Report No. 34233-BG Bulgaria The Road to Successful EU IntegrationThe Policy Agenda Country Economic Memorandum November 2005 Poverty Reduction and Economic Management Unit Europe and Central Asia Region Document of the World Bank TFP Total Factor Productivity VAT Value Added Tax TIMSS Trends inMath and Social Sciences VEC Vector Error Correction U L C UnitLabor Cost WDI World Development Indicators USAID United States Agency for International WIIW Vienna Institute for International Development Economic Studies VAR Vector Autoregression Models WTO World Trade Organization Country Director: h a n d K. Seth, ECCUS 3 CONTENTS ACKNOWLEDGMENTS ............................................................................................................. i THE ROAD TO SUCCESSFULEUINTEGRATION: THE POLICY AGENDA ................i EXECUTIVE SUMMARY ........................................................................................................... i VOLUME 1: SUMMARY REPORT ........................................................................................... 1 A GROWTHDYNAMICS 3 B THE CORE REFORMAGENDA 10 C UPGRADING SKILLS OF THE LABOR FORCE ... ........................................................................................................... ......................................................................................... ............................................................ 11 D UPGRADINGTHE TRANSPORT NETWORK E PROMOTINGEFFICENT AND TRANSPARENT PUBLIC SECTOR .. ................................................................. 12 MANAGEMENT, INCLUDING FURTHERPUBLIC SPENDING RESTRUCTURING ............................................................................................................. 13 F.ENHANCING LABORMARKET PERFORMANCEAND EMPLOYMENT ...............17 G IMPROVE FUNCTIONINGOFTHE JUDICIARY AND REDUCE REGULATORY . COMPLEXITY ..................................................................................................................... 19 H MACROECONOMICPOLICIESTO ADDRESS RISKSAND TO SUPPORT . PROGRESSTOWARDSMONETARY INTEGRATION .............................................. 21 VOLUME 2: MAINREPORT ................................................................................................... 24 1 STRENGTHENING . STABILITY, GROWTH, AND INTEGRATION WITH THE ~ ~ 2 4 A Macroeconomic stability and growth: A short history of transition . ........................... 25 A.l Fromtransition to crisis: The legacy o fthe past..................................................... 26 30 B Bulgaria's performanceina regionalperspective . A.2 From crisis to stability and growth ......................................................................... ........................................................ 35 B.l Structuralreforms: Bulgaria andthe CEEC............................................................ 35 B.2 Selected economic and social dimensions: Bulgaria andthe CEEC....................... 36 39 C Fromstability and growth to EUconvergence . B.3 Trade and investmentintegration: Bulgaria andthe CEEC .................................... ............................................................. 40 C.1. The challenge o f convergence................................................................................ 41 C.2. Bulgaria's demographic challenge and the Lisbon agenda.................................... 41 2 MOVING TOWARDS A LISBON-GROWTHPATH . ....................................................... 45 A Growth dynamics ............................................................................................................. ............................................................................................................. 46 B Driversof growth . . 47 B.1Drivers o fgrowth: Aggregate demand components ............................................... 48 B.2 Sectoral investment andproductivity ...................................................................... 49 C Growth accounting: Drivers of growth on the supply side 55 D Potentialoutput growth and convergence .. B.3 Inter-sectoral restructuring ...................................................................................... 53 ..................................................................... .......................................... 57 E Movingtowards a Lisbon-growthpath: A core reform agenda . .................................. 3 BUILDING ONTRADE INTEGRATIONAND INVESTMENT INTEGRATION . .......61 64 A Trade policy 64 B Trade performance .. ...................................................................................................................... 66 C Competitiveness, trade, andFDIdynamics 72 D Building on trade and investment integration: a reformagenda .. .......................................................................................................... ................................................................... ................................ 74 4. RESTRUCTURING PUBLIC EXPENDITURES ............................................................... 77 A Fiscal adjustment 77 B Publicexpenditurepolicy reform .. ............................................................................................................. 81 C Public sector reform 89 D Restructuringpublic expenditures: A reform agenda .. ........................................................................................................ ................................................................................... ................................................. 92 5. ENHANCING LABOR MARKET ADJUSTMENT AND EMPLOYMENT ...................98 A Labor market policies and institutions 98 B Growth. labor market adjustment. and employment 101 C Inter-sectoral labor adjustment D Employment. unemploymentandwages .... .......................................................................... .................................................................................... ................................................. ..................................................................... 103 E Enhancinglabor market adjustment andemployment: A reformagenda . ..............107 109 6. ACCELERATING SYSTEM ............................................................................................................................. THE DEVELOPMENT OF AN EFFICIENT RULES-BASED -112 A 112 B Domesticcompetition: The regulatoryframework . .Institutional developmentand economicperformance .............................................. ............................................................................................................... ..................................................... 116 C Judicial reform 121 D Acceleratingdevelopmentof a rules-basedsystem: A reform agenda . . ..................... 124 7. MACROECONOMIC RISKS AND VULNERABILITIES ............................................. 127 A Addressingrisks andvulnerabilities 127 B Monetaryintegration . . ..................................................................................................... ............................................................................ 130 8 CONCLUSIONS . ................................................................................................................... 135 TechnicalAnnexes ANNEX A: DETERMINANTS OF PRODUCTIVITY: FIRM-LEVEL DATA .................141 ANNEX B: DETERMINANTS OF LONG RUNGROWTH: A GROWTHACCOUNTING EXERCISE .......................................................................................................................... 144 ANNEX C: A MODEL OF BULGARIA'S GROWTH CYCLE ........................................... 147 ANNEX D: CROSS-COUNTRYCONVERGENCEMODEL .............................................. 152 ANNEX E: THE INFORMAL ECONOMY ........................................................................... 158 ANNEX F: STATISTICAL TESTS OF LABOR MARKET FLEXIBILITY ..................... 165 Map 2 ACKNOWLEDGMENTS This Country Economic Memorandum (CEM) was prepared, under the general direction o f Bernard Funck, by a core team comprising Rosalinda Quintanilla (Task Team Leader), Satu Knstiina Kahkonen (executive summary), Juan Carlos Ginarte (macroeconomic stability, labor market policies, and public finances), Stella Ilieva (economic restructuring and investment), Matthew Andrews (public sector reform), Bartelomiej Kaminski and Francis Ng (trade integration and FDI), Simeon Djankov, Caralee Mcliesh, and Gallina Andronova Vincelette (domestic competition), Edgardo Favaro, Mizuho Kida, Stephen Miller (growth dynamics and analysis o f potential output growth), Roberta V. Gatti, Catalin Pauna, Kalin Hristov, and Venelin Boshnakov (cross-country and firm level growth analyses, and analysis o f the informal economy), James Anderson (governance reform) and Irina Kichigina (judiciary reform). Liliana Doudeva prepared the statistical database, Olga Victorovna Vybornaia and Iglika Vassileva provided research support. Amanda Carqani was responsible for processing the document. Albena Samsonova provided research support to the team as well as logistical support to several missions. Ms.Emily Evershed edited the report. The team also drew upon the work o f regional studies o f Europe and Central Asia Region o f the World Bank including, Judicial Systems in Transition Economies (2005, forthcoming), Growth, Poverty, and Inequality in Europe and Central Asia: Past, Present and Future (2005, forthcoming), From Disintegration to Reintegration: Europe and Central Asia in International Trade (2005, forthcoming), European Integration, Regional Policy, and Growth, (2003), and Building Market Institutions in South Eastern Europe (2004), and Expenditure Policies Toward E U Accession, (2002). The team also drew upon the work o f Bulgaria's country team colleagues inECSIE, ECSHD, ECSSD, and ECSPF. The team has benefited from comments and guidance from Kyle Peters and Jaime Jaramillo-Vallejo as World Bank peer reviewers. Ali Mansoor, Harry Broadman, Ardo Hansson, Peter Miovic, and Oscar de Bruyn Kops, provided very useful comments. The team is also grateful to Pradeep Mitra, Cheryl Gray, Anand Seth, Albert Martinez, Myla Taylor Williams, and Oscar de BruynKops for their advice and support. This report was produced in close collaboration with the Bulgarian authorities. The team benefited from the generous amount o f time and support o f senior government officials. The Ministry o f Finance coordinated on behalf o f the Government of Bulgaria. Mr. Lyubomir Datzov, Deputy Minister, Ministry o f Finance, was the Leader o f the C E M Working Group. Discussions and collaboration with the following officials o f the administration mid-2001 to mid-2005 are gratefully acknowledged: Messrs./Mdmes. Iliya Lingorslu, Stamen Tassev, Gergana Beremska, Ana Mihailova and their colleagues at the Ministry o f Finance; Mr.Roumen Borissov and his colleagues at the Agency o f Economic Analysis and Forecasting; Mr. EmilDimitrov and Ms.Mariela Nenova o f the Bulgarian National Bank; Mr.Dimiter Ivanovski, Ministryo f Economy; Ms.Hristina Hristova, Minister o f Labor and Social Policy; Mr.Dimitar Kalchev, Minister o f the State Administration. Discussions and collaboration with officials o f the new administration, which took office in July 2005, are gratefully acknowledged: MessrdMdmes. Plamen Oresharski, Minister o f Finance, Roumen Ovcharov, Minister o f Economy and Energy, Dimiter Ivanovski Deputy Minister, Ministry o f Finance, Emilia Maslarova, Minister o f Labor and Social Policy, Assen Gagauzov, Minister o f Regional Development and Public Works, Nihat Kabil, Minister o f Agriculture and Forestry, Dzhevdet Chakarov, Minister o f Environment and Waters, Nikolay Vassilev, Minister o f State Administration and Reform, Vesselin Blimakov, Minister o f Defense, Kircho Atanassov, Deputy Minister o f Education, Evgenia Koldanova, Deputy Minister, External Affairs; Mdmes. Zinaida Veleva and Petya Vassileva of the Council o f Ministers, officials from line ministries; Messrs. Alexander Hadjiiski, Governor, Todor Todorov, Spass Kostov, and their colleagues o f the National Statistical Institute, and Mr. Pave1 Ezekiev, Bulgarian InvestmentAgency. The team also benefited from consultations with trade unionrepresentatives o fPodkrepa and CITUB, representatives o f business associations-Bulgarian International Business Association and Bulgarian Chamber of Commerce and Industry, and from collaboration with Bulgarian researchers. Mr. Petyo Nikolov, Former Senior Advisor to the Executive Director o f the World Bank (2001-2005) and Ms. Rumyana Kyuchukova, Senior Advisor to the Executive Director o f the World Bank, provided important insights and invaluable assistance inpreparing this report. T h e R o a d to Successful EUIntegration: The Policy Agenda EXECUTIVE SUMMARY 1. Over the last seven years Bulgaria has made impressive progress towards long-term stability and sustained growth. As a result o f sound macroeconomic policies and deep structural reforms, average growth has reached the levels o f eight European Union (EU) New Member States (NMS-8).' Increased growth and low inflation have contributed to increased per capita income and standard of living: per capita income at purchasing power parity (PPP) rose from US$5,502 in 1998 to US$8,260 in2004. The unemployment rate, which has remainedhigh since the early 1990s, started declining in 2003. While part o f the decline was owing to active labor market programs, for the first time since 1996/97 there was net employment creationby the private sector in2004, which contributed to a further reduction inunemployment to 12.7 percent. Stability, growth, and to some extent the extensive social protection systemhave also contributed to a substantial reduction inpoverty, which continues to decline. 2. Despite the overall positive performance, Bulgaria is significantly poorer than NMS-8. The country's per capita income, at PPS in 2003, was 31 and 56 percent o f the average level o f EU-25 and NMS-8 countries, respectively. Given Bulgaria's large income gap with EU25, improving efficiency o f the economy is necessary to ensure that forthcoming EU accession will rapidly narrow the gap andresult insustained improvements inliving standards. 3. T o narrow the income gap and facilitate the convergence with the EU, Bulgaria would need to increase employment and productivity consistent with the Lisbon agenda. These reforms are particularly pressing for Bulgaria, because the country has a negative natural rate o f population growth (about -0.7 percent per year) and out-migration (about -0.1 percent per year). These have resulted in a rapidly declining working age and a growing share o f old people inthe total population. At the same time, the country has the second lowest labor market participation (49.4 percent) and the lowest employment rate (43.6 percent) among Central and Eastern European countries. To catch up with the EU, employment and labor market participation would need to be raised. Also, a large share o f employment i s in unproductive segments o f the economy, or in areas where labor productivity i s largely stagnant. This includes above all agriculture, where productivity has fallen sharply. Without increases in labor market flexibility, there is a risk that the economy will remain locked in a low wage and low productivity growth path. Giventhe demographic trends, Bulgaria could hardly afford that. 4. Raising investment and total factor productivity would be critical for continued growth. The upgrading o fthe capital stock is still at an early stage, and factors ofproduction are underutilized (firms operating at 70 o fproduction capacity). This is a result o f years o f relatively low investment. The average investment to GDP ratio in 1991-2004 was 17 percent inBulgaria compared to the NMS-8 average o f 25 percent. Even though investment has since recovered, partly owing to an improved investment climate, and reached 23.5 percent o f GDP in 2004, the accumulated capital stock gap remains large and needs attention. Also, public and private investment is concentrated in sectors geared to servicing domestic instead o f export markets, ' CzechRepublic,Estonia,Hungary, Latvia, Lithuania, Poland, Slovak Republic, Slovenia. while exports are dominated by unskilledlabor-intensiveproducts. Competition inthe domestic market is, however, limited by an overly complex regulatory environment, and inefficient contract enforcement. While progress has been made inregulatory andjudicial reforms over the last few years, Bulgaria lags behind other countries inthe region inthese areas. 5. As Bulgaria approaches EU accession, it has the opportunity to act on these challenges and set the economy on a higher growth path. This report proposes a reform agenda focused on five areas to narrow the income, productivity, and investment gaps and accelerate convergence with the EU: (i) Improve the quantity and quality of human capital. To promote labor productivity and expand export capacity beyond unskilled labor-intensive products, there i s a need to upgrade the skills o f the labor force. This would call for a reform o f the education system, including universities and vocational schools, to: (a) strengthen the links between the skills acquired in the education system and those needed in the job market through establishment o f links and feedback mechanisms between the education system and labor markets; and (b) improve the efficiency and effectiveness o f public outlays in education by reallocating resources from underused facilities and excess staffing to modernization o f curriculum, textbooks, and teaching materials; implementing a composite per student resource allocation; and linking financing to educational results inaddition to enrollment levels. (ii) Improve the quantity and quality of physical capital. Bulgaria's transport network needs upgrading since maritime ports and close to two-thirds o f roads are in poor condition. Priority in resource allocation would thus need to be given to maintenance rather than new investment. The sequencing o f reforms in this area would need to be based on cost-benefit analysis, and take into account liberalization o f international trucking in the context o f EU integration. Priority reform areas would be as follows: (a) upgrading and addressing bottlenecks along road segments that carry long distance international traffic; and (b) reform o f maritime ports, including introduction o f landlocked port, with the separation o f operational and commercial functions, privatization o f port services, and concessioning o f large container and bulk terminals to private operators. (iii) Promote efficient and transparent public sector management, including further public expenditure restructuring. Bulgaria has implemented a series o f reforms aimed at improving the functioning o f the public sector. A large number o f new laws and regulations have been introduced to achieve this. The next stage o f reform would need to focus on: (a) increasing efficiency o f health and social assistance spending; (b) extending program budgeting to all ministries and linking budget allocation with performance indicators; (c) implementingthe financial management information system; (d) applying merit criteria for hiring, promotion, and salary increases; (e) implementing anti-corruption strategies and enforcing conflict o f interest and asset declaration regulations; and (f) ensuring transparent and competitive procurement for civil works. .. 11 (iv) Improve labor market performance. The current labor framework provides large benefits to those with labor contracts and seniority, leaving behind a large share o f the long-term unemployed with few job opportunities and workers in the informal sector. Key reforms in this areas would include: (a) reducing payroll taxes to decrease disincentives to employment creation; (b) adopting policies for flexible use of fixed-term contracts and working schedules; (c) allowing more flexible terms for hiring and firing owing to fluctuations in production levels, performance, and absenteeism; (d) strengthening the link betweenwages and performance by folding the seniority premium into the basic wage and thereby eliminating it; (e) adjusting unemployment benefits to the average level and duration o f OECD countries to eliminate disincentives to work; and (f) increasing labor force participation by allowing flexible use o ftemporary andpart-time contracts. (v) Improve functioning of the judiciary and reduce regulatory complexity. While the investment climate has improved in the past years, Bulgaria's legislative and regulatory framework remains excessively complex and contract enforcement weak. To promote competition in the domestic market and private sector activity overall, it would be important to streamline the regulatory regimes and implement judicial reform to strengthen enforcement o f contracts. This would include: (a) simplifying contract enforcement procedures; (b) streamlining the licensing regimes; (c) transferring registration out o f courts; (d) eliminating or reducing the level o f minimum capital for private limited companies; (e) simplifying documentation required for new business registration; and (f) introducing statutory response time ("silence i s consent" rules) for business registration and licensing. 6. T o be effective, the above structural reforms would need to be carried out in a stable macroeconomic environment. Hence, the reform agenda needs to be supported by macroeconomic policies that address the risks and vulnerabilities, and make progress towards monetary integration. In the short run, there are risks associated with the pressures from the widening external current account deficit and rapid private sector credit growth. To date, these risks have been partly offset by turning the budget into surplus. Looking forward, however, the fiscal stance needs to remain sufficiently flexible to respond to potential external and internal shocks. ... 111 VOLUME 1: SUMMARY REPORT 1. Over the last seven years Bulgaria has made impressiveprogresstowards long-term stability and sustained growth. As a result o f sound macroeconomic policies and deep structural reforms, average growth reachedthe levels o fNew Member States (NMS-8) at about 5 percent per year in 2000-04. Despite this overall positive performance, Bulgaria i s one of the poorest countries in Central and Eastern Europe. The country's per capita income, at PPS in 2003, was 30 and 57 percent respectively, of the average level of European Union-25 (EU-25) and NMS-8 countries. Given Bulgaria's large income gap with EU-25, and its rapidly aging population, improving and expanding its efficiency gains across the economy i s necessary, ifthe country's accession to the European Union (EU) is to result in sustained and meaningful improvements inits standards o f living. 2. Maintainingmacroeconomic stability and implementingstructural reforms have led to overall improvementsin the standard of living over the last seven years. Following a failed transition which culminated in a severe crisis in 1996-97, macroeconomic stability was reestablished and has been maintained by prudent fiscal policies and strict discipline in incomes policy anchored in the Currency Board Arrangement (CBA) adopted in mid-1997. Fiscal consolidation reduced the overall fiscal deficit from 15.2 percent o f GDP in 1996 to a balanced budget in2003 and generated an overall fiscal surplus o f 1.8 percent in2004. The external and public debt to GDP ratios declined from nearly 100 and 107.4 percent o f GDP, respectively, in 1997 to about 64 and 41 percent, respectively, by 2004. As a result o f these policies, inflation has declined to single digits from hyperinflation levels. 3. A broad structural reform program has contributed to solid economic performance. The severity o f the 1996-97 crisis and the prospect of EU accession changed the political economy in the direction o f reforms. Between 1998 and 2002, most o f the non-infrastructure enterprises and banks were privatized or liquidated, banking supervision was strengthened, trade and prices were liberalized, energy reforms made important progress, and the first steps were taken in regulatory reform to improve the investment climate. In 2003-04, Bulgaria completed the privatization o f the electricity distribution companies and o f telecommunications, and public sector and institutional reforms included the adoption o f new legislation, regulations, and a management system. As a result, growth is led by the private sector which now accounts for 75 percent o f gross domestic product (GDP) with an equal share of total employment. Integration with external markets has expanded with trade in goods and services relative to GDP increasing to 127 percent in 2004 from 94 percent in 1998. Foreign direct investment flows increased to about 37 percent o f total investment in 2003 compared to 25 percent in 1998. Investor confidence improved over time both domestically and externally. Bulgaria's long-term foreign currency debt was upgraded to an investment grade rating inthe summer o f 2004. 4. The resulting low inflation, declining interest rates, and sustained growth have contributed to increases in per capita income and its purchasing power and to overall improvementsin the standardof living. Per capita income at PPPincreased from US$5,216 in 1998 to US$8,007 in 2004 and grew at an average of 6.5 percent per year in real terms. Registeredmonthly average wages and salaries in increased from 183 BGNin 1998 to 294 BGN in 2004-an average o f 2 percent per year in real terms. The unemployment rate, which had remained highsince the early 1990s and which peakedat 18 percent in2000, started to decline in 2003 to 14.3 percent. While part o f the decline in the unemployment rate has been due to the government's active labor market programs, in 2004 there was, for the first time since the 1996- 97 crisis, a net employment creation by the private sector which contributed to a further reduction in unemployment to 12.7 percent. Stability, growth and the country's extensive social protection system contributed to a substantial reduction inpoverty, which continues to decline. 5. The urgent need to implement a core reform agenda centered on productivity, however, more pressing for Bulgaria than for other countries in the region, where the demographic discount is not as large and is not increasing as rapidly. Bulgaria has had a negative natural rate o f population growth (about -0.7 percent per year) which, together with net out-migration (about -0.1 percent per year), has resulted in a negative population growth averaging -0.8 percent per year during the 1990s and early 2000s. As a result, the working age population i s declining rapidly and the aging population is growing rapidly-the demographic discount i s large and is rising. It i s estimated that in 20 years the working age population will have declined by 19 percent and the population over 64 years o f age will have increased by 17 percent relative to the 2005 levels. 6. The Lisbon agenda focuses on increasing employment and productivity: the core agenda proposedinthis reportcovers the reformsneededifBulgariais to move to a Lisbon agenda growth path. In addition to increasing productivity, the Lisbon agenda targets increasing the employment rate to 70 percent by 2010, which implies increasing labor market participation and reducing unemployment rather than discouraging the labor market supply. These objectives cannot be taken lightly by Bulgaria since, in addition to its large income and productivity gap in relation to EU-25 and the demographic discount, its labor market i s perfonning poorly. Bulgaria has the second lowest labor market participation (49.4 percent versus 58 percent in EU-25 in 2003) and the lowest employment rate compared to EU-25 (43.6 percent versus 53 percent in EU-25 in 2003). Simple estimates indicate that if Bulgaria is to meet the Lisbon labor market performance targets by 2010, employment would need to increase by about 8 percent per year in2005-10, and labor market participationwould needto increase by about 4.5 percent per year duringthe same period. Meeting the Lisbon labor market performance targets could be challenging even if these targets were met by 202&employment and labor market participation would need to increase by 3.5 percent and at least 2 percent per year, respectively, over the next 15 years. 7. Convergence implies that Bulgaria needs to increase its productivity and labor marketparticipation. Simpleestimates indicate that inorder to double the per capita income in 20 years so as to reach two-thirds o f the EU-25 per capita level, GDP per worker would need to increase by at least 5.5 percent per year. The trend inthe declining labor force implies that labor productivity needs to increase by about 1 percentage point per year just to compensate for the impact o f the decline in the labor force on per capita income. This increase would need to be larger if labor market participation remained low. Productivity growth cannot compensate for low labor market participation. Even if productivity were to increase, without an increase in labor market participation to 70 percent by 2015, Bulgaria's per capita income would remain below one-third o fthe EU-25 level. 2 8. A broad reform agenda, therefore, lies ahead ifBulgaria's performanceis to improve and the demographic discount is to be compensated by increasing productivity and employment. The large income differences between Bulgaria and the EU-25 and NMS-8 countries reflect gaps inproductivity, inthe physical and human capital stock, inthe functioning o f products and factor markets inthe economy, and, more generally, in the quality o f the policy and institutional frameworks. Bulgaria should build on its achievements to date to accelerate real convergence. An analysis o f the drivers o f growth on the aggregate demand side and the sources o f growth on the factor side outlines the central challenges in growth and competitiveness. 9. From this perspective,five areas are central in defining a core agenda: productivity and growth, trade and investment integration, public expenditure policy, labor market adjustment, and the regulatory and institutionalframework. Chapter 1focuses on stability, growth dynamics, and integration. The analysis o f these three fundamental dimensions over time i s complemented with cross-country comparisons. Chapter 2 examines the foundations o f Bulgaria's growth performance over the last several years in terms o f the drivers o f growth, intra-sectoral restructuring, growth accounting, and potential output growth. Trade policy, trade and FDI dynamics, and recent trends in trade performance are examined in Chapter 3, while Chapter 4 discusses the reforms needed inpublic expenditure policies to support deepening trade integration. Chapter 5 covers labor market performance and labor market policies and institutions. In Chapter 6, selected dimensions o f the institutional framework for growth and competitiveness are examined, and Chapter 7 discusses macroeconomic policies aimed at addressing risks and vulnerabilities and supporting progress in monetary integration. Finally, Chapter 8 presents a summary o f the main conclusions drawn from the report. A. GROWTHDYNAMICS 10. The growth of aggregatedemand components shows a relatively strong bias towards consumptionrelativeto investment. On the aggregate demand side, close to 90 percent o f the average GDP growth in2000-04 was due to total consumption and was driven mainly by private consumption. The growth o f total consumption represented about 4.5 percentage points o f an average growth rate o f 4.9 percent per year during this period. With sustained growth and relatively low inflation resulting in improvements in income levels and purchasing power, the growth o f private consumption has been robust at about 3.5 percentage points o f an average GDP growth rate o f 4.9 percent per year. In contrast, gross domestic investment's contribution to growth averaged o f 2.9 percentage points, despite its rapid growth by 14 percent per year in 2000-04. This is largely due to Bulgaria's relatively low level o f investment o f around 21 percent o f GDP duringthe same period. 11. Bulgaria's investment gap relative to that of other countries in the region is large. The average investment rate to GDP in Bulgaria in 1991-2004 was 17 percent compared to an average o f 25 percent inNMS-8. This reflects a long-standing neglect o f investment inBulgaria despite the rapid increase o f debt up to 1997. Later, the legacy o f highindebtedness limitedboth public and private investment, and it is only recently that investment has started to recover from a trough o f 10 percent o f GDP in 1997. Investment to GDP has been recovering more rapidly since 2003, increasing to 23.5 percent o f GDP in2004. Stabilization and structural reforms have 3 improved the investment climate and as a result the private sector-whose share i s about 73 percent o f total investment-leads in carrying out most of the investment. Similarly, foreign direct investment i s playing a central role in rebuilding Bulgaria's capital stock. An average of about 7 percent o f GDP per year during 2000-04 may be reaching a critical mass needed to improve efficiency and deepen integration with the external markets. Nevertheless, an investment gap o f 8 percentage points o f GDP relative to NMS-8 is indicative that the level o f investment remains an important challenge in Bulgaria. Two aspects are central to addressing this challenge: one i s the mix o f public and private investment, and the other i s the trade orientation and factor intensity o f investment. 12. While private investment has accelerated, it lacks an essential complement in basic public infrastructure,as gross domestic fixed investmentby the public sector has declined sharply. Gross domestic fixed investment by the private sector more than doubled, from 6.4 to 14 percent o f GDP between 1998 and 2002. In contrast, fixed investment by the public sector declined from 6.7 to 4.3 percent o f GDP during the same period. This decline can be explained partly by the legacy o f high indebtness o f the public sector, but for the most part it i s due to growing imbalances between public current and capital expenditures while fiscal consolidation was being implemented. Indeed, fiscal savings from debt reduction were redeployed into social spending, which now accounts for about 37 percent o f the budget. Public capital expenditures and operations and maintainance outlays, on the other hand, bore a heavy brunt in the fiscal adjustment. This trend began to be reversed only in 2003. As a result, private investment lacks an essential complement, owing to deficiencies in basic public infrastructure, particularly in the transport network. 13. Deficienciesin the basicpublic infrastructureessentialto internationaltrade may well be helping to bias investment towards domestic markets relative to external markets. Reaching distant markets i s likely to be more costly, and therefore less profitable, than selling in domestic markets, owing to their proximity to producers in Bulgaria and scope to mark up domestic prices. While higher transport costs o f imports can be transferred to domestic consumers through higher domestic prices, competitive pressures in external markets limit the extent to which Bulgarian exporters can transfer these costs to consumers in foreign markets through higher prices. Moreover, the infrastructure deficiencies, together with the accumulated appreciation o f the exchange rate, are all the more serious, since they are likely to make selling inmore distant markets, such as EUmarkets, more costly compared to producers located closer to EUmarkets or with better infrastructure than producers located inBulgaria. The pattern o f the trade orientation o f investment shows a bias towards servicing domestic markets relative to external markets, although infrastructure deficiencies are not the only factor at play. 14. Investment in sectors oriented toward servicing the domestic market constitutes a disproportionately large share of total investment, one that has expanded faster than outward-orientedinvestmentduring 1998-2003. The analysis o f gross domestic investment in this report distinguishes two categories: (i) outward-oriented investment includes investment in sectors with net export receipts exceeding 30 percent o f total receipts from sales; and (ii) inward- oriented investment includes investment in sectors with net export receipts below 30 percent o f total receipts from sales, and hence these sectors are predominantly oriented to servicing the domestic market, and include nontradables. Investment in sectors oriented to service the 4 domestic market increased by about 9 percentage points o f GDP over 1998-2003, reaching a level of 21 percent of GDP in 2003. In contrast, investments in outward-oriented sectors experienced a modest growth of only 1 percentage point o f GDP during the same period, reaching a level o f 4 percent of GDP in2003. The expansion o f investmentinsectors oriented to servicing the domestic market has been concentrated in wholesale and retail trade, energy, and real estate and business services. The modest increase of investment in outward-oriented sectors took place mostly in the manufacturing o f textiles and textile products and the manufacturing o f basic metals. 15. Similarly,investmentis biasedtowardsnontradablesandresource-intensiveactivities. Investment innontradables and inresource-intensive activities experienced the largest expansion over 1998-2003. Investment in nontradables increased by 6 percentage points, reaching 15 percent o f GDP, and investment in resource-intensive activities increased by 2.2 percentage points, reaching 6 percent o f GDP in 2003. Investment in low-tech-unskilled labor-intensive tradables increased by about 1.5 percentage points o f GDP over the same period, reaching a level of about 3.1 percent o f GDP in 2003. In contrast, the expansion o f investment in medium to high-tech-intensive tradables showed the smallest increase: 0.2 percentage points over 1998- 2003, and the investmentto GDP ratio inthese activities was under 1percent. Therefore, export capacity i s low and i s dominated by unskilledlabor-intensive products. 16. As a result, the economy has yet to move to a more solid path of developing export capacity. As a result of adopting liberal trade policies, trade in goods and services expanded from 95 to 117 percent o f GDP between 1998 and 2003, and reached 127 percent in 2004. At this level, Bulgaria compareswell with the NMS-8 average trade o f 124percent o f GDP in2003. However, this ratio masks Bulgaria's relatively limited export capacity. Bulgaria's imports o f goods were about 50 percent o f GDP which i s not far from the average o f 54 percent o f NMS-8. Nevertheless, the difference interms o f exports of goods to GDP i s twice as large-37 percent o f GDPfor Bulgaria comparedto an average of46 percent for NMS-8. 17. Moreover, the economy hasyet to buildupits export capacitybeyondunskilledlabor- intensive products. Exports are dominated by unskilled labor-intensive products, and final manufactures of consumer goods and automobiles and parts dominate imports. Reflecting the patterns o f investment, exports are dominated by such unskilled labor-intensive activities as textiles, leather, and furniture, which represented about 41 percent of Bulgaria's total exports; another 31 percent consists o f natural resource-intensive products. Capital and skilled labor- intensive exports, incontrast, showed only a small gain of 2 percentage points between 2002 and 2003, reaching a share o f about 29 percent. 18. Imports are dominatedby final manufacturesof consumer goods. On the imports side, final manufactures o f consumer goods, together with imports of automobiles and parts, represented on average 60 percent of total imports in 2000-03 and showed the largest expansion duringthis period. Incontrast, imports o fmachineryand equipment, excluding automobiles, and industrial raw materials represented, on average, 25 percent o f total imports and showed a modest expansion from 24 to 26 percent o f total imports between 2000 and 2003. While more productive activities may be emerging, these advances seem to be at an early stage and have not reached a critical mass that can be observed at aggregate levels. These trends reflect the 5 improved, but still limited ability o f the Bulgarian economy to benefit more fully from trade integration with EUand global markets, and to adjust to changes inmarket conditions. 19. As a result,Bulgarianexportshavebeen less responsive to market opportunitiesthan the exports of other countriesin the region. As the Bulgarian economy was movingtowards a vigorous recovery in 1999, and as Bulgarian exports began a slow recovery in 1999-2000, the world economy entered a downturn in2001. Growth rates inhighincome economies fell sharply from above 3.5 percent in 2000 to less than 1 percent in 2001. This downturn in the world economy and the sluggish recovery since then has had an impact on the export performance o f most o f the CEEC. For example, Hungary and Romania experienced a sizable decline in exports o f goods between 2000 and 2001-1.9 billion euros or about 8.7 percent inthe case o f Hungary, and 1.1 billion euros or 6.1 percent inthe case o f Romania. Hungary's exports began to recover in 2003 and Romania's a year earlier. In contrast, the magnitude of the decline was larger in Bulgaria-about 10 percent-and the recovery has been slower than that in Hungary and Romania. The incentives to producers inBulgaria to turn to the domestic market when external markets weaken seem to be stronger than inHungary or Romania. 20. The Currency Board Arrangement combined with the euro appreciation against the U S dollar compounded a price bias in favor of nontradables which has not been compensated by a more adaptable economy that reallocatesresources from lower to higher productivity activities. Between 2001 and the middle o f 2004, the lev to euro exchange rate appreciated in real terms by nearly 30 percent when nominal rates were deflated by consumer price indices (CPIs). At the same time, the producer price index (PPI) based measure appreciated less than 10 percent duringthe same period. 21. The flip side of the coin is that the productivityof tradables is increasingfaster than that of nontradables. The mechanics o f this effect take place as the productivity of tradables rises faster than that o f nontradables and the relative price o f nontradables increases faster relative to tradables in Bulgaria. This takes place more rapidly than inthe euro zone, where the gap between the productivity o f tradables and nontradables i s lower. Hence, productivity differentials contribute to the more rapid growth o f prices in Bulgaria relative to the euro zone. However, not all o f the accumulated real exchange rate appreciation o f the lev to the euro can be attributed to productivity differentials, in the presence o f significant impediments to resource reallocation from lower to higher productivity activities. 22. N o more than half of the observed appreciation is due to the effect of productivity differentials; the other half is related to the combined effects of impediments to resource allocation. Based on the available estimates o f the possible magnitudes o f the effect of productivity differentials on real exchange rate appreciation, no more than half of the observed appreciation o f the lev can be attributed to faster increases in the productivity of tradables relative to nontradables (called the Balassa-Samuelson effect). The other half i s related to the combined effects o f important impediments to resource reallocation towards more productive activities. Limited labor market adjustment, deficiencies in basic infrastructure, limited competition inthe domestic markets, and the remaining deficiencies in institutional quality have contributed to a price bias in favor o f nontradables. As will be discussed below, implementing reforms to substantially reduce these constraints i s all the more pressing, given that, as in other 6 CEECs, interest rate differentials between Bulgaria and Western European economies have resulted in larger short-tern capital inflows, which also contribute to the recent acceleration o f real appreciation. 23. Indeed, growth is being driven mostly by total factor productivity. The growth accounting analysis for the period 1998-2003 shows that total factor productivity contributed 3.8 percentage points o f the GDP growth o f 4.1 percent per year, with capital accumulation contributing 0.5 percentage points, while the labor force contribution was -0.2 percentage points. This implies that to a large extent, on average, growth in 1998-2003 was driven mostly by the shedding o f excess capacity and the elimination o f inefficiencies, while upgrading o f capital stock i s at an early stage with the contribution o fthe labor force acting as a brake. 24. Nevertheless, large inefficiencies remain. At a first glance it appears that Bulgaria's inter-sectoral restructuring o f the economy is evolving towards a more efficient structure o f production. Namely, the shares o f agriculture and industry in GVA have been declining, while the service economy has become an increasingly important source o f income. In contrast to NMS-8, however, the share o f employment inagriculture inBulgaria increased from 18.5 to 25.5 percent between 1990 and 2004. The decline inthe share o f employment in agriculture inN M S - 8 was about 7 percentage points during the same period. Clearly, productivity in agriculture in Bulgaria has fallen sharply and a large share of employment i s locked into a sector which has low and declining productivity. 