another 6 were using RIA for some regulations. Serbia does not have any RIA equivalent. The Secretariat for Legislation, established by the Republic Law on Ministries reviews all laws and government decrees for consistency with the constitution, with other laws, and with drafting standards, but there is no corresponding review for economic, business, or SME impacts. 54 CHAPTER 5: FACTOR MARKETS Access to Finance Recent empirical work has shown that access to external financing is shaped by the country's legal and financial environment.44 In countries with weak legal and financial systems, the private sector obtains less external financing and especially less term financing, than in countries with better developed systems. As a result, private sector growth and investment efficiency are reduced and firms are forced to rely mostly on internal funds and retained earnings. Further, when non-bank financial institutions like stock markets and bond markets are not sufficiently developed, financing by banks represent the only viable form of outside financing. The ability to finance activities through the sale of equity on the capital markets will not become available to the vast majority of firms until the institutions supporting such a system are adequately developed. Thus, reforms in this area must extend further than company laws and extend to the institutions underlying firm transparency and accountability that will give investors the confidence to invest. Restructuring and strengthening the banking sector. The Serbian authorities have made significant progress in the restructuring of the banking sector, culminating in the closure of the four largest banks in January 2002. To date 25 insolvent banks representing nearly two thirds of the assets of the system were either liquidated or put under bankruptcy. As a result of the restructuring process and the entry of nine new foreign banks, confidence has begun to return to the Serbian banking system. This is evident in the rapid growth of foreign exchange deposits, which rose from almost none in 2000 to over EUR 1.8 billion at end-2003.45 Despite this progress, however, the banking system remains insufficiently developed to deliver and adequate range of credit and other financial services to the enterprise sector. The current level of financial intermediation in Serbia is not sufficient to support the needs of growing economy. The banking system lacks long-term funds as short-term foreign-currency deposits prevail. As a consequence, the share of all short-term credit from banks to the enterprise sector amounts to only 19 percent of GDP. This is quite low compared with most other transition countries. As noted by Cottarelli et. al.46, banking sector reform in transition countries can be divided into three phases: (i) the recognition that a large share of the loans extended by public banks, mostly to SOEs have to be written off and the losses shifted to the government; (ii) the sale of banks, primarily to foreign investors; and (iii) the beginning of more standard banking operations, including increased lending to truly private enterprises. In that sense, to date the reform has been concentrated on stabilization and urgent first stage restructuring in the banking sector and 44Laporta, R.; F. Lopez-de-Silanes, A. Shleifer, and R. Vishny, "Law and Finance," Journal of Political Economy 106, 1113-1155 (1998); Demirgüç-Kunt, A. and V. Maksimovic, (1999), "Institutions, Financial Markets And Firm Debt Maturity"; or Demirgüç-Kunt, A. and V. Maksimovic, (1998), "Law, Finance, and Firm Growth" 45Serbia Financial Sector Note. This note examines the banking sector issues and issues of access to and cost of finance from the financial sector perspective in greater detail. 46 Cottarelli, Carlom Giovanni Dell'Ariccia and Ivanna Vladkova-Holar "Early Birds, Later Risers, and Sleeping Beauties: Bank Credit Growth to the Private Sector in Central and Eastern Europe and the Balkans". IMF Working Paper 03/213 (2003) 55 parallel but also incomplete reforms in the financial sector regulatory and institutional framework. 47 Low level of lending to SMEs. Banks still focus on servicing larger enterprises and individuals. Experience from other transition countries suggests that changes in bank culture and attitudes play an important role in increasing the lending to private sector. Some banks are very small with only regional coverage or they act as "agent banks" mainly operating as treasury divisions of single enterprises or groups of enterprises. The lack of interest in SMEs is also due to a number of factors including high fixed costs of loan processing, high level of informality, and reliance on mortgages that are hard for banks to realize. SMEs loans tend to be relatively small, typically ranging from US$2,000 to US$5,000. In turn, this makes the share of fixed costs in the total loan cost large and forces less competitive and higher cost banks to charge high interest rates. Informal operations severely limit banks' ability to assess real income and lend. Enterprises generate a high level revenue from informal or partly informal operations that is underreported in their business in financial statements in order to evade taxes. Based on interviews with entrepreneurs, it is claimed that tax evasion is probably the main source of financing investments and the opportunity cost of not being able to use bank credits for most of them is still lower than profit made through informal operations (PICS reported that formally registered companies estimate that almost one quarter of their sales has not been reported). In addition, financially able entrepreneurs, with some or all of their income derived from the gray economy or from remittances, cannot obtain loans. The ability of banks to lend should improve as foreign banks invest in the financial sector and as specialized credit lines support by international financial institutions are helping to transfer skills in risk assessment and risk management to local banks and to make SME lending more profitable. However, to date the effect of the presence of the newly established foreign banks, though beneficial, has been limited, as they have focused on consumer lending and large or foreign-owned companies. High dependence on internally generated cash flows. As a result of the lack of access to credit through banks, firms are highly dependent on internal funds and retained earnings as their main source of financing. According to PICS, Serbian enterprises continue to rely primarily on their own retained earnings as their means for financing their activities. While there is nothing wrong with this per se, it does serve as a reminder that financing available for growth is constrained. What is of special concern is the substantially higher use of internal funds or retained earnings in Serbia than in other countries (Table 11). 47Another often expressed concern regarding the foreign bank entry has been its potential impact on lending to small and medium enterprises (SMEs). If foreign banks dominate domestic banking systems, this might reduce the access of SMEs to finance, owing to information problems. But this problem is unlikely to be severe because foreign firms tend to enter by acquiring local banks and because competition from more efficient foreign banks may force local banks into new market niches, such as SME lending, where they have a comparative advantage. The detailed evidence available from Hungary indicates that foreign banks are heavily involved in retail banking in both deposit taking and consumer lending. There is also evidence that foreign competition has compelled some domestic banks to seek new market niches. 56 The problem in Serbia is compounded by the fact that firms have not accumulated enough profits to rely on internal financing. In view of this, growth of new projects and expansion are constrained (i.e. projects with positive net present value are rejected). Table 11: Average Share of Internal Funds or Retained Earnings in Firms' Financing Serbia Romania Poland Hungary Bulgaria Croatia PICS 1 BEEPS 2 PICS 2 Investments 72.0 64.8 57.6 69.6 45.2 86.3 92.5 84.9 Working Capital 69.6 68.1 63.6 65.8 52.5 76.7 86.9 83.5 Source: PICS 2003 for Serbia, BEEPS 2002 for other countries Significance of trade credit. Trade credit represents an important source of financing for SMEs. Since the legal system is an important determinant of the availability of external financing, one might expect the amount of trade credit to be positively related to the efficiency of a country's legal system. However, the efficiency of the legal system is likely to have a greater effect on lending by financial intermediaries than on suppliers of trade credit. Recent empirical evidences demonstrate that there is more dependence on trade credit when the country's legal system is inefficient48. Thus, we may expect that the use of trade credit by Serbian firms is wide-spread. The PICS survey indicates that 67 percent of all enterprises make some use of such financing, while 35 percent of enterprise sales are being conducted this way. It is expected that the relative importance of trade to bank credit will decline as the Serbian legal system--and especially creditors rights and contract enforcement--become more efficient. High costs of financing. The average real interest rate is approximately 10-12 percent. While the cost of financing remains high, there was significant improvement in 2003 and interest margin and real interest rate are decreasing. However, Serbia still has significantly higher interest rate than neighboring countries. Nearly all loans that are denominated in the national currency include a foreign exchange clause, transferring the risk to enterprises that cannot hedge such risk. High interest rates are in many ways an appropriate response to the risks inherent in the current system including judicial inefficiency.49 Due to the fact that the cost of borrowing in the securities market (Belgrade Stock Exchange) has been high, it appears that companies have had no particular reason to 48Even though weak creditor protection and imperfect information will affect both formal intermediaries and trade credit providers, trade creditors may mitigate these problems better than formal lenders for several reasons. These include advantages in 1) information acquisition - suppliers have a cost advantage over banks in acquisition of information about the financial health of the buyers. They can receive better signals about the customer's probability of default than do banks; 2) the renegotiation/liquidation process, and 3) enforcement - trade creditors may be able to punish debtors without resort to the legal system by withholding further deliveries. See Demirguç-Kunt, A. and V. Maksimovic "Firms as Financial Intermediaries: Evidence from Trade Credit Data" The World Bank Working Paper (2002); 49Leaven and Majnoni found that judicial efficiency, in addition to inflation, is the main driver of interest rate spreads across countries. (See Leaven, L. and G. Majnoni "Does Judicial Efficiency Lower the Cost of Credit?" World Bank Policy Research Working Paper 3159 (2003) 57 provide financing by borrowing from sources other than the banking sector. It should be noted, however, that the spread between the stock exchange interest rates and those on other short-term credits is narrowing, and that the trend will continue over time with the development of financial market. At this moment, a key benefit of the corporate form of organization--the ability to finance activities through the sale of equity on the capital markets--will not become reality to the vast majority of firms in Serbia. Institutional arrangements for collateral and creditors' rights. In deciding whether to extend credit and at what interest rate, lenders need to know what share of debt they can recover if the borrower defaults. The value of collateral depends largely on the ease of creating and enforcing security agreements and in Serbia this appears to be very problematic. This includes weak arrangement from valuing and enforcing collateral (and especially mortgages) and lack of information on credit history of the client. These deficiencies raise the transaction costs of the banks offering credits and other services and increase the risk of adverse selection. The currently limited access to finance and high cost of financing present a major challenge and the Government has recognized the existing weaknesses and made efforts towards reforming the legal framework (e.g. Law on Registered Charges on Movable Assets or Law on Financial Leasing) and establishing the most important institutions. Developing a collateral registry. Since practically all lending to enterprises is backed by collateral, legislation to facilitate the use of collateral is particularly important. In Serbia banks apply a variety of preventive measures in order to secure their credit portfolios, and the availability of collateral is an important factor in securing access to loans. The banks will generally require either real estate or other tangible assets as collateral or a guarantee from a person or an institution. The PICS survey reports that in 81 percent of cases, creditors required collateral or deposits for the credits, which represents a substantial increase compared to only 50 percent in 2001. The central collateral registry together with the credit registry has been recognized by the banking sector as a critical condition for improving access to finance. Although the Law on Secured Transactions was adopted in 2003, a centralized collateral registry has not yet been established. In some of the neighboring countries the impact of the similar law on lending has been remarkable, while in others the banks do not seem to be using it, although the legal framework is satisfactory.50 The need to reform the mortgage-based system of lending. Mortgages on buildings are the most frequently requested type of collateral and account for 45 percent of all bank collateral. In addition, financing new construction is limited because it is not possible to mortgage a building under construction.51 The development of sustainable mortgage 50In Romania in the first 18 months since the reform was implemented, there have been more than 400,000 loans against which security interests have been registered. On the other hand, in FYROM the framework is not being fully utilized, indicating that there are problems at some point in the four stages of creating a secured transaction : creation ­ establishment of the security interest; priority ­ establishment of the order in which claims are executed; publicity ­ disclosure of the established priority; and enforcement ­ability of the creditor to enforce the secured interest. 51 As noted in the FIC's White Book "Institution of mortgage is stipulated in the Serbian legislature only by few provisions of the Law on the principles of property relations, and in Article 64 of that Law mortgage is conditioned by entry into the public register. The fact is that Article 69 of the same Law still forbids fiduciary transfer of property. Mortgage cannot be instituted on a building in the process of construction; efficient (and less expensive) financing of large projects is thus extremely hard, almost disabled." 