25. A key reasonfor this is that adjustment in sectoralemploymentin Bulgariahas been more limitedthan that observed in the New Member States. The departure indexes provide an estimate o f inter-sectoral labor adjustments relative to two comparators: EU-North and EU- South. The departure index o f Bulgaria, estimated at 24 for 1989, improved by about 2 points with the index declining to 22 by 2003. In contrast, the departure indexes o f selected CEEC improved by 7 points, reflecting significantly larger labor adjustments inthese economies during the same period. Furthermore, the expansion o f services has been more pronounced inBulgaria than inthe NMS-8. Between 1990 and 2004, the share o f services inGVA increased from 31to 59 percent while the share o f employment in services increased from 37 to 47 percent. The expansion of the share o f services in GVA and the share o f employment in the NMS-8 was 20 and 13 percentage points each. 26. A closer analysis of sectoral investment shows that the impact of inward-oriented investment on labor productivity is limited but that of outward-oriented investment is large. Furthermore, small increases in investment in the outward sectors have had a higher payoff in terms o f labor productivity. Outward sectors enjoyed a gain in labor productivity (as measured by the GVA per worker) o f 61 percent between 1998 and 2003, or an increase in labor productivity 4 times higher than the increase in labor productivity associated with inward- oriented investment. Under real exchange rate pressure, investors in outward-oriented sectors are compelled to use the factors of production more efficiently and to adapt new technologies in order to stay competitive in the international markets. Competitive pressures in the domestic market do not seem to be as strong. 7 27. A focus on firm level performance shows that large firms are far more productive than small and medium enterprises. The private sector in Bulgaria is characterized by a relatively small number o f large firms, some o f which are foreign owned, which are growing rapidly and generating high levels o f productivity and are working fairly isolated from small and medium enterprises. Small and medium enterprises exhibit large inefficiencies. Statistical analysis also shows that foreign direct investment (FDI) flows and exports are positively associated with higher productivity but the relation i s not statistically significant for the economy as a whole. However, a closer look at the sectoral composition o f FDI flows shows that a large share o f FDIis in services, and only a small share is inthe tradable sectors. Hence, FDI flows in tradables have not reached a sufficient critical mass to have a robust positive impact on productivity. On average in 1998-2003, about 60 percent o f FDI in Bulgaria was located in services, with more than half o f these flows inthe financial sector and only about 30 percent is in tradables-nearly all in manufacturing. Similarly, exports are dominated by unskilled labor- intensive products, which explain the weak statistical relation between exports and productivity indicators at the firm level. 28. Thus, it is not surprising that there is prima facie evidence that growth is below its potentiallevel and the factors of production are underutilized:(i) firms are operating at 70 percent o f production capacity; (ii) market participation is very low; and (iii)large share labor a o f employment remains in largely unproductive segments o f the economy or in activities where productivity is stagnant. 29. Potentialoutput growth was estimated using a growth cycle modelwhich mimics the growthpath of the economy of Bulgaria. Inessence, estimates based on a growth cycle model aim to outline the interplay o f key determinants o f alternative growth paths. Growth performance is determined by the rate o f capital accumulation, by the accumulation o f human capital (which may be achieved by improvements in the quality or skill content o f the labor force), by total factor productivity, and by the policy and institutional environment that supports an efficient resource allocation inthe economy. The rate o f potential output growth is determined by the rate o f expansion o f new investment and the phasing out o f old, outdated capital stock. The rate o f expansion o f new investment is determined by domestic savings, the capacity to attract FDI flows, and the ability to transform old capital into new capital. FDI flows are a function o f the expectations held by foreign investors about possible future profits, which are largely determined by the profitability o f economic opportunities and the predictability o f the institutional framework. The former i s linked to the flexibility o f the goods and factors markets, and the latter is linked to the extent that Bulgaria can move to a predictable rules-based institutional framework. 30. Estimatesbased on the modelmimic very well the growth cycle of Bulgariaduringthe period1990-2003. The estimates o fthe model follow closely the sharp protractedcontraction o f output up until the 1996-97 crisis and the vigorous recovery since then. On the basis o f these estimates, three alternative potential output growth paths were calculated. The slowest growth path aims at reflecting a scenario that would occur if Bulgaria were to lose momentum in implementingreforms and investment were to slow down relative to recent trends. The second higher path aims at reflecting recent performance. And the highest path aims at reflecting a possible path, if a core reform agenda proposed inthis report were to be implemented. It is only 8 under the third paththat Bulgaria could rebuild its capital stock2at a pace at which the effects o f its lowest point inthe 1996-97 cycle could be eliminated inthe mediumterm. 31. The estimates of the growth cycle model point to several important policy implications. First, for additional capital accumulation to deliver high efficiency gains, disincentives to investment in outward-oriented activities need to be substantially reduced. While FDI flows are to a large extent filling the gap between domestic savings and investment' inthe longterm Bulgaria's low domestic savings represent a constraint to reaching the potential output growth. Second, significant improvements in the skills content o f the labor force are needed to compensate for the demographic discount and to increase labor productivity inorder to develop an export capacity beyond unskilled labor-intensive production. As discussed in Chapters 3 and 4, addressing the deficiencies in the transport network for trade facilitation and significant skills mismatch i s central to these two challenges. A third implication is that the economy needs further restructuring to eliminate the remaining large inefficiencies. The incentive structure resulted in shifting a large share o f employment to largely unproductive segments o f the economy, or areas where labor productivity i s largely stagnant. With a demographic discount on growth, Bulgaria cannot afford to have a large share o f employment locked into unproductive activities or activities where labor productivity is stagnant, or where labor market participation is low; nor can it afford to reduce unemployment by discouraging labor supply (see Chapter 5). Without substantial increases in labor market flexibility, the risks are substantial that the economy will remain locked into low wages and a l o w productivity growth path. Finally, a fourth implication is that institutional quality needs to improve if productivity growth i s to increase (see Chapter 6). 32. Policy and institutionalreforms are central to moving to the higher growth path. A simple cross-country panel analysis shows that growth performance could be improved if Bulgaria were to catch up to the extent o f the reforms already achieved by the strong performers among the New Member States. While structural reforms have improved the policy and institutional framework in Bulgaria, they still lag behind the extent o f the reforms in the New Member States. 33. In particular, as indicatedearlier, competitionin the domestic market is still limited by an overly complex regulatory environment, poor contract enforcement and a poorly performing labor market. These conditions create rents that lock resources into sub-optimal uses. Notwithstanding the important progress made in simplifying the regulatory framework in the last few years, the entry and exit o f businesses in Bulgaria i s not easy, fast, or cheap. The number o f procedures involved in opening a business is larger than in selected countries in the region-1 1procedures inBulgaria compared to 6 in Estonia and 4 in Ireland. Bulgaria has the highest capital requirement as a share o f per capita income in Europe-1 17 percent o f per capita income in Bulgaria compared to 50 percent in Estonia, and there are no capital requirements in Romania and Ireland. The time needed to execute insolvency in Bulgaria has been reduced from 3.8 years to 3.3 years, but it is significantly high compared to 2 years in Hungary, 1.1 years in Latvia and 0.4 years in Ireland. The cost o f closing a business is also higher than in other countries inthe region. The procedures for contract enforcement are significantly more complex inBulgaria than in other countries in the region (34 procedures inBulgaria, compared to 25 in 'In this report the termcapital stock refers to the economic rather than a financial definition. 9 Estonia, 21inHungary, and 16 in Ireland). Complexity in enforcing contracts and inefficiencies inthe court system contribute to the fact that businesses inBulgaria rely more heavily on cash transactions than on the more efficient, sophisticated, and contract-intensive forms of payment employed by other countries inthe region. Bulgaria shows low contract intensive money (CIM), or ratio o f non-cash fraction to broad money (M2),over the last several years compared to other countries in the region, including Estonia and Romania, or the median observed in Eastern Europe. B. THE COREREFORMAGENDA 34. The core reform agenda proposed in this report is centered on achieving successful integrationwith the EUand the global markets. Bulgaria should buildon its achievements to date to accelerate real convergence by expanding efficiency gains and competitiveness more broadly across the economy. To this end, Bulgaria needs to continue to act on several fronts, as it has done since 1998, and define a core agenda of reforms to deepen trade integration with the EUand global markets. For Bulgaria to move towards potential output growth andbecomemore competitive, policies and economic restructuring need to focus on achieving higher economic efficiency and productivity supported by a fiscally sustainable and well-targeted social protection system. 35. This reportproposes a core reformagendathat focuses on five areas. These areas are: (i)improving the quantity and quality o f human capital accumulation by focusing on upgrading labor skills; (ii) improving the quantity and quality o f physical capital by upgrading the transport network; (iii) promoting efficient and transparent public sector management including further public expenditure restructuring; (iv) enhancing labor market performance and the creation o f employment by implementing labor market reforms that balance better labor adjustment, the creation o f employment, and the management o f unemployment risks, and that include the rationalization o f the social protection system to enable the reduction o f payroll taxes; and (v) improving functioning o f the judiciary and reducing regulatory complexity in order to improve competition inthe domestic market, and reliance o f economic transactions on contracts and their efficient enforcement. This core agenda needs to be supported by macroeconomic policies that address risks and vulnerabilities and that make progress towards monetary integration. 36. The urgentneedfor reformthat movetowards a Lisbongrowth pathis morepressing for Bulgaria than for other countries whose demographic discount is not as large. First, Bulgaria's economy needs further restructuring to eliminate the remaining large inefficiencies. The incentive structure had resulted in the shifting o f a large share o f employment to largely unproductive segments in agriculture. With a demographic discount on growth, Bulgaria cannot afford to have a large share o f employment locked into unproductive activities, or into low labor market participation, nor can it afford to reduce unemployment by discouraging labor supply. Second, if additional capital accumulation i s to deliver high efficiency gains, disincentives to investment inoutward-oriented activities need to be substantially reduced. A third implication is that significant improvements in human capital accumulation are needed to compensate for the demographic discount and to increase labor productivity to develop export capacity beyond unskilled labor-intensive production. A fourth implication is that institutional quality needs to improve ifproductivity growth i s to increase. 10 C.UPGRADINGSKILLS OF THELABORFORCE 37. The first core area of reform is to upgrade skills of the labor force. Bulgaria needs to implement a major shift in the paradigm o f the education system, including universities and vocational schools, to address the declining size o f cohorts, to generate the skills needed to compete in the global markets, and to improve access by disadvantaged groups. Improving the efficiency and effectiveness o f public expenditures in education i s particularly important, given the objective o f improving labor productivity, which i s needed for raising the wage income and for expanding the export capacity beyond unskilled labor-intensive products. Despite increases in the level of expenditures, the indicators of the quality of education, such as scores in mathematics and science, are below those observed in other CEECs with similar levels o f expenditure on education. Furthermore, there i s evidence that the education system generates significant skills inadequacies which contribute to a very large unemployment rate among recent graduates. Similarly, a survey conducted for this report indicates that foreign investors are experiencing difficulty in finding workers with the skills they need. Businesses report that the skills gap applies to technical specialists, qualified business managers, and other skilled job streams needed for operating modem competitive businesses. As result o f this skills mismatch, firms must incur the highcost o f training new hires, which inthe case o f software companies and other high-tech companies could include three to six months without any productive input from these workers. 38. The direction of reforms in the education system should be driven by three criteria: (i)improving the efficiency and effectiveness ofpublic outlays on education, including the quality o f education and access to education by vulnerable groups; (ii) strengthening the links between the skills acquired inthe education system and those needed in the job market; and (iii) developing an accreditation and certification system. Key reforms recommended include the following: 0 Reallocating resources from a surplus capacity in teaching staff and underusedfacilities to other essential quality-enhancing inputs, including the modernization o f curricula, textbooks, and teaching materials. The reallocation o f resources i s also needed to address the acute deterioration o f education facilities and teaching equipment. 0 Implementing a composite per student resource allocation. An efficient mechanism finances outcomes, in this case students and the quality o f the education they acquire, rather than financing inputs which, in the case o f Bulgaria, are mostly teachers. Per student allocations should be differentiated to reflect differences inthe cost o f education levels and specialization, to improve access to education for disadvantage groups, and to account for cost differences arising from the learning needs o f students. Differentiated per student financing formulas should be designedcarefully to avoid simply reflecting the current per student costs o f different levels and facilities, as this simply perpetuates inefficiencies. Measures that will increase cost recovery inuniversities to 30 percent, as specified by the law, including raising fees, should be implemented. 0 Moving towards a more advanced approach to education by financing educational results and not just enrollment levels. Expanding the role o f private education providers 11 and strengthening competition based on outcomes can be an important force inimproving the quality o f education. In general, countries using this approach condition their payments to private education providers upon agreed education targets in terms o f learning achievements. Consolidating the stafing and facilities of the education system. Per student resource allocation policy needs to be complemented by a plan to consolidate staffing and facilities. Unless financial policy i s supported by a program to consolidate staffing and facilities in the education system, Bulgaria will continue to have schools which are overcrowded in some urban areas while other facilities are being nearly empty. Consolidating facilities will require providing transportation to students from smaller to larger facilities. Addressing the skills mismatch gap. The education reform strategy needs to shift toward adopting automatic mechanisms and feedback rules to link the education system with labor market conditions, rather than relying so heavily on centralized planning mechanisms. First, the provision o f education services should be demand-driven and based onjob performance. Adopting a composite financing per student policy supports a system in which the demands o f students and parents determine which educational programs expand and which contract. Second, stronger competition among education institutions, including universities, for public resources and private sector funding i s needed. This would also help to attract the best teachers and students, and would improve the academic attainment and job performance o f graduates. Third measure i s that both academic attainment andjob performance should be monitored and disclosed at the level o findividual institutions. D.UPGRADINGTHE TRANSPORTNETWORK 39. The second core area of reform focuses on upgrading of the transport network. AddressingBulgaria's large basic infrastructure needs will take time. The conditions are poor in close to two-thirds o f the road network. According to a recent field survey by the World Bank, bottlenecks are emerging along road segments that carry international traffic. Furthermore, port efficiency is low at about 54 percent o f EU-15 levels; this compares poorly with 84 percent in Romania, 90 percent in Estonia, and an average o f 73 percent in the New Member States. The direction o f reforms needs to be drivenby undertaking cost-benefit analyses o f these investments and addressing the challenge arising from the liberalization o f international trucking in the context o f EU integration. Reforms that would upgrade the transport network include the following, e High priority should be given to long distance international traffic. Investments first need to address the emerging bottlenecks along road segments that carry international traffic. These investments also need to address the problem that the road system for international traffic passesthrough many small towns and villages. 12 Inparallel, reforms of maritimeports are urgently needed since these ports, together with a road system for long distance international traffic, are essential to international trade and to deepening integration with the EU. The best international practice for supporting international trade in this area includes the introduction o f the landlord port, with the separation o f operational and commercial functions, the privatization o fport services, and the concessioning o f large container and bulk terminals to private operators. e I n setting priorities in road investment for long distance international traffic, maintenance has the highest priority and new construction has the lowest. Since road investments compete for limited resources, the preservation and rehabilitation o f the existing assets has a higherpriority than new construction. 0 Bulgaria should adopt transparent and competitive procurementfor all civil works. 0 Public-private partnerships should be entered into only if they offer valuefor money as determined by a public sector comparator. E.PROMOTINGEFFICENTAND TRANSPARENT PUBLICSECTOR MANAGEMENT,INCLUDINGFURTHERPUBLIC SPENDING RESTRUCTURING 40. The third core area of reform is to implement further public expenditure restructuring and to improve the efficiency and transparency of public administration. With about 40 percent o f the resources o f the economy under the command o f the public sector, improving the efficiency and effectiveness o f public expenditure policies is central to Bulgaria's growth prospects. In general, inefficiencies in public expenditure are more detrimental to productivity and to growth prospects because they affect efficiency across a wide spectrum o f activities inthe economy. The strategy o f having the public sector focus on the efficient delivery of public goods, including an efficient and effective social safety net, while the private sector leads in the expansion o f investment, output, and employment, has served Bulgaria well. Bulgaria should remain on this path and should accelerate its progress towards this end to continue improving economic performance. Increasing in the efficiency o f public expenditure and improving the efficiency and transparency o f public administration in general, and in two highpriority areas inparticular, areneeded to improve competitiveness. 41. To make room for spendingto upgrade labor skills and the transport network and to maintain it in the face of an aging population, Bulgaria will need to rein in the rapid growth of expenditures in the pension and health care systems and the social assistance system, Public expenditures on the social protection system increased from 8 percent o f GDP in 1996 to close to 14 percent by 2004 and now account for about 35 percent o f the budget. This i s due to a combination o f demographic trends, with a rapidly aging population, high unemployment, and the broad scope o f the social assistance system. However, the efficiency and effectiveness o f the social protection system has significant room for improvement-8 o f every 10 Bulgarians receive at least one social benefit, including pensions, but only 1 in 3 i s poor or 13 old. With one o f the fastest aging populations inthe region, the financial pressures are large and are increasing rapidly. Reforms inthe social protection system are urgently needed to enable the reduction o f the very high payroll taxes, which have created significant incentives for the informal economy and for tax evasion, and limitjob creation. 42. Deepeningsocial security reform. Despite substantial reforms since 1999 when Bulgaria moved to a three pillar system, its social security system i s not aligned with demographic trends. As a result, payroll taxes are very high and provide incentives to the informal economy and to the under-reporting o f income, both o f which exacerbate the demographic trend o f rising dependency ratios. The reforms recommended include the following: Continue to implement reforms to contain financial imbalances. Containing financial imbalances would entail mainly increasing the retirement ages further or reducing benefits. This reform path requires at least two additional measures: (i) implementing strict discipline inthe use o fthe disability pension systemwhich i s currently still prone to fraud and abuse; and (ii) implementingthe strict collection o f contributions. The former needs both strict certification and steep sanctions for fraud and abuse. Making the National Revenue Agency operational i s essential to improving revenue collection. Under this first reform scenario, reductions inthe payroll tax are likely to remain limited to the extent that the contribution base can be expanded. However, risks to the system's financial and fiscal sustainability may increase over time owing to demographic trends, and may be magnified by periods of low economic growth. 43. Given the substantial financial challenges involved, detailed actuarial work is needed to assess the implications and the feasibility o f alternative reformpaths to deepening social security reform. Clearly, the most risky scenario financially, with potentially serious outcomes, is not to consider reform options to address the rapidly growing financial pressures. 44. Reformingthe health care system. Despite increases inpublic expenditures and attempts at rationalization, improvements in the efficiency o f public expenditures on health care and health indicators have been at best modest. Under the current trends, their fiscal sustainability i s questionable. Reforms are needed to address the concurrent problems o f rapid increases in health expenditures, surplus capacity in health care staffing and facilities, deterioration in the quality o f facilities, and inadequate modernization of equipment. A new strategy is needed to address financial sustainability in general, and informal payments and corrupt practices in particular, as well as to improve quality and access. Several reforms are being discussed including: Developing of a basic health care package. The reform envisions universal coverage underthe basic package. Developing of aprivate voluntary health care insurance supplemental to the basic health carepackage. 14 a Funding high tech and expensive treatments. These treatments would be funded by the state in an amount o f additional payment inversely proportional to the value o f the services. a Providing options in the ownership and management of hospitals. The reform envisions providing free choice o f health care providers so that hospitals compete on the basis o f the price and quality o f their services. a Subsidizing the NHIFfor paying the full cost of the medical services of disadvantaged citizens. 45. As in the case o f pensions, careful analysis of reform options is needed in order to assess their implications in terms o f financial sustainability and the provision o f health care services. However, as the experience o f the reforms o f the last several years shows, health care financial policy reform will need to be supported by measures to restructure and consolidate the surplus capacity inheath care staffing and facilities. 46. Reforming the social assistance system. There are large inefficiencies in the social assistance system, as is illustrated by its negligible impact on poverty. The system is highly fragmented, with more than 34 programs. The largest program is that o f the Guaranteed Minimum Income which in the last few years has improved its targeting and effectiveness. However, for the most part, there i s considerable scope for improving the targeting o f the social assistance system. The recommendedreforms include the following: a Consolidating the social assistance system in the direction of in-cash and income support mechanisms. a Training social workers to better identifi poor households. a Improving information systems to facilitate means testing and reduce the payment of duplicative benejh. 0 Expanding communication activities to inform beneficiaries about eligibility criteria and applicationprocedures. 47. Public sector reform: Over the last two years, Bulgaria has implemented important reforms aimed at improving the functioning o f the public sector. A large number o f laws and regulatory reforms aimed at achieving these objectives were implemented in 2002-03, and the first steps in implementation were taken in 2004 and early 2005. These included amendments to the Civil Service Law and the Law on Administration, the development o f sectoral anti- corruption strategies, and the enactment o f the new Public Procurement Act. Similarly, further progress was made in developing the Medium Term Fiscal Framework (MTFF) by strengthening the links betweenbudget allocations and performance indicators and implementing, for the 2005 budget, guidelines for capital expenditure planning and appraisal. The next stage o f reforms should focus on implementing fully the new legislative and regulatory framework. Key reform measures include the following: 15 0 Strengthening public expenditure management: (i) completing the development and implementation o f the MTFF to strengthen the links between budget allocation and performance indicators-program budgeting should be extended to all ministries; (ii) incorporating the expenditure accounts o f the judicial system into the budgetary payment system; and (iii)developing and making fully operational a financial management information system. 0 Improving public administration: (i)implementing fully the merit criteria for hiring, promotion and salary increases provided by the new legislation embodied in the Civil Service Law and the Law on Administration in order to depoliticize the civil service; (ii) completing functional reviews o f all ministries and agencies supported by the central budget, and, based on its findings, implementing a restructuring program for the public sector; and (iii) increasing the scope and number o f administrative services subject to formal competitive processes, outsourcing these services, and generating analyses o f the savings generated. Improving public sector governance and fznancial accountability: (i) simplifying an overly complex legislative and regulatory framework (see Chapter 6); (ii) implementing national and sectoral anti-corruption strategies; (ii) enforcing fully the new regulations relating to conflicts o f interest and asset declarations and the code o f ethics introduced in 2004, and ensure mechanisms for verification o f asset declaration; (iii) implementing transparent and competitive procurement for all civil works including those emerging from extra budgetary funds; and (iv) implementinginternal and external auditing in all government entities, making evaluation reports publicly available, and implementing remedial actions based on the findings o fthe reports byNAO and PIFCA. 48. Substantialrationalizationof publicexpendituresshouldprovidesolid support for the continued improvement of the efficiency of revenue policy and should reduce the tax and social contributionsburden on the economy. The operationalization o f the National Revenue Agency should not be fbrther delayed and the rationalization o f public expenditures, including expenditure cuts, should be implementedmore decisively, if tax policy reform i s to proceed on more solid ground. Inthe last four year, several tax rates have been reduced consistent with the strategy to reduce reliance on direct taxation and expand the role of indirect taxation in tax revenues. Despite the recent reduction inseveral tax rates, deficiencies intax policy, particularly inrevenueadministration, contribute to a sizable, albeit declining, informal economy, and to tax evasion, and also limit employment creation. Reforms needed to operationalize the National Revenue Agency should not be delayed further. Reforms aimed at reducing payroll taxes, supported by labor market reform, are a high priority for enhancing labor adjustment in order to improve and expand efficiency gains more broadly across the economy. 49. To translate the efficiency gains resultingfrom public expenditure policy and public sector reforms into higher productivity levels in the economy, the Bulgarian economy depends on labor marketadjustment. Upgradingthe basic infrastructure and the skills content o f human capital will have a limited impact on improving productivity more broadly across the economy unless the labor market reallocates labor from low to higher productivity activities. 16 F.ENHANCINGLABORMARKETPERFORMANCEAND EMPLOYMENT 50. The fourth core area of reform is to improve labor market performance and the creation of employment. Reform o f labor market regulations to adopt some o f the best practices inthe EUcan enhance creation o f employment and the expansion o f productivity gains more broadly across the economy, in order to benefit more fully from economic integration with the EU. Bulgaria has some o f the most inflexible conditions o f employment, with tight restrictions on working schedules. The terms o f leave andworking hours are more generous than the average for countries inthe region. The combination o f rules for dismissal and generous sick leave benefits are a major disincentive to work, and hinder the creation o f employment. As discussed in the main report, in Bulgaria, high payroll taxes hinder net job creation and contribute to the informal economy. 51. Labor market reforms are urgently needed to better balance the need to enhance labor performance, create employment, and manage unemployment risks. The current framework provides large benefits to those with labor contracts and seniority, leaving behind a large share o f the long-term unemployed with few job opportunities and leaving a larger share o f workers in the informal sector with almost no coverage o f social risks. This is not equitable, efficient, or sustainable. While labor market reforms are essentially a social compact and hence are determined by the country context reflecting social preferences and tolerance, the experiences o f other countries in the region offer useful points o f referen~e.~The reforms recommendedinthis report include the following: Strengthening the links between wages and salaries, andperformance. Wages should be better aligned to performance and labor productivity rather than seniority. The legislation inBulgaria, however, includes the provision of an annual seniority bonus, which, over time, led to the creation o f a second tier component in workers' remuneration-the cumulated seniority premium that often account for a significant share o f total pay for older workers. Many firms discount the seniority premium from the basic salary, and pay a total wage that is the same for workers in the same occupation and with same skills regardless o f their cumulated work experience. But this practice is not possible in labor intensive industries (e.g. textile) where many workers are at the minimumwage, reducing job creation especially among older workers. The seniority premium should be folded into the basic wage in a once-and-for-all change, and the practice and portability o f the seniority premium should be eliminated. Similarly, given the wide variance in productivity by sectors and firms, the links between wages and salaries and performance should be strengthened at the firm level rather than being mandated these at the sectoral or economy-wide level. The 2004 Report of the EUEconomyby the EuropeanCommission shows that the labor market reforms o fthe UnitedKingdom, Denmark, the Netherlands, and Irelandhave been successful interms o f employment and labor market participation. 17 Adopting policies for a flexible use of fixed-term contracts and working schedules. The current Bulgarian labor code includes fairly restrictive rules on working time. In particular, there are precise rules regarding daily hours and weekly holidays and only limited possibilities are offered to firms for adjusting working time during the year. Moreover, overtime work is forbidden by the law, except for certain cases enumerated by the law. Creating greater flexibility in working time is important for firms to face fluctuation in demand. One option would be to follow the experience o f several EU countries with the annualization of working time. Under annualization, employers set normal weekly hours at varying levels over the year, subject to a fixed annual total. Only when these limits are exceeded do overtime premia become payable. The modulation o f hours over the year may be organized in different ways. The full schedule o f hours may be fixed in advance, or there may be a working-time corridor o f minimum and maximum hours o f work, with overtime payments being paid when the average level o f weekly or daily hours, taken over the year, exceeds an agreed limit.4 e Increasing entry and exit from employment based on performance and an economic rationale beyond firm closure and allowing more flexible terms for hiring and firing owing to fluctuations in production levels, performance, and absenteeism. The current rules are highly protective o f employment. As a result, net job creation is limited and firms carry large costs due to idle capacity (see Chapter 2). The benefits o f current rules accrue to those with labor contracts and seniority at the expense o f workers and young graduates seeking jobs, and these rules lead to low levels o f competitiveness. Rules for entry and exit from employment should be linked to economic rationales, including adjustment inproduction levels, performance, and absenteeism. Eliminating disincentives to work. Two areas of reform need to be implemented. Relatively large unemployment benefits may have been justified during the period o f large privatizationprograms. However, now reforms should aim at taking advantage o f a growing economy. Unemployment benefits should be adjusted (currently 60 percent o f average wages) towards a level no higher than the average for OECD countries (the average is 45 percent o f average wages). The duration o f unemployment benefits should also be reduced-the current duration o f 4 months for those with up to 3 years o f service, and 12 months for those with more than 25 years o f service, is too high. The second area o f reform pertains to the rules for dismissal and sick leave benefits. The current rules combined result in a worker's receiving a higher net income while absent from work on sick leave than while working. The number o f sick days should be limited, and the taxes on wage income should be the same whether the wages are earned while working or while on sick leave. Reducing disincentives to employment creation on the demand side. Very high payroll taxes are a major disincentive to employment creation. For type 3 jobs, payroll taxes are 43 percent o f the wage income, o f which 75 percent is paid by the employer. As 4 Annualization may be thought o f as a special case o f "averaging" inwhich the reference period i s a year. Ingeneral, three key elements inthe averaging o fworking hours are the "unit" o f work time employed as the basis o f the averaging procedure, the "reference period" over which the averaging is done, and the "limitations" that apply. 18 discussed in the main report, reducing payroll taxes i s highly desirable not only for net job creation but also to reduce the size o f the informal economy, and to reduce tax evasion. However, several reforms need to precede payroll tax reduction, including a major rationalization o f public expenditures, including a reform o f social spending in particular. Increasing labor participation by bringing a large percentage of the working-age population back into the labor market. Disincentives to work beyond the retirement age should be eliminated by allowing flexible use o f temporary and part-time contracts, and temporary work arrangements. To reincorporate the long-term unemployed, the government should consider relying on effective retraining programs, including on-the- job retraining. G. IMPROVE FUNCTIONINGOF THE JUDICIARY AND REDUCE REGULATORYCOMPLEXITY 52. The fifth core area of reformis to enhancethe policy and institutionalenvironmentin which firms operate, focusingon streamliningthe rules of the game and enforcingthem in a more predictable manner. Inthe last two years, Bulgaria has moved towards implementing its public sector reform program, which spans over the period 2002-07. The program aims to improve public administration, to strengthen public expenditure management, and to reduce corruption. A number o f legal and regulatory reforms in this direction were implemented in 2002-03. And first steps towards implementing these measures were taken in 2004 and early 2005. 53. Simplifying regulatory regimes and reducing the costs of doing business. While regulatory reforms have improved the investmentclimate in Bulgaria, Bulgaria's legislative and regulatory system still remains excessively complex, and this places producers in Bulgaria at a disadvantage relative to their peers in other countries in the region. The reforms recommended inthis report include the following: Simplifiing the procedures for contract enforcement. The procedures for contract enforcement are significantly more complex in Bulgaria than in other countries in the region. The number o f procedures required to enforce a contract in Bulgaria i s 34 compared to 25 inEstonia, 21 inHungary, and 16 inIreland. Completing streamlining of licensing regimes. Important progress was made in streamlining licensing regimes at the national level. The next step is to review and streamline sector-specific regimes and municipal regulations. The system governing construction permits needs to be modernized-businesses report that the current system represents a large burden on the costs o f doing business. The reform should aim at simplifying the regimes by eliminating a multitude o f overlapping regulatory structures, including sector-specific and municipal level regulatory structures. 19 a Making registration an administrative process. The current organization o f Bulgaria's registration system i s high in cost and low in efficiency. Its location in the courts is contrary to good European practices. It also damages economic growth by reducing the capacities o f the commercial courts to deal with legal disputes. A more viable option would be the establishment o f an expert registration agency, as foreseen in the Bulstat law. 0 Eliminating or reducing the level of minimum capital for private limited companies, Minimum capital has beencited ina number o f studies as an outdated legal concept and its problems are further amplified in Bulgaria, where the required amount (5,000 lev) i s the highest in Europe as a share o f average income. The high capital requirement for limited liability companies prevents many would-be entrepreneurs from starting up. Moreover, the ability o f the entrepreneur to withdraw the capital immediately after registration i s completed overrides its designed purpose. As a result, minimum capital requirements serve no useful purpose other than preventing poorer would-be entrepreneurs from doing business. a Reducing the scope and complexity of the required documentation. There i s a need to reduce the scope and complexity o f the documents required to establish new businesses. To this end, several measures can be implemented including the use o f standardized forms, single registration numbers, and notification and self-certification rather than authorization. a SimplifLing bankruptcy procedures. Notwithstanding reforms to date in this area, closing a business inBulgaria i s far more expensive than in other countries inthe region. The number o f years to execute insolvency was reduced from 3.8 to 3.3, and the cost was reduced from 18 to 14 percent o f debt between 2003 and 2004. This places Bulgaria at a disadvantage relative to other countries in the region. The time taken to execute insolvency in Estonia i s 3 years, in Hungary it is 2 years, in Latvia it i s 1.1 years, and in Ireland it i s only 0.4 years. The law needs to be amended and implemented to bring bankruptcy procedures in Bulgaria to a higher level o f efficiency and strengthen exit/entry o f firms inthe economy. a Introducing statutory response times ("silence is consent" rules). As adopted, the Law on Administrative Regulation and Administrative Control on Economic Activities does not provide for the use o f the silent consent rule inbusiness registration and most types o f business licensing. a Strengthening policy making on regulations. Bulgaria adopted the Law on Administrative Regulation and Administrative Control on Economic Activities in 2004. The Law establishes principles for the introduction o f new regulatory regimes. Among other dimensions, the Law requires that proposals for new regulations be based on a clear rationale and to be subject to cost-benefit analysis. 20 54. Bulgaria would also do well to create an independentbody that could champion and lead deregulatory reforms in a more systematic and sustained way. Different options are available for setting up an institutional framework for effective and systematic simplification o f the legislative system and for deregulation. Insome countries this issue is addressed by creating a Better Regulation Unit in the government (as in the UnitedKingdom, Australia and Canada). Inother countries this issue is resolved by establishing an independent agency that screens all proposed legislation (as inDenmark and the Netherlands). 