58 finance, however, is still deterred by problematic enforcement of contracts unsuitable legal framework and underdeveloped practice. A large portion of real estate in Serbia is not properly registered and even when it is registered, it is not unusual for there to be discrepancies between the facts and the data contained in the public register.52 The absence of clarity on these issues makes it riskier and more costly for lenders to assume exposure in housing or commercial real estate. In addition, current practice requires fees of 2 percent charged on the value of underlying property, not just the mortgage value, which translates into a tax on equity that ultimately deters investment. A proposal for a new agency (National Mortgage Insurance Corporation) to provide mortgage guarantee insurance and to act as a secondary market institution for capital market funding currently rests with the MOF. A recent assessment of the mortgage market in Serbia, found that that the government assumption of the major part of the credit risk in mortgage lending--75 or 80 percent according to the proposal--is not an appropriate way to move forward to a private market and can create a large contingent liability.53 The credit registry and credit bureau. Credit information registries and credit bureaus can serve as key institutions to help improve access to finance. They are an institutional response to the problem of asymmetric information in credit markets. Serbia has both a Credit Registry (Solvency Center) and Credit Bureau. The Solvency Center acts as a rating institution and collects financial statements of both financial and non-financial sector, covering the majority of the economy and the submission of the financial statements to the Solvency Center is mandatory. An effective, broad-based credit bureau, with a full scope of credit information, is critically important for developing the means to control and price credit risk. The recently established Credit Bureau is managed by the Serbian Bankers Association (SBA) and was built on the basis of experience in other countries including Germany (the German "Schufa" model), Poland and the Czech Republic. The Credit Bureau is still in the early phase of development. However, it has signed an agreement with nearly all commercial banks in Serbia, under which banks are obliged to provide data on approved credits on the daily basis. The Credit Bureau will collect both positive and negative information and currently is collecting data on the last three years, in order to provide sufficient amount of information to build credit history of clients. Given the current low default rate for credits (3 percent), the report of the Credit Bureau shall, in the majority of cases, serve as a recommendation to the banks to approve a loan request with a shortened procedure. The second stage in the Credit Bureau development will be (i) data gathering on overdue payments of entrepreneurs, companies or natural persons towards other institutions 52According to information from the Republic Geodetic Agency, approximately 80% of real estate is still unregistered. Even when it is registered, in most cases this cannot ensure that a third party gets quality and quantity of ownership title of the subject of entry The good progress has been made related to the privatization process. The real estate of the most of the large companies has already been registered, and the situation regarding real estate involving companies undergoing privatization is progressing. 53"Risk can be shared with the banks through mortgage insurance, but the banks should assume at least 50 percent of the credit risk and preferably a greater share, such as 60 or 70 percent. The proposed approach is also a completely inappropriate way to shore up lending in wobbly state banks" Merrill, S., S. Rabenhorst and S. Butler, "An Assessment of the Mortgage Market in Serbia" (2004), An assessment prepared by the Urban Institute for the USAID. 59 (public utilities and tax authorities) and (ii) offering services to other interested parties. To this end, Serbia must adopt a credit bureau and privacy legislation. Leasing. The passage of the Law on Financial Leasing in May 2003 and accompanying Tax Law amendments have laid the foundation for the development of the leasing industry in Serbia54. The potential demand for leasing services in Serbia is large, and all of the leasing companies plan to expend their businesses in 2004. Thus, many of SMEs should benefit from the increasing competition and the availability of new leasing products. The prospects for the expansion of the leasing industry may be to some extent caused by weak creditor rights, giving leasing an advantage over secured lending.55 The Law on Financial Leasing stipulates a swift repossession procedure. In the case that the lessee does not pay three successive installments, the lessor can request the court to decide on repossession within three days, while the procedure on repossession of the leased object has to be carried out within three days of the court decision. As a result the lessor can effectively repossess the leased object from the lessee in default in six days. Similar provisions contributed to the rapid development of the leasing industry in EU accession countries, where leasing has reached a level of development that is higher than in many EU member states56. A streamlined and publicly accessible leasing registry has not yet been established and amendments that will reflect changes with respect to the introduction of VAT should be prepared. The legal framework is in place and operations related to setting up the registry are in progress. Finally, the Law on Financial Leasing covers movable property (vehicles and equipment/machinery) leaving out real estate leasing. In addition, it should be noted that factoring and other types of assets-based financing practically do not exist. Currently there is a lack of knowledge and skills of financial intermediaries, and legal impediments for assets-based finance. Donor financed credit lines. International donors have played an important role in providing finance facilities and credit lines for SMEs in Serbia. The European Investment Bank (EIB) and European Agency for Reconstruction (EAR) have established facilities to promote access to finance through participating banks together with technical assistance to improve the bank's lending skills. In addition, KfW is managing similar programs with good results. The EBRD has also expressed interest in starting to lend to SMEs through 54To date nine leasing companies started operations (all of them being direct subsidiaries of banks operating in Serbia). The volume of the leasing industry in 2003 was around EUR 54 million. As would be expected at the early stage of development, about 70 percent of leasing transactions was related to purchase of commercial vehicles, and up to 20 percent are machinery and other equipment. Interviews with the leasing companies suggest the volume of leasing transactions will increase to around 170 million euros. In addition, a number of international leasing companies have expressed their interest in operating in Serbia. 55See Baker, Marie-Renee and Alexandra Gross "Development of Non-bank Financial Institutions and Capital Markets in European Union Accession Countries", World Bank Working Paper No. 28 (2004) 56The leasing industry in Serbia enjoys favorable tax treatment. The service tax was eradicated in case of both leasing installments and the management fee that is charged when processing the leasing application, effectively bringing down the cost for lessee to the leasing rate charged by the leasing company. However, as the law and the accompanying tax amendments did not address certain aspects or some issues were not clearly defined, there was a need for clarifications. As noted by the Foreign Investors Council, the majority of issues were clarified in favor of the development of the leasing business including abolition of tax in cases in which leased object is transferred to another lessee and the clarification of accounting treatment of leasing transactions. 60 banks in 2004. By the end of February 2004, the total amount disbursed (including initial grant plus revolving fund) was EUR 23.1 million. The continuation of macroeconomic stability and sound financial sector prudential regulation with improved supervision authorities should provide the opportunity for the increase in the volume of the credit lines, but only at unsubsidized interest rates and only to financial intermediaries that meet minimum eligibility criteria. Data on the number of requests for financing and approved credits shed light on the strong demand for credit. Out of 1,550 received requests, only 450 were eligible and 200 loans disbursed. From interviews with the credit line managers, it turns out that most of the requests for credit were not eligible because of lack of transparency of financial statements, inadequate collateral, incomplete business plans, or unacceptable debt to equity ratios. These figures support the view that there are a limited number of "bankable projects" in Serbia. Development finance institutions. Directing credit, even when it may help meet some social objective, is difficult in practice because it often runs counters to current market forces. Lenders and borrowers want to lend and invest where the returns are greatest, not in sectors deemed a priority by the government. Therefore, state supported schemes should be carefully designed so as not to create distortions in financial markets. Lenders often reclassify loans to comply with the directions, and borrowers surreptitiously use credit for unintended purposes. It is recommended that any government intervention in SME financing to be channeled through banks, with maximum compliance with market mechanisms. As bank financing begins to expand, the need for state interventions should decline. The most active institution in the previous period was the Development Fund.57 Since 2001, the fund has partly changed its role to increase its orientation toward financial support to SMEs. In 2003 it provided more than US$160 million to 1,200 companies, representing 0.75 percent of GDP. It is not clear to what extent directed loans go to unprofitable projects, but recent evidence suggests that the default rate is rather high. The experience of other countries with similar state support schemes suggests that performance of such state institutions demands the establishment and maintenance of the rigorous mechanisms for controlling the funds' activities. Credit guarantees. In 2003 the Law on the Guarantee Fund was adopted. The basic function of this fund will be to guarantee a portion (50 percent) of the credit that is extended by banks to SMEs. The intention is to facilitate credit to prospective businesses, which typically cannot meet conditions for receiving a credit. Credit guarantees offered by governments can encourage more lending to riskier borrowers, including new and small firms. But shifting the risk of default to taxpayers raises several practical challenges. Because guarantees encourage banks to worry less about credit risk and monitoring borrowers, default rates can be high, raising issues of sustainability. 57Since it was established in 1992, the fund was essentially operating outside the regular fiscal system, redistributing budgetary and other governmental resources in a non-transparent way. Although not explicitly prescribed, the purpose was to support strategic enterprises (i.e., to soften the budget constraints of some of the largest loss-makers). Eventually, the fund became a shareholder in some of these enterprises through the debt to equity swap. For more detailed discussion on the role of the Development Fund see Serbia Financial Sector Note and Serbia Private Sector Note. 61 Therefore, prior to substantial increase in the operations of the Guarantee Fund, proper governance must be in place to ensure that the project selection process is influenced strictly by commercial considerations. Policy implications: · The Government should adopt the new Mortgage Law. · Fostering competition in the banking sector through privatization of relatively viable banks to strategic investors. · Streamlined and publicly accessible registry for movable collateral and leasing registry should be established and amendments that will reflect changes with respect to the introduction of VAT should be prepared. · An increase in the volume of the IFIs credit lines at unsubsidized interest rates and only through financial intermediaries that meet minimum eligibility criteria should be considered.58 · Performance of state-support funds needs to be audited and demands the establishment and maintenance of the rigorous mechanisms for controlling the funds' activities. The Government should increase its focus on the provision of public goods. · The government should adopt credit bureau and privacy legislation in order to support an effective, broad-based credit bureau, with a full scope of credit information, 58This should be done only if the conditions of the macroeconomic stability remain and sound financial sector prudential regulation with improved supervisory authorities are developed. 62 Labor Market Rising unemployment. A moderate decline in overall employment and increase in unemployment from the previous decade continued during the period 2000-2003. Despite the growth of GDP in real terms, employment in 2003 dropped by approximately 2 percent compared to 2000. The decline in employment was largely the result of labor shedding from the socially- and state-owned sector, coupled with unfavorable investment climate and labor market rigidities. Average productivity growth measured by GDP per worker in the period 1995-2002 was 2.6 percent per annum. Over the same period, real wages grew much faster than productivity (7 percent), influencing increase in unit labor costs. The hidden unemployment in the socially- and state-owned sector was estimated to be 500,000-700,000. This labor hoarding was caused by protective regulation that prevented layoffs of employees during the UN sanctions. Job security of workers was guaranteed so that the workers were protected from external labor market competition. In recent years, high costs and complex procedures for dismissals have prevented employers from firing redundant workers. Table 12: Key Labor Force Indicators, 2000-2003 1999 2000 2001 2002 2003 Labor force participation rate 70.4 71.5 72.1 72.2 71.8 Employment rate 60.3 62.2 62.6 62.0 60.5 Unemployment rate (LFS- ILO) 14.3 13.1 13.1 14.1 15.8 Share of long term unemployment 79.3 81.7 73.4 74.3 - Unemployment rate (Registered) 25.5 25.6 26.8 31.3 34.5 Source: LFS 1999-2003, Labour Market Bureau, NBS Notes : Labour force participation rate = (Employed + Unemployed / Working age population (Working age population comprises Males 15-64 and Females 15-59) Employment rate = Employed/ Working age population Unemployment rate = Unemployed/ Labour Force (Unemployed are defined as individuals - a) who do not have a job and b) who are actively seeking a job Long Term Unemployment = Duration of unemployment is one year or longer Need for significant employment adjustment. According to the Labor Force Survey (LFS), the unemployment rate was 15.8 percent in 2003, although the official figures are as high as 34.5 percent. The discrepancy between the official and the LFS rates is related to the informal sector. A recent estimate of the extent of excess labor suggests that close to 150,000 workers were on administrative leave.59 Many of the employees on forced leave from state- and socially-owned enterprises participated in the informal economic activity. The estimate suggests that around 8 percent of them had informal sector jobs in 2002.60 These employees had a low reservation wage because they had full health and pension insurance, so that they could accept lower wages in the informal sector. 