55. Movingtoward a rules-basedsystem. Bulgaria has made substantial progress inthe area o fjudicial reform in the last two years. Although a comprehensive legislative framework for a more efficient judiciary has been put in place, regulations and management systems are still needed to implement the enacted changes, particularly inthe areas o f anti-conuption efforts, the improvement o f court organization, the reduction o f court delays, and the enhancement o f the quality o fjustice. The draft on the development o f performance standards for judges is yet to be completed. The most immediate and urgentreforms include the following: e Fully implementing the Rules for Court Administration passed in October 2004. These rules introduce the concept o f a court administrator, and they should help in bringing professionalism and transparency to court administration, as well as relieving judges and court presidents from time-consuming administrative tasks. e Automating case management-a measure that should reduce the length o f court proceedings and bring more transparency and accountability to the performance o f judges. e Establishing specialized administrative courts, provided by the new Administrative Procedure Code (the draft Code is, however, still awaiting the first reading in the Parliament). e Upgradingjudicial training-for newly appointed magistrates and also for continuous education; this measure should include strengthening the new National Institute for Judiciary established inDecember 2003. e Developing a comprehensive monitoring mechanismfor evaluatingjudicial performance, including the re-design o f the court statistics system and the objective and transparent evaluation o fjudges' performance. H.MACROECONOMIC POLICIESTO ADDRESS RISKSAND TO SUPPORT PROGRESSTOWARDS MONETARY INTEGRATION 56. Macroeconomic stability is a pre-condition if structural reforms are to result in sustained high growth. To be effective, the structural reforms o f the core agenda proposed in this report need a stable environment. Feasibility o f other areas o f reforms proposed here depends on a financially robust fiscal stance, including social spending reform and payroll tax 21 reductions. Over the last three years Bulgaria has made impressiveprogress towards long-term stability. In 2004, Bulgaria's macro-financial position was stronger than ever, owing due to sustained large declines inthe debt to GDP ratio, an overall fiscal surplus o f 1.8 percent o f GDP, and an investment grade rating on the country's foreign debt. However, risks and vulnerabilities remain and new challenges to macroeconomic policy emerge as Bulgaria moves towards monetary integration. 57. The fiscal consolidation strategy of the last several years targeted gradual adjustment in the overall fiscal balance. The strategy has aimed at reaching a balanced budget and reducing external debt to meet the Maastricht Treaty target o f a public debt to GDP ratio of under 60 percent by 2005. The gradual adjustment in the overall fiscal balance has been accompanied by reductions in tax rates and significant reductions in public debt. Bulgaria reached the Maastricht Treaty target in 2002 and achieved a balanced budget in 2003. As a result, Bulgaria's macroeconomic financial position has strengthened. Nevertheless, there are risks and vulnerabilities to sustaining the macroeconomic stabilization and reform achievements to date. 58. Short-term and medium-term risks weigh heavily on the fiscal stance. Inthe short run, there are risks associated with the pressures towards the widening o f the external current account deficit and the rapid growth o f credit to the private sector. To date, these have been partly offset by turning the budget into surplus. Looking forward, however, the fiscal stance needs to remain sufficiently flexible to respond to external shocks. 59. Indeed, macroeconomic management could be further complicated if there were greater capital flow volatility. The risk associated with such volatility is magnified by the economy's growing reliance on short-term capital inflows. Short-term capital inflows more than doubled from 649 million to 1,633 million euros between 2003 and 2004. Their share in the financing o f the external current account deficit increased from 27 to 55 percent during the same period. Much o f these inflows represent money attracted to Bulgaria by the relatively high interest rates inrelation to the returns on comparable securitieshvestments in the euro zone. A sharp reduction inthese rates and/or interest rates as in other parts o f the global economy would reduce the differentials, which could produce a rapid capital outflow and could result in a considerable shock to the economy. 60. The rapid expansion of private debt magnifies the potential risks. The foreign indebtedness o f private firms and households more than quadrupled between 1999 and 2004. Similarly, with improvements in the investment climate, and in anticipation o f EU accession, private firms are undertaking investments at a rapid pace. To the extent that these firms have turned to the financial markets to finance productive investments that would contribute to productivity growth and the ability to repay obligations (both domestic and foreign) there would be limited reason to worry. However, while more than one-third o f the foreign debt growth is FDI-related, most o f it i s owed to domestic banks. The capacity o f lenders to correctly assess the risk associated with certain loans, particularly consumer loans, is untested. While statistics on non-performing loans currently indicate that bank portfolios remain solid, there are often time lags before bad credits manifest themselves. It i s important, therefore, to closely monitor the 22 effectiveness o f recent measures to reduce the growth o f credit to the private sector and to ensure that banking supervision continues monitoring closely for signs o f inordinate lendingbehavior. 61. The gradualprogresstowards monetary integrationraises additionalmacroeconomic challenges. Recently, Bulgaria took the initiative to announce its intention to join the exchange rate mechanism (ERM-2) as soon as possible after accession and to adopt the euro in the minimumterm providedfor by EUlegislation-generally interpreted to meantwo years. During this time, Bulgaria will needto maintain the lev within the fluctuation bands provided by ERM- 2. This it will need to do without employing capital or exchange rate controls, Bulgaria will thus have to remove controls on credit operations, real estate transactions, personal capital movements, money market instruments and other capital transactions. Short o f these instruments, the authorities will needto pilot the economy towards full adoption o f the euro and to maintain it within strict parameters: a fiscal deficit to 3 percent of GDP, a public debt ratio o f less than 60 percent o f GDP, low inflation, and interest rates close to the EUaverage. 62. The task will be complicated in view of the large structural differences between Bulgaria and countries in the euro zone. Structural differences, both real and financial, between Bulgaria and other euro zone countries could mean that macroeconomic policies aimed at promoting stability inthe euro zone as a whole neednot correspond to the short-run economic realities inBulgaria. 63. Indeed, response of Bulgaria's economy to shocks differs substantially from those in the Eurozone. The statistical analysis conducted for this report shows that these asymmetries are declining but the gap remains large. Estimates show that asymmetric response i s larger in demand and supply responsiveness and the degree o f co-movement varies considerably across sectors and across major trading partners. 64. Going forward, a cohesive structural reform agenda supported by disciplined macroeconomic policies is central if Bulgaria is to translate EU accession into sustained improvement in standards of living for its population. Bulgaria has gained considerable expertise in recent years in managing its economy under a hard peg. Best o f all, it has demonstrated a willingness to subject its fiscal policy to the exigencies o f the currency board and a flexibility to do so that will hold in good stead as it integrates in European monetary arrangements. Still, these challenges underscore the centrality o f implementing the structural reforms outlined in this report to get the Bulgarian economy in sync with the rest o f the currency area and able to withstand common external shocks. Reforms to substantially enhance labor market adjustment, to improve competition in the domestic markets, to enhance the dependability o f contracts and to upgrade skills and transport network, will help achieve that, while accelerating real convergence. These are central for Bulgaria to translate EU accession into sustained improvement instandards o f living for its population. 23 VOLUME 2: MAINREPORT 1. STRENGTHENINGSTABILITY, GROWTH,AND INTEGRATION WITH THE EU 1.1 Over the last seven years (1998-2004), Bulgaria has taken impressive strides in the direction of long-run stability and growth, based on its progress towards EU accession. During 1998-2004, Bulgaria implemented a broad reform program, including economic restructuring and institutional reforms. The economic mismanagement o f the first nine years o f transition (1989-97) culminated in a severe economic and financial crisis in 1996-97. This crisis, together with the prospect o f EUmembership, changed the political economy inBulgaria infavor o freforms. 1.2 Maintaining macroeconomic stability and implementing structural reforms have been conducive to solid economic performanceas well as to progresstoward EU accession. Macroeconomic stability was achieved, and has been maintained by prudent fiscal policies and strict discipline in incomes policy anchored inthe Currency Board Arrangement (CBA) adopted in mid-1997, and by 1999 the economy was largely stabilized and growth had been re- established. Bulgaria was invited to start negotiations for EU membership in December 1999. Between 1998 and 2002, most o f the non-infrastructure enterprises and banks were privatized, banking restructuring was completed and banking supervision strengthened, trade and prices were liberalized, energy reforms had made important progress, and the first steps had been taken in regulatory reform to improve the investment climate. In its Regular Report of 2002, the European Commission concluded that Bulgaria had a functioning market economy. In 2003-04 Bulgaria implemented major public sector and institutional reforms. The former included measures to improve public administration, public expenditure management, andpublic financial accountability, and the latter focused on judiciary reforms and anti-corruption measures. By June 2004, Bulgaria had provisionally closed all 31 chapters o f the acquis communautaire, and inApril2005, Bulgariasigned the Accession Treaty. 1.3 The challenge for the future is to implement a reform agenda aimed at successful integrationwith the EUand the globalmarkets. A broadreform agenda must lie ahead ifEU accession i s to be translated into meaningful and sustained benefits in the standard o f living of Bulgarians. Bulgaria's progress to date increased its per capita income, at PPP, from US$5,508 in 1998 to US$8,260 in200Lrepresenting 30 and 57 percent respectively, o fthe average level for the European Union-25 (EU-25) and NMS-8. This large income difference reflects gaps in productivity, in the physical and human capital stock, in the functioning o f products and factor markets in the economy, and, more generally, in the quality o f the policy and institutional frameworks. Over the last seven years, Bulgaria has implemented a broad range o f reforms and should buildon its achievements to close the gaps in all these dimensions and, inthis way, catch up with the New Member States (NMS). Clearly, Bulgaria needs to continue to act on several fronts as it has done since 1998. 1.4 Strengthening stability, growth, and integration with the EU and global markets while preserving social cohesion is the cornerstone of identifying a core agenda for the future. Reform priorities are, to a large extent, determinedby the context o f Bulgaria's unique features and constraints, by its achievements to date and its development goals as it approaches EUaccession andbeyond. Bulgaria's uniquefeatures include its demographic characteristics, its legacy o f run-down basic public assets, particularly public infrastructure, the poor performance o f its labor market, and its need to continue to strengthen its institutional framework. Successful integration with the EU implies continuing progress in improving the standards o f living in Bulgaria and reaching convergence with the EUsingle market. 1.5 From this perspective, five areas are central in defining a core agenda: productivity and growth, trade and investment integration, public expenditure policy, labor market adjustment, and the regulatory and institutional framework. Chapter 1 focuses on stability, growth dynamics, and integration. The analysis o f these three fundamental dimensions over time i s complemented with cross-country comparisons. Chapter 2 examines the foundations o f Bulgaria's growth performance over the last several years in terms o f the drivers o f growth, intra-sectoral restructuring, growth accounting, and potential output growth. Trade policy, trade and FDI dynamics, and recent trends in trade perfonnance are examined in Chapter 3, while Chapter 4 discusses the reforms needed inpublic expenditure policies to support deepening trade integration. Chapter 5 covers labor market performance and labor market policies and institutions. In Chapter 6, selected dimensions o f the institutional framework for growth and competitiveness are examined, and Chapter 7 discusses macroeconomic policies aimed at addressing risks and vulnerabilities and supporting progress in monetary integration. Finally, Chapter 8 presents a summary o f the main conclusions drawn from the report. 1.6 Bulgaria moved from being one of the poorest performers in the region during 1989-97, to the position of a solid performer in 1998-2004. The former period was marked by output decline, massive borrowing, large fiscal deficits and very highlevels o f inflation, political instability and civil strife. The later period was characterized by fiscal discipline, the implementation o f major structural reforms, including the privatization and liquidation of unviable enterprises and banks, low inflation, sustained output expansion, political stability and an absence o f civil strife. Bulgaria should build on its achievements to date and accelerate its pace toward matching the quality o f its institutions and its economic performance with those of the stronger performers among the New Member States. Moving from stability and growth to convergence entails implementing a core reform agenda centered on productivity, in order to shift to a Lisbon-growth path. This is central to Bulgaria, since large increases in productivity are needed to compensate for its demographic discount, with a declining labor force and a rapidly aging population, and to close the income gap with the EU-25. This chapter first provides a short summary o f Bulgaria's transition. It then discusses Bulgaria's economic performance in a regional perspective. The last section examines the challenges o f moving from stability and growth to accelerating EUconvergence. A. Macroeconomic stability and growth: A short history of transition 1.7 Bulgaria's transition is characterized by a marked shift to an economy exhibiting solid economic performance over the last seven years (1998-2004), from being one of the 25 poorest performers in the region during the first nine years of transition. The implementation of sound macroeconomic policies and deep structural reforms over the 1998- 2004, with political stability and without civil strife, has resulted in a solid economic performance. In contrast, during the first nine years o f transition, Bulgaria was one o f the poorest performers in the region and experienced increasing social tension and political instability. A comprehensive analysis o f Bulgaria's transition i s beyond the scope o f this report. This section is limited to outlining selected features o f Bulgaria's transformation aimed at defining priorities for the central links between stability and growth while maintaining social cohesion. A.l Fromtransition to crisis: The legacy of the past 1.8 In contrast to the New Member States (NMS-8): Bulgariabeganits reforms after a severe crisis in 1996-97 with a legacy of high indebtness,over-industrialization,and a low stock of human capital. At the start o f political transition in 1989, most dimensions o f Bulgaria's initial conditions were within a range between the Baltic countries and the NMS-8 (see Table 1.1). Interms o f the share o f Council for Mutual Economic Assistance (CMEA) trade in GDP, Bulgaria was in a better position, with a lower share of 15 percent o f GDP compared with the Baltic countries with 31 percent and the NMS-8 with 21 percent. However, Bulgaria was worse o f f relative to the Baltic countries and the NMS-8 in terms o f the legacy of over- industrialization, low stock o f human capital, and high indebtness. The index o f over- industrialization is indicative o f the fact that Bulgaria inherited larger economic distortions than the Baltic countries and the NMS-8. Similarly, Bulgaria's secondary school enrollment o f 75 percent is indicative of a significantly lower stock o f human capital compared to 94 percent in the Baltic countries and 90 percent in the NMS-8. Furthermore, by 1990 Bulgaria had reached an unsustainable level o f debt to GDP ratio o f 57 percent compared to 1 percent in the Baltic countries and 27 percent in the NMS-8. In April 1990, Bulgaria declared a moratorium on its external debt. 1.9 By the early 1980s, the growth potential of Bulgaria's centrally planned economy had been exhausted and additional time was bought with rapid and unsustainable debt accumulation. The centrally planned economy based on industrialization and high investment followed for more than three decades had exhausted its growth potential by the early 1980s. In the late 1950s, Bulgaria implemented a massive transformation program based on large-scale investment for the relocation o f industrialization and labor from agriculture to industry.6 Between the mid-1960s to the mid-l970s, the economy grew at 5 percent and industry grew at 7 percent. Inthis reportthe NewMember States (NMS-8) include Estonia, Latvia, Lithuania (or the Baltic countries), the Czech Republic, Hungary, Poland, the Slovak Republic, and Slovenia. Employment inagriculture fell from 55 percent in 1960to under 25 percent by 1980 while industrial employment increased from 17 to 27 percent during the same period. Estimates o f fixed investment show an increase o f more than 400 percent inreal terms during this period. 26 Table 1.1 Indicatorsof Initial Conditions:Bulgariaand Selected New Member States a/ Bulgaria Baltics bi NMS-8 bi PPP adjusted GDP per capita, 1989 (US dollars) 5,874 5,094 6,7 19 External debt to GDP in 1990 (percent) 57 1 27 External debt to GDP in 1997 (percent) 100 46 44 Share o f CMEA trade in 1990 GDP (percent) 15 3 1 21 Share o f agriculture inGDP in 1990 (percent) 18 22 13 Index o f over-industrialization `' 0.23 0.10 0.11 Natural resource endowment (O=poor, 1=moderate, 2=rich) 0 0 0.25 Secondary school enrollment inpre-transition year (gross; percent) 75 94 90 Average growth, 1985-89 (percent) 2.7 3.0 2.0 Repressed inflation, 1987-90 18 26 10 Distance o f Capital from Brussels (km) 1698 1508 1177 Years under communism 43 51 47 Notes: ai Selected indicators of initial conditions are taken from the three papers cited inthe Source. Most indicators are usedin at least two of the three papers; b/ Simple averages; c/ Defined as the difference between the actual and predicted share of industry, where the latter i s derived from regressions estimatedby Syrquin and Chenery (1986); di Calculated as the difference betweenthe growth o f real wages and real GDP growth over 1987-90. Sources; World Bank staff calculations; and Fischer and Sahay (ZOOO), Berg et al. (1998),de Melo et al. (1997). 1.10 Duringthis period,however, total employmentgrowth averagedless than 1percent per year with economic growth beingdriven mainly by large investments and the sectoral reallocation of labor. As labor productivity and growth stagnated, trade integration with the Soviet economy-the mainmarket of Bulgaria's exports and a source of cheap energy imports- deepened and the energy intensity o f the economy increased. By the early 1980s, the sustaining of growth through capital accumulation and labor reallocation across sectors was exhausted and economic distortions in the economy widened-growth in 1975-85 was less than 1 percent per year and labor productivity stagnated. Unable to change course to improve productivity, the country bought additional time by rapid debt accumulation inconvertiblecurrencies: the increase inexternal debt between 1985 and 1989tripled to US$9.2 billion (or 57percent ofGDP). 1.11 The economic mismanagement of the first nine years culminated in a severe economic and financial crisis in 1996-97 whose legacy burdens Bulgaria's economy even today (see Table 1.2). The initial transformational recession was prolonged and deepened by expansionary budgetary and quasi-fiscal deficits sustained by the rapid expansion of debt accumulation and monetary aggregates. In contrast to New Member States, which had relied heavily on foreign direct investment (FDI) flows to implement economic restructuring since the start of transition, FDI flows to Bulgaria were negligible averaging around 1 percent o f GDP during 1989-97. Instead, Bulgaria had relied heavily on the rapid expansion o f debt accumulation since the mid-1980s, andexpansion of monetary aggregates in 1993-95.' By 1997, external debt and public debt to GDP were close to 100 and 105 percent, respectively. 'Unsustainable macroeconomic imbalances exacerbated the costs o f the external shocks o f the Gulf and Yugoslav crises during 1989-97. 27 Table 1.2 Bulgaria:Selected EconomicIndicators,1989-2004 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 GNI per capita (Altas method, USS) 2850 2260 1620 1430 1250 1250 I370 1210 1200 1270 1450 1590 1670 1790 2310 2740 GDPper capita, PPPcurrent international IF 5623 5700 5590 4896 4981 5216 5566 5202 4990 5216 5399 5990 6483 6906 7501 8007 (Percent) Povertyrate .. 5.5 .. 36 .. 12.8 Unemployment rate, annual average n.a. 1.7d l l . l a / 15.2 15.8 14.1 11.4 11.1 14.0 12.1 13.8 18.1 18.1 17.7 14.3 12.7 RealGDPgrowth -3.5 -9.1 -8.4 -7.3 -1.5 1.8 2.9 -9.4 -5.6 4.0 2.3 5.4 4.1 4.9 4.5 5.6 CPI inflation(p.a.) n.a. 64.0 4190 91.3 72.8 96.0 62.1 121.61058.4 18.7 2.6 10.3 7.4 5.8 2.3 6.I Lendinginterestrate, nominal. e.0.p d! 64.1 83.6 117.1 51.1 481.3 13.8 13.4 12.4 12.2 13.2 10.2 9.5 8.7 Deposit interestrate, nominal, e.0.p. el 45.2 53.6 72.4 25.3 211.9 3.0 3.3 3,3 3.3 3.4 3.2 3.2 3.0 (Percent ofGDP) Total revenues andgrants b/ 64.5 58.0 44.8 43.2 39.4 43 7 39.9 36 6 35.6 37.7 38.5 38.5 38.6 38.4 40.0 41.2 Total expenditure andnet lendingb/ 65.4 62.5 47.9 48.9 50.7 48.8 45.0 51.8 35.9 36.4 38.3 39.1 38.4 39.0 40.0 39.5 Interestpaymentb/ 3.7 5.3 6.5 6.9 9.8 14.6 14.6 20.1 8 3 4 3 3.8 4.0 3.7 2.2 2.1 1.8 Primqbalanceb/ 2.8 0.8 3.4 1.2 -1.5 9.5 9.5 4.9 8.0 5.6 4.0 3.4 3.9 1.6 2.1 3.5 Overallfiscalbalanceb/ -0.9 -4.5 -3.1 -5.7 -11.3 -5.1 -5.1 -15.2 -0.3 1.3 0.2 -0.6 0.2 -0.6 0.0 1.7 Public andpublicly-gtd.debt b/ c/ .. 181.0 146.0 146.0 181.0 113.0 302.0 105.0 83.0 86.8 77.1 69.9 56.3 48.4 40.9 Gross DomesticInvestment 33.1 30.4 22.6 19.9 15.3 9.4 15.7 8.1 9.9 16.9 17.9 18.3 20.7 19.8 21.7 23 5 Gross Domestic Savings 31.4 22.0 26.9 14.1 7.7 8.8 14.1 13.5 14.5 17.1 12.1 12.9 13.1 13.2 12.0 13.2 Externalcurrentaccountbalance c/ -3.7 .. -1.0 -4.2 -10.1 -0.3 -1.5 1.7 10.1 -0.5 -5.0 -5.6 -7.2 -5.3 -9.3 -7.5 ForeignDirect Investment c/ ma. 0.0 0.7 0.5 0.4 1.1 0.8 1.4 4.9 4.2 6.2 7.9 5.9 5.6 10.4 8.4 Workers'remittances c/ n.a. n.a. 1.4 1.3 2.6 3.6 1.8 2.0 1.4 1.5 1.9 2.2 3.1 3.2 3.5 4.2 External debt c/ 46.6 52 8 154.0 116.0 114.5 101.7 81.2 106.2 99.9 83.2 89.2 86.9 78.6 65.1 60.7 63.0 Official reserves(mos.od Imports, G&S) .. 0.9 2.0 1.3 2.3 2.2 0.9 4.5 5.6 6.3 5.3 5.2 5.6 5.7 6.1 Domesticcredit .. 323.3 126.5 132.7 103.4 68.8 108.7 20.8 15.6 15.3 17.8 20.2 23.7 29.8 36.2 Broadmoney ., 73.4 74.6 76.9 77.8 66.8 74.8 34.1 29.6 31.7 36.8 41.7 42.9 48. I 53.7 (AnnualPercentageChange) Realcredit growth .. -58.4 -16.0 50.4 36.9 16.8 -23.2 -70.7 -5.0 -2.8 17.8 20.2 22.7 26.7 29.0 O.W.to the non-government .. -68.8 -23.5 69.0 31.6 31.8 -23.3 -77.4 43.2 13.5 5.1 26.1 38.7 40.4 42.9 Merchandiseexportgrowth, real 8.7 -30.9 11.8 -46.6 -5.4 2.4 15.0 -5.2 8.9 -8.6 -3.7 20.6 11.5 11.6 11.8 12.0 Merchandiseimportgrowth, real -1.1 3.0 63.0 -82.8 11.1 -15.7 15.6 -11.4 2.7 11.1 11.0 19.6 15.3 6.9 18.0 11.9 Soureex NSI, MoF, BNB. EmploymentAgency, WorldBank and IMFdatabases; Poverty Assessment (1995, 1997 and 2001) ai Endofperiod. b/ Dataaccordingto the GFS Mannual2001with the exceptionof 1989(GFSM86) c/ Since 1999externalcurrentaccount balance, externaldebt andpublic andpubliclyguaranteeddebt are calculated form Eurovalues. d/ Refers to interestrateon short-termcredits in Leva. e/ Refersto interestrate on time deposits m Leva. 1.12 Despitethis massive injectionof financialresourcesinto the economy, a rapidfall in output and investment ensued. The cumulative fall in real GDP in 1989-97 was above 45 percent. Gross domestic investment fell by 23 percentage points o f GDP to under 10percent by 1997. In addition to the rapid debt accumulation, the growth o f credit in real terms was 35 percent per year during 1993-95. As a result, inflation accelerated to around 122 percent in 1996 28 and reached over 1,000 percent per year in 1997. * The large macroeconomic imbalances and the debt overhang led to a crippling run on the banks as the economy approached hyperinflation. With the loss o f confidence in the banking sector, real credit growth turned sharply negative, maturities shortened, and interest rates soared. The sharp increase in interest rates was fiscally unsustainable and did not halt the collapse in money demand and capital flight. The banking sector collapsed, and by the end of the first quarter of 1997 the international reserves o f the Bulgarian National Bank (BNB) were virtually depleted at less than the equivalent needed to cover one monthof imports. 1.13 The costs borne by the Bulgarians were large, contributing to increases in social tensions. Per capita income declined by 12 percent, while the hyperinflation period eroded the purchasing power o f wage incomes and wiped out the savings of a significant proportion o f the population. The unemploymentrate remained highat an average o f 14percent during the 1990s. And poverty increased substantially, reaching its highest level in 1997. During this period Bulgaria experienced political instability and increasing social tensions, which culminatedinthe fall of the government inthe beginning o f 1997. 1.14 The first years of transition contributedto the depletion of Bulgaria's capital stock andthe loss of skills of its stock of humancapital. Inaddition to the sharp deterioration on the macro-financial side, duringthis period Bulgaria's capital stock was being depleted, as the value of the assets and profits o f state-owned enterprises and banks were being stripped by vested interest groups and widespread corruption practices. By the time Bulgaria started its privatization program in 1998, the value of the assets of many state-owned enterprises andbanks had been substantially reduced. Furthermore, for the most part, public investment had neglected investment needs in basic public infrastructure, and assets in the education and health care systems had deteriorated. Similarly, the prolonged period o f poor economic performance had contributed to the loss of skills o f Bulgaria's human capital as a large share o f its labor force had become long-term unemployed, discouraged workers, and labor market participation had declined. The unemployment rate had reached a high level o f 15.8 percent by 1993 and remained highuntil2003. 1.15 The legacy of indebtnesshas weighed heavily on Bulgaria's economic restructuring by limiting both public and private investment. During the first nine years o f transition, Bulgaria's investment level to GDP, at around 14 percent, was one o f the lowest among Central and Eastern European countries, where investment to GDP averaged about 27 percent. Cross- country research indicates that in a highly indebted country, as in the case o f Bulgaria in 1997, the cost interms of growth has been estimated at an average 2 percentage points per year. These estimates seem to be consistent with Bulgaria's experience compared with the Baltic countries and the NMS-8. The effects and transmission mechanisms o f high indebtness and growth are complex andbeyond the scope o f this report. However, a simple comparison, while incomplete, illustrates that the costs o f Bulgaria's indebtness have beenhigh. 1.16 Average annual growth of per capita income at PPP in 1991-2001 was around 1 percent in Bulgariacomparedwith 3 percent in Estonia and 2.7 percentin the NMS-8. At the start o fthe transition, Bulgaria's per capita income at PPP was US$5,874 or about 15 percent 8 During the first part of 1997 the average inflation rate was 240 percent. 29 higher than the average in Baltic countries and 14 percent lower than inNMS-8 (see Table 1.1). However, at the start o f the transition (1989) the average debt to GDP was about 1 percent for the Baltic countries and 27 percent for the NMS-8, compared to an unsustainable level o f 57 percent in Bulgaria. During the transition, the Baltic countries and the NMS-8 were able to leverage their debt headroomwhich together with larger FDIflows resulted inhigher investment levels over a long period o f time. Higher investment levels driven mostly by the private sector, particularly foreign investors, under an improved policy environment have contributed to increases inproductivity and growth. In contrast, the legacy o f high indebtness in Bulgaria has limited both public and private investment levels in Bulgaria. By 2000, the average per capita incomes at PPP for the Baltic countries and that for NMS-8 were 26 and 67 percent higher, respectively, than inBulgaria. A.2 Fromcrisis to stability and growth 1.17 The combination of the severe 1996-97 crisis and the prospect for EU membership changed the political economy in Bulgaria in favor of reforms. Since 1997, Bulgaria has been implementing a comprehensive stabilization and structural reform program anchored in its process for EU accession. Macroeconomic discipline has been supportive o f providing momentum for structural reforms by enforcing budgetary discipline. Figure 1.1 illustrates selected characteristics o f Bulgaria's macroeconomic adjustment and the results o f structural reforms in terms o f investment and growth. Two observations can be derived from these facts. One is that given the severity o fthe crisis, Bulgaria hadto implement a huge fiscal adjustment to restore stability as a first step towards EU membership. The other is that enforcing budgetary discipline provided momentum for the rapid implementation o f structural reforms, particularly the privatization and liquidation o f unviable state-owned enterprises and price liberalization, reforms which had for many years stalled under soft budgetary and financial constraints. 1.18 Macroeconomic stability was achieved and has been maintained by prudent fiscal policies and strict discipline in incomes policy, anchored in the Currency Board Arrangement (CBA) adopted in mid-1997. With the adoption o f the CBA, the state limited the role o f the Central Bank to bankingsupervision, economic monitoring and reporting, and the management o f international reserves. Without the possibility o f using monetary and exchange rate policies, fiscal and incomes policies became the only route for macroeconomic adjustment. The cumulative fiscal adjustment between 1996 and 1998 was on the order o f 12.8 percent o f GDP which was brought about by eliminating soft budget and financial practices. This fiscal adjustment resulted in a near balanced external current account through which, together with large FDI flows related to privatization and official external financing, Bulgaria re-established international reserves to about 5.6 months o f imports by 1998. By 1999, the economy was largely stabilized and growth had been re-established. Bulgaria also implemented a strict incomes policy by controlling wage increases in the public sector, particularly in loss-making state-owned enterprises, and by limiting increases in minimum wages. As a result, inflation declined sharply from above an annual rate o f 1,000 percent in 1997 to under 19 percent by 1998. With a more stable macroeconomic environment and deep structural reforms, investment began to recover ledby a growing private sector and growth performance has been robust. 30 Figure 1.1 Bulgaria: Stability and Growth CurrentAccount DeficitandnetFDI FiscalBalance (% o fGDP) (%o fGDP) P 4 8 -1 4 -6 0 -11 -4 0 Current Account Deficit -B -8 FDI -21 -P 895 896 897 898 899 2000 2001 2002 2003 2004 895 896 897 898 899 2000 2001 2002 2003 2004 CPIInflation ExternalDebtandDeMService (%change, annual average) PO-- (%ofGDP) r T 1DO 14 7 ID DO- 90- 80- 70- 60- 50- -- 2 4 0 - / , , , , , , , , I 10 895 896 897 898 899 2000 200120022003 2004 -DO A895 896 897 898 899 2000 2001 2002 2003 2004 -External Debt &External Debt Service (RHS) 25 Investment l T o RealGDP 23 -D (YOchange) 21 -20 8 12 12 aE "a -30 ~ m 8 8 -405 4 4 s i ) -50' 0 0 0 11 -4 -4 -60 9 -8 -8 7 -70 -12 -12 5 -80 -16 -16 895 896 897 898 899 2000 20012002 2003 2004 +GDP annualgrowth -GDP growth, 895=1 I G D l % O f GDP -GDF1(889=1) Source: World Bank staff estimates based on data from SI, BNB, MOF (GFS2004,andNEA. 1.19 Fiscal discipline has been at the center of Bulgaria's progress towards long-term stability and growth, and has provided momentum for structural reform8 by enforcing 31 budgetary and financial discipline. Between 1998 and 1999, the government consolidated hundreds o f extrabudgetary funds into a single Treasury account and adopted strict budgetary controls and a strict timetable in its budgetary process. Budgetary and financial discipline provided the momentum for implementing structural reforms, including privatization and the reduction o f subsidies. Bulgaria also introduced a Fiscal Reserve Account (FRA) for which the government kept sufficient financial resources to cover at least 90 percent o f the gross debt service o f the following year. Inaddition, privatization receipts were deposited inthe FRA with withdrawals limited to external debt reductionpurposes. 1.20 While the macroeconomicfinancial position has been strengthened, contributingto the acceleration of economic growth, the investment-savings gap has been widening rapidly (see Figure 1.2). With an improvedpolicy environment, investment grew rapidly at 16 percent per year in 2003 and 2004 and is rapidly approaching CEEC9 levels. Savings to GDP are lower in Bulgaria than in the NMS-8, reflecting to some extent per capita income differentials. Inmost countries savings rise with income but do so more slowly than per capita income growth. Research on the potential effects o f growth in per capita income on savings shows that on average, an increase inthe income growth rate by 1percentage point results in an increase in the private savings rate o f as much as 0.45 percentage points in the medium to long term. However, as shown in Chapter 2, growth in Bulgaria exhibits a strong bias in favor o f consumption, which explains the relatively stagnant levels o f savings. 1.21 As illustratedby the experience of the Cohesion Countries," growth and stability and improvementsin the standard of living do not come about through EU membership alone (see Figure 1.2). Rather, improvements in the standard of living result from reforms that enhance the policy and institutional frameworks so that they sustain growth with stability over a long period o f time. Irelandjoined the EUin 1973 with a per capita income at 60 percent o f the EU average and with savings to GDP o f about 18 percent. With Ireland's impressive growth performance, its per capita income at PPP surpassed the EU average in 1997 and its savings to GDP increased to 34.5 percent. In contrast, improvements in per capita income in Spain, Portugal, and Greece have been more modest and their trends vary. Spain is converging to the EUaverage per capita income, albeit at a slower pace than Ireland. The per capita incomes of Portugal and Greece have shown a slight decline relative to the EUaverage inrecent years. Central and EasternEuropean Countries (CEEC) are defined here to include Bulgaria, Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Republic and Slovenia. loCohesion countries include Greece, Ireland, Portugal and Spain. 32 Figure 1.2 Investment,Savings, and Growth: Bulgaria,CEEC, and Cohesion Countries __ GrossDomestic Investment(YOof GDP) SavingsRates(YOof GDP) 30 ". -"" " ..... ..~ - . .- ,.. " " "-.......".. ....... "^ ~ I ,, - 25 20 m 15 10 i l 5 T~ 1995 1997 1999 2001 2003 1995 1996 1997 1998 1999 2000 2001 2002 2003 __ -Bulgaria +-8 Cohesioncountries: GDP per capita inPPP CohesionCountries: GrossDomestic Savings (1980=100) (Yo of GDP) 350 ' 350 40 45 40 300 / 300 35 30 35 30 25 25 20 20 25 15 %I r) Ireland Spain Greece *I' l*lPortugal Ireland Spain ~~, I-Portugal Greece Source: World Bank staff estimates based onN S I and Eurostat. 1.22 Over the last four years Bulgaria's macroeconomic financial position has strengthened because of fiscal consolidationand sustained reductions in its external debt; however, the external current account is high. The fiscal consolidation strategy o f the last several years targeted gradual adjustment in the overall fiscal deficit aimed at reaching a balanced budget and an external debt reduction so as to meet the Maastricht Treaty target o f a public debt to GDP ratio of under 60 percent by 2005. The gradual adjustment in the overall balance has been accompanied by reductions in tax rates and significant reductions in public debt. Bulgaria reached the Maastricht Treaty target in 2002 and achieved a balanced budget in 2003. Bulgaria's medium-tern macroeconomic framework over the last four years implied a large current account deficit o f around 6 percent o f GDP as projected imports surpassed Bulgaria's improving export capacity. This reflects the fact that investment needs in Bulgaria are being addressed at a more rapid pace than can be financed by domestic savings. However, there have been episodes during which the external current account deficit reached very high levels. 33 1.23 Under the CBA a prompt reaction in the fiscal stance is needed to respond to external developments. The sharp downturn in the world economy in 2001, with growth declining from 3.5 percent in 2000 to 1 percent in2001 and with slow recovery inthe EU since then, has exerted pressures on Bulgaria's external deficit. The external current account deficit increased from 5.6 to 7.2 percent o f GDP between 2000 and 2001. The government responded prudently by tightening its fiscal stance and developed contingency measures to be applied should the external environment deteriorate further. In2002, the external deficit narrowed to 5.6 percent. Under the CBA, Bulgaria's macroeconomic framework requires a more immediate response to external developments than do most other New Member States, to keep external imbalances within prudent levels. 1.24 As in other countries in the region, a large interest rate differential between Bulgaria and Western Europe has accelerated the expansion of credit-creating pressures on externalimbalances. The external current account deficit widened to 9.3 percent o f GDP in 2003. The tightening o f the fiscal stance was complemented by measures taken to reduce the growth o f credit in2004. Preliminary estimates indicate that the external deficit declined to 7.5 percent inthat year." Since 1998, external current account deficits have been financed by large FDI flows and international reserves have increased every year since then (see Table 1.2 and Figure 1.1). By 2004, international reserves reached a record level o f 6.7 billion euros, representingabout 6.1 monthso f imports, about 34 percent o f GDP, and more than 60 percent of broad money. FDI flows and other private capital flows are expected to finance the projected external current account deficits. Given Bulgaria's recent experience with rapidly changing external conditions, these large external imbalances imply that Bulgaria is vulnerable to external shocks and adjustments. 1.25 Maintaining macroeconomic discipline and keeping up the momentum in implementing structural reforms resulted in solid economic performance and overall improvementsin the standard of living (see Table 1.2 and Figure 1.1). Since 1999, annual average economic growth has been 4.8 percent and accelerated to 5.6 percent in 2004. The average annual rate o f inflation has been around 6 percent with a lower average end-of-period rate o f inflation o f around 4.5-with most o f the difference attributed to the one-off effects o f the price liberalization and tax adjustments implemented in 1999-2004. Interest rates have declined from around 13 percent to 9 percent. Sustained growth, low inflation, and declining interest rates have contributed toward increases in per capita income and growth in the purchasing power of wage income. Per capita income, at PPP, increased fromUS$5,502 in 1998 to US$8,260 in2004 and at an average o f 6.5 percent per year in real terns. Registered monthly average wages and salaries increased from 183 BGNin 1998 to 294 BGNin2004-an average of 2 percent per year in real terns.'* The unemployment rate, which had remained high since the early 1990s and which peaked at 18 percent in2000, started to decline in2003 to 14.3 percent. While part of this decline was due to the active labor market programs o f the government, in 2004, for the first "Finalestimates of the external current account deficit in2004 may result ina higher deficit, since preliminary balance of payments estimates show a sizable estimate under errors and omissions of close to 2 percent of GDP. '*While wages and salaries inBulgaria are l o w compared to most other countries inthe CEEC, they seem to b e in line with labor productivity, which is also low compared to other countries in the region. It should also be noted that registered wages and salaries are grossly under-reported as a way to minimize highpayroll taxes. 34 time since the 1996-97 crisis, there was net employment creation by the private sector which contributed to a further reduction o f unemployment to 12.7 percent. Thus stability, growth and Bulgaria's extensive social protection system contributed to a substantial reduction in poverty, which continues to decline. B. Bulgaria's performance ina regionalperspective 1.26 Despite positive developments over the last seven years, important gaps in reforms and performance relative to the New Member States (NMS) remain. Over the last seven years, Bulgaria has made important progress in structural reforms. As a result, its trade and investment integration has expanded. Despite a well diversified trade partnership and a liberal trade regime which provide producers in Bulgaria with easy access to global markets, its economy has yet to benefit more fully from its integration with the EU and global markets. Bulgaria's export capacity remains limited compared to the NMS. Competitiveness is limited to unskilled labor-intensive and energy-intensive activities. This due to a large extent to gaps in structural reforms relative to N M S . Bulgaria's relatively solid growth and stability performance is tempered by having one o f the lowest per capita incomes in the region. Rising per capita income have contributed to improvements in social indicators, but sizable gaps in social indicators relative to the N M S remain. Moreover, further macroeconomic adjustment i s needed to support growth. Bulgaria needs to implement a core reform agenda to move to a Lisbon- growth path and accelerate convergence, if EU accession i s to be translated into sustained and meaningful improvements to its standard o f living. B.l Structuralreforms: Bulgariaand the CEEC 1.27 Selected indicators of the progress of structural reforms show important advances in 1999-2004. Between 1998 and 2002, most o f the non-infrastructure enterprises and banks were privatized, many loss-making state-owned enterprises were liquidated and failed banks were closed. Inaddition, trade and prices were liberalized; banking restructuring was completed and banking supervisions strengthened; reforms in the energy sector were implemented in both the electricity and district heating sectors; and the first steps were taken in regulatory reform to improve the investment climate. In 2003-04, Bulgaria implemented major public sector and institutional reforms. The public sector reforms included measures to improve public administration, public expenditure management, and public financial accountability, and the institutional reforms focused onjudiciary reform and anti-corruptionmeasures. 1.28 Structural reforms have experienced a strong momentum since 1998 but gaps relative to the New Member States remain. Figure 1.3 distinguishes between two phases o f reforms. The initial phase relates to the changes in basic economic liberalization typical o f the early stages o f transition: price liberalization, foreign exchange and trade liberalization, and small-scale privatization. The second-phase o f reforms includes deeper and more complex reforms: the privatization o f large-scale enterprises, governance and enterprise restructuring, competition policies, infrastructure reforms, banking and interest rate liberalization, and non- banking financial institutions. The ratings range between 1 and 4+, with a higher rating indicating a more advanced stage o f reform. 