59The World Bank "Serbia and Montenegro: Poverty Assessment" Report No 26011 YU (2003) 60Registration data overstates the actual level of unemployment in Serbia. High incentives to register with the Labor Market Bureau (LMB) both for first-time job seekers and for previously employed are significant. Unemployment 63 According to PICS, only 25 percent of the respondents state that they have the optimal staffing level (Figure 12). As expected, the largest percentage of enterprises claiming workers overhangs is among state- and socially-owned companies (78.8 percent in 2003), followed by privatized companies where 64.4 percent claim to have redundant workers. The surplus of workers in the new private sector is predominantly the result of labor hoarding, as the owners are reluctant to lay off workers either because of scarcity of skilled workers or because of past investments in specific training. Figure 12: Optimal level of employment as a percent of existing work force 100% 16.9% 9.3% 80% 11.9% 42.2% 18.6% 60% 40% 36.4% 78.8% 64.4% 20% 21.4% 0% New private Privatized SOEs Surplus of workers Current level optimal Shortage of workers Source: PICS 2003. Introduction of more flexible labor regulations. A 2001 Labor Law introduced three significant changes to labor legislation: it increased the scope for employers to dismiss employees, introduced rules on how to structure a redundancy program, and specifically recognized the role of short term and temporary work. Employers were given the right to dismiss employees for non-performance, or as a result of technical, economic or organizational change, thus significantly increasing flexibility of firing. Most importantly, firms with fewer than 50 employees were exempted from the burden of transferring redundant workers to another part of the firm.61 Followed by the new Employment Law, new labor regulation placed Serbia ahead of most other countries in the region, with flexibility of hiring and flexibility of firing close to average and low employment conditions rigidity (Table 13). benefit is not the main reason for registration, as 10 percent of registered unemployed in Serbia actually receive such benefit. The incentives to register come largely from the fact that only registered unemployed can qualify for a broad range of other benefits. It is estimated that one out of four registered unemployed has informal sector job. If self- reported unemployed who perform occasional, informal jobs were treated as unemployed, unemployment rate would increase by 4 percentage points. see Krstic, G. "Labor Markets in Serbia and Montenegro ­ Background Paper for Country Economic Memorandum" (2003) 61The law placed an obligation on the firm to produce a redundancy programme, after consultation with trade unions. The programme specifies whether workers will be transferred, placed on short time working, or whether they will benefit from retraining. The law reduced the ceiling on redundancy pay to a maximum of 5 months wages. Though this new ceiling is comparable to the situation in many EU countries, it is more burdensome than in most transition countries. 64 Table 13: Rigidities in Labor Regulation* Difficulty of Hiring Difficulty of Firing Rigidity of Employment Bulgaria 33 10 28 Croatia 61 50 57 Macedonia FYR 33 40 38 Poland 11 30 34 Germany 44 40 55 Romania 78 50 63 Serbia 28 40 23 Source: The World Bank "Doing Business in 2005 ­ Removing Obstacles to Growth" * Indices range from 0 to 100, with higher values indicating more-rigid regulation. aPart-time contracts and Fixed-term contracts bGrounds for firing, Firing procedures, Notice and severance payment and Job security cHours of work, Leaves and Minimum wage Flexibility differs across ownership sectors. Newly private companies did not report major problems with firing procedures. In practice, to avoid any potential dispute with employees, very often, private companies ask newly employed to sign resignations in advance as a condition for a job, in order to avoid dispute. However, flexibility in SOEs and privatized firms has been limited and should be seen as a part of the privatization/restructuring program. According to PICS, out of the 58 percent of SOEs (and 57 percent privatized companies) that wanted to fire people only 43 percent (37 percent) reported labor shedding. In addition, on average newly private firms had to pay a worker 1.5 monthly salaries when terminating their employment for poor performance, while privatized and SOEs had to pay 5.5 and 3.6 months. A risk of reversal. Recently proposed amendments to the Law on Labor posed a serious risk and could undermine SMEs development and increase the informal economy.62 A recent EC report shows that economies with the most protective labor laws and the most rigid labor policies (former socialist economies) also have the highest level of informal economy.63 The introduction of firstly proposed changes would put Serbia back on the list of countries with the most rigid labor market while the strict limitations on terminating (for economic reasons) of regular employees, including generous severance pay will have as a consequence lower labor turnover rates, lower aggregate employment levels and labor force participation rates and longer average unemployment durations. Although the current draft law brought most important provisions (e.g.minimum wage regulations) to a reasonable level, it is too early to say what will be the impact of the new law. Trade unions. The strong presence of trade unions is evident primarily in privatized and state- and socially-owned companies. According to PICS, on average only 6.2 percent of 62At the moment of writing the draft Labor Law has been submitted to the Parliament. The first draft has included the obligation to conclude a collective agreement for employers who have more than 10 employees, administratively setting the level of minimum wages to 50 percent of the average wage, making dismissal for economic reasons more difficult or costly, etc. 63European Commission - Directorate-General for Employment and Social Affairs, "Undeclared work in an enlarged Union: An Analysis of Undeclared Work: An In-Depth Study of Specific Items." Brussels, May 2004 65 workers within the new private sector (excluding privatized companies) are members of the trade unions, while the network of trade unions remained present across SOEs and privatized companies (83.0 percent and 66.4 percent respectively). Lack of flexible forms of employment. Part-time jobs comprised only one percent of total wage employment in 2002. In the EU and OECD countries, the share of part-time employment represented around 15 percent of total employment. Similarly, temporary work with temporary or fixed term contracts is not common type of employment and accounted for only 5 percent of total employment in Serbia in 2002. However, the PICS survey indicates significant increase in the use of temporary workers in 2002, where approximately one third of firms use this opportunity compared to only one fifth in 2001. Feeble development of entrepreneurship and self-employment activity in the formal sector. The proportion of employers and self-employed in total employment is very low at approximately 5 percent. Self-employment and entrepreneurship are critical for the economy and providing employment opportunities. The share in the overall employment of the formal private sector was 43 percent in 2003. Although private sector employment in manufacturing and service sector rose by 21 percent in 2000-2002 period, this was mainly the outcome of privatization, rather than the result of fast growth of new private sector jobs. Weak job creation in the formal private sector indicates that there are entry barriers for small businesses. Education and skills. According to LFS, the unemployment rate for those with secondary education is twice as the rate for university degree holders (Table 14). Poorly educated long-term unemployed, are faced with scarcity of low-skill jobs, lowering the probability of escaping unemployment. A large number of redundancies makes filling vacant positions fast. According to PICS on average it takes 22 days for a private firm to hire a skilled worker and about two weeks to fill a vacancy for an unskilled position. The abovementioned discrepancy between the labor supply and demand can be also observed through unemployment to vacancies ratio. The registered number of unemployed was 22 times higher than the registered number of vacancies (potential proxy for labor demand). The BEEPS confirms relative ease of filling the position compared to the other transition countries, placing Serbia among top five countries with the shortest time necessary to fill the vacancy. Table 14. Unemployment by Educational Level 2002 in Percent Unemployment rate Share in unemployment Share in employment All 11.1 100.0 100.0 Less than primary 2.9 1.9 8.1 Primary completed 10.1 19.9 22.2 Secondary completed 13.5 69.1 55.2 Higher 8.9 4.8 6.1 University 5.9 4.2 8.4 Source: Labor Force Survey, 2002 66 Skills mismatch. According to the Labor Market Bureau data, there was a significant discrepancy between the registered number of unemployed and job vacancies. About 17 percent of vacancies were unfilled in 2002. This is to some extent explained by different skills of unemployed as compared to the skills required for vacant jobs. The predominant factor was low territorial labor mobility, which is primarily the result of an uneven regional development and a poorly developed housing market. However, given the unemployment rate in 2002, the estimated skill gap indicates that only small portion of unemployment is attributable to observed skill gap. On average, the unemployed, and especially the long-term unemployed, have lower educational attainment and lower skills than the employed. In other words, there is an "excess supply" of poorly educated persons among the unemployed in the sense that, all else being equal, there is not enough low skilled jobs to eliminate unemployment. The index of skill mismatch in Serbia is relatively low by transition economies standards.64 This index for Serbia amounts to 14 percent (i.e. 14 percent of all unemployed will not find job due to skill differential under an assumption that the number of vacancies equals the number of job seekers.) The current labor market does not have adequate information sharing system. For example, the National Labor Office lacks complete database coverage which undermines mobility of labor (i.e., job seekers from Belgrade are not aware of vacancies offered in other places). Need for increase of the on-the-job and formal training. PICS reports that the proportion of the companies offering formal training fell from 56 percent in 2001 to 31.5 percent in 2003. There is a difference between small companies (only 21 percent of which offered training) and medium and large-sized companies (41 percent). Even in companies which offered training, a relatively small number of employees received it and for relatively short periods. In addition, there is a significant problem with the adult education and training related to the certification and quality control. Courses in non- profit institutions usually do not lead to vocational qualifications that are recognized in the labor market. The demand for on-the-job training and adult education will rise as Serbian firms have engaged in technology upgrading over the last couple of years, and because of the failure of the education system to provide students with applicable skills. A significant proportion of firms are in the process of adapting their technologies. Using the new equipment requires skilled workers, given the skills technology complementarities. Openness to foreign trade and technology is likely to induce skills- biased technological change by raising the relative price of skill-intensive products and raising the wage of skilled workers relative to unskilled workers. In 2003 24.3 percent of firms introduced new technologies. Similarly, the share of firms using advanced process information technology rose from 48 percent to 83 percent over the same period. As a consequence, the need to operate new machineries/technologies has led firms to look for and value firm specific training. 64see Krstic, G. "Labor Markets in Serbia and Montenegro ­ Background Paper for Country Economic Memorandum" (2003) and Rutkowski, J., "Rapid Labor Reallocation with the Stagnant Unemployment Pool: The Puzzle of the Labor Market in Lithuania". Policy Research Working Paper No. 2946; Rutkowski, J., "Why is Unemployment so High in Bulgaria". Policy Research Working Paper No. 3017; Rutkowski, J., "Does Strict Employment Protection Discourage Job Creation? Evidence from Croatia". 67 Higher unit labor costs. Real earnings grew faster than productivity, increasing unit labor costs. The benefits of economic growth over the last three years have been translated into higher wages. However, a rise in euro wages contains real wage growth and real appreciation of dinar. High wages and consequently high unit labor costs undermine competitiveness of the economy. Given that Serbian euro wage is higher than the one in Bulgaria or Romania, this could seriously erode its ability to attract needed foreign investments (Table 15). Table 15: Gross Monthly Wage in Euros (average) Bulgaria Romania FYROM Croatia Serbia Serbia Serbia 2003 2003 2003 2003 2001 2002 2003 Gross average monthly wage 138.2 179.4 326.9 743.1 146.2 218.7 254.4 Source: Statistical Offices, National Banks, authors calculations Taxation and labor cost. High payroll taxes, raise the cost of labor, thereby undermining competitiveness and creating incentives for tax evasion and by hiring workers informally. Serbia may be considered relatively competitive in the region from the point of view of income taxation and social security contributions. In 2001 tax reform, the total fiscal burden went down to 0.72 dinars on 1 dinar of net wage. Recent repeal of the 3.5 percent payroll tax reduced the overall fiscal burden on wages and salaries down to 0.69 dinars on 1 dinar of net payment, although the existence of the prescribed minimum base for social security contributions are effectively increasing the overall burden especially in low wage and SOEs dominated sectors such as textile. However, initial tax cuts have not yet brought a reduction in informal employment and an increase in overall employment and labor taxation is still perceived as one of the major sources of discouraging legally based conduct of business. The attractiveness of the informal labor market lies in the fact that the average hourly wage in the informal sector is 28 percent higher than in the formal sector, as a result of tax evasion. Table 16: Overview of the Tax Rates and Social Security Contributions in Percent (2003) Social security taxes total of separate titles Country Personal income tax rates Paid by employer Paid by employee Czech Rep. 15-32 35 12.5 Estonia 26 33 none Hungary 20-40 (18-38) 29 11.5 Poland 19-40 19.83­22.72 18.71 3.57­6.46 2.45 Slovenia 17-50 15.9 22.1 plus additional payroll tax 0 - 14.8. Serbia* 14 16.8 16.8 Croatia 15 - 35 16.6 20.6 plus 10 - 30 local surtax FYROM 23 ­ 35 30.1 30.1 Bulgaria 18 - 29 32.2 10.5 Romania 18-40 37.33 19.67 Slovakia 10-38 38 12.8 68 Source: KPMG, National authorities (rates expected from 2004 in brackets) Notes: *Additional payroll tax 3.5 removed in 2004. For this reason a combination of a carefully devised, revenue neutral and phased reduction in payroll taxes and stronger and non-selective tax and labor inspection should be considered. Policy implications: · The Government should resist attempts to significantly change the 2001 Law in ways that would introduce distortions and by imposing a collective bargaining on companies above 10 and by intervening in the bankruptcy proceeds. · To generate sustainable employment by encouraging growth of the private sector, privatization and restructuring of the large SOEs should be accelerated. · To generate sustainable employment by encouraging growth of the private sector, barriers to entry should be removed as recommended in preceding chapters. · To improve labor mobility, National Labor Office should improve the information sharing system. 69 Land Issues Land ownership. State or social ownership of urban lands is presently enshrined in the Constitution of the Republic of Serbia. Social ownership is essentially state ownership of land which is under the control or in use of public enterprises and state- owned organizations. Privatization has not included urban land, which remains a public property with occupants possessing land use rights. Urban construction land is still state-owned, implying that a foreign investor (as well as a domestic investor) may be given only the right to use it, for which a charge is to be paid65. A new Law on Urban Planning and Construction took effect in 2003, under which urban land for construction or development can be leased for up to a 99-year term.66 The land is developed through the construction of buildings according to plans and regulations set by the municipality council, which is responsible for land management and the provision of utility installations. Much land outside of urban settlements, consisting primarily of agricultural land, was privately owned even during the Socialist period and remains so today. However, starting in the late 1980s, Serbia began restituting agricultural areas, and the process is close to completion. It was estimated that 85-90 percent of agricultural land is privately held. The only exceptions are the large Agricultural "Kombinats" and cooperatives. Nationalized property. A major topic on the Government's agenda is restitution of property nationalised by the former regimes. In the case of assets that are subject to privatization, the Privatization Law adopted in June 2001 established the principle of compensation for nationalized property in money rather than in kind (Article 15) and allocated 5 percent from privatization proceeds for compensation. The lack of resolution of the issue continues to create uncertainty and deters many real estate and other investors. The new government has yet to announce its plan regarding this law. While the details are still to be worked out, the first draft of the law was based on the following general principles: (i) compensation will be done, as a matter of general policy, through bonds, and will be substantially in accordance with principles of the 2001 Privatization Law; (ii) compensation will be done in a fiscally prudent manner; and (iii) to avoid the creation of new injustices, the legally owned non-state property shall not be restituted (in kind) to original owners; other forms of compensation will be used. Burdensome construction permitting procedures67. It takes from 6 to 12 months to begin construction in Belgrade depending on the need to obtain permits. The process involves urban plan updates, detailed planning, tender for the land user right, compensation negotiations with occupants and more than 15 different authority consents and permits (demolishing permit, construction permits, various building infrastructure permits issued by different authorities - electricity, sewerage, traffic, 65The only restriction concerning foreigners rights to acquire the title of ownership (to buy a real estate ­ business premises and apartments) is the reciprocity. 66At this moment it is too soon to tell if the new law will have an effect on the amount or cost of land available for development. 67The impact of the new Law is difficult to assess, as the survey refers to the period before the Law was adopted. 70 parking, seismic, heating, telecom, fire. In the past this has caused delays and cost increases in investment projects, and in some cases, planned investments have been cancelled. The land administration system. The Serbian land administration system is in a poor state. In addition to almost 80 percent of apartments not being registered, real property transactions are often not registered, substantial illegal building has taken place, and there is a significant backlog in civil courts (about 20 percent of cases are land or real estate related). Land records often do not match those in the legal registers kept in municipal courts, which results in a long waiting period for the resolving of such problems. In addition, real estate registers are incomplete and outdated. In large number of cases there are no proper records on the nationalization of the land, which may slow down restitution efforts. The Real Estate Cadastre and Registration Project. The objective of the World Bank supported project is to increase confidence and lower transaction costs by building a more efficient property registration and cadastre system, with the purpose of contributing to the development of effective real property markets. This will be achieved by improving the real estate cadastre system that is operated by the Repuplic Geodetic Authority (RGA) and its branch and local offices. More specifically the project will support the phased transfer of the real property registration system-- currently operating in some of the municipal courts--into the RGA, as stipulated in the Law on State Land Survey, Cadastre and Registration of Real Estate Rights. At the moment, the RGA has established the real estate cadaster in 67 percent of the country, though not yet in Belgrade. However, in addition to providing technical mapping, the project focuses on the registration of property rights by establishing the real estate cadastre throughout the whole country and improving service delivery for secondary transactions (sales, mortgages). Only by fully implementing the real estate cadastre of the reform will the Government reduce the uncertainty over real estate ownership. Property transactions. Transaction costs can be significant and may be out of line with costs in developed market economies. For the court's contract certification function, a fee of 0.5 percent of contract value is charged, which is equivalent to notary charges, but there is a 5 percent transfer tax on residential sales. There is some concern, based on experience in other countries, that excessive transaction costs may discourage registration of transactions and ultimately lead to long term problems for the cadastre system. Legalization of buildings. The combination of problems with licensing and land acquisition process, the lack of effective public control, and the war has resulted in a tremendous increase of illegal buildings. The Law on Urban Planning and Construction provided the owners of an estimated one million illegal dwellings with the possibility to legalize. Although it is too early to see results, the law will contribute to the development of the real estate market, as 375,000 citizens as well as more than 7,000 enterprises applied for legalization. On average, the construction costs of a one-family illegal building is approximately EUR 30,000­40,000, which corresponds to more than 15 years of the average net salary. Although the illegal 71 construction is not a phenomenon solely related to marginalized social groups such as refugees and internally-displaced persons, it will enable these groups easier access to finance and eventual entry of new firms. The mortgage market. Real property mortgaging exists on a modest scale and predominantly in Belgrade. The total number of mortgages on immovable properties in Belgrade was only a few hundred. Only registered properties qualify for mortgaging, which rules out majority of real properties and construction project financing. Industrial properties qualify for 40 percent, while residential and office premises can qualify for 40-70 percent of loan value depending on bank policy and risk analysis. Often additional guarantees are required in addition to the mortgage. At the moment, mortgage creditors are subordinated to tax authorities and other state social obligations. The best practice would be if the mortgages were registered with the real right itself (ie. the where the property is registered) which is in the real estate cadastre maintained by RGA. Policy implications: · The Government needs to introduce legislative amendments that are required to improve the perception of security of tenure and the real property market financing and to attract foreign investors (these should include the change of the urban land concept provided by the current Constitution). · The Government should draft and adopt the Law on Denationalization, based on the four principles discussed in the text above. · The Government should consider decreasing the transfer tax that would facilitate real estate transactions. · The Government should prepare a study on improving administrative procedures in the process of acquiring construction land and suggest necessary measures. The Government should evaluate the current Law on Urban Planning and Construction and recommend amendments if necessary · The Government should develop a model database of remaining available construction land parcels, corresponding plans, potential in-process denationalization requests and previous exchange of rights. 72 CHAPTER 6: KEY POLICY RECOMMENDATIONS While some macroeconomic reforms require major policy changes at the outset of transition (e.g. price and trade liberalization), almost all the reforms of the enterprise and financial sectors take several years to implement and even longer for their full impact to be observed. An economy in transition will expect significant structural change in the face of market prices, ownership relations and other incentives. The reforms of the enterprise and financial sectors described above, if implemented on a consistent and transparent basis and in a timely fashion, could facilitate this reallocation and speed up achieving sustainable growth. Over the past three years, the Serbian authorities have made impressive progress in reforming the socially-owned enterprise sector and banking system. However, significant further policy decisions are needed--many urgently--in order to fully utilize Serbia's potential for economic growth and employment creation. Drawing on the analysis above, and given its limited institutional capacity, the Government should focus its attention on a few areas of reform that are most urgent and are likely to provide benefits at reasonable cost. Macro-stability should remain a fundamental goal; as Serbia's own experience has shown, none of the goals listed below can achieved without macro-stability as a prerequisite. No wonder entrepreneurs worldwide as well as in Serbia point to macro-instability as their number one concern. Political support for reform in Serbia is still fragile, and hence these benefits should also be visible to the public. Based on the ICA as well as a range of other surveys, empirical studies, and the sectoral work supported by the World Bank and other donors, following are priorities for improving the investment climate in Serbia: · Improving contract enforcement The court system in Serbia is markedly pro- debtor. Weak judicial capacity also increases the backlog of cases, and encourages firms to use courts only as a last resort. Accordingly, firms rely more on personal relationships and are less willing to deal with new customers and suppliers, which in turn, limits the number of transactions and limits market dynamism. Weak contract enforcement contributes to uncertainty and non-transparency that create fertile ground for corruption. The following steps are recommended (i) simplifying contract enforcement procedures as provided for by the recently adopted Law on Execution Procedure; (ii) strengthening the performance of bailiffs through the adoption of the relevant by-laws and training; (iii) strengthening court resources through education, improved court statistics and information sharing, to enhance overall court administration and management capabilities and to identify courts primary users and bottlenecks. · Improving access to financing and monitoring the use of state funds. Lack of access to credit limits businesses to using only their own funds, retained earnings, or money borrowed from friends and family. This constraint limits market dynamism in a post-conflict and post-socialist environment in which 73 retained earning have not yet been accumulated. The lack of access to credit through banks and other formal financial institutions removes an important motive for businesses to formalize--thus contributing to the persistence of the informal economy. The following steps are recommended: (i) improvement of the legal and institutional framework for lending including the introduction of the new Mortgage Law; (ii) an increased focus on the provision of public goods (contract/collateral enforcement) rather than directed state support; (iii) a continuation of the development of the publicly accessible leasing and pledge registry; and (iv) audit of the performance of state-support funds. · Reducing regulatory burden. The "regulatory burden" is the consequence of a series of government actions and polices such as a legacy of inefficient and anti-market laws and ministerial regulations, complex company registration, frequent multiple and overlapping inspections or lengthy certification. This regulatory structure distorts and hinders market development, and, for those businesses who succeed in becoming viable, create opportunities for corruption (the bribe tax) and wasting of businessmen's time (time tax). Administrative harassment and corruption as well as the unpredictability of business rules, and the unfairness in their application represent a serious problem. Improvement of the quality of regulation by streamlining, improving transparency and market-friendliness, and consistent application should also reduce the so-called "time and bribe taxes" mentioned above. In addition, increasing the tax administration capacity to support rule-based tax collection rather than punishing successful businesses is an important goal. The following steps are recommended: (i) implementation of the Amendments to the Rules of Operation of the Government requiring responsible ministry to prepare a justification statement for a new law, a decree and an order in line with the OECD Regulatory Impact Analysis, including a justification statement for all laws and other regulations that explains the expected benefits and costs of the actions; (ii) identification and removal of overlapping functions of various state authorities; (iii) adaptation of the national system for accreditation and certification to meet EU standards and technical norms. · Resolution of all outstanding urban land market issues. An ongoing unresolved issue is the failure to adopt a legal framework for dealing with nationalized property and claims for restitution. The failure to resolve the restitution issue regarding both land and other properties, although understandable in view of the recent political history (the Law was ready just before the assassination of PM Zoran Djindjic), continues to create uncertainty and deters many real estate and other investors from investment. The restrictions on use and state ownership and control of urban land (particularly construction land) are major obstacles for the construction and services sectors. The Serbian land administration system is in a very poor state, with most of the apartments in Belgrade and other cities not registered. Analysis and removal of the remaining burdensome construction permits procedures deserve the immediate attention of the authorities. The following steps are 74 recommended: (i) constitutional reform that will change the treatment of urban land converting it to at least modern leases with defined durations and market rents or to private land ownership (ii) removal of various administrative barriers related to the access to land; (iii) resolve the issue of denationalization (restitution) by adoption of a long overdue Law on Denationalization; (iv) phased decrease and enhanced enforcement of the property transfer tax, with compensating increases in other taxes; and (v) enhancement of the currently limited human and physical resources in various components of the land management system, such as cadastre. · Continued restructuring and privatization of SOEs and establishment of an effective bankruptcy mechanism. Schumpeterian "creative destruction" facilitating entry and exit is especially important in the post socialist industrial structure, in which restructuring and bankruptcy are channels for a dynamic entry of small new enterprises. The process whereby bad firms exit and new ones replace them is an important condition for the sustainability of growth in Serbia. Moreover, the although the new private firms are better performers, their scope for growth, and the possibility of entry by new entrepreneurs, are both severely limited by the investment climate and crowded-out by the unrestructured enterprises. The following steps are recommended: (i) continuation with the tender and auction privatization of SOEs remaining in the Privatization Agency pipeline; (ii) Ammendment of the Privatization Law to faciliatate acceleration of the resolution of remaining companies in the restructuring program and readiness to exercise the companies' public debt restructuring in cases when an attractive bid is conditional on partial debt write-off; (iii) sale of minority state stakes and divestiture of assets of the Share Fund; (iv) movement of the unsaleable loss-making conglomerates or parts to bankruptcy procedure; (v) establishment of the requisite institutions; (vi) identification of the resources to train personnel and build the capacity of the Supervisory Agency and the unit that will serve as administrator of socially owned-companies; and (vii) initiation of the bankruptcy proceedings by the state in its role as creditor of SOEs through the tax administration, the electric power company (EPS), and state-owned banks. ICA in the Growth Context We conclude with an effort to put the ICA in the context of the wider picture of growth, which has been fully elaborated in the World Bank's Serbia Economic Memorandum (SEM), prepared in parallel to the ICA. The findings of the ICA are summarized in chapter 4 of the SEM. As stressed in the SEM, the post-socialist transition experience confirms the importance of government policies in determining growth performance.68 First, and not surprisingly, macroeconomic stabilization is essential: growth has generally been stronger and come sooner in countries that have contained inflation. Second, countries 68See for instance, World Bank (2002) Transition: The First Ten Years; Fischer and Sahay (2004), Havrylyshyn (2001), as well as Havrylyshyn and Wolf (1999). 