35 3.6 3 8 4 4.2~ 4.4 _ _ 0 1 2 3 4 1.29 Two observations should be noted. One observation is that while Bulgaria has accelerated its implementation o f structural reforms since 1998, it lags behind the N M S . By 2004, Bulgaria's second phase o f reforms were at the stage o f the Slovak Republic in 1999 and were lagging behind the stage o f second phase reforms in Estonia, Poland, Hungary, and the Czech Republic in 1999. The other observation is that progress in second phase reforms in 1999-2004 was still at an early stage. Furthermore, the N M S continue to make progress inboth phases o f structural reforms. Seemingly small differences in the ratings reflect large qualitative differences in the policy environment once the ratings surpass the level o f 4 in the first phase index and 3.5 in the second phase index. For example, in January 2004 the Polish Parliament adopted the Law on Economic Freedom, which reduces the number o f licensing regimes from 9 to 5, introduces silent consent rule in business registration and other areas o f administrative regulation, and limits the number and duration o f various inspections. In contrast, in Bulgaria, notwithstanding an important regulatory reform in 2001-04, there are 39 licensing regimes de juro (but 60 licensing regimes de facto), the silent consent applies only to a handful o f licenses, and there i s no legislation on inspections. Institutional and regulatory reforms are discussed in Chapter 6. B.2 Selected economic and social dimensions: Bulgaria and the CEEC 1.30 Bulgaria's economic performance over the last few years has been characterized by robust and sustained growth with solid macroeconomic stability. As discussed in the previous section, the economy shows a vigorous recovery from the protracted economic contraction o f the first nine years, which culminated inthe 1996-97 crisis. Inaddition, risingper 36 capita income has contributed to improvements in social indicators (see Box 1.1). Despite these overall positive developments, the country's economic performance i s mixed compared to the NMS-8 (see Table 1.3). Table 1.3 Bulgaria and the CEEC: Selected Economic Indicators, 2000-04 Interest Private Rates Overall Broad Sector Net GDP per (avg. Fiscal External Money Output, FDI capitain GDP CPI long Balance CAB (M2) 2004 (avg. 2003, PPP Growth Inflation term, (avg, YO (yoof (Yo of (Yoof Yo of (US$) (avg. YO) (avg. YO) ("/.)a/ ofGDP) GDP) GDP) GDP) bl GDP) Bulgaria 7,501 4.9 6.4 10.4 0.0 -7.0 45.0 75 7.7 Croatia 11,139 4.1 2.8 11.6 -5.5 -5.5 61.9 60 5.3 Czech Rep. 16,448 3.1 2.7 3.9 -6.7 -5.8 70.0 80 7.0 Estonia 13,348 6.5 3.5 7.6 0.9 -9.6 , 39.1 80 5.1 Hungary 14,572 3.9 7.1 7.8 -5.6 -8.0 47.1 80 3.9 Latvia 9,981 7.5 3.2 3.4 -1.9 -8.6 33.7 70 3.5 Lithuania 11,250 6.7 0.5 8.0 -2.0 -3.1 28.9 75 3.0 Poland 11,623 3.1 4.4 14.3 -5.5 -3.0 42.5 75 2.9 Romania 7,222 5.2 28.7 22.6 -2.7 -5.4 24.6 70 3.0 Slovak Rep. 13,469 4.1 7.7 6.8 -5.6 -4.7 64.7 80 7.4 Slovenia 19,300 4.1 7.7 9.7 -5.6 -4.7 64.7 65 7.4 NMS-8 13,749 4.9 4.6 7.7 -4.0 -5.9 48.8 76 5.0 Notes: ai Ratespertainto the following: Bulgaria,commercial bank lending (weightedaverage); Croatia,prime lending,CzechRepublic, one- year interbanklending; Estonia,commercial lending; Hungary, IO-year governmentbond auction; Latvia,prime lending -refinancing; Lithuania, prime lending; Poland,refinancing;Romania,interbank lending; Slovakia, one-yearinterbanklending; and Slovenia, discount rate. b/ Mid-year estimate, EBRD TransitionReport,2004. Sources: InternationalFinancialStatistics, IMF; World Bank ECA RegionalTables (2005), Datastream,EBRD. 1.31 Bulgaria's relatively solid growth and stability performance is tempered by one of the lowest per capita incomes in the region. As a result o f reforms, Bulgaria reached the average growth performance o f the NMS-8 during 2000-04. However, this solid performance i s tempered by the fact that Bulgaria has the second lowest per capita income among the CEEC. Moreover, Bulgaria's growth performance i s below that observed in Estonia, which also has a CBA. This indicates that growth may be below its potential. (This question i s addressed in Chapter 2). At 75 percent o f GDP, the share o f the private sector in the economy i s about the same as the average share inthe NMS-8 but falls short when compared to strong performers such as Estonia whose share is 80 percent. To the extent that resources are used more efficiently in the private sector, further expansion o f the private sector inthe economy can be expected to have a positive impact on productivity and growth. Bulgaria had the highest FDI flows to GDP among all comparators. However, as shown in Chapter 3, cumulative FDI flows have yet to reach a critical mass to improve Bulgaria's productivity and competitiveness performance. 1.32 Moreover, further macroeconomic adjustment is needed to support growth and accelerate convergence. The higher inflation rate relative to the average rate o f the N M S - 8 i s partly due to the fact that price liberalization and tax harmonization have been taking place in Bulgaria more recently than in most o f the NMS-8. Nevertheless, further fiscal adjustment is neededto converge to a lower inflation rate. Monetary and credit aggregates have been growing rapidly reflecting the rapid expansion o f credit to the private sector and a vigorous increase in money demand in Bulgaria. As a result, broad money in Bulgaria i s close to the average of the 37 NMS-8. However, broad money to GDP in Bulgaria at 45 percent i s now considerably higher than the 39 percent in Estonia, which also has a CBA and a sizable external current account deficit. However, in contrast to Bulgaria, Estonia's productivity levels are high and public expenditures are much more flexible-as i s illustratedby their ability to generate a fiscal surplus o f close to 4 percent o f GDP in 2003 to respond to the widening o f the external current account deficit. Considering that the world economy i s moving from an accommodating monetary stance to a more neutral position, changes in financial conditions are not without risks. If recent monetary measures are effective in reducing the growth o f credit to the private sector, vulnerabilities could be reduced. 1.33 The relative size and efficiency of the public and private sectors poses important challenges. As noted earlier, the share o f the private sector inBulgaria's economy has increased to 75 percent o f GDP but it i s below that f more competitive economies in the region such as Estonia and the Czech Republic. This poses the challenge o f trade-offs regarding the size o f the public sector and the efficiency with which it uses the resources at its command relative to the private sector. As shown in Chapter 4, the tax burden in Bulgaria has been declining, but it remains high and there i s scope to improve the efficiency o f public expenditures. At the same time, the pressures on the external account deficit and the rapid growth o f credit to the private sector entail macroeconomic adjustment. With a rapidly growing share o f social expenditures in total public expenditures, and with their relative inefficiencies, budgetary inflexibility i s increasing. 1.34 While social indicators have been improving, sizable gaps relative to the NMS remain (see Box 1.1). The improvement in social indicators i s remarkable considering the severe and protracted economic contraction o f the first nine years since transition. The growth and stability o f the last seven years contributed to rising incomes and improvements in social indicators. In addition, the extensive and sizable social protection system in particular, and, more generally the large and increasing shares o f social expenditureshave played a central role inthe improvements in social indicators. However, the rapidly aging population together with important inefficiencies, including governance issues, in social expenditures have resulted in growing budgetary pressures. Reforms that will improve the efficiency and fiscal sustainability o f social expenditures need to be implementedif Bulgaria is to catch up in social development with the New Member States while safeguarding its stability. 38 IBox 1.1 Trends in Selected Social Indicators:Bulgariaand the CEEC Based on GDP per capita at PPP, Bulgaria is the second poorest country with the fastest aging population in the region. Nevertheless, in addition to Bulgaria's making progress in poverty reduction, the country's social indicators have also been improving. Adult illiteracy, infant mortality and the mortality o f children under 5 years o f age, school enrollment and life expectancy at birthhave improved since the start o f transition. These improvements are remarkable taking into account the length and severity o f the protracted contraction o f 1989-97 and the crisis in 1996-97, and the differences in per capita income. The improvements in school enrollment are noteworthy. Net school enrollment increased substantially from 63 to 88 percent by 200314, surpassing the levels o f Estonia and Croatia, with the latter having started from the same low level as Bulgaria. It can also be observed that for the CEEC increases inper capita income are associated with considerable improvements in social indicators. Bulgaria's growth-driven agenda i s supportive of continued improvement inits social indicators. Bulgariaand the CEEC: Selected SocialIndicators GDPper Adult School Mortality Mortality rate, Life expectancy capita, Illiteracy enrollment, rate, infant under-5 (per at birth, total PPP ("/.I secondary(YOnet) (US$) (per 1,000) 1,000) (years) 2003 1990 2004 1990al 2002bl 1990 2002~1 1990 2002~1 1990dl 2002el Bulgaria 7,807 2.8 1.4 63 88 14.8 12.3 16 13 71.3 72.1 Croatia 1,139 3 1.9 63 84 12 7 13 8 72.2 73.8 Czech Republic 6,448 86 89 10 4 11 5 71.7 75.0 Estonia 3,348 0.2 0.2 82 85 15 10 17 12 69.5 70.6 Hungary 4,572 0.9 0.7 75 92 15 8 16 9 69.3 72.3 Latvia 9,981 0.2 0.3 77 87 16 17 20 21 69.3 70.4 Lithuania 1,250 0.7 0.4 81 93 17 8 13 9 71.3 72.7 Poland 1,623 .. 0.3 76 89 16 8 19 9 70.9 73.8 Romania 7,222 2.9 2.7 73 79 27 19 32 21 69.7 70.1 Slovak Republic 13,469 .. 0.3 86 14 8 15 9 70.9 73.3 ' Slovenia 19,300 0.4 0.3 89 92 8 4 9 5 73.3 75.9 Notes: a/ Datafor the Czech Republic, Latvia and Romaniarefer to 1993, Estonia, 1992, Lithuania, 1994, and Slovenia, 1997. b/ Data refer ~ to 2001-02 with the exception of Latvia and Poland (2000102), and Bulgaria (2003/04). c/ Data for Bulgaria refer to 2003. d/ Data for Bulgariarefer to 1988-90. e/ Data for Bulgariarefer to 2001-03. Source: UNESCO (Adult illiteracy rates for 2002 refer to 2002-2004, and are available for Censusyears), World Bank database (DDP), B.3 Trade and investmentintegration: Bulgariaand the CEEC 1.35 As a result of progress in structural reforms, Bulgaria's trade and investment integrationhas expanded,but its economy hasyet to benefit more fully from its integration with EUand global markets (see Table 1.4). Bulgaria has an open economy with total trade in goods and services representing about 122 percent o f GDP and compares well with the average ratio o f 124 percent o f the NMS-8. At 29 percent, the share o f trade in services to GDP in Bulgaria i s significantly higher than the average o f 20 percent for the NMS-8. This reflects the rapidly growing tourism industry in Bulgaria. On the side o f merchandise trade, however, Bulgaria's share to GDP at 93 percent i s lower than the average share o f 103 percent of the NMS-8. These ratios, however, mask larger and more fundamental differences than may appear. Bulgaria's merchandise imports are about 30 percent higher than its merchandise exports. While the share o fmerchandise imports to GDP is comparable to other countries inthe region, its share 39 o fmerchandise exports to GDP i s very low. Bulgaria's export capacity is improvingvery slowly and, as shown in Chapter 3, exports are dominated by unskilled labor-intensive and energy- intensive activities. Table 1.4 Bulgaria and CEEC: Selected Indicators of Integration, 2003 Imports Imports Imports Imports Exports to Exports to FDI from plus plus from EU from EU(Yo Of CEEC-10 EU(Yoof Exports Exports of (% of CEEC-10 total (YOof total total FDI) of Goods Services to Total (YOof total exports) exports) (2002) to GDP GDP(Y0) imports) imports) (YO) Bulgaria 92.9 28.8 49.6 8.4 56.5 6.8 66.1 Croatia 70.6 n.a. 56.5 16.7 54.9 12.0 71.2' Czech Rep. 111.3 16.6 59.2 12.6 69.8 18.0 86.1 Estonia 149.6 39.6 51.9 10.2 58.3 14.1 81.5 Hungary 108.4 20.0 56.1 9.7 73.5 11.3 80.2"/ Latvia 73.5 22.0 51.0 24.6 61.8 17.5 56.4 Lithuania 93.1 17.0 44.5 11.6 42.1 18.8 59.5 Poland 58.0 10.4 60.6 8.3 68.4 12.9 82.5 Romania 73.1 10.4 57.7 10.6 67.9 8.1 61.1"/ Slovak Rep. 137.0 19.4 51.4 22.9 60.7 25.7 79.7 Slovenia 96.2 17.8 67.1 12.8 58.4 18.6 78.1 NMS-8 103.4 20.4 55.2 14.1 61.6 17.1 74.8 Source: World Bank estimatesbased on data from UNCOMTRADE,Eurostat, and BNB . 1.36 With a well diversified trade partnership and a liberal trade regime, producers in Bulgaria have easy access to global markets. Bulgaria has a more diversified trade partnership than other countries in the region. The EU i s Bulgaria's largest and most important trading market. Close to 50 percent of its imports and 57 percent o f its exports are with the EU. In addition, more than two-thirds o f FDIinflows to Bulgaria are from the EU. Germany, Italy, and Russia are the most important trading partners on the import side, with a combined share of about 40 percent of total imports. Italy, Germany, Turkey, and Greece are the most important trading partners, representing about 43 percent o f its total exports. Imports from the United States, Canada, and Japan together represent about 4 percent o f total imports, and exports to these countries combined are about 5 percent o f total exports. Trade with the Balkan countries is 6 percent on the export side and less than 1percent on the import side. C. Fromstability and growth to EUconvergence 1.37 Bulgaria should take advantage of the momentum of its economic progress to increase productivity and growth performance and accelerate real convergence. Bulgaria's gap inper capita income with EU-25 levels has been declining since 2000-per capita income, at PPS, increased from 25 to close to 31 percent by 2004. This improvement is underpinned by the structural transformation o f its economy. The share o f the private sector in the economy increased to around 75 percent o f GDP by 2004 from 60 percent in 1998. Trade in goods and services expanded to over 117 percent o f GDP in 2003 and to 127percent o f GDP in2004 from 40 94 percent in 1998. And FDI flows have increased to an average o f 7 percent o f GDP per year since 1998 compared to an average o f 1 percent o f GDP in the first nine years o f transition. Investors' confidence has improved over time, both domestically and externally, with spreads o f Brady and Euro bonds declining since 2002. During the summer o f 2004, S&P and Fitch upgraded Bulgaria's long-term foreign currency debt to investment grade rating. If macroeconomic discipline i s maintained and the pace o f structural reforms is continued, Bulgaria can build on its achievements to date to accelerate real convergence. C.l. The challengeof convergence 1.38 Despitethis overall positiveperformance,Bulgariais one of the poorestcountriesin CentralandEasternEurope. Bulgaria's per capita income at PPS in2004 i s 31percent and 56 percent, respectively, o f the average level o f the EU-25 and the NMS-8 (see Figure 1.4). The large income differences between Bulgaria and the EU-25 and NMS-8 reflect large gaps in productivity, in physical and human capital stock, in the functioning o f products and factor markets in the economy, and more generally in the quality o f the policy and institutional frameworks. Closing these gaps is the central challenge o f convergence. Figure1.4 GDPper Capita, PPS,EU-25=100 :Bulgaria andthe New Member States Bulgaria1 Romania Latvia 2004 Poland 0 2000 Lithuania Estonia Slovakia kngary Czech Rep. Slovenia 0 20 40 60 80 I00 Source: World Bank staff estimates based on Eurostat. C.2. Bulgaria's demographic challenge and the Lisbonagenda 1.39 The Lisbon agenda focuses on increasing productivity, growth, and employment. Bulgaria expects to join the EU in January 2007. To this end, Bulgaria provisionally closed all 31 chapters o f the acquis communautaire in June 2004. This commitment implies that Bulgaria needs to make significant progress in institutional reforms in the medium term. It also implies that Bulgaria i s signing on to central EU principles. The EU Single Market and the Lisbon agenda are particularly relevant in defining Bulgaria's core reform agenda for the future. The 41 Single European Market principle implies a commitment to the free movement o f goods, services, capital and persons within the EU. It also implies that Bulgaria cannot undertake policies that foster-for example, through subsidies or other more subtle means o f state aid- domestic production at the expense o f undermining producers elsewhere in the EU. Another implication particularly relevant to Bulgaria i s a commitment to highly competitive domestic markets. Furthermore, in addition to increasing productivity and growth performance, the Lisbon agenda targets increasing the employment rate to 70 percent by 2010 which implies increasing labor market participation and reducing unemployment rather than discouraging labor market supply. 1.40 These commitments cannot be taken lightly by Bulgaria given its large income and productivity gap with the EU-25 and its large demographic discount. As shown in Chapter 2, there is evidence that productivity is low and output growth is below its potential. labor productivity and investment in inward oriented activities, representing the lion's share o f total investment, are stagnant. And while productivity in outward oriented activities i s rising rapidly, employment and investment inthese activities has not reached a critical mass in the economy so as to be reflected at aggregate levels. Moreover, the factors o f production are underutilized. As a result, competitiveness is limited to unskilled labor-intensive and energy-intensive products, and export capacity is developing very slowly (see Chapter 3). Figure 1.5 Convergence and Demograr>hics GDP per capita, PPP 2000 constant$ Trend due to decline in labor force 6200 6000 5800 5600 5400 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 lource: World Bank staff estimates based on World Bank population ~~ projections. 1.41 Bulgaria faces a larger demographic discount than most countries in the region. The urgency o f implementing a core reform agenda centered on productivity and growth is more pressing for Bulgaria than for other countries in the region where the demographic discount is not as large and is not increasing as rapidly. Bulgaria has had a negative natural rate of population growth (about -0.7 percent per year), which, together with net out-migration (about - 0.1 percent per year) resulted in a negative population growth averaging -0.8 percent a year 42 during the 1990s and early 2000s. As a result, the working age population is declining rapidly and the aging population is growing fast-the demographic discount i s large and rising (see Figure 1.5). It is estimated that in 20 years' time the working age population will decline by 19 percent and the population older than 64 years will increase by 17 percent relative to 2005 levels. Figure 1.5 shows the trend inthe GDP per capita at PPP, 2000 constant US dollars if it i s driven solely by the decline inthe labor force as a result o fthe demographic trends. Figure 1.6 Convergence,Labor Market P nticipal In, and Productivity Labor market participation: I from 50 to 70 percent by 2015 18000- I 16000- Trend increasing productivity and labor market partiupation 14000- 12000~ 10000- 8000- 6000- 4000~ 2000- 0 1 1 2 3 4 5 6 7 8 9 1011 12 13 14 15 16 17 18 19 20 Source: World Bank staff estimates based on databases o f the World Bank and World Bank population projections. 1.42 Convergence implies that Bulgaria needs to increaseproductivity and labor market participation (see Figure 1.6). Simple estimates indicate that ifper capita income is to double in20 years, which wouldresult inits reachingtwo-thirds ofthe EU-25 per capita level, GDP per worker would need to increase by about at least 5.5 percent per year. The trend inthe declining labor force impliesthat labor productivityneeds to increase by about 1 percentage point per year just to compensate for the impact of the decline in the labor force on income per capita. Increasing labor market participation appears to be insufficient to compensate the impact. As shown in Figure 1.4, without increases in labor productivity to compensate for the declining trend inthe labor force, the effects o fthis trend on per capita income can be sizable. 1.43 As shown in Figure 1.6, convergence depends strongly on increasing labor market participation. Productivity growth cannot compensate for a low labor market participation o f 50 percent. Even if productivity increases, without an increase in labor market participation to 70 percent, Bulgaria's per capita income will remain below one-third of the EU-25 level. Figure 1.6 (right-hand panel) shows two per capita income paths; the lower path i s an estimate o f the trend that per capita income would take if labor market participation remained at the low current level o f 50 percent. Under this scenario, Bulgaria's per capita income would grow but the gap with the EU-25 would remain at the current large level of close to 70 percent. Alternatively, if labor market participation should increase from 50 to 70 percent by 2015 and productivity should also increase then per capita income could double in about 20 years and the gap with the EU-25 could be reducedto about 30 percent. This i s the higher path shown inFigure 1.6. 1.44 Reallocating resources from low to higher productivity activities is seriously constrained by labor market rigidities. Bulgaria's labor market is performing poorly. As 43 shown in Chapter 5, Bulgaria has the second lowest labor market participation (49.4 percent compared to 58 percent in the EU-25 in 2003) and the lowest employment rate (43.6 percent compared to 53 percent in the EU-25 in 2003). Simple estimates indicate that if Bulgaria i s to meet the Lisbon labor market performance targets by 2010, employment would need to increase by about 8.0 percent per year in 2005-10; and labor market participation would need to increase by about 4.5 percent per year during the same period. Meeting the Lisbon labor market performance targets i s challenging even if these are met by 2020: employment and labor market participation would need to increase by 3.5 and by at least 2 percent per year, respectively, over the next 15 years. 1.45 In addition, there is evidence suggesting that the regulatory and institutional framework for competition in the domestic markets has a large scope for improvement. Entry and exit o f businesses in Bulgaria is not easy, fast, or cheap. These impediments often tend to result in substantial temporary rents. Limited competition inthe domestic markets does not exert pressures to cut costs or improve productivity. Therefore, the central challenge o f Bulgaria's core reform agenda is to move towards a Lisbon-growth path-mainly reforms that enable the market mechanism to reallocate resources from lower to higher productivity activities. 44 2. MOVINGTOWARDS A LISBON-GROWTHPATH 2.1 To move to a Lisbon-growthpath, the key drivers of convergence are productivity and growth, Bulgaria faces important reform challenges in improving its productivity and growth performance. Indeed, there i s evidence that growth i s below its potential level and the factors o f production are underutilized: (i) firms are on average operating at 70 percent o f their production capacity; (ii) market participation i s very low; and (iii)considerable share of labor a employment remains in largely unproductive segments o f the economy or where productivity i s stagnant. 2.2 Despite positive developments in terms of reforms and economic performance, Bulgaria's economy has yet to move to a more solid growth path. The growth o f aggregate demand components shows a relatively strong bias in favor o f consumption rather than investment. This i s a troubling trend as Bulgaria's investment gap compared to other countries inthe region is large. Although productivity is risingrapidly inactivities oriented to servicing external markets, productivity in activities oriented to servicing domestic markets is stagnant, and investment shows a strong bias in favor o f sectors that mainly service the domestic market. This reflects the combined effects o f important deficiencies in the policy and institutional frameworks, whose costs can be transferred to domestic consumers via a combination o f higher prices and lower quality but not to consumers in external markets, where there i s intense competition. On the supply side, growth i s being drivenmainly by total factor productivity with only small contributions due to capital accumulation and the declining labor force acting as a brake. Intra-sectoral restructuring shows that substantial inefficiencies remain because of the large share of employment that i s in largely unproductive segments o f the economy or where labor productivity i s stagnant. As a result, further restructuring and relatively sizable labor adjustments will be needed to move towards a Lisbon-growth path. 2.3 In addition to increasing productivity and growth, the Lisbon agenda targets increasing the employment rate to 70 percent by 2010. This target implies that reforms should increase labor market participation and reduce unemployment rather than discouraging labor market supply. These objectives cannot be taken lightly, since, in addition to Bulgaria's large income and productivity gap with the EU-25, its labor market is performing poorly (see Chapter 5) and its demographic discount i s large. 2.4 This chapter, then, is concernedwith the issuesthat are involvedin movingtowards a Lisbon-growthpath. First the chapter examines growth dynamics, and then it assesses the trends in the drivers o f growth on the aggregate demand side. It then examines the patterns o f sectoral investmentand productivity. The chapter then analyzes the shifts that have taken place through inter-sectoral restructuring. The findings o f an analysis o f the general characteristics of economic Performance at the firm level are discussed. The growth accounting analysis examines the growth drivers on the supply side. The mechanics o f alternative growth paths and their key determinants are discussed in the context of an estimated growth cycle model that mimics the growth path o f Bulgaria's economy. The chapter thenpresents the key findings of a cross-country dynamic panel analysis used to estimate Bulgaria's growth in the two periods 1995-99 and 2000-2003 relative to the performance o f other countries. The chapter ends by outlining the main components o f a core reform agenda for moving towards a Lisbon-growth path. A. Growth dynamics 2.5 Growth dynamics over the last seven years reflect deep economic restructuring. The economic transformation o f Bulgaria's economy began 7 rather than 16 years ago. A failed transition through 1997 magnified the problems o f resource allocation. Indeed, some o f the structural shifts that took place in the first half o f the 1990s had to be reversed once market- oriented reforms began in 1997. As a result, the economy suffered a longer period o f output decline, with an initial collapse inoutput followed by a second and deeper economic contraction in the middle of the decade. In fact, as shown in Figure 2.1, the contraction in output in Bulgaria was more severe than in other transition economies. Bulgaria's economic recovery since 1998 has taken place in the context o f impressive economic reform and prudent macroeconomic policies. The growth rate averaged 4.2 percent per year between 1998 and 2003, about the same rate as the average seen in New Member States. However, trends in productivity will need to catch up with those in strong performers o f the New Member States, if Bulgaria's competitiveness i s to be improved. Figure 2.1 GDP Dynamics, 1990-2003 1990=100) -Bulgaria - - 'Early Reformers 'Late Reformers ource: World Bank staff estimates based onDDP database. 2.6 Overall economic growth and rising per capita income in real terms closely follow labor productivity and investment developments (see Figure 2.2). As inother CEEC, during the first few years o f political transition, output and per capita income experienced a deep fall, reaching a first trough in 1993. This was followed by a transitory recovery in 1994-95 driven by a quasi-fiscal deficit and an accommodating monetary policy, which resulted inhighrates o f money creation and consequently was unsustainable. The stripping o f the value o f assets and hyperinflation resulted in a dramatic fall in investment and hence in labor productivity. The 46 decisive stabilization and structural reforms initiated in the second half o f 1997 provided momentum for a vigorous recovery which has been sustained since 1998. The large income differences between Bulgaria and the EU-25 and NMS-8 reflect gaps inproductivity, inphysical and human capital stock, inthe functioning o f products and factor markets inthe economy, and, more generally, in the quality o f the policy and institutional frameworks. The analysis o f the drivers o f growth on the aggregate demand side, inter-sectoral economic restructuring, and the sources o f growth on the factor side outlines the central challenges in growth and competitiveness. Figure 2.2 Growth, GDP per Capita Growth, Labor Productivity, and Investment, 1990-2004 1900 56?' - 25% 1700 20% ~ , I~~~~ 15% 1300 10% 1100 8 1% 1 900 L..- - 2 0% __ GDP per capita (constant 2000 US) ~- ~~~ -e- Growth of GDP +Growth of GDPper employ I .. .. .. Gross capital b m t i o n (%ofGDP) - - ~ ~ _ _ _ _ Source. World Bank staff estimates based on NSI, MOLSP, andWorld Bank databases. B. Drivers of growth 2.7 The analysis of growth in this section reveals some troubling trends. The growth o f aggregate demand components shows a bias in favor o f consumption relative to investment. Despite the fact that the productivity o f sectors oriented mainly to servicing the external markets i s rising rapidly while the productivity o f sectors oriented to servicing the domestic markets is stagnant, investment i s biased in favor o f servicing the domestic market. Inter-sectoral restructuring shows that large inefficiencies remain. As a result, growth on the supply side i s being driven mostly by total factor productivity, with only small contributions from capital accumulation and the labor force acting as a brake on growth. Moreover, there is evidence that growth i s below its potential and the factors o f production are underutilized. This section examines the drivers o f growth on the aggregate demand side o f the economy and includes a sectoral analysis o f investment, employment and productivity. The overall trends o f inter- sectoral restructuring are then discussed and compared to those observed in the NMS. The results o f estimates o f sources o f growth on the supply side are then presented, followed by a discussion o f the implications o f two models: one focusing on potential output growth as estimated by a growth cycle model, and the other using a model o f conditional convergence to compare the growth performance o f Bulgaria with that o f other countries inthe region. 47 B.l Drivers of growth: Aggregate demandcomponents 2.8 The growthof aggregatedemandcomponents shows a relativelystrongbiastowards consumption relative to investment (see Table 2.1 and Figure 2.3). On the aggregate demand side, close to 90 percent o f the average GDP growth in 2000-04 was due to total consumption, driven mostly by private consumption. The growth o f total consumption represented about 4.5 percentage points o f an average growth rate o f 4.9 percent per year during this period. With sustained growth and relatively low inflation resulting in improvements in income levels and purchasing power, the growth o f private consumption has been robust at about 3.5 percentage points o f an average GDP growth rate o f 4.9 percent per year. Incontrast, the contribution o f gross domestic investment to growth was at an average o f 2.9 percentage points despite its rapid growth o f an average 14 percent per year in 2000-04. This was largely due to Bulgaria's relatively low level o f investment o f around 21 percent o f GDP during the same period. 2.9 The challengeof investmentin Bulgariaincludesseveral dimensionsin both its level and its composition. As a result o f the protracted transition, Bulgaria lost more physical and human capital than most o f the other transition countries. Investment fell more rapidly than output; the decline was particularly severe between 1994 and 1997, at the time o f the second and deeper economic contraction. Only one other country in CEE registered a lower investment ratio during the transition (Albania, in 1991-92). Similarly, Bulgaria's human capital has fallen precipitously during the transition years. Only recently has investment in Bulgaria started to recover. As a result, the investment gap in Bulgaria relative to other countries in the region is large. The average investmentto GDP inBulgaria in 1991-2004 was 17 percent compared to an average o f 25 percent in the NMS-8. This reflects a longstanding neglect o f investment in Bulgaria despite the rapid increase in debt up to 1997. Later, the legacy of high indebtness limited both public and private investment. It i s only recently that investment has started to recover from a trough o f 10percent o f GDP in 1997. Table 2.1 Bulgaria: Drivers of Growth of Demand Components, 1990-2004 Growthrates (YO) Sharesof GDP (YO) 1990 to 1996 to 1998 to 2000 to 1990 to 1996 to 1998 to 2000 to 1995 1997 2004 2004 1995 1997 2004 2004p GDP -4 -8 4 4.9 100 100 100 100 Exports of GNFS -14 12 6 10.9 42 57 52 55 Importsof GNFS -16 4 13 13.5 44 52 59 63 Investment 12 -22 25 14.2 18 9 20 21 FixedInvestment -10 -20 18 14.6 16 12 17 19 Gov. Consumption -9 -15 6 5.7 18 12 18 18 Priv. Consumption -4 -7 5 5.1 66 74 69 69 Source: World Bank staff estimates based on NSIdata residing in World Bank databases, and latest NSI estimates for 2003- 2004; 2004 data are preliminary data. 48 Figure 2.3 Contributions to Economic Grc vth, 2000-2004: Bulgaria and NMS-8 (YO) Bulgaria Bulgaria and New Member States 5.5 4 5 3.5 2.5 1.5 0.5 -0.5 -1 5 -2 5 -3.5 ~~ Household cons- ~e&ports lnports Investment Net Exports ~ Source: World Bank staff estimatesbased onN S I and Curostat. 2.10 Investment to GDP has been recovering more rapidly since 2003, increasing to 23.5 percent of GDP in 2004. Stabilization and structural reforms have improved the investment climate and as a result the private sector leads in carrying out most o f the investment with a share o f about 73 percent o f total investment. Similarly, foreign direct investment i s playing a central role in rebuilding Bulgaria's capital stock, at about an average o f 8 percent o f GDP per year during 2000-04, but it has yet to reach the critical mass needed to improve efficiency and deepen integration with the external markets. Nevertheless, an investment gap o f 8 percentage points o f GDP relative to the NMS-8 indicates that the level o f investment remains an important challenge in Bulgaria. Two aspects are central to addressing this challenge: one is the mix o f public and private investment, and the other i s the trade orientation and factor intensity o f investment. B.2 Sectoral investmentand productivity 2.11 Since 1998, solid economic recovery has been supported by the recovery of investment Figure2.4 GrossDomestic FixedInvestmentby and rising labor productivity (see Figures 2.4 Sector 1990-04 (percent GDP) and 2.5). The stabilization o f the economy and the 45 40 implementation o f a number o f structural reforms 35 have improved the environment for private sector 30 development and investment. Private sector output 25 grew at 8 percent per year between 1998 and 2003, 20 while private investment increased from 5 percent 15 10 o f GDP in 1996-97 to almost 8 percent in 1998. 5 By 1999, private sector investment exceeded 01990 . . . . . . . . . . . . investment by the public sector (general 92 1994 1996 1998 2000 2002 2004 government, SOEs and state-owned banks), and HPrivateinvestment 0Public investment 0GDFl three years later private investment was already Source: World Bank staff estimates based on NSI 49 more than two-thirds o f total investment. Labor productivity grew faster during 1998-2002 than GDP growth. The re-entry in low-productivity jobs o f about 100,000 long-term unemployed under the active labor market program "From Social Assistance to Employment" contributed to the decline inlabor productivity growth, which fell below the rate o f GDP growth in2002-04. 2.12 While private investment accelerated, it lacked an essential complement in basic public infrastructure as gross domestic fixed investment by the public sector declined sharply. Gross domestic fixed investmentby the private sector more than doubled, from 6.4 to 14 percent o f GDP, between 1998 and 2002. In contrast, fixed investment by the public sector declined from 6.7 to 4.3 percent o f GDP during the same period. This decline is partly explained by the legacy o f high indebtness o f the public sector but, as shown in Chapter 4, for the most part it was due to growing imbalances betweenpublic current and capital expenditures while fiscal consolidation was being implemented. Public capital expenditures and operations and maintainance outlays bore a heavy brunt in the fiscal adjustment. This trend began to be reversed only in 2003. As a result o f the long-standing neglect o f basic public infrastructure, private investment lacks an essential complement, owing to deficiencies in basic public infrastructure, particularly inthe transport network. 2.13 Investment in sectors oriented to servicing the domestic market dominate total investment,while outwardorientedinvestmentshows a modest expansion(see Figure 2.5). An analysis o f gross domestic investment distinguishes two categories. Outward-oriented investment includes investmentin sectors with net export receipts o f at least 30 percent o f total receipts from sales. And inward-oriented investment includes investment in sectors with net export receipts below 30 percent o f total receipts from sales, and hence these sectors are predominantly oriented to servicing the domestic market and include nontradables. Solid growth perfonnance relies on both types o f investments and activities driven by their corresponding levels o fproductivity. The question i s ifthe policy framework supports resource allocation to highproductivity activities. Investment oriented to servicing the domestic market (or inward-oriented investment) has seen an increase o f close to 9 percentage points o f GDP over 1998-2003, while investment oriented to servicing outward markets (or outward-oriented investment) has experienced only a modest growth during the same period. The expansion o f inward-oriented investment was concentrated intrade, energy, real estate and business services. This expansion made important contributions in developing basic services for the economy. However, with the exception o f trade, these sectors have not contributed much to the overall growth o f the gross value added o f the domestic sectors as a whole, despite the fact that these had a cumulative growth o f 18 percent in 1998-2003. Outward-oriented investmentwas driven by the manufacturing o f textiles and textile products, and the manufacturing o f basic metals. These sectors have also been driving output growth in the outward sectors, together with the manufacturing o f machinery and equipment and chemical products. Overall gross value added inthe outward sectors showed a growth o f44 percent in2003 comparedto 1998. 2.14 However, the labor productivity growth of outward oriented investment markets outpaces the labor productivity of inward oriented investment (see Figure 2.5). The outward-oriented sectors enjoyed a gain inproductivity as measured by a gross value added per worker o f 61 percent between 1998 and 2003-or about four times higher than the productivity gains of the inward-oriented sectors. The productivity o f the latter even declined in 2003 50 compared to 2002, which was associated with the increase in public employment due to a government sponsored employment program.' The sizable labor productivity growth o f the outward-oriented sectors is associated with the positive effects o f international competition, which exerts pressures to the use o f factors o f production as efficiently as possible in order to remain competitive in the international markets. Outward-oriented sectors in general also tend to import and adapt new technologies in order to remain competitive with other players in the external markets, which inturn increases labor productivity. Fieure2.5 Trade Orientationof InvestmentandLabor Productivitv.1998-2003 Investment, 1998-2003 GVA per worker, 1998-2003 ( % o f GDP) (BGN, at 2000 prices) 13000 9000 8000 7000 6000 5000 4000 c 3000 2000 899 2000 2001 2002 2003 I 898 898 899 2000 2001 2002 2003 Domestic sectors b o u t w a r d sectors Domestic -Outward I Source: World Bank staff estimates based on N S I data. 2.15 After the economic transformation started in 1998, productivity growth was accompanied by labor shedding, with outward-oriented sectors exhibiting momentum in employment expansion in 2002-03 (see Figure 2.6). The trade orientation o f employment resembles that o f investment, with inward-oriented sectors having five to six times more employment than outward-oriented sectors. Employment in both categories declined in 1998- 2000 as a result o f deep economic restructuring. However, the economic restructuring was deeper in the outward-oriented sectors. Employment in the outward-oriented sectors declined by about 15 percent in 1998-2000 as part o f the major economic restructuring associated with the privatization o f SOEs and the liquidation o f unviable firms. Surplus labor was highest inthe manufacturing o f transport equipment, chemicals and rubber, and electrical and optical machinery. At the same time, employment in textile manufacturing was growing steadily throughout the period, while sectors such as the manufacturing o f wood and wood products, rubber and basic metals experienced significant increases in employment between 2001 and 2003. Labor shedding was less pronounced in inward-oriented sectors in 1998-2000 and the sharp increase in employment in this category in 2002-03 reflects the re-entry o f the long-term unemployed into low productivity jobs. l3The program "From Social Assistance to Employment" started inlate 2002 as a pilot and proceeded at full speed in2003. Itprovidedpublic sectorjobs to around 100,000 long-term unemployed. 51 2.16 Although starting from a very low base, investment per worker and labor productivityare catchingup to the NMS-8 (see Figure2.6). With the depletion o fassets, and o f physical and human capital, and the large distortions in the economy with many sectors having considerable surplus labor, Bulgaria began its economic transformation with a very low level o f investment per worker. With the deep economic restructuring since 1998, both investment per worker and labor productivity are catching up with the levels observed in the NMS-8. However, productivity gains in Bulgaria during 1998-2004 remain below the NMS-8 average. Figure2.6 Trade Orientationof Employmen andInvestmentand Output per Worker Outward-oriented Employment, 1998-2003 lnward-o riented Employment, 1998-2003 (thousand) (thousand) 2700 540 520 2650 500 2600 480 2550 460 2500 440 420 2450 400 2400 1998 899 2000 2001 ' 2002 2003 898 1999 2000 2001 2002 2003 GVA per Worker PO0 - (Euro, at 1995 prices) T 3000 4000 1 (Euro, at 1995 prices) PO00 moo - -- 2500 3800 - 3600 - DO00 800 - -- 2000 3400 - 8000 3200 600 - -- a00 3000 -- 6000 400 - -- 000 2800 - 2600 - 4000 200 - -- Romania -- 500 -- A NMS-8(RHS) Romania A N M S - 8 (RHS) 2000 0 Source: World Bank staff estimates based onN S I and Eurostat data. Note: 2004 are preliminary data. 2.17 The productivity gap is widening compared to the Baltic countries and Poland, whichhavebeencatchingup rapidlywith EUlevels over the last few years. In2002-04, the Baltic countries experienced productivity growth rates o f more than 6 percent per year, whereas productivity gains in Bulgaria slowed down in 2002 and declined in 2003. This slowdown and decline are associated with the public sector sponsored low productivity jobs for the long-term 52 unemployed and the introduction o f the mandatory registration o f labor contracts. Both government policy measures, together with some private job creation, resulted in an unprecedented increase in employment by 6 percent in 2003 compared to 2002. Adjusting overall productivity levels for low-productivity public employment does not change the slump inoverall productivity, although it renders it less pronounced. The imposition of stricter rules for labor registration have increased formal employment but have not translated to higher productivity levels. While labor productivity recovered to around 4 percent growth in 2004, it remains below the 4.5 percent growth attained in2002. 2.18 An analysis of economic performance at the firm level shows that large firms are far more productive than small and medium enterprises and that on average firms operate at 70 percent of production capacity. Annex A presents a model for estimating total factor productivity at the firm level. Economic performance indicators, including growth, labor productivity, and total factor productivity, were examined at the firm level using regression analysis to identify the major correlates o f growth and structural characteristics. The statistical model is a basic specification with a parsimonious set o f correlates. The analysis indicates that the private sector in Bulgaria i s characterized by a relatively small number o f large firms, some o f which are foreign owned, which are growing rapidly and generating high levels o f productivity, and are fairly isolated in their operations from small and medium enterprises. Small and medium enterprises exhibit large inefficiencies. On average, firms operate at 70 percent o fproduction capacity. 2.19 The statistical analysis also shows that foreign direct investment flows and exports are positively associated with higher productivity but the relation is not statistically significant. There are a limited number o f foreign owned firms in the sample used in the exercise, which may be responsible for the weak statistical link between FDI and exports, and productivity. However, as shown in Chapter 3, a closer look at the sectoral composition o f FDI flows shows that a large share o f FDI i s in services and only a small share is in the tradable sectors. Hence, FDI flows in tradables have not reached a sufficient critical mass to have a robust positive impact on productivity. On average, in 1998-2003 about 60 percent o f FDI in Bulgaria was located in services, with more than half o f these flows in the financial sector, and only about 30 percent o f FDI was in tradables, nearly all in manufacturing. Similarly, exports were dominated by unskilled labor intensive products, which explains the weak statistical relation between exports and productivity indicators. Moreover, analysis o f inter-sectoral restructuring indicates that large inefficiencies remain. B.3 Inter-sectoral restructuring 2.20 The economic restructuring of the Bulgarian economy thus far has been remarkable, considering that it has been taking place for only seven years (see Table 2.2). In addition to this short period o f economic transformation, the initial conditions for restructuring were unfavorable compared to those in the NMS-8. The legacy o f indebtness and the collapse o f the banking sector in 1996 severely limited the amount o f available financing. 14 l4The restructuring of the banking sector was implementedquickly and by 2002 the majority o f banks were foreign and were well capitalized. However, financial intermediation was l o w and credit growth declined from above 6 53 Nevertheless, the Bulgarian economy today is substantially different interms o f its structure and functioning from that o f 1990. The share o f services in GVA shows a large expansion which nearly doubled from 31to 59 percent between 1990 and 2004. The re-establishment o fbanking, the development o f the tourism industry, business services, transport, and so on have contributed to the rapid expansion o f services. As a result, Bulgaria's share o f services in GVA compares well with the average for the NMS-8 which stands at 63 percent. 2.21 The expansion of services in the economy was accompaniedby a sizable increasein the share of employment in services. The share o f employment in services to total employment increased from a low level o f 37 percent, compared to 42 percent in NMS-8, in 1990, to 47 percent in2003. Bulgaria's share o f employment in services is lower than the share o f 56 percent for the NMS-8. However, the extent to which the labor increase in services reflects the efficient allocation o f human capital or employment o f last resort inlow productivity and low paying jobs i s not clear. The latter scenario cannot be excluded, since labor productivity in inward-oriented activities, which include a large share o f the services sector, is stagnant. 2.22 The trends in agriculture, however, are in sharp contrast to those observed in the New Member States. InBulgaria the share o f employment inagriculture intotal employment increased sharply, from 18.5 to 25.5 percent between 1990 and 2003. Inthe NMS-8 the share o f employment in agriculture declined from 19 to 13 percent (see Table 2.2). Furthermore, the share o f agriculture in GDP inBulgaria declined from 18 to 11percent during the same period. As a result, productivity inagriculture has declined sharply. 2.23 Given Bulgaria's much higher level of over-industrializationin 1990, the economic restructuring in industry has been more pronounced than in the NMS-8. The share o f industrial employment inBulgaria dropped from 44.8 to 27.2 percent, or about 17.6 percentage points, between 1990 and 2003 (see Table 2.2). In the NMS-8 the share o f industrial employment declined from 38.6 to 31.5 percent, or about 6.5 percentage points, duringthe same period. On the output side, the share o f industry in GVA in Bulgaria declined by 21.3 percentage points while in the NMS-8 it declined by 12.7 percentage points. Despite the deep restructuring in industry, its share in GVA at 30 percent i s significantly lower than the share o f 34 percent inthe NMS-8. 2.24 The implication of these trends for growth potential and convergence cannot be understated. With large inefficiencies inthe agricultural sector and stagnant labor productivity inservices, Bulgaria faces a considerable challenge inproductivity. With labor beingshiftedto low or stagnant labor productivity activities, aggregate productivity levels could grow too slowly (as it seems to be the case) for Bulgaria to move towards a more solid convergence path as it approaches EU accession. Indeed, the analysis o f growth accounting shows that growth is beingdrivenmostly bytotal factor productivity. percent in 1999 to under 1percent in2002 (see Figure 1.7 inChapter 1). Credit started to grow inlate 2002 andhad accelerated to a highrate o f 43 percent in2004. 54 Table 2.2 Inter-sectoralRestructuring:Bulgariaand NMS-8,1990 - 2004 A. Average RealGDP Growth (percent) Index of Real GDPin2004 1990-2004 1995-2004 (1990 = 100) Bulgaria -0.4 1.9 101.1 NMS-8 1.9 4.1 137.8 B. Structure ofProduction(percent) 1990 1995 2004 Bulgaria Agric, forestry & fishing 17.7 13.4 10.9 Industry 51.3 32.7 30.0 Services 31.0 54.0 59.1 NMS-8 31741 Agric, forestry & fishing 10.7 6.3 3.3 Industry 46.8 37.4 34.1 Services 42.5 56.3 62.6 C. Structure of Employment (percent) 1990 1995 2004 Bulgaria 100.0 100.0 100.0 Agric, forestry & fishing 18.5 23.9 25.5 41 Industry 44.8 33.8 27.2 41 Services 36.7 42.3 47.3 41 NMS-8 31 Agric, forestry & fishing 19.1 16.3 12.6 Industry 38.6 34.2 31.5 Services 42.3 49.5 55.9 Memorandum: Bulgaria NMS-8 Average growth in2000 2004p (percent) - 4' 4.9 4.8 Notes: NMS-8includes new member states, excluding Cyprus and Malta. l / Averages are based on population weights. Data weighted averages ofthe earliest available years, which are as follows: Production data: Estonia, Latvia, Slovenia - 1991, for 2004 are preliminary. 2/ Share of gross value added at basic prices. 3/ Data are for 2003.41 Observations for 1990 are others, 1990. Employment data: Latviaexcluded, Hungary - 1992, Czech Republic, Slovenia - 1993, Slovak Republic - 1994, others is 1990; 5/ NMS-8 productioncalculation uses 2003 data for Estonia, Hungary, Slovenia. 6/ NMS-8employment calculation for 1995 uses 1997 data for Latvia. Sources: World Bank staff based on data from N S I Bulgaria, EUROSTAT, ILO, WDI 2004, ECA Regional Tables (2005). C. Growth accounting:Driversof growth on the supply side 2.25 The sources of growth indicatethat the economy shifted from a negativetotal factor productivity in 1989-97 to a positive rate in 1998-2003. Table 2.3 presents a growth accounting decomposition estimate based on different sets o f assumptions regarding the initial 55 capital stock and the rate o f depreciation (see Annex B).15 Given the poor quality o f the data, the conclusions can be indicative only o f the direction o f changes during the period. These results, however, are robust compared to alternative assumptions about the initial capital stock and depreciation rates, and are broadly consistent with other estimates o f total factor productivity (TFP) for Bulgaria and with the prevailing trends inthe CEE reported elsewhere in the literature.16 Most o f the fall in GDP in 1990-97 was the result o f a drop in (TFP). In particular, the rate o f growth o f TFP was highly negative during the earlier years o f transition, and the decline accelerated during the second deeper contraction o f 1996-97. The decline in TFP implies that the marginal productivity o f capital fell abruptly following the transition. In contrast to the productivity trends in 1989-2003, the recovery o f economic growth since 1998 has ledto a rapid increase inTFP. 2.26 The contributionto growth of capitalaccumulationfollows closely the trend in total factor productivity. Inadditionto a large positive TFP in 1998-2003, the rate of growth o fthe physical capital stock followed a similar trend: capital per worker fell between 1990 and 1997 and has increased since 1998. Finally, the rate o f growth o f the labor force was negative in 1990-97 and continued to benegative during 1998-2003. Table 2.3 Bulgaria:GrowthAccounting,1990-2003 (depreciation rate = 0.06) Contributionof: RealGDP Physical Physical Physical Growth Labor capital' TFP' capital' TFP' capital3 TFP3 1990-03 -0.8 -0.2 0.2 -0.8 -2 -0.4 -0.4 -0.2 1990-97 -4.6 -0.3 -0.4 -3.8 -0.8 -3.5 -1.0 -3.3 1998-03 4.1 -0.2 1.o 3.3 0.6 3.7 0.3 4.0 'ko=1.5,2'ko=2.0,3/ko=2.5 (depreciation rate = 0.08) Contributionof: Real GDP Physical Physical Physical Growth Labor capital' TFP' capital' TFP' capital3 TFP3 1990-03 -0.8 -0.2 0.2 -0.4 -0.6 0.0 -0.9 0.3 1990-97 -4.6 -0.3 -1.1 -3.2 -1.4 -2.8 -1.7 -2.6 1998-03 4.1 -0.2 0.9 3.4 0.5 3.8 0.1 4.2 "ko=1.5, 2'ko=2.0,3'ko=2.5. I5 The estimated Cobb-Douglas production fbnction is: Y = AKaL1-".,where Y = total output, L = labor force, K = capital, a = capital share o f output, A= total factor productivity. The capital stock is estimated as K, = (1-6)K,-, I,,where 6 isthe depreciationrate andassumed equal to 6percent and 8percent to generate + alternative estimates. I,is gross capital formation. (for 1987) i s assumed equal to 1.5*GDP, 2.0*GDP, and 2.5*GDP to generate alternative estimates. The values o f the depreciation rate used are higher than the usual benchmark o f 4 percent. This i s to reflect the fact that capital stock used to produce output before the transition had became rapidly obsolete post transition. The assumptions o f KOinclude higher ones than the benchmark for less developing countries (usually 1.5*GDP) to reflect the fact that most Eastern European transition economies had been characterized by "over-industrialization'' in the pre-transition era compared to their counterparts in other developing regions. See Annex B. l6 See, for example, EBRDTransition Reports, Campo and Coricelli, 2002. 56 2.27 The growth accounting analysis for the period 1998-2003 shows that growth has been driven mostly by TFP. This implies that to a large extent growth has been the outcome o f shedding excess capacity and eliminating inefficiencies. There has been a small positive contribution o f capital accumulation to growth, reflecting the fact that the upgrading o f the capital stock is at an early stage. At the same time, a shrinking labor force imposes a demographic discount on growth. Limited improvements in upgrading skills inthe labor force may be partly compensating for the negative growth o f the labor force. 2.28 The results indicate that the contribution of TFP to growth was lower after 2000 when comparedto the estimates without the human capital adjustment. Estimates can be calculated adjusting for the quality o f human capital using the educational attainment o f the labor force and the existing estimates o f returns to education as proxies to estimate a measure o f human capital (H). Two alternative methods were used. One method follows Loayza et al. (2004) and the other Gosh and Kraay (2000). During 1992-97 the negative growth resulting ina sizable contraction o f output can be attributed mostly to sharp declines in TFP more than to changes in the labor force once it i s adjusted for the quality o f human capital. This is due to a large outflow o f workers with a lower education level from the labor force, which changed the educational structure of employment substantially during this period. In contrast, from 1998 to 2003 the structure o f the labor force changed less dramatically, with only a small improvement in the share o f workers with a higher education going from 14 percent in 1998 to about 18 percent in2003. 2.29 Adjustmentsfor the skill contentof the labor force resultin a slightly lower estimate of TFP. As a result, the growth attributable to TFP changes is lower with human capital adjustment than with unadjusted changes in the labor force. Using a depreciation rate o f 8 percent and an initial capital stock to GDP o f 2.0, TFP contributes 3.8 percentage points o f the GDP growth o f 4.1 percent per year, with capital accumulation contributing 0.5 percentage points, while the contribution o f the labor force, unadjusted for educational attainment, is -0.2 percent. Adjusting for the quality of human capital, TFP is about 3.7 percentage points o f the GDP growth o f 4.1 percent per year, with capital accumulation contributing 0.5 percentage points, while the contribution o f the labor force, adjusted for educational attainment, is estimated at -0.01 percentage points. This implies that the upgrading o f skills can have a significant impact incompensating for the demographic discount o f the declining labor force. 2.30 A growth cycle model that mimics the growth path o f Bulgaria's economy was estimated to outline more clearly the mechanics o f alternative growth paths looking ahead, and their key determinants. D.Potentialoutput growth and convergence 2.31 To the extent that there are per capita incomedifferentials,higher growth rates are needed by countries with lower per capita incometo achieve real convergencewith higher per capita income countries. Bulgaria's current per capita income, at PPP, is about 30 percent o f the EU-25 level and about 57 percent o f the levels o f the NMS-8. Economic growth in Bulgaria in2000-04 was on average 4.9 percent per year or about the same level as inthe N M S - 8 during the same period. At this rate o f growth Bulgaria i s converging to EU-25 at the same 57 speed as the average rate o f the N M S - 8 but Bulgaria has a wider per capita income gap with the EU-25. However, there is evidence that Bulgaria's current rate o f growth i s below its potential output growth level. 2.32 The evidence that growth is below its potential level and the factors of production are underutilized lies in the following: (i)firms are operating at 70 percent o f production capacity; (ii)labor market participation i s very low; and (iii) a large share o f employment remains in largely unproductive segments o f the economy or in activities where productivity i s stagnant. O n the side o f potential output, it should be noted that growth refers to the average growth rate obtained by the efficient uses o f resources, under a supportive and predictable policy environment that can be sustained over a long period o f time. Year to year growth rates fluctuate around a long-run trend which approaches potential output growth following the business cycle. Hence, while growth accelerated to 5.6 percent in 2004, the experience o f 2000- 03 shows that the economy remains vulnerable to external shocks since the labor market and investment are unable to absorb the shocks fully. Consequently, potential output growth estimates are based on average growth rates over several years. 2.33 Potentialoutputgrowth was estimatedusinga growth cycle modelwhich mimicsthe growth path of the economy of Bulgaria. Annex C presents a growth cycle model which mimics closely the growth path o f Bulgaria since 1990. Ln essence, estimates based on a growth cycle model aim to outline the interplay o f the key determinants o f alternative growth paths: the relationship between growth and investment, savings, labor, FDI, the rate o f phasing out old capital, policies and institutions. Growth performance i s determined by the rate o f capital accumulation, by the accumulation o f human capital (which may be achieved by improvements inthe quality or skill content of the labor force), by total factor productivity, and by the policy and institutional environment that supports an efficient resource allocation inthe economy. The rate o fpotential output growth i s determinedby the rate o f expansion o f new investment and the phasing out o f old outdated capital stock. The rate o f expansion o f new investment is determined by domestic savings, the capacity to attract FDI flows, and the ability to transform old capital into new capital. FDI flows are a function o f expectations held by foreign investors about possible future profits, which are largely determined by the profitability o f economic opportunities and the predictability o f the institutional framework. The former is linked to the flexibility o f the goods and factors markets and the latter to the extent that Bulgaria can move to a predictable rules-based institutional framework. 2.34 Estimatesbased on the model mimic very well the growth cycle of Bulgaria during the period 1990-2003(see Figure2.7). The pathpredictedby the model mimics the U-shaped output cycle that characterizes Bulgaria's actual growth path. The trough and speed o f recovery, and over many years o f convergence, depends on the technological parameters and the productivity o f factors o f production. It should be noted that the simulations are based on parameters which aim to capture possible paths o f better performance but do not measure directly the impact o f the proposed reforms. Instead, the simulations are only indicative o f possible outcomes subject to alternative key parameters. The estimates o f the model follow closely the sharp protracted contraction o f output up until the 1996-97 crisis, and the vigorous recovery since then. Based on these estimates, three alternative potential output growth paths into the future were calculated. The slowest growth path, labeled C in Figure 2.7, aims at 58 reflecting a scenario if Bulgaria were to lose momentum in implementing reforms and if investment were to slow down relative to recent trends. The second higher path, labeled B in Figure 2.7, aims at reflecting recent performance. And the highest path, labeled A inFigure 2.7, aims at reflecting a possible path if a core reform agenda proposed in this report were implemented. It i s only underthe third, highest, growth path that Bulgaria can rebuild its capital stock at such a pace that the effects o f its lowest point inthe cycle in 1996-97 can be eliminated inthe medium term. Figure2.7 Bulgaria's Growth( xle Model: Three Scenarios New capital Old capital 2000 , 1000 1500 800 600 1000 400 500 200 0 ' I 0 0 10 20 30 40 50 60 0 20 40 60 GDP Investment 100 80 60 40 200 J I I 0- 0 20 40 60 0 20 40 60 Source: World Bank staff estimates based on simulations o f a growth cycle model; see Annex A. 2.35 Transforming old capital into new capital in the economy is very limited, hence phasingout old capitalcan take some time (see Figure2.7). A large share of old investments are irreversible and embedded technologies are obsolete. Moreover, institutional and regulatory constraints limit the extent to which the available factors o f production are being reallocated from low to higher productivity activities. This implies that in potential output growth paths there will be no new investment in old capital inputs, and the stock o f old capital will decline over time at a pace that depends on depreciation and the ability to release resources associated with old capital for use in new investments. However, given the legacy o f over-investment in largely inefficient production, reflected by a large over-industrialization index at the start o f the transition, the phasing out o f old capital could take a longtime. 2.36 For a given rate of new investment,the more productivethe new technology is and the higher the institutional quality is compared to current levels, the higher the rate of potential output growth will be. High investments, domestic and foreign, and high growth depend on the extent to which new investments in Bulgaria are linked to the increasing 59 efficiency and productivity o f the economy and the development o f export capacity. This implies that reforms need to be implementedto substantially reduce the constraints to expanding outward-oriented investment. Among these reforms, improving institutional quality is central to achieving potential output growth. 2.37 The estimations of the growth cycle model point to several important policy implications. First, for additional capital accumulation to deliver high efficiency gains, disincentives to investment in outward-oriented activities need to be substantially reduced. While FDI flows are to a large extent filling the gap between domestic savings and investment, inthe long term Bulgaria's low domestic savings represent a constraint to reaching the potential output growth. Second, significant improvements in the skill content o f the labor force are neededto compensate for the demographic discount and increased labor productivity is needed to develop export capacity beyond unskilled labor intensive production. As discussed in Chapters 3 and 4, addressing deficiencies in the transport network for trade facilitation and addressing the significant skills mismatch are central to these two challenges. A third implication i s that the economy needs further restructuring to eliminate the remaining large inefficiencies. The incentive structure resulted in shifting a large share o f employment to largely unproductive segments in the economy or to areas where labor productivity was largely stagnant. With a demographic discount on growth, Bulgaria cannot afford to have a large share of employment locked-in in unproductive activities or activities where labor productivity i s stagnant. Nor can it afford to have low labor market participation or to reduce unemployment by discouraging labor supply (see Chapter 5). Without substantial increases in labor market flexibility, the risks o f the economy remaining locked into low wages and a low productivity growth path are substantial. A fourth implication is that the institutional quality needs to improve if productivity growth i s to increase (see Chapter 6). Indeed, as shown in Figure 2.8, the following cross-country analysis indicates that economic performance can improve substantially if Bulgaria catches up with the extent o f the institutional reforms achieved by strongperformers among the NMS-8. 2.38 Controlling for the quality of its policy framework, Bulgaria's growth performance is about average but i s below its potential compared to other countries (see Figure 2.8). Annex D presents a cross-country dynamic panel model using a parsimonious set o f regressors to capture conditional convergence. This model was used to estimate Bulgaria's growth in the two periods 1995-99 and 2000-03 relative to the performance o f other countries. Inparticular, Figure 2.8 plots the unexplained part o f growth on the unexplained part o f per capita initial income. The unexplainedpart o fthe growth regressions is the residual o f a regression o f growth on a parsimonious set o f explanatory variables, except income. And the unexplained part o f per capita initial income i s the residual o f a regression o f income on the complete set o f explanatory variables. As such, the slope o fthe plotted line is the coefficient on log income. 2.39 Policy and institutional reforms are central to moving to that higher growth path. The analysis also shows that growth performance can be improved ifBulgaria were to catch up to the extent o f the reforms already achieved by strong performers among the NMS-8. While structural reforms have improved the policy and institutional framework in Bulgaria, these nevertheless lag behindthe NMS-8. 60 Figure 2.8 Residuals from the Unpredicted Part of Growth on the Unpredicted Part of Initial per Capita Income 5 Reference line of gap with NMS Model simulation OCZE 00 0 POL00 0 ROM 95 e BGR95 I I I I I -2 -1 e( IcldPPC I x ) 0 1 2 coef = -.97662859, (robust) se = 1.263799,t = -.77 Source: World Bank staff estimates. E. Movingtowards a Lisbon-growth path: A core reform agenda 2.40 An analysis of growth shows that the economy has yet to move towards a Lisbon- growth path. As has been mentioned, there i s evidence that growth i s below its potential level and factors o f production are underutilized: (i) firms are on average operating at 70 percent o f their production capacity; (ii)labor market participation is very low; and (iii) large share o f a employment remains in largely unproductive segments o f the economy or where productivity is stagnant. The growth o f aggregate demand components shows a relatively strong bias in favor o f consumption relative to investment. These are troubling trends as the investment gap o f Bulgaria relative to other countries in the region i s large. To some extent this is partly due to the fact that private investment, while accelerating in the last few years, lacks an essential complement in basic public infrastructure, as public capital expenditures bore a heavy brunt in the fiscal adjustment. Similarly, labor market inflexibility and skills mismatch limit the scope o fprofitable investment opportunities, since to compensate for employment rigidities firms need to operate significantly below production capacity, and lack o f adequate skills further increases the costs ofoperations as workers needto betrained onthejob. 2.41 On the supply side, growth is being driven largely by TFP, while the contribution of capital accumulation is small and declines in the labor force are acting as a brake. TFP in 1998-2003 contributed 3.8 percentage points o f a GDP growth o f 4.1 percent per year, with capital accumulation contributing 0.5 percentage points and the labor force contributing -0.2 61 percentage points. This implies that growth during this period was driven mostly by the shedding o f excess capacity and the elimination o f inefficiencies. while upgrading the capital stock i s at an early stage and the demographic trends o f the labor force act as a brake. Indeed, investmentand FDIflows intradables have not reached a critical mass that would have a robust positive impact on productivity observable at an aggregate level. At the firm level, efficiency gains are driven by a relatively small number o f large firms, some o f which are foreign owned, are growing rapidly and generating high levels o f productivity, and are operating fairly isolated from small and medium enterprises which, in general, are relatively inefficient. Inter-sectoral restructuring shows that large inefficiencies remain. As a result, further restructuring and relatively large labor adjustments will be needed to move towards a Lisbon-growth path. 2.42 Trade integration with the EU and global markets is the productivity driver in Bulgaria's economy. Productivity in investment and labor i s growing rapidly in outward- oriented activities while productivity in inward-oriented sectors i s stagnant. The sizable labor productivity growth o f outward-oriented sectors i s associated with the positive effects o f international competition, which exerts pressures to use the factors o f production efficiently in order to stay competitive in the international markets. However, the combined effects o f important impedimentsto resource allocation towards more productive activities are limitingthe expansion o f investment and employment in outward-oriented sectors. Limited labor adjustment, deficiencies inbasic infrastructure, limited competition inthe domestic markets and deficiencies in institutional quality have contributed to a price bias in favor o f investment and employment in activities oriented mainly to servicing the domestic markets. These deficiencies translate into costs (or temporary rents) that producers can transfer to domestic consumers via a combination o f higher prices and lower quality. In contrast, the scope for transferring these costs, or taking advantage o f temporary rents, to consumers in the external markets is highly limitedbecause o fthe intense global competition. 2.43 The analysis of growth and productivity implies a core set of reforms to move towards a Lisbon-growth path. This report proposes to deepen trade and investment integration with the EU and global markets (see Chapter 3) by implementing a core reform agenda inthree intertwinedareas: (i)Improvingthequantityandqualityofphysicalandhumancapitalaccumulation, by restructuring public expenditures to upgrade skills and the transport network, and improving efficiency and transparency inpublic administration (see Chapter 4) (ii)Enhancing labor market adjustment and the creation o f employment by implementing labor market reforms that better balance labor adjustment, the creation o f employment, and unemployment risks, and that include the rationalization o f the social protection system to enable the reduction o fpayroll taxes (see Chapter 5) (iii)Implementinginstitutional reforms intwo areas-reducing legislative and regulatory complexity to improve competition in the domestic market, and moving toward a rules- based system by implementing judicial reform to increase the reliance o f economic transactions on contracts and their efficient enforcement (see Chapter 6). 62 2.44 This core agenda needs to be supported by macroeconomicpolicies that address risks and vulnerabilities and that make progress towards monetary integration (see Chapter 7). Macroeconomic stability is apre-condition for sustaining a highgrowth path. As a result o f implementing prudent macroeconomic policies and reducing debt levels, Bulgaria has strengthened its macro-financial position. However, there are a number o f risks and vulnerabilities, and new macroeconomic challenges are emerging. Inthe short run, these risks and challenges include the rapid growth o f credit to the private sector and the resulting rapid growth o f debt by the private sector, and the widening o fthe external current account deficit and the larger role o f short term capital flows infinancing this deficit. The economy is vulnerable to external shocks including further oil price and interest rate increases. There are also risks related to the adjustment from an accommodating to a more neutral monetary stance in high income economies and their growth prospects. As shown in Chapter 3, weak growth performance in the EU has a significant economic and financial impact on the economy o f Bulgaria. The rapid growth o f expenditures in the social protection system and the rapidly aging population represent a threat to budgetary stability and fiscal sustainability inthe medium to long term. The preparation for and eventual adoption o f the euro will increase the need for enhanced macro-financial discipline, given the relatively large asymmetries between the economy o f Bulgaria and the economies o f the Eurozone. 2.45 The urgency of this agenda is greater than may appear given the comprehensive reformprogramimplementedthus far, the strengthenedmacro-financialposition, and the solid growth outcomes achieved. However, as shown in this chapter, for Bulgaria to move towards a Lisbon-growth-path, its core reform agenda needs to be anchored in deepening its trade and investment integration, since rising productivity gains are concentrated in outward- oriented activities. As discussed in the next chapter, Bulgaria's trade performance reveals that there are a number o f challenges involved indeepening trade and investment integration. 63 3. BUILDINGON TRADE INTEGRATIONAND INVESTMENT INTEGRATION 3.1 Despite positive developments in expanding trade and investment relative to GDP, Bulgaria has yet to benefit fully from its integration into the EU and Pan-European markets. In large part thanks to the EU accession process, Bulgaria has implemented a broad package o f structural reforms, with the acquis communautaire providing templates, and has dealt successfully with macroeconomic stabilization following the crisis in 1996 and 1997. As a result, Bulgaria has an open economy that has become integrated into global markets, although it has yet to develop an export capacity that would move the country's specialization beyond unskilled labor intensive products. These products still dominate Bulgaria's exports. With the exception o f textiles and clothing, Bulgarian firms-in marked contrast to those in most o f the NMS-8-have not become part o f global value chains driven by international outsourcing despite some recent progress. Furthermore, the low level o f imports o f capital goods indicates that establishing more sophisticated export capabilities has been proceeding slowly. W h i l e these trends overall an improving capacity o f the Bulgarian economy to tap benefits from trade integration into Pan-European and global markets, this capacity remains limited in terms o f adjusting to changes inmarket conditions. This chapter first provides an overview o f Bulgaria's trade policy framework and then examines recent trends in its trade performance and seeks to explain its weaknesses and strengths. The chapter ends with a reform agenda for deepening trade and investment integration. A. Trade policy 3.2 The impact of EU accession goes beyond the trade liberalization of the Bulgarian economy. The severe crisis o f 1996-97 and the prospects for EUmembership in 1996 when the European Commission (EC) invited five Central European countries to start the EU accession process changed the political economy in Bulgaria in favor o f market-oriented reforms. Bulgaria was invited to begin negotiations for EU accession in December o f 1999, once the economy was largely stabilized and growth was re-established. Since then, the reforms in Bulgaria have been anchored to EU accession. As a result, Bulgaria has implemented a broad reform program and economic performance has improved. In its Regular Report o f 2002, the EC has assessed Bulgaria as a functioning market economy. By June 15, 2004, Bulgaria provisionally closed all 31 chapters o f the acquis communautaire. The Accession Treaty was signed inApril 2005. Bulgaria expects to join the EUinJanuary 2007. 3.3 Thanks to the EU accession process, Bulgaria has become part of one of the world's largest single free trade area for industrial products. The EU associate status, or more exactly the EU Eastern Enlargement project, has driven Bulgaria's foreign trade policy. It encouraged bilateral trade liberalization initially with the EUand EFTA and, subsequently, with other European preferential partners o f the EU (Le., signatories o f the Pan-EuropeanAgreement on the Cumulation o f the Rules o f Origin" and participants o f Stability Pact for Southeastern " Theconceptofpan-EuropeancumulationintroducesdiagonalcumulationacrossaEurope-widefreetradearea encompassing the EU-25, EFTA, Bulgaria, Romania and Turkey. Diagonal cumulation allows materials originating inany o fparticipants to count towards fulfillment o fthe rules oforigin. Europe). The Pan-European Agreement combined with the removal o f tariffs on all industrial products and the harmonization o f technical standards among its participants has led to the emergence o f a de facto single free trade area for industrial products. In consequence, Bulgarian producers face intensive competition from imports and also have preferential duty- free access to imported inputs, both o fwhich contribute to their competitiveness. Pan-European partners account for around three-quarters o f Bulgaria's industrial trade. 3.4 Since multilateral liberalization has not gone hand-in-hand with vigorously pursued regional liberalization, the extent of reverse discrimination has increased significantly.' Together with Romania, Bulgaria has the most protectionist MFNtariff schedule in the region (Table 3.1). While the coverage o f WTO bound rates i s 100 percent, Bulgaria has set bound rates at a much higher level than it applies. Hence, Bulgaria has kept its options open to increase MFNtariffs up to the levels set by the bound rates. In fact, the applied rates are lower than bound rates but significantly higher than in Croatia, Turkey and the EU. Both simple and weighted average MFN applied tariff rates on total goods in Bulgaria are significantly larger than those in the EU, Croatia, and Turkey. The average MFNtariff on total goods in Bulgaria o f 11.5 percent i s twice as high as in the tariff rate o f 4.2 percent in the EU. The difference is greater on agricultural than on industrial goods. Bulgaria's tariff rate on agricultural goods is 22.9 percent compared to 7.6 percent inthe EU. Inthe case o f industrial goods, the tariff rate in Bulgaria i s 8.6 percent compared to 3.7 percent in the EU. Tariff rates in all categories in Croatia and Turkey are closer to the EUlevels than are those o f Bulgaria. Table 3.1 Average Applied Tariffs and BoundRates: Bulgaria and RegionalPartners (percent) Average ADDlied MFNTariff Rate Total Goods Agricultural Industrial Goods Bound Rate of Binding Goods All goods Reporter Year Simple Weighted Simple Weighted Simple Weighted Simple Weighted Coverage Average Average Average Average Average Average Average Average in percent Bulgaria 2004 11.5 2.8 22.9 16.2 8.6 2.5 24.2 20.0 100 Croatia 2002 6.0 4.7 10.9 11.3 5.6 4.2 5.8 4.8 100 Romania 2001 11.6 8.3 24.8 30.7 10.6 6.3 40.4 40.3 100 Memorandum: European Union (15) 2004 4.2 3.0 7.6 4.5 3.7 2.3 3.9 3.0 100 Turkey 2003 5.2 4.0 20.7 12.3 3.7 2.3 28.6 Note: Agricultural goods are basedon the WTO classificationofHS 01-24and industrial goods are HS 25-96.* 19.8 50 Indicatesnon- WTO member as of August 2003. Source: World Bank staff based on UNCTAD TRAINS database and WTO boundtariff rates from WTO IDBdatabase. 3.5 High reverse discrimination is redundant and is a bad trade policy. Such reverse discrimination i s redundant as Pan-European suppliers already have duty-free access, and government resorts to tariff exemptions if these products are not available in Pan-European '* Reverse discrimination occurs when differences between MFNtariff rates inthe home country and in the trading partners' domestic market, ceteris paribus, result in placing producers in the home country at a disadvantage compared to producers in the preferential trading partners' domestic markets. The result is similar to providing an import subsidy on the final product imported by the home country since the MFNtariff rate o n the imported input in the home country i s substantially larger than the MFN tariff rate on the same imported input faced by producers in the preferential trading partners' domestic markets. Hence, these differences place producers of the final product in the home country at a disadvantage compared to their peers inthe trading partners' domestic markets. 65 countries. But if they are available, high MFNtariffs merely protect F T A suppliers from MFN competition, which leads to rents going to Pan-European suppliers. The overall result i s that domestic producers remain exposed to competition from imports, and the government i s deprivedo f customs duty revenues. Bulgaria should thus converge to the EUCommon External Tariff on industrial products evenbefore accession. B. Trade performance 3.6 Despite an unfavorable external environment, Bulgaria's trade expanded rapidly in 2000-03 (see Figure 3.1). The growth cycle discussed in the previous chapter follows closely the trade developments in Bulgaria. After the 1996-97 crisis, the economy was largely stabilized and growth was re-established by 1999. As shown in the previous chapter, trade has been expanding at a rapid rate since then (see Figure 3.1;A). The value o f the imports o f goods and services in euro terms in 2004 was 218 percent higher than in 1999. And the exports o f goods and services were not far behind, at 209 percent in2004 relative to 1999. 3.7 As the Bulgarian economy was turning towards a vigorous recovery in 1999, and Bulgarian exports began a slow recovery in 1999-2000, the world economy entered a downturn in 2001. Growth rates inhighincome economies fell sharply from above 3.5 percent in2000 to less than 1percent in2001. The downturn in the world economy in2001 and the sluggish recovery since then have had a negative impact on the export performance o f most CEEC. Figure 3.1; D shows the sizable decline in exports o f goods o f Hungary, which declined between 2000 and 2001 by 1.9 billion euros or about 8.7 percent and the slow growth in the world economy; inRomania the decline was by 1.1billion euro or about 6.1 percent during the same period. However, the magnitude o f decline was higher in Bulgaria-about 10 percent (Figure 3.1: D). The sluggish recovery o f the world economy continued to have an impact in Hungary and Bulgaria, as both economies experienced further declines intheir exports o f goods in 2002 and started to recover only in 2003. Romania's recovery, however, started one year earlier. 3.8 Trade performance has been increasingly shaping Bulgaria's GDP growth performance. The links between GDP growth and the ratio o f exports to imports have become stronger, with GDP growth responding to trade developments with a lag o f one year (Figure 3.1: B and C). GDP growth is highlycorrelated with changes inthe ratio o f exports to imports with one lagged period. The simple correlation coefficient between GDP growth and the ratio o f exports to imports is 0.72, with GDP growth responding to developments in exports to imports one year later. This implies that increasing the adaptability o f the Bulgarian economy to adjust to changes in external market conditions is becoming critical to growth performance. Specifically, reforms o f the policy and institutional framework need to improve the ability o f firms located inBulgaria to adjust production levels, cut costs, or increase productivity growth, as well as to exit sunset industries and enter into activities that will take advantage o f new opportunities emerging in the external markets. The centrality o f these reforms i s well illustrated by the path of export adjustment o f Bulgaria relative to other countries in the region to the downturn inthe world economy in2001 and the slow recovery since then. 66 A. Imports and exports of goods and services B. Growth of GDP, imports, and exports (Dercentl 14000 -1 120 25 i 1 6 2000 118 20 0000 116 8000 114 9 15 112 $ 6000 110 2 10 4000 108 2000 106 5 1 0 104 0 1999 2000 2001 2002 2003 )&I9 2000 2001 2002 2003 20041 rzzaImportsG&S nExportsG&S -5 0 0Eportgrowth EZZZiImportgrowth -Ratio of Exports to Iworts +GW growth RHS C. Correlation GDP growth (at t-1)and ratio D.Exports of goods (Mil.Euro) of exports to imports (at t)=0.72 120 25,000 , 118 116 20,000 114 15,000 112 110 10,000 108 5,000 106 0 104 1999 2000 2001 2002 2003 1999 2000 2001 2002 2003 -Ratio of Exports to lh-ports -GDP growth RHS 0Hungary -Romania +Bulgaria Source: World Bank staff estimates based on NSI, BNB. 3.9 Economic growth and the ratio of exports to imports exhibit relatively large fluctuations around their average rate of expansion (see Figure 3.1: B). One important implication o f output and trade volatility i s that output and trade perfonnance has been lower compared to strong performers in the region, such as Estonia, whose economies have sufficient flexibility and adaptability to smooth out adjustments to changes in market conditions. These fluctuations are partly explained by external developments but for the most part are the result of the limitedadaptability o fthe Bulgarian economy to changes inexternal market conditions. 67 3.10 Bulgaria's trade integration into EU and Pan-European markets has proceeded at fast pace (see Figure3.2). Duringthe period 2000-03, Bulgaria expanded its trade with the EU and the Pan-European markets, with imports expanding more rapidly than exports o f goods. The deceleration of growth in the EU-15 in 2000-02 had a significant impact on the slowing down of exports to these markets, while imports from these markets kept up a rapid pace of growth (see Figure 3.2: A and B). Similarly, the recovery of growth inthe EUin2003 provided new momentum to the expansion of exports to the EU and Pan-European markets and accelerated the growth o f imports from those markets. Figure 3.2 Geographical Orientation of Export5 IndImports, 2000-03 IA. ExDorts of goods B. ImDorts of goods 9000, c 5: 4500 E 0 3 = 15000 -f I 1500 2000 2001 2002 2003 2001 2002 2003 -__ _-_____ ____ _ - . -- 2000__ ~ _ ~ _ _ _ _ 1 -World &Pan-European +W15 ~ -World -Pan-European + N 1 5 I-- __- ~_ C. ExDorts of goods: relative change index D. PmDorts of goods: relative change index 50 40 40 U CI C C 30 30 n $ na 20 20 10 10 0 EU-15 W 1 5 NMS-8 EFTA SE-5 CIS (11) I ____0 NN6-8 EFTA SEE-5 US(11) I - 2000 2003 ia 2000 0 2003 I Source: World Bank estimates based on UNCOMTRADE c abase as reportedby Bulgaria. 3.11 The rapid expansion of preferential trade in products with the EU and Pan- European markets indicates that competitive pressures on domestic productive structure 68 have increased. With the Bulgarian economy growing twice as fast as the EU and undergoing major economic restructuring, the share o f imports from EU and Pan-European markets increased more than that o f exports o f goods to these markets (Figure 3.2: A and B). The share o f imports o f goods from EU-15 and Pan-European markets increased from 44 and 58 percent each in 2000 to 50 and 65 percent, respectively, in 2003 (Figure 3.2: D). Not surprisingly, the expansion was mainly due to industrial imports-the share o f Pan-European suppliers rose from 58 percent in 2000 to 77 percent in 2003. In contrast, the share o f exports o f goods to the EU and Pan-European markets expanded from 51and 67 percent each in2000 to 57 and 73 percent, respectively, in 2003. Two observations can be made. One i s that deeper integration with EU markets implies significant competitive pressures on producers located in Bulgaria. These competitive pressures are contributing to increasing the productivity gains o f outward oriented activities. Indeed, as shown in Chapter 2, labor productivity and the productivity o f investment in outward-oriented activities are growing faster than in inward-oriented activities. The other observation i s that the deepening trade integration into Pan-European and global markets is the key determinant for expanding productivity gains more broadly across Bulgaria's economy. 