75 that liberalized prices early and comprehensively experienced the earliest output recoveries. Third, output has also risen more rapidly in countries with high average growth rates of exports, underlining that the openness and export orientation are important determinants of growth. Fourth, initial growth is triggered by the efficiency gains resulting from reforms such as hard budget constraints and liberalization, which strengthen incentives for enterprises to become more productive.69 Investment alone does not guarantee early growth and recovery. Also, higher investment often takes two to three years to produce a strong output response. However, as economies reach a more advanced stage of development, higher investment becomes increasingly important if growth is to be sustained. Fifth, foreign direct investment (FDI) appears to play an important role, which can be linked to the quality of the business environment. In addition, the share of the private sector in GDP is distinctly larger for countries with rapid and consistent growth than for those with slow and uneven growth. Finally, investments in human capital through more training and education will allow labor to produce more output with the same work effort, thereby raising overall productivity. In light of the above, achieving more rapid and sustainable growth in Serbia will require the coherent implementation of reform, institutional strengthening and human capital development measures aimed at ensuring a stable macroeconomic environment and creating a competitive economy with solid export performance and a climate hospitable to foreign and domestic investment. To this end, the report focuses on the following policy areas: · Ensuring a macroeconomic environment conducive to growth as Serbia's high, poorly allocated and inefficient government expenditures, coupled with large public and external debts and imbalances, emerge as key impediments for growth and macroeconomic stability. · Improvements in trade policy and regional integration aimed at increasing the export orientation of the economy, enhancing its efficiency through greater competition, and making it more attractive for FDI. · The development of a vibrant real economy driven by the private sector. This requires policies focused on increasing economic efficiency through continued privatization and harder budget constraints and the creation of a business- environment conducive to investments and improved productivity, supported with measures to create a stable and efficient financial sector. · Promoting a well-functioning and flexible labor market which enhances employment by increasing the demand for labor, improves productivity by supporting the reallocation of labor to more productive uses, and encourages the formalization of employment. 69Doyle et al. (2001). 76 · Improvements in education and training systems aimed at enhancing human capital (a key determinant of productivity), improving employment prospects, and facilitating the movement of labor to more productive, higher paying and more secure jobs. As can be seen above, the reform agenda detailed in the ICA is integral to the general strategy of the SEM. 77 BIBLIOGRAPHY ABA-CEELI, Judicial Reform Index for Serbia, (May, 2002) Baker, Marie-Renee and Alexandra Gross (2004), Development of Non-bank Financial Institutions and Capital Markets in European Union Accession Countries, World Bank Working Paper No. 28 Bank Austria Creditanstalt Group, Investment Guide for Serbia (November 2002) Batra, Geeta, Daniel Kaufmann and Andrew H.W. Stone, Investmen Climate Around the World ­ Voices of the Firms from the World Business Environment Survey The World Bank (2003) Business Environment and Enterprise Performance Survey (BEEPS) Interactive Dataset, World Bank and the European Bank for Reconstruction and Development, Carlin, W., S. Fries, M. Schaffer and P. Seabright, Competition and Enterprise Performance in Transition Economies: Evidence from a Cross-Country Survey, EBRD Working Paper No. 63 (2001) Center for Liberal-Democratic Studies, Competition Policy In FR Yugoslavia (2002) Center for Liberal-Democratic Studies, Corruption at the Customs: Combating Corruption at the Customs Administration Center for Liberal-Democratic Studies, Corruption in judiciary (2004) Center for Liberal-Democratic Studies, New Competition Policy ­ A Policy Paper (2003) Center for Liberal-Democratic Studies, New Law on Enforcement Procedure (2004) Cottarelli, Carlom Giovanni Dell'Ariccia and Ivanna Vladkova-Holar (2003), Early Birds, Later Risers, and Sleeping Beauties: Bank Credit Growth to the Private Sector in Central and Eastern Europe and the Balkans. IMF Working Paper 03/213 Davidson, R. and J.G. MacKinnon,, Estimation and Inference in Econometrics, Oxford University Press, 1993 Demirguç-Kunt, A. and V. Maksimovic (2002), Firms as Financial Intermediaries: Evidence from Trade Credit Data The World Bank Working Paper Demirgüç-Kunt, A. and V. Maksimovic, (1998),, Law, Finance, and Firm Growth Journal of Finance 53, 2107-2137 Demirgüç-Kunt, A. and V. Maksimovic, (1999),, Institutions, Financial Markets And Firm Debt Maturity, Journal of Financial Economics 54 (3), 295-336 Djankov, Simeon, Rafael La Porta, Florencio Lopez-de-Silanes, and Andrei Shleifer, Courts, Quarterly Journal of Economics 118(2): 453-517 (2003) Dollar, D., M. Hallward-Driemeier, and T. Mengistae (2003b),, Investment Climate, Infrastructure and Trade: A Comparison of Latin America and Asia The World Bank: Washington DC EBRD, (2001), Transition report 2001: Energy in transition European Bank for Reconstruction and Development, London EBRD, (2002), Transition report 2002: Agriculture and rural transition European Bank for Reconstruction and Development, London EBRD, (2003), Transition report 2003: Integration and regional cooperation European Bank for Reconstruction and Development, London 78 Economics Institute, Deregulation And The Improvement Of The Business Climate Belgrade, Study prepared for GTZ (September 2002) Economics Institute and SMMRI, Serbia Investment Climate Conditions, Regulatory Costs and Productivity, Study prepared for FIAS (2002) Economist Intelligence Unit, Serbia and Montenegro - Country Risk Service" various issues European Commission, FR Yugoslavia Country Strategy Paper 2002 -2006, European Commission, The Western Balkans in Transition Occasional Papers N°1 (January, 2003) European Commission, Transport and Energy infrastructure in South East Europe, (October, 2001) F. Schneider, The Size and Development of the Shadow Economies and Shadow Economy Labor Force of 22 Transition and 21 OECD Countries: What do We Really Know? Paper presented at the Institute of Public Finance, Zagreb, Croatia, (October 2002) Factis, New Investment Incentives Regulation in Serbia, Study prepared for GTZ (September 2002) FIAS, Republic of Serbia - The Climate for Foreign Direct Investment (January, 2002) Foreign Investors Council (2003), White Book - Proposals for Improvement of the Investment Climate in Serbia Foreign Investors Council (2004), White Book - Proposals for Improvement of the Investment Climate in Serbia, Fries, S. T. Lysenko and S. Polanec, The 2002 Business Environment and Enterprise Performance Survey , EBRD Working Paper 84. London, November 2003 G17 Institute,, Cost of doing business in Serbia (December, 2003) G17 Institute,, Cost of doing business in Serbia (March, 2002) Hellman, Joel and Geraint Jones, Daniel Kaufmann,, Far From Home: Do Foreign Investors Import Higher Standards of Governance in Transition Economies? World Bank Discussion Paper, (August, 2002 IMF, Serbia and Montenegro: First Review under the Extended Arrangement, Country Report 03/151, Johnson, Simon, John McMillan and Christopher Woodruff, Contract enforcement in transition, EBRD Working Paper 45 (1999) Klein, Michael, and Bita Hadjimichael,, The Private Sector in Development ­ Enterpreneurship, Regulation and Competitive Disciplines The World Bank (2003) Kraay, Aart, Daniel Kaufmann, and Pablo Zoido-Lobatón,, Governance Matters II: Updated Indicators for 2000/01 World Bank, Policy Research Working Paper 2772, (February 2002) Krstic, G. (2003), Labor Markets in Serbia and Montenegro ­ Background Paper for Country Economic Memorandum La Porta, R. Lopez-de-Silanes, F., Shleifer, A., Vishny, R., (1998)., Law and finance. Journal of Political Economy 106, 1113-1155. Lim, Ewe-Ghee,, Determinants of, and the Relation Between, Foreign Direct Investment and Growth: A Summary of the Recent Literature International Monetary Fund, Working Paper 01/175, (November, 2001) McMilllan John, Simon Johnson, Christopher Woodruff, Property Rights and Finance,: American Economic Review 92 (5): pp.1335-1356, (2002) 79 Morriset, Jacques and Olivier Lumenga Neso,, Administrative Barriers to Foreign Investment in Developing Countries Policy Research Working Paper 2848, World Bank, (May, 2002) O'Driscoll, Gerald and Edwin J. Feulner, Mary Anastasia O'Grady,, 2003 Index of Economic Freedom Heritage Foundation, 2002 OECD, Federal Republic Of Yugoslavia Economic Assessment, (2002) Pissarides, Francesca, Financial structures to promote private sector development in south-eastern Europe, EBRD Working Paper 64 (2001) Quintin Erwan, Contract Enforcement and the Size of the Unofficial Economy, Federal Reserve Bank of Dallas, Dallas, (2002) Raiser, Martin , Alan Rousso and Franklin Steves, Trust in transition: cross-country and firm evidence, EBRD Working Paper 82 (2003) Ruud, P.A., An Introduction to Classical Econometric Theory, Oxford University Press, 2000. Sader, Frank, A View on One-Stop Shops, World Bank Working Paper, Washington DC. (2002) Sanders, Gerard and David Bernstein, Law in Transition: Ten Years of Legal Transition European Bank for Reconstruction and Development, (August 2002 " Schiffer, Mirjam and Beatrice Weder,, Firm size and the business environment : worldwide survey results Discussion Paper 43, International Finance Corporation, (August ,2001) World Bank, Private Sector Development Strategy ­ Directions for the World Bank Group, April 2002, World Bank, Transition: The First Ten Years, Washington D.C. (2002) World Bank, Doing Business Database, Private Sector Advisory Service, World Bank (2002), World Bank, FR Yugoslavia Interim Poverty Reduction Strategy Paper, World Bank , World Development Indictors 2003, Washington, DC. (2003) World Bank, Breaking with the Past ­ The Path to Stability and Growth - Volume I: The Economic, Social and Institutional Reform Agenda (2001) World Bank, Breaking with the Past ­ The Path to Stability and Growth- Volume II: Assistance Priorities and Sectoral Analyses, (2001) World Bank, Country Profile: Serbia and Montenegro, Document prepared for the WB/IFC/OECD Workshop on Investment Climate Reform (June, 2003) World Bank, Doing Business in 2004: Understanding Regulation, Washington, DC. (2003) World Bank, Doing Business in 2005: Removing Obstacles to Growth, Washington, DC. (2004) World Bank, Federal Republic of Yugoslavia: Transitional Support Strategy, Report No. 22090- YU (June, 2001) World Bank, Federal Republic of Yugoslavia: Transitional Support Strategy Update, World Bank, World Development Report 2005: A Better Investment Climate For Everyone, Washington, DC. (2004) World Economic Forum, Global Competitiveness Report 2003-2004, Geneva (2004) 80 Appendix 1: Factor Analysis The World Development Report 2005 analyzes the following issues, among others related to the investment climate: · Property rights · Regulatory frameworks · Taxation · Infrastructure (transport, energy and telecom) · Financial sector (regulation, information to creditors etc) · Labor markets (regulation, hiring and firing, standards and social protection) Respondents in the Serbia PICS 2003 survey were asked to rate 18 dimensions of the investment climate in terms of whether they were important obstacles to doing business. These dimensions were rated from 0 (factor is no problem at all) to 4 (factor is a major obstacle), and can be mapped fairly straightforwardly to the WDR typology described above. Here the ICA addresses the question of whether this is reasonable in the Serbian case. That is, do these responses of managers reflect their views that the investment climate they face has a small number of key dimensions? The WDR framework is a product of many studies of the investment climate; do Serbian managers use a similar framework? This question can be answered using the technique of factor analysis. Factor analysis, in essence, takes data on correlations--here, the views of Serbian managers on 18 different possible dimensions of the investment climate--and reduces the original number of dimensions to a smaller number of components or "factors". After identification of the factors and a standard transformation to make the results understandable, it is possible to interpret the structure of the data. For each factor and each of the 18 original dimensions, we find what are called the "rotated factor loadings". The loadings represent the correlation between a variable (one of the original 18 dimensions) and a factor. If, say, 4 of the original 18 dimensions have large loadings for the first factor, it suggests that these 4 are driven by the respondents underlying attitude to a basic feature of the investment climate, and the first factor represents this basic feature. The results are reported in the table below. When conducting factor analysis, it is important not to try to identify too many or too few factors. Too many and several factors will be trying to catch the same underlying phenomenon; too few, and several phenomena may merge in the data under one factor. Here we report the results for identifying 4 factors. The Stata statistical package was used to perform the analysis. The extraction method was maximum likelihood method, with a minimum eigenvalue of one. For the rotation, the Horst modification of the varimax rotation was used. For a technical discussion with examples, see the Stata 8 Reference Manual. 