3.12 There has been significant geographical reorientation in Bulgaria's merchandise trade. The dynamics of the geographical reorientation o f Bulgaria's merchandise trade are illustrated in Figure 3.2: C and D by the shares in total exports and imports o f the goods of trading partners in 2003 compared to 2000. Only exports to the EU and NMS-8 markets expanded in 2003 relative to 2000 while shares o f exports to EFTA, SEE-5, and CIS-11 declined. The largest expansion o f exports was to EUmarkets, with the share growing from 51 to 57 percent between2000 and 2003. By 2004, exports to the EU-15 represented 54 percent of total exports and the share o f exports to Pan-European markets was 61 percent. 3.13 However, as shown in Chapter 1, Bulgaria has a relatively limited export capacity compared to the NMS-8. During 1999-2003, exports o f goods from Bulgaria were 37 percent o f GDP compared to the average for the NMS-8 of 46 percent of GDP. Two observations can be made. One i s that exporters located in Bulgaria have yet to develop and export capacity in the highly competitive markets o f Europe. The other i s that Bulgaria needs to maintain the reform momentum in order to catch up with its peers in the NMS-8 to compete in the EU-15 market. 3.14 In contrast to export geographical dynamics, imports from markets other than the EU-15 expanded except from the SEE-5 and CIS-11. Incontrast to exports o f goods, imports are more diversified across trading partners. The largest expansionwas inimports from the EU- 15, with an increase o f 6 percentage points between 2000 and 2003. The shares o f imports from the NMS-8 and EFTA increased by 1 percentage point each. The share o f imports from the CIS-11, where Russia remains the most important trading partner, intotal imports declined to 19 percent in2003 compared to 30 percent in2000. 3.15 Unskilled labor intensive products continue to be the levers of Bulgaria's export performance, with Bulgaria's revealed comparative advantage strongly anchored in unskilled labor and natural resource intensive activities. These two groups accounted for 66 percent o f total exports in 2003, up from 62 percent in 2000. The increase in their aggregate share was exclusively due to unskilled labor intensive exports whose share increased from 27 to 69 32 percent. The share o f other groups declined, with capital-intensive products registering the largest contraction. 3.16 The dominance of unskilled labor intensive products is even more pronounced in Bulgaria's exports to the EU-15, accounting for 40 percent of EU-destined exports-up from 36 percent in 2000 (Figure 3.3: A). Together with natural resource intensive products, theyjointly accounted for 71 percent of these exports in2003-up from 70 percent in 2000. In contrast, the aggregate share o f skilled labor and capital intensive products contracted slightly from 30 percent in 2000 to 29 percent in 2003. Hence, industrial restructuring has yet to produce shifts in specialization towards skilled labor and capital intensive products similar to those already experienced by the NMS-8 economies. A. Factorintensityof exportsto EU-15 B. Comnositionof total imnorts I4O0 I 70 02000 132001 02002 02003 t.-=50 E s 40 c 's 30 * c 0 20 n 10 0 2000 2001 2002 2003 Agricultural lndustnalRaw Machinery, Automobiles Oiher Food & Feeds Materials n Labor exc &Paris Consumer __ -Ca~ital izzzmNaturalIntensive Resource -Skilled Unskilled Labor Automobiles Goods Source: World Bank staff estimates based onUNCOMTRADE database as reportedby Bulgaria. 3.17 The compositionof importsindicateslimitedactivityin expandingindustrialexport capacities (see Figure3.3: B). While imports o f final manufactured consumer goods strongly expanded between 2000 and 2004, the share o f machinery (excluding automobiles and parts) has been stuck at around 20 percent o f total imports from the EU since 2001. Imports o f capital goods may be contributing to developing more productive activities, but, as shown inChapter 2, these seem to be at an early stage o f development and have not reached a critical mass so as to be observed inproductivity measures at aggregate levels. To a large extent, the composition o f imports has yet to contribute to the building up o f an export capacity beyond unskilled labor intensive exports. These trends reflect the fact that Bulgaria has yet to benefit more fully from its integration with the EUand global markets. 3.18 The limited progressin enteringnew forms of the global division of labor based on the outsourcing of manufacturing activities provides another illustration of untapped benefits offered by integration into Pan-European markets. Participation in international networks-in both traditional (clothing and footwear) and more advanced (automotive, Information Communications Technology [ICT] and electro-engineering parts)-brings outside managerial and technological expertise to a local company and offers a "cheap way" to market 70 products, as firms do not incur marketing costs. Last, but not least, since parent companies sell their products in many different markets, the suppliers are no longer dependent on the vicissitudes o f import demand in any single market. While Bulgarian firms have become parts of traditional global value chains, with textiles and clothing and footwear accounting for 35 percent o f EU-destined exports, their involvement in more sophisticated supply chains remains limited. The share of exports o f automotive network products has remained below 1 percent. But intwo other networks there are signso fchange, albeit from avery modest base: the share o f I C T network products rose from 0.8 percent o f EU-destined exports in 2000 to 1.3 percent in 2003 and that o f engineering parts rose from 3.5 percent to 5.4 percent over the same period. Figure 3.4 Bulgaria: Indicators of Competitiveness,2000-04 CENTRAL EUROPEANREALEFFECTIVE EXCHANGERATES (IMF; 1997 = 100) 160 1160 140 140 120 120 - 100 100 80 Bulgaria - - - - - --Croatia Hungary -Romania &*rm =Poland 60 2000 2001 2002 2003 2004 UNIT LABOR COSTS (January2000= 100) SHARE INEUIMPORTSOF SELECTED COUNTRIES (December 1998= 100; 12-monthrolling average) I / 130 1 130 180 - 180 120 120 -Bulgaria 160 .......Czech Republic 110 110 Hungary 100 100 140 90 90 120 120 80 80 100 -100 701 -In Euros 80 60 2000 2001 2002 2003 I/Bulgaria, CzechRepublic,andHungary:Authorities' d a;andRomania:DirectionofTradeStatistics. Source: IMF Staff Report on Bulgaria, August 2004. 71 C. Competitiveness,trade, and FDI dynamics 3.19 Assessing and maintainingcompetitiveness under a Currency Board Arrangement (CBA) is central to policymakersin Bulgaria. According to the European Competitiveness Report o f 2003,19 Bulgaria shows low productivity levels and competes on the basis o f lower labor costs. Indeed, as discussed above, Bulgaria has a strong revealed comparative advantage in labor intensive products, albeit only in unskilled labor intensive products. These findings prompt concerns about the risk that Bulgaria may be locked into low labor cost, low productivity specialization patterns. However, as discussed in the previous chapter, Bulgaria i s undergoing significant economic transformation and i s following a growth cycle dominated by TFP while rebuilding its capital stock. It is therefore important to further the trends in the fundamentals o f Bulgaria's competitiveness. 3.20 Competitivenesshas been definedin several different ways. The EC has defined it as "the ability o f an economy to provide its people with high and rising standards o f living and highrates o f employment on a sustainable level." The OECD has defined it as "the degree to which a country can, under free trade and fair market conditions, produce goods and services which meet the test o f international markets, while simultaneously maintaining and expanding the real incomes o f its people over the long term." And the World Economic Forum defines competitiveness as "a country's ability to maintain high rates o f growth and employment inthe medium term." Analytically, in the short run, an appropriate level o f competitiveness i s associated with the value o f the exchange rate, given certain domestic policies, that ensures both sustainable internal and external balances. In this sense, exchange rate misalignment and competitiveness problems are interchangeable. In the short run, improvements can be made in competitiveness by reducing costs. However, in the medium term competitiveness is determined by the trend o f productivity growth. Hence, competitiveness is determined by the ability of firms to increase productivity. 2o 3.21 The CBA combined with the euro appreciation against the U S dollar has compounded a price bias in favor of nontradableswhich has not been compensated by a more adaptable economy that reallocates resources from lower to higher productivity activities. Based on recent estimates o f the IMF, the real effective exchange rate based on the CPI has continued to appreciate. Hence, there has been some loss o f competitiveness compared to several CEEC countries. However, this trend i s tempered by declining unit labor costs in manufacturing and an increasing share in EU imports (see Figure 3.4). While wages and salaries are grossly under-reported, there i s no evidence that actual unit labor costs are not declining. However, as shown inChapter 2, investment and economic restructuring show a bias infavor ofnontradables. Furthermore, as discussed inthe previous chapter, no more than half o f the accumulated real exchange appreciation o f the lev to the euro can be attributed to the l 9 Commission Staff Working Document, 2003, European Competitiveness Report 2003, European Commission, *'Indicators SEC (2003)1299. o f competitivenessrange from relative price and cost indicators, and development o f trade market shares-produced bynational statistical offices, OECD, and the IMF-to survey indicators produced by the World Economic Forum and the Lausanne Institute for Management Development. Survey indicators take into account a wide range o f characteristics o f countries and countries' competitiveness is rated relative to other countries. 72 productivity differentials o f tradables to nontradables. The other half i s related to the combined effects o f important impediments to resource reallocation towards more productive activities. These impediments are mainly deficiencies in basic infrastructure (discussed in Chapter 4), limited labor market adjustment (discussed in Chapter 5), limited competition in the domestic markets, and the remaining deficiencies in the institutional quality (discussed in Chapter 6), which contribute to'a price bias in favor o f nontradables. In this context, FDI flows have a central role indeepening trade integration. Figure3.5 StructuralReformsandFDI:Bulgariaand SelectedCEEC A. Compositionof FDIinBulgaria B. StructuralreformsandFDIpercapita - FDI in Bulgaria ( % o f GDP) T6o 8 53 7 6 40 5 30 4 3 20 2 x) 1 0 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Privatizationflows INon-privatizationflows 80 70 60 50 40 30 20 10 0 1998 1999 2000 2001 2002 2003 -1 00 -050 0 00 0 50 100 I Governance Index Outw ard-oriented o/w manufacturing Inw ard-oriented o/w financial sector ~~ _____ Source: World Bank staff estimates based on NSI, BN 3BRD transition indicators database and World Bai governance indicators. 3.22 Followingthe progressin structuralreforms, foreign direct investmentinflows have increased rapidly and play a central role in building up the assets of the economy (see Figure 3.5). Stability and progress in structural reforms have contributed to a rapid and sustained increase in FDI flows to Bulgaria. These flows increased from less than 1percent o f GDP in 1995-96 to an average o f close to 6 percent in 1997-2004 (see Figure 3.5: A). On a cumulative basis, FDI per capita in Bulgaria has now surpassed the U S $500 mark, indicating 73 that new FDI-driveninvestments are reaching a critical mass needed in order to have a sizable impact on Bulgaria's improved capacity to compete in external markets. However, the gap vis- d-vis strong trade performers among the N M S remains large-cumulative FDI per capita in Hungary i s about U S $3,700 and i s close to U S $3,800 in the Czech Republic (see Figure 3.5 B). To improve competitiveness, Bulgaria needs to implement reforms to continue to attract FDI flows. Among these reforms, improving the policy and institutional framework is central. A simple correlation between governance and FDI flows among the CEEC shows that improvements in governance have a non-linear correlation with FDI performance (see Figure 3.5: C). This illustrates the large potential benefits related to improving the quality o f institutions. While the overall governance index o f Bulgaria improved from -0.15 in 1996 to 0.26 in 2002, as illustrated in Figure 3.5: C, substantial progress is needed to provide an environment on a par with the NMS-8. The institutional and regulatory environment is discussed inChapter 6 o f this report. 3.23 Thus far, a large share of FDIflows to Bulgariais in inwardoriented sectors. Figure 3.5: D shows the large share o f FDI flows concentrated in inward-oriented activities. During 1998-2003, about 60 percent o f FDI flows were to inward-oriented activities. To some extent this i s explained by the fact that some of the best opportunities inBulgaria for foreign investors have beenin sectors that primarilythe domestic market, such as banking, business services, and electricity distribution. This has resulted in a strong inward orientation o f FDI. However, it also reflects impediments to outward-oriented investment. As discussed in the previous chapter, costly inefficiencies inBulgaria's economy can be transferred to domestic consumers via higher prices but not to consumers in external markets where there i s stronger competition. This underscores the centrality of upgrading basic infrastructure, reducing regulatory complexity, and developing a reliable rules-based institutional framework. D.Buildingon trade andinvestmentintegration:a reformagenda 3.24 Deepeningtrade integrationwith the EUand globalmarketsis a key determinantin expandingproductivitygains more broadly across Bulgaria's economy. The expansion o f trade relative to GDP i s exerting competitive pressures that contribute to increases in productivity in outward-oriented activities. However, investment in the outward-oriented sectors i s low. As a result, export capacity i s improving very slowly and exports are dominated by unskilled labor intensive products. To improve competitiveness, Bulgaria needs to implement a core reform agenda, which includes upgrading basic infrastructure and skills, enhancing labor market adjustment and improving the regulatory and institutional frameworks. 3.25 These reforms are needed to address the challenges of the present trends in trade performance and the composition of FDI flows. As a result of the process toward EU accession, Bulgaria has relatively liberal trade policies and exporters located in Bulgaria have easy access to global markets. Despite this supportive trade policy framework, the building of an export capacity beyond unskilled labor intensive and resource intensive commodities is progressing slowly. Moreover, the trends inthe composition o f imports indicate that these have yet to contribute to building up export capacity, given the relatively small share o f imports o f machinery and equipment in total imports. These factors are an indication that government 74 policies are not supportive o f competitive markets that would reallocate resources from sectors with low productivity to industrial sectors with apotential comparative advantage. 3.26 The sustainabilityof the trends in trade performanceis questionable. First, Bulgaria has a very large gap between imports and exports. Moreover, the fact that EU quantitative restrictions on textiles and clothing imports under the WTO Agreement on Textiles and Clothing ended on January 1, 2005 represents a significant challenge to the sustainability o f export growth since these exports are dominated by unskilled labor intensive exports. In addition, while in most o f the NMS-8, FDI inflows have been.closing the gap between endowments in skilled labor and the dominance o f unskilled labor-intensive products in their exports to EUmarkets, this however, i s not yet the case with Bulgaria. 3.27 Reforms need to be implemented in a highly coordinated manner. Without a reduction in the regulatory complexity and improvements in the quality o f the institutional framework, including trade facilitation measures, FDI flows are unlikely to move to outward- oriented sectors in Bulgaria. Similarly, without increased labor adjustment, the market mechanism cannot reallocate resources from low to higher productivity activities in the economy. Moreover, without reorienting public expenditures towards upgrading Bulgaria's skills and its transport network so as to facilitate trade, the competitiveness o f the country's producers will continue to depend on the availability of low-wage unskilled labor. Consequently, competing with suppliers from developing countries will become increasingly difficult. Reforms that would address these deficiencies in the policy framework are strong complements o f each other. Thus, their effectiveness in improving the competitiveness of producers inBulgaria depends on a highly coordinated approach to their implementation. 3.28 The key areas of reforms aimed at deepening trade and investment integration includethe following: Upgrading the transport network and tradefacilitation system to reduce the costs of competing in external markets (see Chapter 4) Upgrading skills to enable the development of export capacity beyond unskilled labor intensive exports (see Chapter 4) Increasing labor marketflexibility tofacilitate the reallocation of resourcesfrom low to higher productivity activities in the economy (see Chapter 5) Reducing regulatory complexity and improving the quality of institutions to attract FDIflows to outward-oriented sectors (see Chapter 6) Maintaining macroeconomic stability to provide an environment supportive of the private sector undertaking investments needed to compete in the global markets (see Chapter 7). 3.29 To upgrade its transport network and trade facilitation system and skills of the labor force, Bulgaria will need to restructure public expenditures and reorient them 75 towards these two high priority areas. As shown in the next chapter, for the restructuring o f public expenditures to be effective these efforts need to be supported by public sector reform to improve efficiency and transparency o fpublic administration. 76 4. RESTRUCTURINGPUBLICEXPENDITURES 4.1 The public expenditure policy reform underpinningfiscal adjustment is central to Bulgaria's productivityand growth prospects. With 40 percent o f GDP in the command o f the public sector, improving the efficiency and effectiveness o f public expenditure policies is a central consideration ifBulgaria's growth prospects are to be realized. First, inefficiencies in public expenditures are particularly detrimental to productivity and growth prospects, since they affect efficiency across a wide spectrum o f activities in the economy. Second, given such a large share o f public expenditures in the economy, improvements in efficiency in the private will not compensate for the costs associated with inefficient public expenditures. The strategy o f having the public sector focus on the efficient delivery o f public goods, including an efficient and effective social safety net, while the private sector leads in the expansion o f investment, output, and employment, has served Bulgaria well. Bulgaria should remain on this path and should accelerate progress towards this end if economic performance is to continue do improve. The size and the role of public finances are at the center o f the reforms, specifically, the overall economic efficiency with which revenues are collected and the cost-benefits o f budgetary allocations (inputs) and their outcomes (results). This would include the efficiency and sustainability o fthe social protection system aimed at preserving social cohesion. 4.2 Bulgaria needs to build on its fiscal achievementsto date to increase the efficiency and effectivenessof public expenditures. As shown in the previous chapters, Bulgaria faces an investment challenge, particularly as regards improving basic public infrastructure, and the challenge o f improving competitiveness, particularly as regards expanding its export capacity beyond unskilled labor intensive activities. Public expenditure policy reform i s at the center o f upgrading skills, upgrading basic public infrastructure, and laying a solid foundation for sustainable reductions in the tax burden on the economy. Similarly, tax policy reform has a significant impact on resource allocation and efficiency. As shown in the next chapter, implementingreforms so as to enable sustainable reductions inpayrolltaxes, together with labor market reform, can play a central role in enhancing labor adjustment and employment. Furthermore, improvements are needed inkey dimensions o f public sector reform. This chapter first provides an overview o f Bulgaria's fiscal adjustment over the last eight years. The second section discusses public expenditure policy reform focusing on basic infrastructure and education. The next section examines the key areas o f public sector reform needed to support the effectiveness o f public expenditures, including public expenditure management, public administration, public governance and financial accountability. The chapter ends with a summary o f a reform agenda. A. Fiscal adjustment 4.3 Although Bulgaria has strengthenedits fiscal position the last few years, challenges still remain. Figure4.1 shows the annual average overall fiscal balance relative to GDP for the periods 2000-02 and 2000-04. Bulgaria's overall fiscal balance compares favorably with the NMS-8. Both Bulgaria and Estonia registered overall fiscal surpluses in 2004. Latvia and Lithuania have overall fiscal deficits o f around 2 percent o f GDP, followed by Romania with a deficit o f around 3 percent o f GDP. Other CEEC countries face overall fiscal deficits larger than 5 percent of GDP. As a result, considerable fiscal adjustment will be necessary in countries with relatively large fiscal deficits, with potentially dampening effects on growth. As discussed in the previous chapters, large fluctuations in macroeconomic imbalances carry a heavy toll on growth in particular and on economic performance in general. This experience underscores the importance to Bulgaria o f maintaining a prudent fiscal stance while improving the quality o f fiscal adjustment. Public expenditure policy reform supported by public sector reform is central to reorienting expenditures towards high priority areas needed to deepen trade integrationwith the EUand to use large EUfunds efficiently while preservingstability. Figure4.1 AverageAnnual, OverallFiscalBalanceto GDP of Bulgariaand SelectedCountries 2000-02 and 2000-04 (in percent) Slovenia fi Slomk Rep Romania Poland 2000-2004 Lithuania 0 2000-2002 Latvia Hungary Estonia Czech Rep Croatia n Bulgaria -7.0 -5.0 -3.0 -1.o 1.o Source: WorldBank staffestimatesbasedon Eurostatand MOF. 4.4 Following the 1996-97 crisis, fiscal stabilization was achieved by revenue expansion rather than the rationalization of expenditures (see Table 4.1). Public revenues increased from 33 percent o f GDP in 1996 to over 41 percent in 2000. Public expenditures to GDP declined sharply from 43 to 38 percent between 1996 and 1998,but later increasedto about their 1996 level by 2000. Inthe last four years, however, public expenditures to GDP declined from 42 percent in2000 to 40 percent in2004 largely because o f containment inthe growth o f public expenditures relative to GDP growth and to the some limited rationalization o f public expenditures. 4.5 Over the last four years Bulgaria made progress towards tax policy reform while safeguarding stability and growth. In the last four years several tax rates were reduced consistent with the strategy o f reducing reliance on direct taxation and expanding the role o f indirect taxation in tax revenues (see Table 4.2). Tax rates were reduced to become more supportive o f growth and investment. 78 Table 4.1 Bulgaria:Summary of GeneralGovernment Operations,"/1996-2004 (% of GDP) 1996 1997 1998 1999 2000 2001 2002 2003 2004p Total revenue andgrants 32.9 37.5 39.7 40.7 41.4 39.8 38.7 40.7 41.7 Tax revenues 26.8 28.9 31.6 31.4 32.5 30.9 29.7 32.2 33.6 Non-taxrevenues 6.1 8.0 7.5 8.4 8.0 7.6 8.1 7.7 6.9 Grants 0.0 0.6 0.6 0.9 0.8 1.2 0.9 0.9 1.2 Total expenditures 43.2 38.6 38.4 40.5 42.0 40.4 39.4 40.7 40.0 Current expenditures 41.8 35.2 34.3 35.6 37.6 36.2 35.6 36.8 35.8 Capital expenditures 0.7 2.7 4.1 4.9 4.3 4.2 3.8 3.9 4.2 Primary balance 9.2 7.1 5.5 3.9 3.4 3.0 1.6 2.1 3.6 Overall fiscal balance -10.3 -1.2 1.3 0.2 -0.6 -0.6 -0.6 0.0 1.8 Financing 10.3 1.2 -1.3 -0.2 0.6 0.6 0.6 0.0 -1.7 External -2.8 0.3 -0.6 1.2 -1.5 -0.3 1.5 0.0 -1.6 Domestic 13.2 -2.3 -2.2 -3.5 0.9 0.3 -1.7 -0.5 -3.2 Privatization revenues 0.0 3.1 1.6 2.2 1.3 0.6 0.9 0.5 3.1 From 1998 onwards, follows the National Classification, cashbasis; 2004 are preliminary estimates Source: Ministry o f Finance, National Classification, cashbasis. Notes: a/ 1996-1997 data based on PEIR (2002). 4.6 While several tax rates have been reduced during the last four years, the reduction of the tax burden on the economy has stalled. Despite progress to date, the tax burden on the economy i s high at about 34 percent o f GDP. To a large extent the rise in tax collection from indirect taxes is due to the rapid rise o f imports. In addition, a minimum social insurance threshold matrix setting minimum wages by activities (17) and by professions (9) increased the number o f registeredcontracts in the first few months o f its introduction in 2003. Hence, the collection o f social security contributions rose sharply between 2002 and 2003. Their sustainability, however, is unclear, as employees may redefine their contracts seeking the lowest possible wages inthe minimumlevels o fthe threshold matrix. Table 4.2 Selected Tax Rates in Bulgaria, 1997-2005 (percent)' 1997 1998 1999 2000 2001 2002 2003 2004 2005 Corporate tax 40.2 37.0 34.3 32.5 28.0 23.5 23.5 19.5 15.0 Income tax: Lowest rate 20.0 20.0 20.0 20.0 20.0 18.0 15.0 12.0 10.0 Highest rate 40.0 40.0 40.0 40.0 38.0 29.0 29.0 29.0 24.0 Socialinsurance contributions (3rd category ofjobs) 42.0 41.5bl 45.7~1 45.7 42.7 42.7 42.7 42.7 42.2 VAT 22.0 22.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 Tax revenues/GDP 28.9 31.6 31.4 32.5 30.9 29.7 32.2 33.6 Notes: ai Statutory tax rates for corporate taxation (combined rate for local and central budgets for 1997- 2002, single rate for central budget since 2003), personal income tax, social security contributions rates for both employee and employer (including pensions, health insurance, and unemployment), and VAT; bi Refers to July-December 1998. c/ Refers to July-December 1999;health insurance introduced since July 1, 1999. Source: Ministry o f Finance. 79 The size and trend o f the informal economy reflect, to large extent, the performance o f the policy and institutional frameworks. There are many factors that create incentives for the development o f an informal economy, including an excessive and cumbersome regulatory framework, weak or costly contract enforcement, widespread corruption, and tax policy. The focus here i s on the last. A l o w tax base and weak tax collection capacity are typically accompaniedby an excessive tax burden which creates substantial incentives for a large informal economy. At the same time, a large and growing informal economy undermines fiscal stability, as a number o f non-tax payers benefit from public services but do not share with tax payers the burden o f the costs. Annex E presents a model used to estimate the informal economy in Bulgaria. The key elements of the model are time series data on currency holdings and an error correction model. This approach i s based on the pioneering work o f Cagan (1958), which was further developed by Tanzi (1980, 1983). The underlying assumption i s that transactions relatedto the informal economy are undertakeninthe form o f cash payments in domestic currency. In turn, individuals and firms underreport economic activity to avoid taxation. T w o alternative specifications o f the opportunity cost o f money were used. Model 1 (continuous line inthe graph) used the weighted interest rate and Model 2 (dotted line inthe graph) used the Lev/US$ exchange rate. The results indicate that Model 2 provides a better statistical fit o f the data, and estimates o f the informal economy under this model are inline with other estimates for Bulgaria. Estimates obtained with Model 2 indicate that the size o f the informal economy i s declining, but at 32 percent o f GDP it remains large. These results are slightly lower than those recently obtained by Doing Business (2004) that estimates the informal economy at about 36.9 percent of GDP and i s consistent with earlier estimates from Enste and Schneider (2000). YO o f G D P 45 40 3 5 3 0 25 , e - - -=---- 15 10 1998 1999 2000 2001 2002 2003 2004 - - - . M o d e l l -Model 2 Information on tax compliance from the FIAS firm survey o f 2004 i s used to analyze the determinants and correlates o f informality at the firm level. The results indicate that (i) higher tax rates (ii) differential enforcement o f tax law for small and large firms and in different areas o f the country and (iii) systemic bribingpractices give firms an incentive not to comply with taxation law. Firm level analysis indicates that incorporated firms subject to a uniform 19.5 percent corporate tax comply substantially more with tax requirements. However, smaller firms and variances in tax policy and enforcement across regions are associated with substantially lower shares of their turnover to tax authorities. 80 4.7 To support a rapid increase in social security outlays, Bulgaria has one of the highestpayrolltaxes, before income tax, in the regionand comparedto the average in the EU (see Figure 4.2). Bulgaria's payroll tax, at 51 percent before income tax, is excessively high compared to the much lower rates in the Czech Republic (43 percent), Hungary (41 percent) or Poland (39 percent), or the EUaverage o f 37 percent. Figure4.2 Taxes Tax burdensto GDP 1999-2004(percent) Social ContributionRates 2003 (percent) a' % ofGDP 7 60 - ~ / * 50 40 l2t i 30 20 10 I 0 Bulgaria Czech Hungary Poland EU Republic Note: a/ simple averages o f social security, health care, and unemployment contribution rates; 43 percent is for 3rd category o fjobs in Bulgaria. - 4.8 Despite reductions in tax rates, deficiencies in tax policy--particularly in revenue administration-contribute to a sizable, albeit declining, informal economy, to tax evasion, and limiting employment creation. Research on the informal economy carried out for this report indicates that while the informal economy i s declining, its size remains high, largely because o f deficiencies in tax policy and revenue administration (see Box 4.1). There are six major pieces o f legislation pending approval inthe Parliament that are aimed at broadening the revenue base and improving its administration, to enable a sustainable reduction o f the tax burden. As a result, the operationalization o f the National Revenue Agency (NRA), a central piece inthe broadening o f the revenue base, has been delayed. B. Publicexpenditurepolicy reform 4.9 The recordof public expenditure restructuring,on the other hand, has been mixed. Indeed, there has been a large increase inprimary expenditures (Le., non-interest expenditures) compared to 1996. Interest payments were close to 20 percent o f GDP in 1996 and declined to under 2 percent o f GDP by 2004. While public expenditures to GDP declined by only 3 percentage points of GDP, from 43 percent in 1996 to 40 percent in 2004, most o f the savings from debt service were redirected towards social spending. As a result o f restructuring inpublic expenditures, the largest expansion has been in the social protection system and the fastest growing expenditures are on health care and medical assistance. Public expenditures on the 81 social protection system increased from 8 percent o f GDP in 1996 to more than 16 percent by 2004 and now account for about 39 percent o fthe budget. 4.10 Given the rapidly aging population, there is an urgent need to reform public expenditure policies in the social protectionsystem to eliminate inefficiencies,and in some cases abuse, and to makethem fiscally sustainable. The rapid expansion of social spending i s due to a combination of demographic trends and a rapidly aging population, high unemployment, the broad scope o f the social assistance system and increases in health care expenditures. However, the efficiency and effectiveness o f the social protection system has significant room for improvement-8 o f every 10 Bulgarians receive at least one social benefit, including pensions, but only 1 in 3 is poor or old. Pensions are the largest social protection expenditure component, including a variety o f contributory and non-contributory benefits. With one o f the fastest aging populations in the region, the financial pressures are large, are increasing rapidly, and have contributed to the very highpayroll taxes as shown in the previous section. At the same time, recent research2' shows that the social assistance system22i s largely ineffective in poverty alleviation. As a result, poverty alleviation in Bulgaria is heavily dependent on pensions while the social assistance system has a negligible impact. This implies that pensions, which are not means-tested and therefore not an efficient or effective instrument for poverty alleviation, are compensating for costly inefficiencies in the social assistance system. At about 5 percent o f GDP, the costs o f the social assistance system are too high for its negligibleimpact on poverty alleviation. 4.11 That beingsaid, the balancebetweencurrent andcapitalexpenditureshas startedto improve (see Figure 4.3). This i s an important first step towards addressing the legacy o f rundown infrastructure and public assets. However, there has been a tendency to cut public capital expenditures and supporting expenditures on operations and maintenance when fiscal tightening has been necessary. Public capital expenditure expanded rapidly from under 1 percent o f GDP in 1996 to 4.9 percent in 1999. Their share to GDP has followed a declining trend since 1999: public capital expenditures to GDP declined from 4.9 percent in 1999 to 3.8 percent in 2002. This trend, however, was reversed only in 2003 and the share to GDP has continued to increase since then with a projected allocation o f close to 5 percent o f GDP in 2005. 23 As result, private investment lacks an essential complement owing to deficiencies in basic public infrastructure, particularly in the transport network. While cutting public capital expenditures when fiscal tightening is needed i s common in many developing countries, this practice is particularly detrimental to investment and growth in the case o f Bulgaria because o f the poor state o f public infrastructure and public assets. Furthermore, for public capital expenditures to be effective in addressing these long-standing needs, capital outlays need to be supported by appropriate levels o f expenditures inoperations and maintenance. 21Tesliuc, Cornelia (2004), "Social Protectionand Poverty Reductionin Bulgaria- an Update," mimeo. 22 The Guaranteed MinimumIncome program (GMI) is the mainnational safety net andis complemented bythe energy benefit program and numerous other programs targeting specific groups. 23 2005 figures as per IMFprogram. 82 Figure4.3 GovernmentExpenditures,1999-2005 '1 8 1999 2000 2001 2002 2003 2004 2005p budgeted 0Maintenance andoperations 0Capitalexpenditure Source: World Bank staff estimates based on MOF data 4.12 Indeed, there is evidence that Bulgaria needs to upgradeits transport network and trade facilitationfacilities (see Figure4.4). Infrastructure data suggest that Bulgaria's growth rate is being constrained by low levels o f investment in basic public infi-astr~cture.~~ While more up-to-date information i s lacking with which to evaluate current circumstances, Figure 4.4 relate the level of real per capita GDP inBulgaria, Romania and the New Member States o f the EUto the stock o f roads, electricity capacity, and telecommunications development. The first panel shows the stock o f infrastructure in 1994, and the second shows the situation in2001.25 In the case o f electricity, Bulgaria has retained a relatively favorable position for meetingthe needs o f both households and producers. In fact, Bulgaria has the highest electricity per worker relative to GDP among the CEEC. The number o f main and cell lines are indicative that Bulgaria compares well with the other CEEC. However, between 1994 and 2001 the number o f new main and-cellphone lines increased faster inthe CEEC than inBulgaria. Recent reforms in telecommunications inBulgaria will help the country to catch up. Interms o f road systems, the country's relative position inrelation to the other CEEC has deteriorated (see Figure 4.4). The number o f kilometers per square kilometer of land is indicative that Bulgaria compares poorly with the average inthe CEEC and also inrelation to the size of its economy. 24The linkbetween public capital outlays and economic growth is not conclusive. Many studies have analyzed this relationship from several different angles and have reached different conclusions. For a review of this literature, see Ter-Minassian and Allen (2004). However, tests o f the relationship betweenphysical infrastructure and growth have been more robust inpointing towards a positive association. See, for example, Calderon and Serven (2004). 25The data are adjusted for country characteristics. A panelregression was used to "extract" the effect o f the size of the labor force, and land area, and level o f urbanization on each type o f infrastructure. The residuals from this regression constitute the adjusted values. The trend lines inthe charts are derived using OLS regression. 83 Figure4.4 SelectedIndicatorsof Basic Infrastructure. 1994 and2001 1994 2001 a V) 0.8 -1 I @ 1.4 ; j- e 0.6 1 LVA 1.3 1 a 5 5 0.4 - 1.2 - z C 0.2 i + 1.1 - SVN 0u 5 30 0.0 n -0.2 1I C .-J -0.4 ~ -0.6 1 0.8 L __ 8.2 8.6 9.0 9.4 8.4 8 8 9.2 9.6 10.0 ~ Real GDP Real GDP U lh ao 0.5 1 0.5- ROM -I EST 6 0.4 - 9 0.3- CZE '2U 0.2- SVK L V n 0.1- LTU SVN U ; m EST -1.0 -1 8.4 8.8 9.2 9.6 10.0 8.4 8.8 9.2 9.6 10.0 Real GDP Real GDP c5 0.2 HUN P BGR L 0.1 - CI CI 0 -01. a CZE ROM 8.8 9.2 9.6 10.0 Real GDP Real GDP Source: National Statistical Institute, Calderon and Serven (2004) database, and UnitedNations Energy Statistics Yearbook. 4.13 Indeed, the road network is ill equipped to support long distance international traffic and is in generalin poor condition (see Figure4.5). The road system in Bulgaria is not supportive o f long distance international traffic and i s in poor condition. The network consists mostly of roads other than motorways and Class Iroads. The percent in total 84 kilometers o f motorways i s less than 2 percent and the share o f Class Iroads is under 16percent o f the total. The share inkilometers o f others roads i s about 83 percent (see Figure 4.5: A). For the most part, the condition of the limitedkilometers o f motorways i s good, while the condition o f Class Iroads i s good in about half o f them. In contrast, the conditions are poor in about 68 percent o f the roads other than motorways and Class Iroads (see Figure 4.5: B). All together, the conditions are poor in close to two-thirds of the road network (see Figure 4.5: C). According to a recent World Bank field survey, bottlenecks are starting to emerge along road segments which carry international traffic and in other cases the traffic volume seems to be higher than that shown by other available data. Furthermore, the road system for international traffic passes through many small towns and villages, which raises potential conflicts between long distance international traffic and local traffic. Figure 4.5 Road Network and Trade Facilita on, 2002 A. Road network in Bulgaria B. Conditions of the road network in Bulgaria 100 80 Total Km=19,008 82.5 70 ---oGood nBad ~ 0 Fair 80 -Y5 E 60 60 c C 50 u- c 40 40 al d2 30 20 20 10 0 Mtorways Class Iroads Other roads 0 Mtorways aass Iroads Other roads C. Overall condition of roads inBulgaria D. Trade facilitation infrastructure index relative to EU-15 I Total Road Network 80 1.0 7 ~ 70 64.4 60 B ::50 20 10 Ports Custom Regulation IT 0 infrastructure Good Badto fair OBulgaria URomania OEstonia OLithuania 0CEEC-8average Source: Figures A, B, and C are World Bank staff estimates based on national sources; Figure D is World Bank staff estimatesbasedon information on the roadnetwork compiledby ECSIE of the World Bank; and Wilson, Mann, Otsuki database (2004) andWilson, Luo and Broadman(2004). 85 4.14 Ports is the area of trade facilitation in which Bulgaria seems to lag further behind the other CEEC. Port efficiency measures the quality o f the infrastructure o f maritime ports and airports relative to the efficiency levels o f the EU-15 (see Figure 4.5: D). The development o f port efficiency i s about 54 percent relative to the EU-15, compared with 84 percent in Romania, 90 percent inEstonia and an average o f 73 percent inthe New Member States (NMS- 8 or CEEC-8). Reforms inport management started in 2004. Bulgaria's Ministry o f Transport and Communications issued a port operator's concession program notice in 2004.26 The program invited interested parties to participate in the future bidding process for the operations and maintenance of port terminals, as well as for the investmentprogram for the reconstruction and development o f the existing and new terminals. To date, however, the focus has been almost exclusively on airports leaving aside maritime ports which are far more central to international trade and hence to competitiveness. A competitive tender, announced in August 2004, is limited to two airports in Burgas and Varna. As o f the end o f January 2005, the concession rights for Burgas and Varna airports, which are expected to attract investments o f nearly 130 million euros over the next couple o f years, have not been allotted. The best international practice for supporting international trade includes the introduction o f the landlord port, with the separation o f operational and commercial functions, the privatization o f port services, and the concessioning o f large container and bulk terminals to private operators. 4.15 Deficiencies in the basic public infrastructure essential for international trade are ` probably contributing to a bias favoring activities that service domesfic rather than external markets. The above analysis shows that there are important deficiencies in the basic infrastructure needed to reach international markets. Inaddition, as shown inprevious chapters, Bulgaria's economy shows a bias favoring activities that mainly service the domestic rather than external markets. These features are closely related, since deficiencies in the transport network needed to reach international markets, together with the accumulated appreciation o f the exchange rate, are all the more serious since they are likely to make selling inmarkets involving longer distances (such as the EU markets) more costly than for producers located closer to EU markets or with better trade facilitation infrastructures than the producers located in Bulgaria. Addressing these deficiencies i s therefore central to improving Bulgaria's competitiveness. 4.16 Bulgaria also needs to improve the efficiency of public expenditures on education in order to upgrade its human capital (see Figure 4.6). The efficiency and effectiveness o f public expenditures on education are particularly important, given the objectives o f improving labor productivity. Better labor productivity is needed to raise wage income and to expand export capacity. However, there i s evidence o f serious issues concerning the efficiency and effectiveness o f public expenditures on education. Specifically, expenditures per student are rising but the quality o f education and the adequacy o f skills relative to market needs are not improving. Education expenditures as a percentage o f GDP have been rising gradually over the last few years to reach a level o f about 4.4 percent in 2003, or about the average for selected CEEC. At the same time that outlays for education have increased, the size o f each age cohort has continued to decrease. Even with this level o f expenditures and the declining student population, there is evidence that the education system i s performing poorly. Despite increased public expenditures, the quality o f education i s below the levels o f other CEEC with similar 26See www.mtc.govemment.bg 86 levels o f spending. Figure 4.6: A shows that Bulgaria spends about the same amount as other CEEC on education but, as shown in Figure 4.6: B, Bulgaria's average scores in math and science (476 and 479, respectively) are below the average scores for these countries. Moreover, as shown in the next chapter, there i s strong evidence that the education system is producing significant skills inadequacies which contribute to a high unemployment rate for recent graduates. Figure4.6 PublicExpendituresin EducationandTIMSS 8th Grade StudentAssessment Resultsfor Mathematicsand Science A. Publicexpenditures on education, 2003 IB. TIMSS assessmentresults for Mathand Science I 7 , 560 540 520 503 480 460 440 420 Average Poland Hungary Bulgaria Czech Rep Slovak Rep Slovak Rep Slovenia Hungary Labia Lthuania Buigaria hternational CEEC oMath IScience Ave I I I Source: World Bank staff estimates based on databases o f the World Bank and TIMss2003: International Mathematics Report, International Association for the Evaluation o f InternationalAchievement, December, 2004, and TIMSS2003: International Science Report, International Association for the Evaluation o f International Achievement, December, 2004. 