81 Rotated Factor Loadings 1 2 3 4 Communality Telecommunications 0.04 0.11 0.16 0.54 0.33 Electricity 0.12 0.20 0.20 0.48 0.32 Roads and ports 0.12 0.11 0.15 0.72 0.57 Access to Land 0.13 0.02 0.03 0.50 0.27 Tax rates 0.22 0.20 0.74 0.12 0.65 Tax administration 0.16 0.14 0.73 0.21 0.62 Customs and trade regulations 0.18 0.22 0.45 0.38 0.43 Labor regulations 0.33 0.16 0.46 0.29 0.43 Skills and education of available workers 0.30 0.23 0.27 0.22 0.26 Business licensing and registration 0.29 0.18 0.29 0.31 0.30 Access to Financing 0.22 0.74 0.14 0.24 0.67 Cost of Financing 0.17 0.81 0.22 0.13 0.75 Economic policy uncertainty 0.39 0.50 0.41 0.14 0.58 Macro-economic instability 0.41 0.34 0.34 0.09 0.41 Corruption 0.80 0.11 0.20 0.10 0.71 Crime, theft and disorder 0.87 0.13 0.13 0.16 0.82 Unfair or informal competition 0.50 0.32 0.22 0.17 0.44 Legal system/conflict resolution 0.51 0.32 0.25 0.26 0.49 Each column represents one of the four identified factors. · The highest values in the first column are for obstacles relating to corruption, crime, unfair competition and the hidden economy, and the legal system. The underlying factor that is driving the responses can be described as the rule of law, and maps closely to the WDR heading of "property rights". · The highest values in the second column are for access to and cost of financing and for economic policy uncertainty. This maps straightforwardly to the WDR heading of "financial sector". · In the third column, the highest figures are for tax issues and regulatory issues, with the first set taking higher values. Here we have a case where a single factor appears to be picking up two different ­ but in the eyes of Serbian managers, apparently closely related ­ phenomena, namely the WDR headings of "taxation" and "regulatory frameworks". · The fourth column gives the highest factor loadings to the WDR heading of "infrastructure": telecommunications, electricity, roads and ports and access to land. The last column reports the "communality" of the variable: the percentage of the variance explained by the factor structure. The variables that are particularly well explained by the identified underlying factors are those of corruption and crime as obstacles to business, followed by financing and taxation obstacles. Several conclusions follow from this analysis. First, managers in Serbia perceive the obstacles to doing business in terms that are internally consistent--for each factor, the variables with the highest loadings are logically connected. It is therefore possible to identify these factors as underlying basic features of the investment climate. Second, these underlying basic features are very similar to the framework of analysis of the investment climate that is used for the WDR and that has been identified in numerous studies of the business environment elsewhere in the world. Serbia, in this sense, is not "different" or a "special case". 82 Appendix 2: Comparing TFP across countries using firm-level data A Cobb-Douglas production function for firm i in country j can be written in log form as logYij = 0 + j + LlogLij + KlogKij + uij (1) where Y is output, L and K are labor and capital inputs, 0 is an intercept term shared across all firms and countries, and j is a country-specific intercept term. The level of TFP in country j is (0+j), where 0 is the average level of TFP for all firms and countries, and j as the level of TFP in country j relative to this cross-country average. The error term uij in (1) can be interpreted as the TFP of firm i in country j relative to other firms in country j, i.e., relative to the country j average. The goal of Section 3.1 is cross-country comparisons of average TFP levels of firms, i.e., comparisons of j. One way to obtain estimates of the average TFP level by country is to estimate equation (1) on country-averages. For each variable in (1), calculate the average value for all the firms in each of the J countries. This yields J observations on each of the country-averaged variables. Denoting an average by a superscript bar, the estimating equation becomes logYj = 0 + LlogLj + KlogK j + XjX + ej (2) which is estimated across the J countries, i.e., there are J observations. The estimator using country-averages regression is known as the "between estimator", so-called because it relies on the variation between groups and ignores the variation within groups. The TFP level of country j, j, is now incorporated in the composite country error ej. In an unbalanced panel (such as the PICS-BEEPS dataset, where each survey covers a different number of firms), the standard between estimator is heteroskedastic. The results reported here use a weighted-least-squares (WLS) version of the between estimator; the results are similar when other versions of the between estimator are used.70 When equation (2) is estimated using country-averages, the residuale^ j for country j can be interpreted as the estimate of TFP for country j relative to the average of the countries in the sample. The TFP gap between country j and Serbia is simply the difference between this residual and the Serbia residual, e^ j - e^Serbia . The estimations in Chapter 2 used the consolidated PICS-BEEPS dataset of 59 surveys from 50 countries.71 The main focus of BEEPS was on the business environment, but some basic productivity and performance measures were also collected; PICS collected a much more detailed range of productivity-related data. The BEEPS survey was conducted in 2002; the PICS survey years vary from 2000 to 70The simple WLS version of the between estimator used here uses weights that are proportional to the number of firms in each country; see the discussion of xtreg in the Stata 8 Manual. The results were very similar when using the more sophisticated WLS between estimator of Biørn (2001), which uses a weighting scheme that generates a homoskedastic error term. 71 An additional four surveys (Kenya 2003, Montenegro 2002, Nigeria 2001 and Uganda 2003) lacked basic data and were not used for the analysis. 83 2003. BEEPS covered 27 countries: 26 European and Central Asian transition countries plus Turkey. PICS covered 29 countries, including several transition countries also covered by BEEPS. The total number of firms in the dataset was about 26 thousand, of which about 6.7 thousand were from BEEPS. The estimation of equation (2) was limited to the roughly 15 thousand manufacturing firms from 55 surveys with data on fixed capital and numbers of employees; no other controls were included in the regression. Output was measured using sales rather than the preferred measure of value added, a limitation driven by the absence of value- added data in the BEEPS datasets, but the clear correlation between estimated TFP and GDP per capita obtained using this imperfect measure of output suggests that the regressions are nevertheless generating useable results. All variables were measured US dollars at current exchange rates. Because several countries were surveyed in more than one year, the regression was estimated on 55 country-survey-averages rather than 47 country-averages. The residuals from this regression are the estimates of average firm TFP levels in each of the 55 country surveys. Table 1, column (1), reports the log difference between TFP in Serbia based on the BEEPS 2002 survey and a selection of other countries, i.e., e^ j - e^Serbia . 2002 An alternative estimator is obtained by replacing the country-survey-average TFP level j in equation (1) with a set of 55 country-survey dummies D1 through D55, and then estimating the equation on firm-level data from the 55 country-surveys; the coefficients on the dummy variables are estimates of TFP in country j relative to the "average country". For the purposes of this section, however, it is more convenient to measure TFP relative to the Serbian level. We therefore choose a normalization that sets the intercept to be TFP in the Serbian BEEPS survey of 2002, and the 54 dummy variables are estimates of average firm TFP in the 54 other surveys relative to Serbia: logYij = Serbia2002 + D2D2 + D3D3 + ...+ D55D55 + LlogLij + KlogKij + uij (3) Each of these country-survey dummies is an explicit estimate of the country-survey average TFP level relative to Serbia 2002, and can be used in formal hypothesis testing. This estimator is known variously as the fixed effects estimator, the least- squares dummy variable estimator, and the within estimator, the last so-called because it relies on the variation within groups and ignores the variation between groups. Table 1, column (2), reports the log difference between TFP in Serbia based on the BEEPS 2002 survey and a selection of other countries, i.e., ^ Dj - ^Serbia . 2002 What types of firms perform better than others? Additional performance measures In equation (3), the residual uij can be thought of as the TFP of firm i in country j after accounting for capital and labor, and measured in relation to the average level of productivity of other firms in the same country. Chapter 2 reported how much of the residual can be attributed to various measurable characteristics of firms, by estimating equation (3) augmented by dummy variables for various characteristics of firms and employing the full PICS-BEEPS dataset. The results showed that being private, 84 foreign-owned, and (for manufacturing only) export-oriented, were all associated with high firm TFP levels. Table A.1 below extends this simple analysis to additional measures of performance: whether or not the firm reinvested any profits, whether it invested in new plant in the preceding three years,72 and the log growth rate in its level of employment. The factors of production included in the TFP analysis are dropped, but a control for size (an SME dummy variables) is included. The estimations reported used manufacturing firms only. The results reinforce the results of the TFP analysis: privatized firms, and especially new private firms, perform more strongly compared to other firms in the same country. Table A.1: Performance regressions using firm-level data with country-survey dummies; manufacturing firms only Employment growth Profit reinvested Invested in new plant Privatized 0.006 0.085** 0.060** New private 0.094** 0.167** 0.136** Foreign-owned 0.013 -0.021 0.014 Exporter 0.010 0.033** 0.032* Capital city -0.006 0.005 -0.005 SME -0.025** -0.015 -0.092** Country-survey dummies Yes Yes Yes Number of observations 7,804 firms from 45 6,818 firms from 47 7,071 firms from 44 country country surveys country surveys surveys * significant at the 5 percent level; ** significant at the 1 percent level; heteroskedastic- and cluster-robust standard errors The main text of Chapter 2 extends this analysis to Serbia in particular and reports TFP estimates for Serbia alone (based on the two PICS and BEEPS Serbia surveys) and in comparison to the other country surveys. Table A.2 does the same for the other three performance measures.73 Because of the relatively small number of Serbian firms, the estimations pool manufacturing, construction and services firms and employ sector dummy variables; the results are similar if the regressions are done separately by sector. The results are generally consistent with the TFP findings: exporters, foreign-owned, privatized and especially new private firms perform better around the world. Fewer coefficients are significantly different from zero for the Serbia regressions, again probably because of the smaller Serbian sample; but there are no statistically significant differences between the Serbian coefficients and those for the rest of the sample. The exception is investment in new plant by exporters, which is more likely in Serbia than in the rest of the sample, and hence strengthens the general picture. 72 A "linear probability model" is reported instead of a logit or probit for this indicator, both for ease of interpretation and to simplify the calculations. Estimations of panel logits or probits generated similar results, but were occasionally subject to computational difficulties and are not reported here. 73 The results comparing Serbia to the other PICS-BEEPS countries here and in Chapter 2 were obtained by interacting a Serbia dummy variable with all regressors and estimating a single equation. 85 Table A.2: Performance regressions using firm-level data with country-survey dummies ­ Serbia vs. other PICS-BEEPS countries Employment growth Profit reinvested Invested in new plant 43 45 42 country Serbia Diff? country Serbia Diff? country Serbia Diff? surveys surveys surveys Privatized -0.007 0.014 No 0.098** -0.043 No 0.041** 0.035 No New private 0.071** 0.104** No 0.200** 0.252** No 0.082** 0.132** No Foreign-owned 0.013* -0.040 No -0.119 -0.032** No 0.018 0.018 No Exporter 0.011* -0.012 No 0.040** 0.035 No 0.011 0.103** Yes Capital city -0.005 -0.002 No 0.009 0.004 No -0.001 -0.032 No SME -0.025** -0.003 No -0.033** -0.044 No -0.103** -0.069 No Country-survey dummies Yes Yes Yes Sector dummies Yes Yes Yes Number of 12,908 firms 121,63 firms 12,382 firms observations from 45 from 47 from 44 country surveys country surveys country surveys * significant at the 5 percent level; ** significant at the 1 percent level; heteroskedastic-robust standard errors How does the investment climate affect TFP? To try to detect the impact of the (country-level) investment climate on TFP, the country-survey-averages equation (2) was used, again using a WLS version of the between estimator. As discussed in the text, the motivation for using the between estimator is to reduce endogeneity problems. The equation was augmented with a single investment-climate measure and then estimated. There were 23 investment climate variables available (and hence 23 regression results are reported): 17 investment climate constraints reported by firms in the PICS questionnaire,74 and six composite IC indicators created from these indicators using polychoric principal components analysis, a method designed for ordinal variables such as these. For a discussion of polychoric correlation and the Stata package that implements it, see Kolenikov (2004). The six composite indicators related to infrastructure, taxation, regulation, finance, macro, and property rights. All these indicators were (prior to creating the country averages) measured on a scale that ranged from a minimum of 1 (no constraint perceived) to a maximum of 4 (a major constraint is perceived). The results reported below in Table A.3 included (country averages of) dummy variables for services and construction as well as for privatized and new private firms; a country survey with, for example, a high percentage of privatized firms in the sample, would have a high value for the corresponding country-survey-average variable. Coefficients in bold were significant at the 5 percent level; standard errors are reported in parentheses. The estimations were generally satisfactory. The sample size was 50-51 country surveys based on about 14k firms. The coefficients on capital and labor were plausible and statistically highly significant. The coefficient on services was positive, which was expected ­ services firms typically have a low ratio of value added to turnover, and the use of sales instead of value added as an output measure would tend 74An additional three IC questions that appear in the BEEPS survey were not used, as was one PICS question that was not available for a large number of countries in the consolidated PICS-BEEPS dataset. 86 to make services firms appear more productive. Private ownership, and especially new private ownership, had a positive impact on TFP, consistent with the findings reported in Chapter 2 using the within estimator (regression with country dummies) and also discussed above. With respect to the investment climate, the only category of IC indicator with a consistently significant coefficient was infrastructure: infrastructure constraints have a consistently negative impact on TFP. The coefficient on the composite infrastructure indicator in Table A.3 is about ­0.5 and significant at the 1 percent level. This was one of the lower estimates obtained. A number of variations on this specification were also tried, e.g., regressions that included as regressors foreign ownership and export activity (which were often insignificant), and separate estimations for manufacturing and services. In these estimations, the coefficients on the infrastructure measures were typically in the range of ­0.5 to ­0.9. The only notable difference, as mentioned in the main text, is that the estimation for just services firms suggested other IC constraints also had a negative impact on TFP. These results were generally less satisfactory overall, however, because of the estimated coefficient on labor (frequently insignificant) and the smaller sample size (fewer than 40 country surveys), and are not reported. 87 Table A.3: The impact of the investment climate on TFP. Estimations using country-survey- averages IC New N coefficient Privatized Private Fixed K Labor Services Constr. Infrastructure 13,668 -0.523 2.314 3.281 0.677 0.531 0.823 0.243 standard error (50) 0.159 1.098 0.733 0.067 0.094 0.331 1.141 Telecoms 14,432 -0.506 0.912 2.150 0.654 0.543 0.597 1.239 standard error (51) 0.157 1.322 0.746 0.069 0.081 0.358 1.465 Electricity 14,492 -0.321 1.693 3.011 0.651 0.511 0.601 0.973 standard error (51) 0.090 1.163 0.690 0.066 0.083 0.332 1.348 Transport 14,183 -0.087 2.637 3.481 0.687 0.476 1.051 0.886 standard error (50) 0.238 1.139 0.828 0.076 0.111 0.365 1.120 Land Access 14,027 -0.336 1.513 2.687 0.645 0.515 0.540 1.781 standard error (51) 0.116 1.242 0.838 0.072 0.079 0.378 1.346 Taxation 14,244 0.061 1.730 2.930 0.660 0.468 0.815 2.031 standard error (50) 0.168 1.255 0.865 0.078 0.104 0.375 1.354 Tax rates 14,352 0.084 1.694 2.884 0.656 0.469 0.821 1.916 standard error (50) 0.154 1.280 0.887 0.080 0.107 0.380 1.373 Tax admin. 14,276 0.031 1.679 2.914 0.659 0.469 0.804 2.053 standard error (50) 0.167 1.251 0.859 0.078 0.102 0.373 1.357 Regulation 12,846 0.282 1.770 2.632 0.636 0.420 0.832 1.486 standard error (50) 0.195 1.142 0.764 0.079 0.104 0.345 1.290 Customs 13,891 -0.095 1.892 2.960 0.652 0.484 0.765 1.867 standard error (51) 0.172 1.182 0.810 0.075 0.088 0.346 1.270 Licensing 13,986 0.135 1.809 2.826 0.669 0.446 0.790 1.780 standard error (50) 0.201 1.226 0.835 0.080 0.107 0.373 1.292 Labor reg. 14,058 0.375 1.724 2.381 0.586 0.437 0.857 1.627 standard error (50) 0.094 1.087 0.701 0.078 0.084 0.346 1.227 Finance 14,027 0.013 1.710 2.836 0.656 0.461 0.843 1.626 standard error (51) 0.176 1.349 0.888 0.082 0.101 0.389 1.436 Finan. access 14,162 -0.018 1.780 2.862 0.663 0.460 0.788 1.770 standard error (51) 0.195 1.279 0.866 0.080 0.097 0.380 1.343 Finan. cost 14,231 0.035 1.568 2.788 0.652 0.462 0.868 1.612 standard error (51) 0.150 1.317 0.883 0.080 0.102 0.387 1.421 Macro 14,016 0.014 2.388 3.294 0.678 0.455 0.994 1.272 standard error (50) 0.160 1.087 0.753 0.076 0.102 0.343 1.131 Uncertainty 14,099 0.039 2.352 3.273 0.677 0.457 0.984 1.281 standard error (50) 0.155 1.094 0.753 0.076 0.103 0.355 1.143 Instability 14,083 -0.015 2.484 3.401 0.681 0.459 1.011 1.323 standard error (50) 0.162 1.106 0.787 0.076 0.102 0.341 1.104 Property rights 13,893 -0.103 2.725 3.596 0.688 0.465 1.008 0.920 standard error (50) 0.158 1.195 0.855 0.078 0.096 0.350 1.086 Corruption 14,219 -0.139 1.927 3.147 0.672 0.459 0.745 1.637 standard error (51) 0.174 1.265 0.879 0.078 0.090 0.357 1.333 Crime 14,046 -0.045 2.550 3.477 0.684 0.465 1.008 1.233 standard error (50) 0.128 1.170 0.818 0.077 0.100 0.350 1.092 Other Skills shortages 14,398 0.226 1.693 2.723 0.628 0.435 0.791 1.919 standard error (51) 0.166 1.133 0.795 0.086 0.104 0.356 1.229 Unfair compet. 14,297 0.068 1.786 2.888 0.662 0.458 0.793 1.895 standard error (51) 0.170 1.158 0.822 0.082 0.107 0.364 1.239 Notes: bold indicates significant at the 5 percent level; heterskedastic-robust standard errors; column (1) reports number of firms used in the regression, and in parentheses, the number of country surveys. 88 What types of firms report investment climate constraints? A simple linear regression analysis was used to explore the relationship between characteristics of firms and the perceived investment climate constraints they report. The method used was a series of least squares regressions with country dummies. The dependent variables in these regressions were the 17 basic investment climate constraints and the six composite IC indicators described above. The regressors were ownership variables (the benchmark was state- and domestic- owned), export orientation, and dummies for location (capital city) and sector (services and construction ­ the benchmark was manufacturing), and a full set of country-survey dummies. The results are reported in Table A.4 below. Each box starts with the result for the composite IC indicator. Bold indicates a value that is significantly different from zero at the 5 percent level. Standard errors are also reported. The first column in the table reports the number of firms used in the estimation, and in parentheses, the number of country-surveys used. The interpretation of the reported coefficients is as follows: a coefficient b means that firms with that characteristic, ceteris paribus, report an IC constraint that is b larger than firms without that characteristic. For example, one of the largest coefficients in the table is for new private firms in relation to constraints from corruption, at 0.268; the standard error is small, at 0.035. The estimate means that, on average and everything else equal, new private firms in the PICS-BEEPS sample report a perceived IC constraint score for corruption that is 0.268 points higher than the perceived constraint reported by state-owned firms. Privatized firms report a score that is 0.157 points higher than for state-owned firms, and also with a small standard error of 0.045. Foreign-owned firms report a score that is 0.026 points higher, but this is not significantly different from zero, i.e., not significantly different from that reported by domestically-owned firms. The results are straightforward and are discussed in the main text. Private ownership, and especially new private firms, are more likely to report that they face constraints, almost across the board. Foreign-owned and exporting firms are more likely to be constrained by poor infrastructure and especially by regulation, but not by finance. Larger firms are also more likely to be constrained by infrastructure, regulation, and skills shortages. The procedure was then repeated, but including one of four performance measures as an additional regressor: a dummy variable for whether firm had reinvested any profits, TFP (the residual from a simple 2-factor country-dummy production function estimation), log employment employment growth, and a dummy variable for whether or not the firm had opened a new plant in the preceding three years. The coefficients on the other covariates were very similar to those reported in Table A.4 below, and are not reported for these augmented regressions as well. Instead, Table A.5 reports just the coefficients for the performance measures for each of the four sets of regressions. 89 These results are also discussed in the main text of Chapter 2. The main result is the performance measures are usually positively correlated with perceived investment climate constraints, particularly in the areas of infrastructure and regulation, and to a lesser extent with property rights and skills shortages. For the remainder of the IC categories there are no strong correlations, with the exception of financial constraints, which show a negative correlation with various performance mesaures. The interpretation of this pattern is that growing, investing firms are more likely to be constrained by problems in the investment climate than poorly performing firms, and that poorly performing firms are more likely to complain that they need cash than are strong firms. 90 Table A.4: What types of firms report investment climate constraints? Regressions with country-survey dummies New Foreign Capital N Privatized Private Owned Exporter SME City Services Constr. Infrastructure 15,455 0.020 0.090 0.067 0.040 -0.041 -0.014 0.019 -0.022 standard error (50) 0.028 0.023 0.020 0.017 0.016 0.014 0.019 0.026 Telecoms 16,427 0.027 0.054 0.098 0.063 -0.021 -0.025 0.179 0.024 standard error (51) 0.036 0.030 0.025 0.021 0.020 0.018 0.024 0.033 Electricity 16,522 0.010 0.066 0.042 0.010 -0.044 -0.098 -0.046 -0.123 standard error (51) 0.038 0.031 0.025 0.022 0.021 0.019 0.025 0.035 Transport 16,419 0.016 0.062 0.099 0.060 -0.090 0.024 -0.009 -0.055 standard error (51) 0.036 0.030 0.025 0.022 0.021 0.019 0.024 0.034 Land Access 15,627 0.037 0.183 -0.019 -0.007 0.034 0.068 -0.077 0.073 standard error (50) 0.038 0.031 0.026 0.023 0.022 0.020 0.025 0.040 Taxation 15,939 0.163 0.211 0.036 -0.029 -0.040 0.060 0.015 0.113 standard error (49) 0.039 0.032 0.025 0.022 0.020 0.019 0.025 0.039 Tax rates 16,105 0.221 0.235 0.012 -0.064 -0.038 0.056 -0.003 0.091 standard error (49) 0.043 0.035 0.027 0.023 0.022 0.021 0.027 0.042 Tax admin. 15,976 0.119 0.208 0.073 0.024 -0.038 0.064 0.021 0.139 standard error (49) 0.044 0.036 0.028 0.024 0.023 0.022 0.028 0.043 Regulation 14,360 0.093 0.175 0.152 0.159 -0.124 0.059 -0.011 -0.020 standard error (49) 0.031 0.025 0.021 0.018 0.017 0.016 0.020 0.033 Customs 15,357 0.158 0.304 0.285 0.322 -0.195 0.113 -0.096 -0.163 standard error (50) 0.043 0.033 0.027 0.025 0.023 0.022 0.028 0.044 Licensing 15,660 0.142 0.238 0.065 0.093 -0.030 0.075 0.103 0.097 standard error (49) 0.042 0.033 0.027 0.024 0.023 0.021 0.027 0.043 Labor reg. 15,743 -0.009 0.018 0.088 0.110 -0.161 0.010 -0.017 -0.014 standard error (49) 0.037 0.030 0.025 0.022 0.021 0.019 0.024 0.037 Finance 15,512 0.040 0.075 -0.223 -0.003 0.075 0.041 -0.152 0.000 standard error (50) 0.040 0.032 0.025 0.022 0.021 0.020 0.025 0.040 Finan. access 15,725 -0.030 0.059 -0.260 0.005 0.118 0.076 -0.167 -0.001 standard error (50) 0.046 0.037 0.029 0.025 0.024 0.023 0.029 0.047 Finan. cost 15,874 0.095 0.085 -0.184 -0.002 0.022 0.009 -0.141 0.002 standard error (50) 0.044 0.036 0.028 0.024 0.023 0.022 0.028 0.044 Macro 16,010 0.106 0.116 0.016 0.015 -0.031 0.059 0.047 0.046 standard error (50) 0.038 0.031 0.024 0.020 0.020 0.019 0.024 0.037 Uncertainty 16,129 0.081 0.107 0.008 0.021 -0.053 0.066 0.065 0.064 standard error (50) 0.042 0.034 0.026 0.022 0.022 0.021 0.026 0.041 Instability 16,128 0.129 0.126 0.042 0.026 -0.028 0.050 0.029 0.029 standard error (50) 0.042 0.034 0.026 0.023 0.022 0.021 0.027 0.041 Property rights 15,721 0.128 0.221 -0.031 -0.051 -0.010 0.092 0.099 0.103 standard error (50) 0.037 0.029 0.024 0.021 0.020 0.019 0.024 0.038 Corruption 15,846 0.157 0.268 0.026 -0.034 0.010 0.103 0.036 0.136 standard error (50) 0.045 0.035 0.028 0.025 0.023 0.022 0.029 0.046 Crime 16,195 0.089 0.166 -0.071 -0.076 -0.044 0.084 0.181 0.085 standard error (51) 0.041 0.033 0.026 0.024 0.022 0.021 0.026 0.041 Other Skills shortages 16,167 0.040 0.032 -0.020 0.092 -0.125 0.082 -0.054 0.019 standard error (50) 0.040 0.032 0.027 0.023 0.022 0.020 0.026 0.041 Unfair compet. 15,982 0.262 0.232 -0.054 -0.099 -0.022 0.063 0.049 0.056 standard error (50) 0.044 0.035 0.028 0.025 0.023 0.022 0.028 0.043 Notes: bold indicates significant at the 5 percent level; heterskedastic-robust standard errors; column (1) reports number of firms used in the regression, and in parentheses, the number of country surveys. 91 Table A.5: How is the performance of firms related to investment climate constraints? Regressions with country-survey dummies TFP Employment Any profit Invested in growth reinvested new plant N Coeff Coeff Coeff Coeff (s.e.) N (s.e.) N (s.e.) N (s.e.) Infrastructure 11,281 0.010 12,288 0.062 10,108 0.086 11,773 0.083 standard error (48) 0.007 (45) 0.036 (46) 0.017 (44) 0.020 Telecoms 11,764 0.010 12,758 0.063 10,548 0.042 12,238 0.016 standard error (48) 0.009 (45) 0.049 (46) 0.021 (44) 0.025 Electricity 11,825 -0.001 12,844 -0.001 10,623 0.078 12,318 0.048 standard error (48) 0.009 (45) 0.049 (46) 0.023 (44) 0.028 Transport 11,776 0.021 12,789 0.052 10,581 0.069 12,254 0.112 standard error (48) 0.009 (45) 0.047 (46) 0.021 (44) 0.028 Land Access 11,375 0.001 12,410 0.183 10,219 0.136 11,911 0.149 standard error (48) 0.009 (45) 0.049 (46) 0.024 (44) 0.030 Taxation 11,594 0.006 12,531 0.013 10,331 0.023 12,075 0.059 standard error (47) 0.009 (44) 0.043 (45) 0.024 (44) 0.026 Tax rates 11,689 0.006 12,675 0.017 10,475 0.024 12,225 0.052 standard error (47) 0.009 (44) 0.045 (45) 0.026 (44) 0.028 Tax admin. 11,618 0.007 12,552 0.009 10,357 0.035 12,102 0.070 standard error (47) 0.010 (44) 0.050 (45) 0.027 (44) 0.030 Regulation 10,396 0.026 11,406 0.114 9,300 0.123 10,903 0.166 standard error (47) 0.007 (44) 0.039 (45) 0.019 (44) 0.023 Customs 11,264 0.066 11,630 0.186 9,519 0.176 11,138 0.220 standard error (48) 0.010 (44) 0.057 (45) 0.027 (44) 0.032 Licensing 11,329 0.022 12,489 0.145 10,292 0.102 12,038 0.171 standard error (47) 0.009 (44) 0.050 (45) 0.026 (44) 0.031 Labor reg. 11,404 0.002 12,526 -0.015 10,322 0.083 12,071 0.094 standard error (47) 0.009 (44) 0.044 (45) 0.023 (44) 0.028 Finance 11,419 -0.042 12,205 -0.154 9,992 -0.042 11,624 -0.012 standard error (48) 0.009 (45) 0.047 (46) 0.025 (44) 0.027 Finan. access 11,527 -0.045 12,384 -0.169 10,159 -0.066 11,802 -0.027 standard error (48) 0.010 (45) 0.053 (46) 0.029 (44) 0.032 Finan. cost 11,613 -0.037 12,501 -0.157 10,277 -0.004 11,936 -0.011 standard error (48) 0.010 (45) 0.051 (46) 0.027 (44) 0.029 Macro 11,657 0.002 12,616 -0.087 10,371 0.008 12,062 0.027 standard error (48) 0.008 (45) 0.042 (46) 0.023 (44) 0.025 Uncertainty 11,720 0.005 12,697 -0.100 10,458 0.003 12,156 0.045 standard error (48) 0.009 (45) 0.047 (46) 0.025 (44) 0.027 Instability 11,718 0.005 12,703 -0.075 10,465 0.023 12,158 0.025 standard error (48) 0.009 (45) 0.047 (46) 0.026 (44) 0.027 Property rights 11,494 -0.011 12,274 0.055 10,061 0.053 11,736 0.081 standard error (48) 0.008 (45) 0.045 (46) 0.023 (44) 0.026 Corruption 11,560 -0.008 12,384 0.029 10,171 0.057 11,851 0.101 standard error (48) 0.010 (45) 0.052 (46) 0.028 (44) 0.030 Crime 11,643 -0.009 12,532 0.070 10,318 0.043 12,002 0.063 standard error (48) 0.009 (45) 0.050 (46) 0.026 (44) 0.030 Other Skills shortages 11,733 -0.006 12,720 0.109 10,503 0.121 12,188 0.182 standard error (48) 0.009 (45) 0.052 (46) 0.025 (44) 0.030 Unfair compet. 11,641 0.007 12,523 0.001 10,323 0.040 12,003 0.143 standard error (48) 0.010 (45) 0.051 (46) 0.027 (44) 0.030 Notes: bold indicates significant at the 5 percent level; heterskedastic-robust standard errors; column (1) reports number of firms used in the regression, and in parentheses, the number of country surveys. 92 Table A.6 reports the results of the same exercise applied to just the firms from the 2003 Serbia PICS dataset. The specifications reported do not include any performance variables; as noted in the main text of Chapter 2, these were never significant, probably because of multicollinarity between the other regressors in the equation and the performance measure, and because of the relatively small sample size of just 408 firms. There were some minor differences in specification compared to the results reported in Table A.4 above: (i) the composite indicators for different categories of the IC were calculated using the standard principal components components rather than the more sophisticated polychoric method (to maintain the parallel with the factor analysis reported in Appendix 1 and in the main text); (ii) an additional location category was used, Vojvodina (the omitted category was central Serbia); (iii) the benchmark industrial cateogry is services (because of the large number of services firms in the survey); (iv) an additional IC indicator was available, quality of the legal system, making 18 raw indicators and 6 composite indicators in all. The results are discussed in the main text of Chapter 2. Foreign-owned firms perceive greater constraints than domestically-owned firms across almost the entire investment climate, except for finance and taxation. Privatized firms face greater skills shortages and bigger constraints from business licensing, and new private firms also perceive licensing as an important constraint. Exporters complain about excessive customs and trade regulations. 