4.17 The relatively poor outcomes for public expenditureson education are mainly due to inefficienciesin outlaysand deficiencies in curricula.*' The education system inBulgaria has an overcapacity in staffing and facilities, as i s illustrated by the very low student to teacher ratios. Inaddition, average student to teacher ratios mask a serious deficiency inthe system o f a large variance across schools. While many schools have very low levels o f students, others, particularly in urban areas, are overcrowded. Inprimary education student to teacher ratios are about one-third below the average OECD levels; in secondary education Bulgaria's ratio i s 30 percent below the OECD average; and in higher education Bulgaria's ratio i s 55 percent below the OECD average. As a result, a large share o f outlays o f about two-thirds o f total public expenditures on education goes to wages and salaries for staff. Because o f this, limited resources are available to keep curricula and textbooks updated and to provide adequate teaching materials and maintain facilities. The Government i s taking first steps to strengthen the links between budget allocations to education and its performance in terms o f outcomes. This report recommends that an in-deptheducation reform i s formulated to address (i) inefficiencies in education spending; (ii)declining student cohorts, and (iii)upgrade curricula. And it supports the initiative to expand private sector participation inthe sector. 27 See World Bank, Public Expenditure and Institutional Reform," 2002; OECD, "Education at Glance," 2004; " and OECD, "Review of National Policies for Education: Bulgaria," 2004. 87 4.18 To make roomfor such spending and maintainit in the face of an agingpopulation, Bulgariawill need to rein in the rapidgrowthof expenditureson pensions, the healthcare system, and the social assistance system. Public expenditures on the social protection system increased from 9 percent o f GDP in 1996 to close to 14 percent by 2004, and now account for about 35 percent o f the budget. This i s due to a combination o f demographic trends, a rapidly aging population, high unemployment, the broad scope o f the social assistance system, and increased health care expenditures. However, the efficiency and effectiveness o f the social protection system has significant room for improvement-8 o f every 10 Bulgarians receive at least one social benefit, including pensions, but only 1 in 3 i s poor or old. With one o f the fastest aging population in the region, the financial pressures are large and are increasing rapidly. Reforms in the social protection system are urgently needed to enable the reduction o f very highpayroll taxes, which create significant incentives for the informal economy and for tax evasion, andwhich also limitjob creation. 4.19 Recent reforms of the social security system. Inthe last few years important reforms have been implemented, including increasing the retirement ages, increasing the contribution- identity o f the pension formula, and tightening the eligibility criteria. Contribution rates and benefit indexation rules are now specified in an amended Social Insurance Code rather than in the Budget Law. This has increased the system's predictability. However, despite recent amendments to the Social Insurance Code, the disability pension system is still open to fraud and often provides benefits without duejustification. The social security system is a key long- term source o f fiscal risk, owing to low coverage and low contribution levels into the system and rising dependency ratios. The rising dependency ratios are mostly driven by demographic trends but are also the result o f high numbers o f beneficiaries and low labor participation rates in the formal economy. Financial pressures are magnified by the common practice o f under- reporting wages, which erodes the contributions to the system and jeopardizes the adequacy o f future pension benefits. Without further reform, the system imbalances will continue to grow and to pose a serious threat to macroeconomic sustainability, particularly if the economy grows slowly. The urgency for reform i s increasing owing to the demographic challenge o f a rapidly aging population. 4.20 Recent reforms in the health care system. The financing o f the health care system, particularly hospital care, has been a source o f unexpected budgetary outlays and represents a central source o f fiscal risk. The health care system is based on the separation o f service provision and financing. Financing includes allocations from the central budget, a health insurance premium collected via payroll taxes, and out o f pocket payments. The National Health Insurance Fund (NHIF) pools risks and purchases health care services through contracts with physicians, group practices, and hospitals. The system faces challenges in several areas including its financial sustainability, its quality and its ease o f access. 4.21 Despite increases in public expenditures and attempts at rationalization, improvementsin the efficiencyof publicexpenditureson healthcare and healthindicators havebeen at bestmodest. This is due to several factors, including the costly excess capacity in both facilities and health care specialists inherited from the socialist past; an inefficient use o f resources, as manifested through both the excess capacity and the low utilization rates o f 88 facilities; and the neglect o f maintenance o f facilities and equipment, as more than 90 percent o f outlays represent current expenditures. 4.22 Reforming the social assistance system. Public expenditures in the social assistance system represent about 5 percent o f GDP. However, recent research shows that the outlays o f the social assistance system are largely ineffective in alleviating poverty. Instead poverty alleviation in Bulgaria i s heavily dependent on pensions, while the social assistance system has a negligible impact. This implies that pensions, which are not means-tested and therefore are not an efficient or effective instrument for poverty alleviation, are compensating for costly inefficiencies in the social assistance system. At 4.4 percent o f GDP, the costs o f the social assistance system are too highfor its negligible impact on poverty. 4.23 Deepening public sector reform is needed to support the restructuring of public expenditures. To achieve higher levels o f efficiency and effectiveness inpublic expenditures, Bulgaria needs to deepen public sector reform. As shown in this section, larger budgetary allocations in education do not by themselves translate into better education outcomes. Similarly, fiscal consolidation is not supportive o f investment and growth if it relies heavily on cuts in public capital expenditures and in outlays on operations and maintenance when fiscal tightening i s needed, since such cuts result indeficiencies inbasic public.infrastructure. N o r do reductions in tax rates achieve their potential impact on investment, growth, and the creation of employment if there are major deficiencies in tax and revenue administration. From this perspective, several areas o f reform are particularly important, including public expenditure management, public administration, and public sector governance and financial accountability. These areas are discussed inthe next section. C.Publicsector reform 4.24 Progress in public sector reform plays a determining role in the extent to which public expenditure policy reform is translated into actual improvements in the efficiency and effectiveness of public resources. Inthe last two years Bulgaria has taken the first steps towards implementingits public sector reform program, which spans 2002-07. The program i s aimed at strengthening public expenditure management, and improving public administration and public sector governance. A large number o f legal and regulatory reforms aimed at these objectives were implemented in2002-03. And the first steps intheir implementation were taken in2004 and early2005. 4.25 Public expenditure management reform. In2002, Bulgaria adopted a comprehensive reform program aimed at strengthening public expenditure management. The program has included actions to increase allocative efficiency and transparency in budget formulation; to strengthen budget execution and management; to strengthen the effectiveness o f internal and external accountability arrangements; and to strengthen the system o f inter governmental finances. The reform program aimed at adopting direct links between programs, budgeted allocations, and performance targets both planned and actual. The first pilot ministry was 89 created in 2002 and was extended to include seven other ministries. 28 Performance indicators and targets for pilot ministries were published in the 2005 budget. The authorities have also moved ahead with the development o f the MediumTerm Fiscal Framework (MTFF) inthe 2005 budget. By end-2006, for the 2007 budget preparation cycle the objective is that the public expenditure management system will be based on the MTFF. By 2007, the objective is to see that the center o f government decision-making and public expenditure management systems in place, to enable the Council o f Ministers to prioritize and allocate resources more effectively. 4.26 A number of measures were taken to improve the management of capital expenditures and funding in intergovernmental finances. The 2003 draft guidelines for capital expenditure planning and appraisal were developed according to EU standards and distributed to all line ministries. These draft guidelines served as the basis for the guidelines approved by the MOF in2004 and were used for the first time in analyzing and prioritizing the limited number o f new capital expenditure proposals in the 2005 budget. Intergovernmental finances were also improved by fully funding, in the 2004 and 2005 budgets, all centrally mandated delegated functions based on agreed expenditure standards. The government has also adopted a transparent formula for allocatingrevenue equalization grants for the 2004 budget and has fully funded this equalization transfer inthe 2005 budget. 4.27 Despite the progress to date in public expenditure management, important challenges remain. Bulgaria relies heavily on capping discretionary expenditures to achieve fiscal targets. Discretionary expenditures are capped below their budgeted levels for the first three quarters o f the year as a way o f providing flexibility in the budget to respond to potential inflationary pressures and external developments. The cap was 88-90 and 95 percent o f the total planned budget in 2002-03, and 2004, respectively. The cap in 2005 is 93 percent o f the total planned budget. While this practice has been an effective approach to achieving fiscal targets, ingeneral it does not support the efficient use o fresources. Inthe last few years a large amount of resources, reaching close to 2 percentage points o f GDP, has been spent during the last quarter o f the year. Furthermore, program budgeting has yet to cover all line ministries. Consequently, not all line ministries are subject to the discipline o f the links between budget allocation and planned and actual performance outcomes. Moreover, the expenditure accounts o f the judicial system have yet to be incorporated in to the budgetary payment system. And finally, Bulgaria cannot as yet count on a financial management system to support a more efficient approach to public expenditure management. 4.28 Public administration reform. The reform began with the adoption in 2003 o f the Strategyfor State Administrative Modernization: From Accession to Integration. This strategy i s aimed at: adopting merit criteria for, and depoliticizing the hiring of, public employees; introducing Performance and service criteria in the delivery o f public services; optimizing government structures; and strengthening accountability and transparency in service delivery. The legislative framework for a merit-based and depoliticized civil service became effective in January 2003. This entailed introducing amendments to the Civil Service Law and the Law on 28 These are the Ministry of Environment and Waters (MEW; first pilot in 2002), the Ministry of Transport and Communications(MTC), the Ministry o f Labor and Social Policy (MOLSP), the Ministry of Educationand Sciences (MES), the Ministry of Economy, the Ministry o f Energy and Energy Resources, and the Ministry of Youth and sports. 90 Administration. The authorities introduced a performance assessment system in all core ministries in January 2003. The share o f competitive recruitment increased to 100 percent in 2004 from under 15 percent in 2002-03. The restructuring o f all ministries is being carried out via functional reviews which started with three reviews in 2003 and increased to 13 reviews (7 ministries and 6 regional administrations) in 2004. On the e-governance action plan, on-line forms for tax and customs services have been introduced, an electronic registry o f state institutions including the database o f regulatory regimes and civil service positions was established; and one-stop shop pilots for public service delivery were put inplace in more than half o f the local administrations. 4.29 The challenge in public administration reform is to fully implement the new legislative and regulatory framework. To depoliticize civil service, hiring, promotion, and salary increases o f public employees needs to implement fully the merit criteria o f the new legislation embodied in the Civil Service Law and the Law on Administration. The functional reviews o f all ministries and agencies supported by the central budget need to be completed, and, based on the findings, a restructuring program and the consolidation o f the public sector need to be implemented. The new principles and regulatory framework also need to be fully implemented to achieve a significant increase in the extent to which administrative services are subject to formal competitive processes, in the outsourcing o f services, and in the analyses of savings generated. Efficiency principles should guide decisions as to which services should be paid for and provided by the public sector and which should be provided by competitive private suppliers. 4.30 Public governance reform and financial accountability. Based on 2002 information, Bulgaria was characterized as a country with high levels o f state capture and a medium level o f administrative corruption.29 The areas most affected have been customs, health services, higher education, tax administration, and the courts. In 2001 the authorities adopted the National Anticorruption Strategy which aims at combating corruption in all areas o f the public sector. In 2002-2003, specific sectoral anticorruption strategies were adopted for education and health, and the Supreme Judicial Council (SJC) approved a linked anticorruption strategy for the judiciary. Measures were taken to address weaknesses relating to conflicts o f interest and asset declarations. A code o f ethics intended to improve administrative procedures, reduce inappropriate discretion on the part of civil servants, and clarify accountability arrangements was introduced in2004. 4.3 1 Financial accountability and procurement reform. Accountability is being strengthened by improving evaluation: the first evaluation report by the National Audit Office (NAO) on the 2002 program budget o f the Ministryo f Environment and Water was released in 2004. Improvements in procurement include the enactment o f a new Public Procurement Act and related secondary legislation. In addition, a public procurement register is being established, and rules and procedures for arbitration in procurement-related disputes are in place. Similarly, internal and external audit practices are being improved by locating internal audit teams in all central and local government entities, ensuring that the annual internal audits were carried out in all government entities, and that reporting on internal audits was also carried out. The new legislative and regulatory framework needs to be implemented fully to make the 29World Bank,Anticorruption in Transition 2, 2004. 91 procurement system more efficient and transparent; and appropriate action by the MOF, including remedial, system, and management improvements, should follow from the reports by the NAO and the Public Internal Financial Central Agency (PIFCA). It is particularly important that transparent and competitive procurement be implemented for all civil works, including those that emerge from extrabudgetary funds. 4.32 To improvepublic sector governanceand financial accountability,Bulgarianeeds to focus on implementingcertain strategies,regulationsand frameworks.These include: (i) its national and sectoral anti-corruption strategies; (ii) the new regulations relating to conflicts o f interest and asset declarations; (iii) code o f ethics introduced in 2004; and (iv) the new the financial accountability framework. At a more fundamental level, as discussed in Chapter 6, Bulgaria needs to simplify its overly complex legislative and regulatory framework. This complexity impinges on economic efficiency and provides fertile ground for administrative corruption. These reforms will help reduce the levels o f state capture and administrative corruption compared to 2002, and will contribute to achieving stronger external accountability and transparency inpublic administration. D.Restructuringpublicexpenditures:A reformagenda 4.33 The first core area of reform is to implement major public expenditure restructuring supported by deepening public sector reform, to ensure the increasing efficiency of public resources and the provision of basic public services. Inefficiencies in public resources are far more detrimental to productivity and growth than are inefficiencies in private activities, since the former affect a wider spectrum o f activities in the economy. Moreover, their sustainability and flexibility are central to macroeconomic stability. Consequently, public expenditure restructuring reform needs to be driven by cost-benefit criteria for budgetary sources and allocations and their outcomes. First, the rationalization o f public expenditures is needed to provide a solid foundation with which to continue to improve the efficiency o f tax policy and reduce the tax burden on the economy. Inthis context, substantial reforms to the social protection system are needed to improve their efficiency and to ensure their fiscal sustainability. These reforms are essential to make it possible to reduce payroll taxes, which i s central to growth and to net job creation. Second, public expenditure restructuring should give high priority to addressing deficiencies in the transport network and to upgrade skills o f its human capital. As shown in the previous chapter, without upgrading the country's transport network and without upgrading skills, Bulgarian producers will continue to face difficulties in competing in the global markets. Similarly, Bulgaria will find it increasingly difficult to attract FDI flows into outward-oriented sectors in particular if deficiencies in basic infrastructure are not addressed, or if the profitability o f economic activities remains limited mainly to unskilled labor intensive products. These reforms are all the more pressing as FDI flows drivenby privatization will decline as the privatization program i s coming to an end. 4.34 Upgrading skills. Bulgaria needs to implement a major shift in the paradigm o f the education system, including universities and vocational schools, to generate the skills needed to compete in the global markets, to address the demographic challenge o f the declining size o f cohorts, and to improve access to education by disadvantage groups. Improving the efficiency and effectiveness o f public expenditures on education i s particularly important given the 92 objective o f improving labor productivity, which i s necessary to raise wage incomes and to expand export capacity beyond unskilled labor intensive products. The direction o f reforms in the education system should be driven by three criteria: (i)improving the efficiency and effectiveness o f public outlays on education, including enhancing the quality o f education and ensuringaccess to education byvulnerable groups; (ii) strengthening the links betweenthe skills acquired in the education system and those needed in the job market; and (iii) developing a reliable accreditation and certification system. Key reforms recommended include the following: Reallocating resources from the surplus capacity in the teaching staff and underused facilities towards other essential quality-enhancing inputs, including the modernization of curricula, textbooks, and teaching materials. The reallocation o f resources is also needed to address the acute deteriorationin education facilities and teaching equipment; Implementing a composite per student resource allocation. An efficient mechanism finances outcomes-in this case students and the quality o f the education they acquire- rather than financing inputs which, in the case o f Bulgaria, are mostly teachers. Per student allocations should be differentiated to reflect the differences in the costs o f education levels and specialization, to improve the access to education o f disadvantaged groups, and to account for cost differences arising from the learning needs o f students. Designingdifferentiated per student financing formulas should be undertaken carefully to avoid simply reflecting current per student costs o f different levels and facilities-which perpetuates inefficiencies. Measures that will increase cost recovery in universities to 30 percent, as specified by the law, including raising fees, should be implemented. Moving towards a more advanced approach to education by financing educational results and notjust enrollment levels. Expanding the role o f private education providers and strengthening competition based on outcomes can be an important force inimproving quality o f education. In general, countries using this approach condition payment to private educationproviders to agreed educationtargets interms o f learning achievements. Consolidating the stafing andfacilities of the education system. A per student resource allocation policy needs to be complemented by a plan to consolidate staffing and facilities. Unless financial policy i s supported by a program to consolidate staffing and facilities in the education system, Bulgaria will continue to have schools that are overcrowded in some o f the urban areas and nearly empty in other facilities. Consolidating facilities will require providing transportation to students from smaller to larger facilities. Addressing the skills mismatch gap. The education reform strategy needs to shift toward adopting automatic mechanisms and feedback rules to link the education system and labor market conditions. First, the provision o f education services should be demand- driven and based on academic attainment and job performance. Adopting a composite financing per student policy supports a system in which the demands o f students and parents determine which educational programs expand and which ones contract. And second, stronger competition among educational institutions, including universities, for 93 public resources and private sector funding i s needed. This would also help to attract the best teachers and students to the best schools and would improve the academic attainment and job performance o f graduates. Both academic attainment and job performance should be monitored and disclosed at the level o f individual institutions. 0 Developing a reliable system of accreditation and certiJication based on objective criteria. Given the rapid changes in labor market conditions, developing a system o f accreditation and certification should be driven by outcomes in terms o f academic attainment and job performance. Basic standards, particularly for technical and higher education, should be definedinconsultation with the private sector. 4.35 Upgradingthe transport network. Addressing the network's large basic infrastructure needs will take time. The direction o f reforms needs to be drivenby undertaking a cost-benefit analysis o f these investments and addressing the challenge arising from the liberalization o f international trucking in the context o f EU integration. Reforms intended to upgrade the transport network include the following: 0 High priority should be given to long distance international traffic. Investments first need to address the emerging bottlenecks along road segments that carry international traffic. These investmentswould also need to address the problem that arises because the road system for international traffic passesthrough many small towns and villages. 0 I n parallel, reforms of maritime ports are urgently needed. These ports together with a road system for long distance international traffic, are essential to international trade and deepening integration with the EU. The best international practice for supporting international trade in this area includes the introduction o f the landlord port, with the separation o f operational and commercial functions, the privatization o f port services, and the concessioning o f large container andbulk terminals to private operators. 0 I n setting priorities in road investment to support long distance international trafic, maintenance has the highest priority and new construction has the lowest. Since road investmentscompete for limited resources, the preservation and rehabilitation o f existing assets has a higher priority than new construction. 0 Bulgaria should adopt transparent and competitiveprocurementfor all civil works. 0 Public-private partnerships should be entered into only if they ofer valuefor money as determined by apublic sector Comparator. 4.36 Addressingthe challenges of the social protectionsystem. Bulgaria needs to develop a strategy to rein in the rapid growth o f expenditures inpensions, the health care system, and the social assistance system. The direction o f reforms should be driven by the efficiency o f these expenditures and their fiscal sustainability. 4.37 Deepening social security reform. Despite substantial reforms since 1999 when Bulgaria moved to a three pillar system, its social security system is not aligned with 94 demographic trends. As a result, payroll taxes are very high and provide incentives to the informal economy, as well as to under-reporting o f income both o f which exacerbate the demographic trend of rising dependency ratios. There are two broad reform scenarios. One reform path is to continue to implement reforms to contain financial imbalances, which would entail mainly further increasing further retirement ages or reducing benefits. This reform pathrequires at least two additional measures: (i) implementingthe strict discipline inthe use o f the disability pension system which i s currently still prone to fraud and abuse; and (ii) implementing strict collection o f contributions. The former needs strict certification and steep sanctions for fraud and abuse. The operationalization o f the National Revenue Agency could help with the latter. Under this first reform scenario, reduction inthe payroll taxes are likely to remain limited to the extent that the contribution base can be expanded. Risks to their financial and fiscal sustainability are increasing over time due to demographic trends and are magnified by periods o f low economic growth. At the other end of the reform spectrum, a more substantive reform could be to consider increasing the linking of benefits to contributions to the system. At the limit, the reform could aim at a fully funded system supplemented by a citizen's pension or basic pension to avoid extreme poverty in the old age population. This option will imply large fiscal costs to honor commitments accrued to retired and active cohorts who have contributed to the system. 4.38 Given the substantial financial challenges involved, detailed actuarial work is needed to assess the implications and feasibility o f alternative reform paths to deepening social security reform. Clearly, the most risky scenario financially, with a potentially serious outcome would be not to consider reform options to address the rapidly growing financial pressures. 4.39 Reforming the health care system. Despite increases in public expenditures and rationalization attempts, improvements in the efficiency o f public expenditures in health care and health indicators have been, at best, modest. Under current trends, the fiscal sustainability of these expenditures is questionable. The direction o f reforms should be aimed at addressing the concurrent problems o f the rapid increases in health expenditures, the surplus capacity in health care staffing and facilities, the deterioration in the quality o f the facilities, and the inadequate modernization o f equipment. The government i s developing a new strategy aimed at addressing financial sustainability in general and informal payments and corrupt practices in particular, as well as improving quality and access. The reforms being considered include the following: Developing of a basic health carepackage. The reform envisions universal coverage under the basicpackage. Developing a private voluntary health care insurance supplemental to the basic health carepackage. e Funding high tech and expensive treatments by the state in an amount of additional payment inversely proportional to the value of the services. 95 Offering options in the ownership and management of hospitals. The reform envisions providingfree choice of health careproviders so that hospitals competefor them on the basis of the price and quality of their sewices. Subsidizing the NHIFfor paying thefull cost for medical services of disadvantaged citizens. 4.40 As in the case o f pensions, careful analysis o f the reform options is needed in order to assess their implications in terms o f financial sustainability and provision o f health care services. However, as the experience with reforms over the last several years has shown financial policy will need to be supported by decisive measures to restructure and consolidate surplus capacity inheathcare staffing and facilities. 4.41 Reforming the social assistance system. There are large inefficiencies in the social assistance system, as illustrated by its negligible impact on poverty. The system i s highly fragmented with more than 34 programs. The largest program is that o f the Guaranteed Minimum Income (GMI) which in the last few years has improved its targeting and effectiveness. However, for the most part the social assistance system's targeting has considerable scope for improvement. The recommendedreforms include the following: Consolidating the social assistance system moving towards in-cash and income support mechanisms Training social workers to better identifi poor households Improving information systems tofacilitate means-testing and reduce the payment of duplicative benefits Expanding communications activities so as to inform beneficiaries about eligibility criteria and application procedures. 4.42 Public sector reform. The way in which the public sector functions determines the extent to which public expenditure policy i s translated into actual improvement in the efficiency and effectiveness o f public expenditures. Over the last two years Bulgaria has implemented important reforms aimed at such improvement. With this objective in view, a large number o f legal and regulatory reforms were implementedin2002-03, and the first steps to implement than were taken in 2004 and early 2005. These measures included amendments to the Civil Service Law and the Law on Administration, the development o f sectoral anti-corruption strategies, and enactment o f the new Public Procurement Act. Similarly, further progress was made in developing the Medium Term Fiscal Framework (MTFF) by strengthening the links between budget allocations and performance indicators and implementing, for the 2005 budget, guidelines for capital expenditure planning and appraisal. The next stage o f reforms should focus on the full implementation o f the new legislative and regulatory framework. Key reform measures include the following: 96 e Strengthening public expenditure management: (i)completing the development and implementation o f the MTFF to strengthen the links between budget allocations and performance indicators-program budgeting should be extended to all ministries; (ii) incorporating the expenditure accounts o f thejudicial system into the budgetary payment system; and (iii)developing and making fully operational a financial management system. e Improving public administration: (i)implementing fully the merit criteria for hiring, promotion and salary increases provided by the new legislation embodied in the Civil Service Law and the Law on Administration, to depoliticize the civil service; (ii) completing functional reviews o f all ministries and agencies supported by the central budget and, based on the findings, implementinga restructuring program for the public sector; and (iii) increasing the scope and number o f administrative services subject to formal competitive processes, outsourcing these services, and generating the analyses o f the savings generated. e Improving public sector governance and financial accountability: (i) simplifying an overly complex legislative and regulatory framework (see Chapter 6); (ii) implementing national and sectoral anti-comption strategies; (ii) enforcing fully the new regulations relating to conflicts o f interest and asset declarations and the code o f ethics introduced in 2004; (iii)implementing transparent and competitive procurement for all civil works including those emerging from extra budgetary funds; and (iv) implementing internal and external auditing in all government entities, making evaluation reports publicly by the N A O and PIFCA. 4.43 T o translate the efficiency gains resultingfrom public expenditure policy and public sector reforms into higher productivity levels in the economy and improved competitiveness, the Bulgarian economy depends on labor market adjustment. Implementing public expenditure restructuring supported by public sector reform can substantially increase the overall economic efficiency with which revenues are collected and the cost-benefits o f budgetary allocations and their outcomes. These reforms are essential in order to upgrade the country's transport network and the skill content o f its human capital stock- actions that are needed if Bulgaria is to compete in the global market. An improved transport network and a well trained labor force would provide a solid foundation for increasing productivity in the economy. However, for these reforms to be translated into enhanced competitiveness, the Bulgarian economy would be dependent on the ability o f the labor market to reallocate labor from lower to higher productivity activities. The extent to which labor market policies and institutions are supportive o f these adjustments is examined in the next chapter. 97 5. ENHANCINGLABORMARKETADJUSTMENT AND EMPLOYMENT 5.1 Enhancing labor market adjustment and employment is central if Bulgaria is to move towards a more solid path of convergence and improve its competitiveness. As discussed in the previous chapters, productivity i s low in most activities in Bulgaria, except in unskilled labor-intensive activities which dominate exports. The deepening o f Bulgaria's integration with the EU and global markets will be determined by its ability to reallocate resources from lower to higher productivity activities across the economy. Success will depend to a large extent on the functioning o f the labor market. Without the enhancement o f labor market flexibility, it i s unlikely that the economy will be able to respond to the emerging challenges discussed in Chapters 2 and 3-mainly to improve competitiveness and move towards a Lisbon-growth path. Nor will the economy be able to improve the prospects for higher labor productivity andhence better labor income earnings. 5.2 As is shown in the following sections, there are signs of slow improvement in labor market performance since 2002. 30 However, the labor market remains performing poorly in terms o f labor market participation and employment and unemployment rates, and inter-sectoral labor adjustment has been limited. A considerable share of employment is in largely unproductive segments in agriculture. Reform o f labor market regulations to adopt some o f the best practices in the EU can enhance creation o f employment and the expansion o f productivity gains more broadly across the economy, in order to benefit more fully from economic integration with the EU. Bulgaria has some o f the most inflexible conditions of employment, with tight restrictions on working schedules. The terms of leave and working hours are more generous than the average for countries inthe region. This chapter first presents an overview o f labor market policies and institutions. 31It then assesses labor market adjustment on the basis o f performance indicators over time and the extent o f labor adjustment at the sectoral level compared to other countries inthe region. The chapter ends with a summary o f reforms needed to enhance both labor market adjustment and employment. A. Labor marketpoliciesand institutions 5.3 Bulgaria's labor market policies and institutions need to be examined carefully given the labor market's poor performancerelative to other countries in the region. Over the last several years, Bulgaria has implemented labor market reforms by enacting amendments to laws and changes to regulations which are in general consistent with EU directives. However, the persistence o f high unemployment rates despite the strong sustained economic growth and rapid expansion o f investment, and the poor performance o f the labor market relative to other countries point to the country's need to examine its labor market policies and institutions. 30As inother transition economies, labor statistics inBulgaria have shortcomings such as under-reporting o f wages and salaries and the presence o f a large informal sector also affects labor indicators. The information presented in this chapter is indicative o f stylized trends and composition o fthe labor market inBulgaria and other countries inthe region. I 31For a comprehensive analysis o f labor market performance inCEECs, see Jerald Schiff et.al., 2005, Labor Market Performance in Transition: the experience of Central and Eastern European Countries, IMF, Washington, D.C. 5.4 The wage-setting mechanisms include collective bargaining, rules applied to the public sector, and a minimum wage. Collective bargaining is based on a tripartite system which includes the government, labor unions, and representatives o f employers. Negotiations are undertaken at the national, branch, regional, and enterprise levels. Wages in the central government are completely centralized. The State Budget Act specifies the wages for budget institutions and companies. The wage bill growth regulation ties the financial performance o f state-owned enterprises to their financial Performance, including debt service. Collective agreements typically extend to non-union members, and in several sectors agreements are applied to the whole sector. A minimum monthly wage i s set at the national level by the government in consultation with social partners. Union membership has declined and now its level o f around 30 percent i s comparable to that in the Slovak Republic and higher than the levels inEstonia, Latvia, and Polandwhich range between 10 and 20 percent. 5.5 High labor taxes are a serious impediment to labor adjustment and employment creation. Wages are low but labor costs are relatively high, since labor taxes are high to compensate for widespread tax evasion. As shown in Chapter 1, Bulgaria has one o f the highest payroll taxes in the region and the tax burden on the economy contributes to a sizable informal economy. To avoid further erosion o f revenue collection, in 2003 the authorities introduced a minimum social insurance threshold matrix setting minimum wages by activities (17) and by professions (9). This measure increased the number o f registered contracts and, hence, the registered employment, in the first few months o f its introduction in 2003, and the collection of social security contributions rose sharply between 2002 and 2003. The sustainability o f these increases, however, i s unclear, as employees may redefine their contracts, seeking the lowest possible wages at the minimum levels o f the threshold matrix. To raise the employment and labor market participation o f the labor force, Bulgaria needs to restructure public expenditures towards high priority areas, particularly towards upgrading skills and the transport network, implementingthe structural reforms to its social protection system to increase its efficiency and fiscal sustainability, and expanding the tax and revenue collection base in order to be able to reduce labor taxes inparticular and the tax burdenon the economy ingeneral. 5.6 Reforms of labor market regulations are needed to improve labor market performanceand the creationof employment. Bulgaria has implemented reforms to its labor laws and regulations over the last several years3* While current labor market regulations seem broadly in line with EU directives, the regulatory framework faces more serious challenges, in view o f the poor performance o f the labor market. The poor performance in the investment climate can no longer be blamed for the labor market, poor performance. Investment, exports, and imports are all growing at a faster rate than GDP growth. Incontrast, labor adjustment and creation are very limited. 5.7 Regulationsfor work schedules are quite restrictive. Regulations regarding working schedules are quite restrictive, more so than the average for countries in the region. These regulations set precise rules regarding daily hours, and weekly holidays, and forbid overtime work except under certain circumstances. Since these regulations are inconsistent with the new 32Bulgaria's Labor Code dates back to 1985 and had major revisions in 1992, and other more modest revisions have beenmade since then. 99 realities o f Bulgaria's market economy, compliance i s unclear and thus they tend to provide incentives to informality. The terms o f working hours are more generous than the average for countries inthe region.33 The combination o f over prescriptive rules for dismissal and generous sick leave benefits is a disincentive to work that hindersthe creation o f employment. 5.8 Regulations link wage increases to seniority rather than performance and labor productivity. The law mandates automatic increases inwages by seniority, which is applicable even incases where previous work experience i s unrelated to an employee's present job. This is a major constraint to employment creation in general and to the employment o f the long-term unemployed in particular. Furthermore, it provides incentives to new businesses to offer low wages in anticipation o f mandated wage increases by seniority. Together, these regulations result in low turnover, limiting labor adjustment to higher productivity jobs. This is a significant problem in such labor-intensive industries as textiles, where turnover is l o w and salaries continue to increase even though labor productivity i s stagnant. 5.9 The combination of over prescriptive rules for dismissal and generous sick leave benefits is a major disincentiveto work and hindersthe creation of employment. The law provides for conditions for dismissal but dismissal is prohibited in the case o f illness. At the same time, the law specifies that employees receive 80 percent o f their wage income while on sick leave. But because they are not required to pay taxes on sick days, the net salary while on sick leave i s higher than while at work. Moreover, there i s no limit on the number o f sick days an employee may take. Untilrecently, the employer paidthe first three days o f every sick leave. Recent budget legislation for 2004 reduced this period to one day, with the remainder being paid out of the government health insurance fund. These monetary incentives are a significant disincentive to work and are likely to consume a substantial portion o f governmental resources. In addition, the fact that dismissal is prohibited in case of illness, combined with the rules on sick leave, creates substantial problems in managing absenteeism and de facto makes dismissal conditions quite restrictive. Amendments to the Health Act in 2004 are aimed at reducing the abuse o f sick leave provisions. However, it i s still too early to assess the full impact o f these measures. 5.10 Unemploymentinsurancebenefits are above the average of those in other countries in the region. Bulgaria is the only transition country that increased the maximum duration o f unemployment insurance payments from 6 to 12 months between the beginning and the end of the 1990s. 34 Furthermore, the duration o f unemployment benefits depends on the length o f service and increases from 4 months o f payment for the minimumservice period (up to 3 years) to 12 months o f payment for more than 25 years o f service. Bulgaria decreased the initial replacement rate of unemployment insurance payments significantly over the 1990s, but at 60 percent it remains highcompared to other countries inthe region. 5.11 The minimumwage is amongthe highestinthe region. The minimumwage was close to 40 percent o f average wages in 2003, prior to the 25 percent increase adopted on January 1, 2004. By mid-2005, the minimumwage was about 53 percent of average wages. The measure o f increasingminimumwages by 25 percent as of January 2004 was aimed at keeping minimum ~~~ 33 World Bank, Bulgaria s Labor Market, July 2004, mimeo, Labor Market Team, HDNSP. 