93 Table A.6: What types of firms in Serbia report investment climate constraints? New Foreign Privatized Private owned Exporter SME Manuf. Agric. Beograd Vojvodina Infrastructure 0.127 0.215 0.617 0.037 -0.299 0.243 0.558 0.022 -0.027 standard error 0.170 0.133 0.205 0.125 0.124 0.121 0.207 0.129 0.122 Telecoms 0.024 0.297 0.205 -0.139 -0.374 0.107 0.536 0.304 0.041 standard error 0.203 0.171 0.242 0.147 0.157 0.145 0.271 0.165 0.149 Electricity 0.296 -0.002 0.639 -0.211 -0.256 0.549 0.712 -0.027 0.049 standard error 0.224 0.174 0.289 0.156 0.162 0.152 0.260 0.164 0.162 Transport 0.059 0.345 0.943 0.414 -0.276 0.076 0.434 -0.170 -0.155 standard error 0.196 0.171 0.309 0.172 0.158 0.160 0.265 0.169 0.166 Land access -0.115 -0.037 0.694 -0.040 -0.029 0.120 0.522 -0.148 -0.229 standard error 0.175 0.159 0.305 0.140 0.147 0.131 0.222 0.148 0.141 Taxation 0.225 0.204 0.252 -0.012 0.099 -0.085 0.481 0.160 0.305 standard error 0.191 0.189 0.217 0.149 0.167 0.151 0.250 0.164 0.158 Tax rates 0.505 0.264 0.262 -0.062 0.094 -0.071 0.475 0.355 0.478 standard error 0.216 0.202 0.238 0.161 0.180 0.165 0.261 0.190 0.168 Tax admin. -0.055 0.143 0.241 0.038 0.105 -0.100 0.487 -0.035 0.132 standard error 0.220 0.209 0.270 0.169 0.187 0.167 0.274 0.174 0.177 Regulation 0.203 0.286 0.660 0.464 -0.064 0.096 0.367 0.045 0.349 standard error 0.153 0.150 0.251 0.136 0.134 0.137 0.200 0.136 0.142 Customs 0.282 0.394 0.623 0.685 0.023 0.111 0.366 0.065 0.272 standard error 0.201 0.189 0.299 0.169 0.168 0.164 0.261 0.169 0.172 Licensing 0.497 0.333 0.462 -0.023 -0.108 0.007 0.295 0.005 -0.085 standard error 0.210 0.162 0.287 0.147 0.154 0.149 0.233 0.165 0.150 Labor reg 0 111 0 161 0.703 0 205 -0 166 0 077 0 367 0 021 0.440 standard error 0.198 0.173 0.277 0.154 0.157 0.150 0.222 0.159 0.165 Macro 0.126 0.089 0.522 -0.061 0.002 0.262 0.426 0.272 0.423 standard error 0.192 0.163 0.203 0.143 0.149 0.143 0.216 0.155 0.150 Uncertainty 0.102 0.063 0.414 0.074 -0.025 0.203 0.452 0.404 0.483 standard error 0.205 0.177 0.228 0.153 0.159 0.156 0.219 0.173 0.160 Instability 0.149 0.115 0.628 -0.192 0.029 0.320 0.401 0.142 0.364 standard error 0.214 0.178 0.244 0.163 0.164 0.157 0.259 0.172 0.164 Finance -0.283 -0.353 0.126 0.121 0.300 0.460 0.710 0.348 0.037 standard error 0 205 0 186 0 303 0 164 0 169 0.163 0.261 0 181 0 163 Finan access -0 338 -0 327 0 322 0 119 0 230 0.520 0.874 0.383 0 008 standard error 0 227 0 208 0 341 0 177 0 188 0.178 0.277 0.195 0 181 Finan cost -0 225 -0 380 -0 082 0 124 0.375 0.397 0.536 0 310 0 069 standard error 0 221 0 199 0 331 0 175 0.182 0.172 0.267 0 193 0 178 Property rights -0 019 0 052 0.538 0 034 -0 043 0 227 -0 021 0 277 0.292 standard error 0 180 0 165 0.221 0 142 0 146 0 144 0 209 0 145 0.151 Corruption -0 045 -0 084 0.577 0 028 0 045 0 212 -0 265 0 106 0 308 standard error 0 233 0 204 0.277 0 177 0 183 0 175 0 250 0 183 0 183 Crime -0 095 0 083 0.566 0 019 -0 153 0 162 -0 100 0 063 0 319 standard error 0 213 0 198 0.289 0 170 0 173 0 169 0 254 0 180 0 178 Other Skills shortages 0.414 -0 061 0 207 0 017 -0.336 0 059 0 211 0 158 0 078 standard error 0.212 0 184 0 276 0 158 0.168 0 152 0 236 0 165 0 154 Unfair compet 0 151 0 126 0.626 0 045 0 156 0 291 -0 085 0.368 0 238 standard error 0.239 0.202 0.254 0.175 0.179 0.174 0.271 0.187 0.188 Legal system -0.082 0.087 0.378 0.047 -0.218 0.250 0.382 0.590 0.303 standard error 0.222 0.212 0.322 0.173 0.193 0.173 0.259 0.187 0.186 Notes: bold indicates significant at the 5 percent level; heterskedastic-robust standard errors; sample size is 408 for all estimations. 94 References Erik Biørn, "How is Generalized Least Squares Related to Within and Between Estimators in Unbalanced Panel Data?", Memorandum No. 6/2001, Department of Economics, University of Oslo. Stanislav Kolenikov (2004), "polychoricpca" Stata program and help file. Available on-line from http://www.unc.edu/~skolenik/stata/ "xtreg", Stata 8 Manual: Cross-sectional Time-series, 2003. 95 alci la DI) ,5-1 istat (W es -2 gener (2 er werlo St B edatlulcac of ) fo 2001 32/ ofsec of d an )99. deam ). -54(d etar ) W omfr pri eht si sse Sco ng al . o vo ertt so rale sdar lim18. tuoba o o are nio itiiltu,y rge o o pre be o eerf o truction egr IMF, catidiin Kot Fed onw cor 10 cons/ onten hou m 00 wards 20 byPDGl egr egr ro,ten to s ectr (en etc.sntema 1-5 Re lidated deficit egr egr ther egr egr nuna earsy onten onten nteo ferre esi dicem conso fiscal onten onten onten omfreul ylsto M onten eluav M wit fro on ina M M M ia ritap M M 00 ­ gheiheht M )99. 9)9.3-3( M egaerva tsap and 9991 omrf ionat deduclxE.s dis 20 ice ces,iv and and 4, and -1 in unfree and iab -2000 pc nfli pr hichwrof ntemn 1- (1 werol Data Description Ser cein(s ser ce re il iab iab iab iab vagnkitaxe lyst iab edhtgi 9951 Offi PDG nomgnidivdi and and dna Serb iab iab Ser Serbia Ser Serbia ardswno sdo Co reta go tedriehni herot Serbia Ser overG Serbia Ser Ser Score Ser Ind eerF Mo Ser We ionatflni th teranoital wi inf e I)E T, Officl m MA F IMF MI e e Statistica I,D ate(m randuome Data source F. W IMF EBRD *esti NBS EBRD NBS M IMF IMF NBS NBS, EBRD EBRD Heritag Heritag 2003 20,729 2,560 3* 7.8 .. 5.4 46.1 3.8 3,550 14,241 4.0 4.28 5 2002 15,662 1,933 4 14.8 21.4 6.0 46.7 4.2 2,280 11,839 4.0 4.21 5 2001 11,577 1,429 5.5 40.7 91.3 22.9 40.2 1 1,169 11,948 4.0 .. .. 96 2000 8,670 1,035 5.0 111.9 60.4 116.0 37.6 0.2 524 11,403 2.3 .. .. 1999 10,090 1,205 18- 45.4 37.1 51.8 .. .. 297,4 10,744 2.3 .. .. 1998 15,487 1,459 1.9 44.4 29.5 49.5 .. .. 326,1 10,539 2.3 .. .. 9 1997 1641 1,549 10.1 9.0 21.3 9.2 .. .. 395,7 9,770 2.7 .. .. 7 Data 1996 1647 1,558 7.8 .. 94.3 .. .. .. .. 11,477 2.7 .. .. Policies 5 1995 1528 1,449 6.1 .. 78.6 .. .. .. .. 11,058 2.7 .. .. mico Reference mil. US$ e total,e of US$) GDP rvese riodep il.m ICA of of eedomfr 3: Macroecon current periodfo averag %)( nditure % rese (in end bt exp in de pricefo ic licy GDP, growth end period chang A3.1 ationl poy pc GDP ion, inf GDP) ficit,ed ex tion econom Index (mil.US$), of of alisa Appendix Table Nominal GDP Real Inflation, Inflat Core Government (% Fiscal reignoF foreignla NBS Tot EBRD liber Index Monetar t and hou dna eth , 03 a by in of the ffi 5 1- liat in 20 dna ionti emst dna tar wit 01 ot debt n- rbia 20 ed Se 9991 0, rmof Serbiehtr ad ngeah edz reo no .rse impar abro encyr reali the befd Score to m 200 cludni In dem cur Serbia d is ilehw fost en from excnigreof add . su an as ludec s s n te,ar s.omst barrieadrt en ent refer Fro. woflin ber , s'nt ee includes con riffta uc ge tw op ta ovo em ec ficited nce rtap bank ignreof an be ertt stme da e emti itta ch Serbia credits be era ni Kos adrt s cit eht th of eg eht werlogn rem ex of av tinga inv 1999 gnid 1 ngie - cluin er-DbotcO paymfoec - wi ssion.i icati `s dicin ro sionim n under eht eg SBN rog ther domes ns inde for en om genahcxe sniot M reig era debt fo,s op tohija ng enento rog dethgiew onitpurroc rogen enento nteo M ro d with Me KFOR M gheiheht M of valu ithw 1996 eg In an balanehtylno desirenidteisop ind dehs eratiopo the 4, nda nda 5 oftn 1- ent Mont, s s canig ts ge the entd entm orgen and of nda an pay 1- outstandi ia nda xdein ia ier werlo tmea Data description Serbia* From* Mon ovosoK 0022 error ardswno eretsigre rsefsnatrtner ch is Cur fore accoun ex from esirnon n th rbia ore the Monte Serbia The Kosovo arrival Serbia Serbia Serbia Serbia Serbia Serb Score Serb A barr wi Se Sc tre e e Data source NBS NBS NBS NBS NBS NBS NBS NBS NBS NBS NBS EBRD Heritag Heritag 2003 4,847 2,289 2,419 1,928 1,360 13,482 71.0 387.0 423 2.2 12.1 3.3 4 5 2002 3,908 1,505 2,227 1,731 475 11,162 78.2 383.2 223 1.6 7.7 3.3 4 5 2001 2,834 1,204 1,239 648 165 11,076 103.9 429.5 107 1.0 4.1 3.0 .. .. 97 2000 1,788 848 621 350 25 10,789 .. .. .. .. .. 1.0 .. .. 1999 1,620 668 716 716 112 .. .. .. .. .. .. 1.0 .. .. 1998 1,816 655 660 660 113 .. .. .. .. .. .. 1.0 .. .. 1997 2,070 310 1,279 1,279 740 .. .. .. .. .. .. 1.0 .. .. Sector 1996 2,260 .. 1,317 1,317 0 .. .. .. .. .. .. 1.0 .. .. External ants gr $)SU d liberal. US$) %) an and and foreeb mil.ni( (in adetr mil. goods US$) %) goods and t of index mil.ni( ne, deficit deficit (inkc of US$) k/GDPc (in Trade port mil. rto ersf US$) sto sto (in forexfo xe citi A3.2 trans account account (mil. bted bted bt/exed %)ni( %)ni( ind stmente inv def net al al al Index service service/GDP service/exp policy Table Trade Current Current Current FDI, Extern Extern Extern services Debt Debt Debt services EBRD Trade Foreign ngtines ontin prere rve r r r inte tte tte tte o o o lue nte o be o be o be egr egr egr rewol nroveg va m egr thereh egr thereh egr thereh onten onten onten onten hig onten hig onten hig M M M the M the M the and and M ithw5 oflve and and 4, and 4, and 4, source e iab iab iab 1- ler 1- 1- 1- ore iab ore iab ore iab ore Data description Ser Ser Ser Sc lowe Ser Sc Ser Sc Ser Sc Data NBS NBS EBRD NBS EBRD NBS EBRD EBRD EBRD Heritag e 2003 40.6 34.2 .. 9.00 27.14 .. 2.3 2.3 .. Data source IMF FSO (EBRD) Heritag EBRD EBRD EBRD 12)( 2002 46.5 39.3 18.9 9.50 32.21 50 2.3 1.7 4 2003 15,3 .. 5 3.0 2.3 2.0 ) 2002 16,1 45 5 3.0 2.0 2.0 2001 43.3 37.0 14.0 16.43 55.31 (845 1.0 1.0 4 ) 2001 13,6 40 .. 3.0 1.0 1.0 2000 40.4 33.1 16.9 26.34 117.43 (318 1.0 1.0 .. ) 2000 14,2 40 .. 3.0 1.0 1.0 1999 98 45.3 40.2 21.0 26.26 .. (357 1.0 1.0 .. 1999 12 42 .. 3.0 1.0 1.0 )3( 1998 50.6 36.6 16.6 33.74 .. 104 1.0 1.0 .. 1998 11 36 .. 3.0 1.0 1.0 )3( 1997 54.8 40.9 .. 33.74 .. 106 1.0 1.0 .. 1997 .. 35 .. 3.0 1.0 1.0 )3( 1996 .. .. .. 68.2 .. 103 1.0 1.0 .. 1996 .. 34 .. 3.0 1.0 1.0 )3( 1995 .. 32 .. 3.0 1.0 1.0 1995 .. .. .. 90.2 .. 112 1.0 1.0 .. Development (in (in%) st ) (%) Sector rmofre ectorS c.ed c.ed former GDP) ree int owned non-bank GDP ation of Private are small-scale large-scale Form sh Financial in(2 level)l (foreigns banking former ncea intervention of of A3.3 institutions Capital GDP) sector and index ation index ation seripretnefoxedni M1,/ M2,/ of sector % of Index M marketye circulation circulation ua of of Fin bank in in rate mon A3.4 money e ann of index index of Table Gross (% Private Government EBRD privatis EBRD privatis DRBE Table Currency %) Currency Broad Discount Averag rates(at Number EBRD EBRD financial Banking y. tose byn up ht-ig s) ne sse of freas gsinhth s suc y'rt relat see dnacilb ).rruptcoyhlig w -e ral stras ine ha (CPI)xe asnoti neegeht (h0d bus lla onitaerc ry int onomcela Coun rm fo Ind c o ns uprro dnatssyl )eancly an ontilaugre to ly thes ertt )dareps ountCt rkeamla itsn ng of o be o rmo de pe im deiw thas rm atht5eli foin thar egr tio wh,s gree ghl egr xisti (hi ther egr unifd oni egr rgela deeht sirle,po anak on onten 10 onten onten ntemnr onten thiwy ist M Percepnio M gheiheht M (E1erehw ove uptrroc;s o naem1erehw M weapd and 5, 5, rrupt eenwteb and 4, and and ans rkeamla iab ofsnotip pess 1- 1- (G5 sse 1- onomcet Coe neis es iab iab pplieadnadra ore ileh ine iab ore rkea ugrd rmo Data description Ser Th perce bu rang Ser Score Ser Sc forw w bus Ser Sc m as inf Data source EIA ITU ITU ITU ITU ITU WDI EBRD e e Data source Transparency international EBRD Heritag Heritag 2002 .. 22.3 25.7 2.71 15.83 597.01 .. 2 2003 2.3 1.0 5 5 2001 31,710 22.9 18.7 2.34 14.67 561.80 62 2 3 2002 .. 1.0 5 5 58, 2000 30 22.6 12.3 2.26 14.13 375.9 4 .. 2 2001 .. 1.0 .. .. 99 1999 32,036 21.4 5.7 2.07 62 1.7 2000 1.3 1.0 .. .. 1998 36,184 21.8 2.3 1.88 59 1.7 1999 2.0 1.0 .. .. 1997 38,502 20.6 0.8 1.70 .. 1.7 1998 3.0 1.0 .. .. 1996 36,378 19.7 0 1.61 .. 1.3 1997 .. 1.0 .. .. 1995 35,507 19.1 0 1.42 .. 1.3 1996 .. 1.0 .. .. kwh) il. (b rmsofere 1995 .. 1.0 .. .. Governance of xe 1000) roads) Ind ional (per etitionp 10000) 10000) total Infrastructure Generation nes er of infrastructur Quality rnatet com Index (per (p Telephones (%de of Inyc eptionsc of xe Per Ind A3.5 people) Market Line Telepho people) computers host users pav index A3.6 Electricity 100 100 arenp index y Table Net Fixed (per Mobile (per Personal Internet Internet Roads, EBRD Table Trans Corruption EBRD polic Regulation Informal t nt on their rkeamt a,se nd to ingdr ialcna rreuc o in sovooK o ithoutw data sector and tear o ingd and pr M since*( * sector a hips tive ees aver and rogen em1erehw ndasodgoofs ecprindasgeawt thasna rvicse stomlaerasthe fin)h alci GDP o ontenegr ontenegr xedim shops egr ent bydel accode th 1999 healtd Offini egr wi eary M ithoutw M ed inclu iab iab ivate onten loymp Construction/ calculat an sh ache onten and and includeds age nteo M ontrolc on bli Ser Ser pu for M iab iab 5, e .nte GDP ensi ts aredpmoc in and iab 1- (p` en in)d( Ser­ 0022 Ser­ 0022 sovo)o sectorll ner ofsernw unem nda ploy iab pric rnm iab fficial rbia ore ve rbia ares Data description ts tha nds Ser K A Serbia Social cooper ow Serbia O em Serbia Serbia nnualA Ser O Se Sc se 5 c go lytelepmo Data Description Se Serbia Sh fu ematts Gazettes, prices Ser 1995 2000- 1995 2000- of fo 3 e ) book 2003 -2 M Data source Statistical earY S& caltisiattS 003 ofy bookraeY 200 M cialif 998 nistri S& Heritag Data source PRSP Of Gazettes (1 M Finance EBRD PRSP HDR PRSP 2003 947 .. 3 2003 11.7 .. .. .. .. 2002 2,067 1,677 390 18 843 29.5 3 2002 11.7 6.4 .. 14.5 34.4 2001 2,102 1,752 350 16 769 27.5 .. 2001 9.4 6.1 96.5 .. .. 2000 2,097 1,786 311 14 722 26.5 .. 2000 .. 5.6 96.9 36.5 28.3 100 1999 2,153 1,860 293 13 736 26.1 .. 1999 .. 6.5 99.2 .. .. 1998 2,357 2,043 314 13 769 25.1 .. 1998 .. 6.6 69.9 .. .. 1997 2,359 2,055 304 12 750 24.5 .. 1997 .. 8.1 70.6 .. .. 1996 2,229 1,954 276 12 759 25.7 .. 1996 .. .. 71.9 .. .. 1995 2,242 1,989 253 11 716 24.6 .. 1995 .. .. 72.9 28.9 0.25 0) 00ni( 00)0 %( %) (% (inr (in 000)ni( ltato Sector sector secto in dey nsionsep ealthh tiora (%) on on ent ta)piac Labor ntemy cial ivate ctor Index res res lmo % so pr see te,ar Social in (pert (%) unemplo Price enrl x,e cien A3.7 A3.8 ind effi emplo ment,y ment,y of menty Privat menty and expenditu expenditu schoo co Table Total Emplo Emplo (New) emplo Number Unemplo Wage Table Public GDP) Public GDP) Basic Poverty Gini over of level excluding o o to egr egr takenerew available value only gherih due higher onten onten the to entm population ease M M egro sovooK as Montenegro ithw 1, decr total and and and for 1997, Data description Montend lationu mofr iab iab 0-e Ser Ser taaD mofr . spondinger developnam year Serbia Scor cor hu of cal book 2003 M Data source Statisti office WDI Statistical earY S& UNDP (HDR) 2003 .. .. .. .. .. .. .. 2002 8.1 73 .. .. .. .. .. 2001 8.0 .. 12.2 10.6 1.6 13.1 0.763 2000 8.2 72 11.8 11.1 0.7 13.2 0.731 101 1999 8.2* .. 11.7 10.9 0.8 13.6 0.737 1998 10.61 72 12.1 11.3 1.4 13.9 .. 1997 .. .. 12.4 10.6 1.8 14.3 .. 1996 .. .. 13.0 10.6 2.5 15 0.725 1995 .. .. .. .. .. 0.701 Health and ) hst e bire Valu liv rthib population 1000 Index 1000 (mil) Population talto, aty 1000 population)0 (per per entm anc (per 100 A3.9 ect (per increase Develop exp births mortality Table Population Life Live Deaths Natural population) Infant Human PICS2 Sample details COOPERATIVE 10% 2% AGROBUSINESS CONSTRUCTION COMMERCE SOCIAL/ 12% 23% STATE 34% SERVICES 20% PRIVATIZED NEW PRIVATE 15% 49% 35% MANUFACTURING LARGE LESS THAN 50,000 17% CAPITAL POPULATION 25% 24% SMALL 49% LARGE CITY MEDIUM 10% 34% SMALL CITY 41% 102