34 Estonia introduced three monthextensions considered on an individualbasis. 100 wages above social assistance allowances and at providing incentives for higher labor market participation. However, there are concerns that increases in minimum wages, which are now among the highest in the region, may have a negative impact on the employment opportunities o f recent graduates-whose unemployment rate i s 27 percent-and o f other vulnerable groups including unskilled workers and the long-term unemployed. Furthermore, highminimumwages relative to average wages typically hurt more small and medium enterprises, and there are concerns that informality may increase. While it i s too early to assess the full impact o f this measure, sustainable improvements in wage income are ultimately determined by labor productivity which in turn depends on upgrading skills and labor adjustment from lower to higherproductivity activities. 5.12 Inadequate skills are an important obstacle to employment creation and trade diversification beyond unskilled labor intensive products. The persistence o f unemployment, the large proportion o f discouraged workers, and the highunemployment rate o f recent graduates are consistent with evidence that the skills provided at school, especially vocational skills, are inadequate for the needs o f firms involved in new technologies. A survey conducted for this report, updating the survey o f 2002, indicates that foreign investors are experiencing considerable difficulty infinding workers with the skills they need. The skills gap applies to technical specialists, qualified business managers, and others needed to operate modem competitive businesses. The source o f this problem i s the disconnection between the skills taught by schools, technical institutes, and universities and the skill requirements of businesses, particularly those that need to compete in the external markets. As result of the skills mismatch, firms must incur high costs to train new hires, which in the case o f software companies and other high-tech companies could include three to six months of no productive inputfromthese workers. B. Growth,labor market adjustment,and employment 5.13 Bulgaria has had to bear high rates of unemployment since the early 1990s. As in other transition economies, economic restructuring in Bulgaria has led to considerable employment loss. However, incontrast to other countries inthe region, the employment loss in Bulgaria was exacerbated by the protracted contraction o f 1989-97. Inthe first halfo f the 1990s the initial transitional recession and associated industrial restructuring gave rise to unemployment rates that exceeded 20 percent. Long-term unemployment has been above 50 percent since the mid-l990s, and at 60 percent in 2003 was the highest among the CEEC. The unemployment rate reached a peak o f 19 percent in 2001. Thanks in large part to macroeconomic discipline and in-depth structural reforms, number o f employed persons began to increase in 2002. A trend that has continued since then and observed in2005. Inthe second quarter o f 2005 the employment rate o f persons between 15 and 64 years old increased by 1.1 percentage points to 56.2 percent compared to 55.1 percent inthe second quarter o f 2004. 5.14 Despite the vigorous economic recovery since 1998 and slight improvement in number of persons employed, the labor market continues to face many difficulties in creating employment opportunities. Figure 5.1 shows the trends in reform, real GDP, labor force participation, and the employment rate. Notwithstanding positive developments in recent years, the labor market continues to face difficulties in creating employment opportunities. Structural reforms have advanced since 1997, and GDP has increased as a result by more than 101 one-fifth inrelation to the level in 1996.35 However, this has translated into relatively littlejob creation as indicated by the employment rate and labor participation. Following the 1996-97 crisis, both labor participation and the employment rate declined until 2000 and 2001. This reflects the process o f the labor shedding o f surplus labor by newly privatized enterprises and other firms now facing hard-budget constraints. This has been common to labor markets in transition economies. In general, transition economies have not yet produced sufficient new jobs to absorb the workers released by these enterprises. In Bulgaria, an average annual net employment growth o f about 3 percent since 2000 has helpedto reduce unemployment, but the drop in the rate o f unemployment also reflects lower participation in the labor force, outward migration, and a temporary employment program financed by the government. Figure 5.1 Reforms, GDP, and the Labor Market, 1996-2004 (1996=100) 140 120 100 80 60 40 20 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 ,I GDP 0Reform +LF Particip -1 Rate , Source; World Bank staff estimates based on N S I and EBRD 5.15 Bulgaria's labor market is performing poorly compared to other countries in the region. Bulgaria has low labor market participation, low employment and high unemployment rates, and a weak relationship among wages, employment, and unemployment. By these measures the labor market inBulgaria is performing poorly. A labor market participation o f 49 percent o f the labor force in 2003 i s the second lowest among selected countries in the region (see Table 5.1). The employment rate i s 43.6 percent-the lowest among these countries. The unemployment rate, which has been persistently high since the early 1990s, has now started to decline and in 2004 was 11.8 percent, and about 60 percent o f these unemployed are long-term 35This is the EBRD's measure of progress intransition. Percentage changes inthe index, o fcourse, should not be interpreted as a change inthe "strength" o f the policy effort. 102 unemployed 36andyouth unemployment i s 27 percent. The unemployment rate continued to decline slowly in the first half o f 2005 reaching 10 percent. However, this lower rate partially reflects seasonal variations associated with higher than normal levels o f employment during the summer. As per the Employment Agency information, the number o fregisteredunemployed in August 2005 was 10.77 percent. 5.16 While Bulgarian regions saw decreases in unemployment rates to varying degrees, these seem to have converged by 2003. The employment rate has been increasing slightly in all regions in recent years except in the North-West. Labor force participation has been declining across Bulgaria in recent years, but two regions stand out: the North-West has seen a decrease inparticipation, while the South-West has seen an increase. Table 5.1 Labor market performance, 2003 (in percent) I Unemployment Participation Employment Total Long-Term Youth Bulgaria 49.4 43.6 11.8 60.0 27.0 CzechRep. 59.3 54.3 8.2 51.0 21.6 Denmark 65.5 61.5 5.7 19.5 9.1 France 56.0 51.3 8.9 41.6 20.4 Germany 57.3 51.7 9.8 50.0 11.0 Greece 48.7 47.8 10.2 54.7 26.5 Ireland 60.2 57.4 4.5 34.3 . 8.3 Italy 49.2 45.6 7.9 48.4 24.6 Latvia 57.5 52.5 10.0 45.5 17.4 Lithuania 58.2 51.0 10.6 49.9 23.2 Poland 54.7 44.9 18.2 56.8 37.9 Romania 54.7 51.6 7.6 58.4 21.6 Spain 54.3 49.4 10.9 32.2 22.4 UnitedKingdom 62.7 59.1 4.8 18.8 13.5 Source: National labor force surveys reportedby Eurostat. C.Inter-sectoral labor adjustment 5.17 The share of employment in the private sector has been increasing in recent years, but labor adjustment is slow (see Figure 5.2). By the end o f 2004, the private sector share had risen to about 77 percent, and the public sector accounted for 23 percent as a result o f privatization and new business creation. Figure 5.2 suggests that labor adjustment inBulgaria is currently proceeding slowly. It shows the gradual change in employment structure for workers since 2000, which i s the period that followed the major post-crisis reform drive. Most o f the net job creation has been in the service sector, which was approaching 50 percent o f total employment by mid-2004 as compared with 37 percent in 1990. 36The EmploymentAgency the long termunemployedwas 56.2 percentof total unemployedpersonsinAugust 2005. However,the NSI, which ingeneralprovidesmore accurate information, reportsthat the long term unemployedis about 60 percent. 103 Figure5.2 RecentEmploymentTrends:Privateand PublicSectors Shares of Private and Public Employment 60 898 899 2000 2001 2002 2003 2004 -private - .-public 1e Source: World Bank staff estimates based onN S I data. 5.18 Bulgaria's inter-sectoral labor adjustment shows trends which differ from output restructuring,with some sectors divergingfrom the adjustmentsobservedin New Member States. As shown in Chapter 2, economic restructuring in terms o f output composition is evolving in the direction observed in New Member States. However, inter-sectoral labor adjustments inBulgaria have followed a very differentpath from the labor adjustments observed inNew Member States. The share of employment in industry fell sharply, from 45 percent in 1990 to 27 percent in 2003, and expanded in services from 37 to 47 percent during the same period. This labor adjustment i s similar to that observed in the transition o f the New Member States. Incontrast to the New Member States, however, the share o f employment in agriculture increased from 18.5 to 25.5 percent between 1990 and 2003 while the share o f employment in agriculture inthe New Member States declined on average from 19 to 17 percent. 5.19 Productivityremainslower than inthe New MemberStateswhich i s partly due, to a large extent, to poor labor market performance. InBulgaria, employment is concentrated in services and agriculture which together represent 73 percent o f total employment, producing about 70 percent of gross value added (GVA). In contrast in the New Member States employment is concentrated inservices and industrywhich together represent 87 percent o f total employment, producing 97 percent o f GVA. With 25 percent o f employment in agriculture where the share o f the sector's GVA has declined sharply, a sizable share o f employment is located invery low and declining productivity areas. In addition, the expansion o f services has been more pronounced in Bulgaria than in the NMS-8. Between 1990 and 2004, the share o f services in GVA increased from 31 to 59 percent, while the share o f employment in services increased from 37 to 47 percent. The expansion o f the share o f services inGVA and its share o f employment inthe NMS-8 was 20 and 13 percentage points, respectively. As shown inChapter 2, the labor productivity o f investment in inward-oriented activities, including services, has stagnated. These estimates indicate that Bulgaria faces a significant challenge as regards the inter-sectoral restructuring needed to reallocate investment and labor to more productive 104 activities, given the large inefficiencies in agriculture and the relatively stagnant labor productivity in services. 5.20 Inter-sectoral labor market adjustment and sectoral employment imbalances in Bulgaria show worrisome trends. Table 5.2 presents estimates of departure indexes3' for Bulgaria and selected CEEC for 1989, 2001, and 2003. The departure indexes provide an estimate o f inter-sectoral labor adjustments relative to two comparators EU-North and EU- South. Since comparisons across countries and over time are relative to the same two benchmarks, estimates o f departure indexes allow an examination o f the extent o f labor adjustment in Bulgaria compared to selected CEEC. Interms o f initial conditions, Bulgaria's departure indexes are comparable to those o f Poland in 1989. However, Bulgaria had the largest share o f employment in manufacturing among these countries. These figures indicate that Bulgaria inherited the most distorted employment structure inmanufacturing among the CEEC. 5.21 Estimates of labor adjustment across sectors show limited labor adjustment in Bulgaria compared to the New Member States (see Table 5.2). Interms o f overall trends, inter-sectoral labor market adjustment in 1989-2003 shows very limited labor adjustment in Bulgaria compared to New Member States. The departure index inBulgaria improved from 24 to 22, or about 2 points, duringthis period. Incontrast, the departure index inPoland improved from 23 to 15, or about 6 points, or three times as much as Bulgaria. These differences cannot be explained by differences in initial conditions, since estimates o f departure indexes o f employment show that in 1989 Bulgaria inherited a similar level o f distortion inits employment structure as Poland, compared to EU-North and EU-South. Furthermore, the trends in the structure o f employment show that EU-South and EU-North continue to generate labor adjustment in line with changes in market conditions. Similarly, the departure indexes inNew Member States continue to decline as their labor markets continue to adjust to market conditions. Given the substantial restructuring o f the enterprise sector in Bulgaria, through privatization and liquidation, price and trade liberalization, and improvements inthe investment climate, failure to adopt policies to enhance labor adjustment are taking a double toll on Bulgaria's economy-a toll that i s related to the economic and social costs o f inefficiencies and limited labor reallocation opportunities, and the costs o f lagging behind the New Member States. 3' The departure index measures the proportion o f the workforce ina given country that would need to change sector to attain the same structure o f employment as that o f a comparable Western European economy in 1989. The departure index i s a stylized indicator o f employment structure that would prevail given the same factor endowments, technologies, and prices ina given country as those observed inthe comparator economy. Hence, the departure index i s a solid approximationof the extent o f distortions relative to comparators. 105 c c - N c 5 N 0 K P c c c - N e e N 3 T c c - c N Fi e 0 K ? - ? ? - ? ? ? ? P w m 9 z m r - N N - r - w t - - ??': N N N c c N c r c N P -T c rc c N 7 c c N Q K -z 2 % s W E v: gEz ck v: 0 D.Employment,unemploymentandwages 5.22 Another dimension of labor market inflexibilityin Bulgaria is the weak relationship between wages, and unemployment and employment. To the extent that employment and unemployment have a close relation to wages, the labor market can adjust to changing market conditions. Productive firms offer higher wages than firms with lower levels o f productivity, and higher (lower) wages are offered to highly (lower) productive workers. Hence, a close association between wages and the rate o f unemployment can be observed, which many empirical studies have found to be an inverse relati~nship.~~Statistical tests o f this relationship, conducted for this report, which control for the effects o f other variables such as productivity and inflation, confirm this conclusion (see Annex F). Figure5.3 UnemploymentandRealWages: VAR SimulationResults Response to Generalized One S.D. Innovations & 2 S.E. Response of D(LWAGE) to D(LUNEMP) .05 .04- 1 `3, .04 .03- .02- -.-.,, ,r_-----.._ -----___ .01-_______-----..,' -..__- ---_-. I` .oo-\ j\ ~. -----____-_----___ ----_ -.01- -* ,` I' -.02-..\....,___ --.. ,__- ... -.02 I` `u` -.03 , , , I 1 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Response of D(LUNEMP) to D(LWAGE) Response of D(LUNEMP) to D(LUNEMP) .089 .06- .06--____-. '*.`. .04- .02--. _,/-------_ _ _ _ _ _ _ _ _ _ -- .oo-` -...____---` -_--_----___ --_ / -.02- .,_______-----__-- --__--_______-- \\\ , , , , , , 1 2 3, 4 5, 6 7 8 9 10 Source: World Bank staff estimates (see Annex F). 38The nature of this association is subject to considerable theoretical debate, however. Many economists interpret supply and demand forces to be at play, while others emphasize the importance ofbargaining strength and reservation wages. For a brief overview o f the issues surrounding this so-called "wage curve," see D.Blanchflower and A. Oswald, "An Introduction to the Wage Curve,'' Journal of Economic Perspectives, Summer 1995, pp. 153- 167. 107 5.23 Employment and unemploymentrates exhibit no statistical relationwith realwages in recentyears (see Figure5.3). Annex Fpresents three models usedto examine the statistical relation between unemployment and employment rates, with real wages using quarterly data from 1997 to 2004. Two different Vector Autoregression models (VAR), one unrestricted and one restricted or Vector Error Correction (VEC), were used to test labor market flexibility through the observation o f real wage and unemployment dynamics. The third model, an equation o f demand for employment, was usedto examine the responsiveness o f employment to changes in the real wage. Figure 5.3 shows the predicted impulse response o f the average real wage to a relatively sharp change in the rate o f unemployment with statistical confidence bands indicated by dash lines. In this simple representation, which controls for seasonal factors and lagged effects, changes inthe real wage appear to have no significant relation to changes inthe rate o f ~ n e m p l o y m e n tControlling for the effect on the real wage o fproductivity and inflation . ~ ~ yields a similar conclusion: namely, that the rate o f unemployment has exhibited no connection with the real wage inrecent years. 5.24 Despite relatively low wages which reflect low labor productivity levels, there are important disincentives to job creation. While wages are low, labor costs in Bulgaria are relatively high. There are disincentives on the demand side as well as on the supply side. 5.25 On the demandside, employmenttends to be discouragedby extremelyhighpayroll taxes and generous unemployment insurance payments. Payroll taxes inBulgaria represent 42.7 percent4' o f the gross wage, 75 percent o f which i s paid by the employer. This comprises a 32.7 percent contribution to the pension fund (24.7 percent made by the employer), a 6 percent contribution to the health insurance fund (4.5 percent made by the employer), and a 4 percent contribution to the unemployment insurance fund (3 percent made by the employer). As shown in Chapter 1, social security contributions inBulgaria are among the highest compared to New Member States, owing to some extent, to deficiencies inrevenue administration. 5.26 On the supply side, seniority wage increases, and a relatively generous unemploymentinsurancesystem, combinedwith an over-extended social assistance system discourage labor market participationand mobility. As shown in the first section o f this chapter, there i s a regulation requiringautomatic increases inwages by seniority, even for cases in which previous work experience is unrelated to an employee's present job. As a result, workers are reluctant to change jobs. There are several sectors, particularly labor-intensive sectors, where turnover i s very low and the wage bill i s rising. The relatively generous unemployment insurance system and the over-extended social assistance system may create a disincentive to work. In the case of the unemployment insurance system, employment is discouraged at least for the duration of eligibility for collecting payments. In Bulgaria, unemployment benefits are equal to 60 percent o f the average during the 9 months prior to dismissal. Internationally, the average is around 30 percent. InOECD countries, which tend to 39Controlling for the effect o freforms does not change this conclusion. 40For 3-rd category of labor. 108 have the highest replacement rates, the average i s about 45 percent.41 The length o f benefit ranges from 4 months (minimum service period) to 12 months (more than 25 years o f service). 5.27 In the end, policies aimed at establishing strong income and employment security are likely to have contributed to high unemployment. The 27 percent rate o f youth unemployment in Bulgaria, for example, is double the rate in the United Kingdom and nearly three times the rate in Denmark and Ireland, countries whose labor markets have half the rate o f unemployment in Europe. Furthermore, the percentage o f unemployed without work for a year or more i s over three times the rate in the United Kingdom and Denmark. If the example o f Spain i s any indication, restrictions on temporary work contracts in Bulgaria may needlessly contribute to a significant part o f the unemployment. Reforms that liberalized the use o f such contracts in the early 1990s helped to lower the Spanish unemployment rate from around 20 percent to about 11percent at the end o f 2004. 5.28 Reforms on several fronts are needed to enhance labor market adjustment if Bulgaria is to improve its productivity and competitiveness. If Bulgaria were to meet the Lisbon targets by 2010, employment would need to increase by about 8 percent per year in 2005-2010; and labor market participation would have to increase by about 2 percent per year duringthe same period. Bulgaria would face significant challenges even if it were to meet the Lisbon targets by 2020. Lack o f reform in this area poses considerable risks for the sustainability o f growth and stability. Reforms in labor market institutions and policies are needed to enhance labor market adjustment andnetjob creation. E. Enhancinglabormarket adjustmentand employment: A reformagenda 5.29 Labor market reforms are urgently needed to better balance the enhancement of labor adjustment, the creation of employment, and the management of unemployment risks. The current framework provides large benefits for those with labor contracts and seniority, leaving behind a large share o f long-term unemployed with few job opportunities and a larger share o f workers inthe informal sector with almost no coverage o f social risks. This i s not equitable, efficient, or sustainable. While labor market reforms are essentially a social compact and are hence determined by country context reflecting social preferences and tolerance, the experiences o f other countries inthe region offer usefulpoints o f reference. 42 The reforms recommendedinthis report include, the following: e Strengthening the links between wages and salaries, andperformance. Wages should be better aligned to performance and labor productivity rather than seniority. The legislation in Bulgaria, however, includes the provision o f an annual seniority bonus, which, over time, led to the creation o f a second tier component in workers' remuneration-the cumulated seniority premium that often account for a significant share o f total pay for 41See Wayne Vroman, "International Evidence on Unemployment Compensation Prevalence and Costs," 2004, Urban Institute, United States. 42The 2004 Report o fthe EUEconomy by the European Commission shows that the labor market reforms o fthe United Kingdom, Denmark, the Netherlands, and Ireland have been successful interms o f employment and labor market participation. 109 older workers. Many firms discount the seniority premium from the basic salary, and pay a total wage that i s the same for workers in the same occupation and with same skills regardless o f their cumulated work experience. But this practice i s not possible in labor intensive industries (e.g. textile) where many workers are at the minimumwage, reducing job creation especially among older workers. The seniority premium should be folded into the basic wage in a once-and-for-all change, and the practice and portability o f the seniority premium should be eliminated. Similarly, given the wide variance in productivity by sectors and firms, the links betweenwages and salaries and performance should be strengthened at the firm level rather than being mandated these at the sectoral or economy-wide level. e Adopting policies for aflexible use offixed-term contracts and working schedules. The current Bulgarian labor code includes fairly restrictive rules on working time. In particular, there are precise rules regarding daily hours and weekly holidays and only limited possibilities are offered to firms for adjusting working time during the year. Moreover, overtime work i s forbidden by the law, except for certain cases enumerated by the law. Creating greater flexibility in working time is important for firms to face fluctuation in demand. One option would be to follow the experience o f several EU countries with the annualization o f working time. Under annualization, employers set normal weekly hours at varying levels over the year, subject to a fixed annual total. Only when these limits are exceeded do overtime premia become payable. The modulation o f hours over the year may be organized in different ways. The full schedule o f hours may be fixed inadvance, or there may be a working-time corridor o f minimumand maximum hours o f work, with overtime payments beingpaid when the average level o f weekly or daily hours, taken over the year, exceeds an agreed limit.43 e Increasing entry and exit from employment based on performance and economic rationale beyond firm closure and allowing more flexible terms of hiring and firing owing to fluctuations in production levels, performance, and absenteeism. The current rules are highly protective o f employment. As a result, net job creation is limited and firms carry large costs due to idle capacity (see Chapter 2). The benefits o f current rules accrue to those with labor contracts and seniority at the expense o f workers and young graduates seeking jobs, and lead to low levels o f competitiveness. Rules for entry and exit from employment should be linked to economic rationales including adjustment in production levels, performance, and absenteeism. e Eliminating disincentives to work. Two areas o f reform need to be implemented. Relatively large unemployment benefits may have been justified during the period o f large privatization programs. However, now the focus needs to shift to enhancing labor market supply and net job creation in a growing economy. Unemployment benefits should be adjusted (currently 60 percent o f average wages) towards a level no higher than the average for OECD countries (the average i s 45 percent o f average wages). The 43 Annualization may be thought o f as a special case o f "averaging"inwhich the referenceperiodis a year. Ingeneral, three key elements inthe averaging o f working hours are the "unit" o fwork time employed as the basis of the averaging procedure, the "reference period" over which the averaging i s done, and the "limitations" that apply. 110 duration of unemployment benefits should also be reduced-the current duration o f 4 months for those with up to 3 years o f service, and 12 months for more than 25 years o f service is too high. The second area o f reform applies to the rules for dismissal and sick leave benefits. The current rules combined result in a higher net income while absent from work on sick leave than while working. The number o f sick days should be limited, and taxes on wage income should be the same whether wages are earned while sick or by working. Reducing disincentives to employment creation on the demand side. Very high payroll taxes are a major disincentive to employment creation. For type 3 jobs, payroll taxes are 43 percent of wage income, o f which 75 percent i s paid bythe employer. As discussed in the previous chapter, reducing payroll taxes i s highly desirable not only for net job creation but also for reducing the size o f the informal economy and reducing tax evasion. Several reforms, however, need to precede payroll tax reduction, including a major rationalization, and, inparticular, the reform o f social spending. Increasing labor participation by bringing a large percentage of the working-age population back into the labor market. Disincentives to work beyond retirement age should be eliminated by allowing flexible use o f temporary and part-time contracts and temporary work arrangements. To reincorporate the long-term unemployed, the government should consider relying on effective retraining programs, including on-the- job retraining. 5.30 The restructuring of public expenditure to upgrade the transport network and improve the skill content of the labor force, together with the reform of the labor market, could have a large payoff in terms of improving competitiveness, if these measures are supportedby an efficient institutionalframework. As discussed inthe next chapter, Bulgaria faces substantial challenges inaccelerating the development o f an efficient rules-based system if its policy and institutional framework i s to be supportive o f resource reallocation to improve economic performance. Further improvements in the regulatory framework are needed to bring the investment climate on a par with that o f Bulgaria's peers inthe region. Moreover, as shown in the next chapter simplifying the legislative and regulatory framework and providing an efficient judicial system are essential components o f the core agenda proposed inthis report. To the extent that the regulatory and legislative fi-amework i s simplified to support stronger competition in the domestic markets, and the judicial institutions enforce the law efficiently in terms o f cost and time, Bulgaria will be able to move towards a predictable rules-based system, therebysubstantially enhancing its economic performance. 111 6. ACCELERATING THE DEVELOPMENT OF AN EFFICIENT RULES-BASED SYSTEM 6.1 Bulgariahasbeen implementinga broadinstitutionalreformagenda anchoredin its process for EU accession. On June 15, 2004, Bulgaria provisionally closed all 31chapters o f the acquis communautaire. And in April 2005 Bulgaria's Accession Treaty was ratified by Parliament. Notwithstanding the fact that important progress has'been made over the last few years, these commitments implythat Bulgaria needs to make significant progress in institutional reforms inthe medium term. Furthermore, as shown inthe previous chapters, Bulgaria needs to improve its productivity, growth and employment performance if its economy is to move towards a Lisbon-growth path. This i s unlikely to happen without major improvements in the quality of institutions. The scope o f this chapter is limited to examining a few selected dimensions o f two areas o f institutional development central to Bulgaria's productivity and growth agenda: the regulatory framework for the functioning o f domestic markets and the judiciary system. A. Institutionaldevelopmentandeconomic performance 6.2 Institutional development is of key importance to economic performance. The importance o f institutions and governance to growth and economic development has long been recognized. Much o f the economic development o f Western high-income economies and much of the reason that they have been able to generate high levels inthe standard o f living is due to about their efforts to create institutions that support growth. 44 The market price mechanisms for resource allocation, binding contracts, tax policy and administration, the protection o f property rights, the accountability o fpublic officials, and so on, are institutions which have been developed and adopted to support the efficient use o f resources and growth as the basis for improving the standard of living. 6.3 There is a wide range of possible approaches to examining institutions and their relation to growth. North (1990) 45 defines institutions for economic performance as the formal and informal constraints on political, economic, and social interactions. Institutions for economic performance are thus an incentive structure that reduces uncertainty and promotes efficiency. At the other end o f the spectrum, institutions for growth include organizational principles, procedural mechanisms, and regulatory frameworks aimed at producing better policy choices. In practice, both the incentive framework (i.e., policies) and its management system (Le., regulations, organizations, and procedures) matter since both dimensions are closely related. 46 The effectiveness o f policies depends on the ability o f institutions to implement those policies. For example, tax policy reform may aim at facilitating investment and growth but it i s unlikely to achieve its objectives if tax administration is deficient. Similarly, institutions 44These countries include the UnitedStates, Canada, andWestern European countries. 45North,DouglassC., 1990, Institutions, Institutional Change and Economic Pe$ormance, New York:Cambridge University Press. 46Hall, Robert E., and Charles I. 1999, "Why Do Some Countries Produce So MuchMore Output per Worker Jones, Than Others?" Quarterly Journal of Economics, Vol. 114, February; Easterly, William, 2001, TheElusive Questfor Growth: Economists 'Adventures and Misadventures in the Tropics, MIT Press; Barro, Robert J., 1997, Determinants of Economic Growth: A Cross-country Empirical Study, MIT Press. contribute to reducing the uncertainty o f economic and financial transactions provided the policies being implemented support the efficient use o f resources and growth. 47 For example, a liberal trade regime supported by an efficient customs administration i s the cornerstone o f export-led growth economies. Similarly, a good budgetary process resulting in a solid fiscal performance i s likely to have a limited impact on growth performance if there are important inefficiencies in public finance policies. Or good educational attainment i s unlikely to have a solid impact on growth performance if,in the absence o f goods market distortions and open trade, labor market policies do not support labor market adjustment in the economy, with higher productivity beingcompensated byhigher wages. Figure 6.1 Institutions and Economic Performance: Institutions and per Capita Income,2002, All Countries T .. . Bulgaria NMS8 . l2 * * . 10 tI .-*s E 8 :L a 0 n - 6 i i tI4 M 9 I . - -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 Overall Governance Index Source: World Bank staff estimatesbasedon databases of the World Bank. 6.4 Institutionalquality is positivelyassociatedwith economic perf~rmance~~Figure (see 6.1). There i s evidence that institutional quality and policy quality are positively associated with economic performance. Figure 6.1 shows the relationship between real per capita income, at 1995 prices, and the index of governance in2002. The governance index i s the average o f six governance indicators which are measured in units ranging from about -2.5 to 2.5, with higher values corresponding to better governance outcomes. 49 The relationship between institutional 47Frankel, Jeffrey A., andDavidRomer, 1999,"Does Trade Cause Growth?" AmericanEconomicReview,Vol. 89, June; Berg, Andrew, andAnne Krueger,2003, "Trade, Growth, andPoverty-A Selective Survey," IMF Working Paper 03/30; Temple, Jonathan, 1999,"The New GrowthEvidence," Journalof EconomicLiterature, Vol. 37, March; 48The World Bank, 2003, World Development Report2002: Building Institutionsfor Markets, Washington. 49The six dimensionsof governanceare: voice and accountability,politicalstability, government effectiveness, regulatoryquality, rule of law, andcontrol of corruption. See D.Kaufmann, A. Kraay, andM.Mastruzzi,2003, 113 quality and economic performance shows a strong positive correlation. This positive relationship i s valid whether economic performance i s measured by differences in per capita income, growth rates, or the volatility o f growth. Moreover, the relationship i s robust, as it holds valid across several years and across countries. 6.5 Domestic and external factors play a determining role in the pace of institutional development. The country's context and its underlying social and political realities are central in determining the relative weights given to one or another dimension of institutional development. For example, Bulgaria's demographic challenge includes the fastest aging population inthe region; consequently, reforms that will increase the productivity o f those inthe working age segment o f the population are far more urgent than in countries where aging population i s not growing as rapidly. However, productivity i s unlikely to improve to its full extent without an institutional framework that secures private property rights and contracts efficiently. Inaddition to domestic factors, external anchors are also important factors affecting institutional development and performance. The process o f EU accession has contributed markedly to improvements inthe institutional framework inBulgaria and other CEEC. 6.6 Institutionaldevelopmentis centralto growthprospectsinBulgaria. Perceptions and assessments o f institutional quality play a determiningrole in attracting and retaining foreign direct investment (see Chapter 3) and in the investment climate in general (see section B o f this chapter). This is all the more important for Bulgaria since, as discussed in Chapter 1, its economy depends on foreign direct investment flows to fill the gap between domestic savings and investmentneeds and to improve its competitiveness. Similarly, as discussed in Chapter 2, for Bulgaria to move towards a Lisbon-growth path it needs to create a policy and institutional framework supportive o f attracting investmentinoutward-oriented activities. 6.7 Since 1998 Bulgaria has been implementing reforms aimed at strengthening its institutionalframework. Indeed, in the case o f Bulgaria, improvements in overall economic performance and per capita income have been accompanied by improvements in the country's policy and institutional framework (see Chapter 1). The institutional dimensions o f the reform program have included streamlining its regulatory framework to improve the functioning o f domestic goods markets; completing banking restructuring and improving the legal framework to deepen financial intermediation; strengthening public administration, public expenditure management, accountability and transparency; introducing measures to reduce corruption; and undertaking reforms aimed at improving the judiciary system. To a large extent, however, institutional reform i s at an early stage inBulgaria compared to other countries inthe region. 50 ~~ "Governance Matters 111:Governance Indicators for 1996-2002," World Bank Policy ResearchWorking Paper 3106. 50 A detaileddescription and assessmento fthe reforms implementedsince 1998 are presented inthe PAL 1, PAL 2, and PAL 3 o f the World Bank and the Regular Reports o f the EuropeanCommission. World Bank, Program Documentfor a Proposed Programmatic Adjustment Loan to the Republic of Bulgaria, January 23,2003; Program Documentfor a Proposed Second Programmatic Adjustment Loan to the Republic of Bulgaria, World Bank, May 4, 2004; Program Documentfor a Proposed Third Programmatic AGustment Loan to the Republic of Bulgaria, World Bank, May 5,2005; and European Commission, Regular Reports on Bulgaria's Progress Towards Accession, 2002, 2003,2004. 114 Figure 6.2 Governance Indicators: Bulgaria and NMS-8 A. Overall governance index: Bulgaria and NMS-8,1996-2002 1.00 - m x 0.60 0 0.40 0 8 j 1999 2000 2001 2002 2 d 3 I ~ r r 1 - i .- n i -m Bulgaria Hungary A Czech Republic f Lithuania IEstonia bland 0 Latvia Slovak Republic 0 Slovenia -~ A NMS-8 Linear (Bulgaria) -Linear (NMS-8) , Corruption perception index: Bulgaria andNMS-8,1998-20 ._ 1998 1999 2000 2002 2003 2004 +- Bulgaria +Average NMS-8 Source: World Bank staff estimates based on databases o f the World Bank and Transparency International corruption perception indexes. 6.8 Bulgaria has made progress in improving several dimensions of its institutional framework but lags behind other countries in the region. Bulgaria's overall governance index has been improving over the last few years (see Figure 6.2). Governance inBulgaria was 115 very poor in the 1990s, and Bulgaria's governance index was negative- -0.15 in 1996 and - 0.07 in 1998. Institutional reforms improved governance inthe early 2000s and the governance index had increased to 0.26 by 2002. However, as shown in Figure 6.2: B, Bulgaria faces a large gap inthe quality o f institutions relative to the New Member States. The average overall governance index o f the NMS-8 improved to 0.80 in 2002 from 0.44 in 1996. In 2002, governance indicators for Bulgaria are lower than those for the New Member States in 1996. Similarly, indicators o f perception o f corruption show that Bulgaria has been improving since 1998. Preliminary estimates o f BEEPS 2005 indicate that this trend towards improving has been maintained over the last several years. Further efforts, however, are needed to close the gap relative to NMS-8 average. 6.9 Institutionalreformhas a large potentialimpactin economicperformance. Over the last seven years Bulgaria has adopted a large number o f new laws and regulations covering a wide range o f areas, including licensing and business registration, bankruptcy, civil service, state administration, procurement, public financial accountability, anti-corruption, and the judiciary system. This overarching new legislative and regulatory framework relies on the functioning o f the judiciary system for its implementation. To the extent that the regulatory and legislative framework i s simplifiedto support stronger competition inthe domestic markets, and to the extent that judicial institutions enforce the law efficiently, in terms o f cost and time, Bulgaria will be able to move towards a predictable rules-based system. B.Domesticcompetition:The regulatoryframework 6.10 Bulgaria has been implementing reforms aimed at improving the functioning of domestic markets. These reforms have included measures to simplify and modernize the regulatory framework, measures to reduce entry constraints, and measures to improve the efficiency o f the insolvency regime. These reforms have helped improve economic performance and the business climate. The index o f economic freedom" is indicative of progress in the policy and institutional underpinnings o f the functioning o f domestic markets (see Figure 6.3). The index ranges from 4, which i s the worst rating and represents the most restrictive policy and institutional environment, to 1, which i s the best rating and represents the most supportive environment for well functioning markets. Bulgaria has made significant progress over the last few years: its index o f economic freedom improved to 3.0 in 2004 from 3.6 in 1998 (see Figure 6.3: A). 6.11 However, the pace of reforms needs to accelerate to improve the functioning of domestic markets if Bulgaria is to close the gap with the NMS-8. The New Member States continue to improve their policy and institutional fi-ameworks to facilitate a better performance of their markets. The average index o f economic freedom o f the New Member States improved to 2.3 in 2004 from 2.9 in 1998. The gap o f 0.7 between Bulgaria and the NMS-8 remained unchanged between 1998 and 2004. However, this gap i s larger when compared to highly competitive countries in the region. The difference in the index o f economic freedom is more than 1point with Estonia, Ireland, and Finland (see Figure 6.3: B). 5 'The index o f economic freedom i s produced by the Heritage Foundation and Dow Jones & Company, Inc., on an annual basis covering 161countries and is based on 50 indicators. 116 A. 1995-2004 B. 2004 2 50 2 00 2.00 1 50 1.50 - 1.oo 1.00 - __ 0.50 0.50 - 0 00 0.00 - , -Bulgaria +Nh!S-8 Source: World Bank staff estimates based on data of The Heritage Foundation and Dow Jones & Company, Inc. 6.12 Over the last few years Bulgariaimplementedseveral reforms aimed at simplifying the regulationof business activities. In2002-03, the government addressed reducing the costs o f doing business by applying a two-pronged strategy for regulatory reform. First, the Inter- ministerial Group for Optimization o f Regulatory Regimes reviewed all national-level regulations in 2002-03. The Group reviewed 361 national-level regimes coming up for parliamentary legislation as well as decrees from the line ministries. The Group prepared a program aimed at eliminating 75 and modifying 117 o f these regimes. To date, 52 regimes have been eliminated and 92 simplified. 6.13 Second, the review andsimplificationof regulationswas followedupby the adoption in 2004 of the Law on Administrative Regulation and Administrative Control on Economic Activities. To safeguard the deregulation progress and to ensure that future regulatory regimes are based on a clear rationale, and have been subjected to cost-benefit analysis, the Law on Administrative Regulation and Administrative Control on Economic Activities was enacted last year. It has been viewed by businesses as a positive development. The Law renders all existing procedures for the creation o f new licensing regimes invalid. New regulation can only be passed with parliamentary approval. The Law establishes the basic principles for the introduction o f new regulatory regimes. This would limit new regulation to the initiative o f line ministries and other government agencies. 6.14 The new Law also streamlined the language in existing laws and codes, thereby providing consistency. For example, the working group identified 19 different terms synonymous with "license" used in over 1,500 existing pieces o f legislation. These terms have been replaced with a single term, thereby providing consistency. The Law aims to define general rules and principles to underpin the introduction o f new administrative regulations for economic activities. The Law also requires new legislation to be accompanied by the assessment o f compliance costs. Finally, statutory limits on the processing o f licensing requests have been mandated. An attempt has been made to introduce the "silent consent" principle. All of these actions reduce the administrative burden on business. 117 Figure 6.4 RegulatoryEnvironment for Businesses: Selected Indicators, 2004 Entry: Stai ing a business 4. Number of proceduresto open a business B.Minimumcapital(YOof per capita income) P 140 PO x) $30 8 80 6 60 4 40 2 20 0 0 Exit: Clos ig a business C. Time of insolvency, years D. Cost, inpercent of debt 5 ", I 4 __ E.Enforcing a contract (no. of procedures) F.Percent ofbusinessesidentifyingregulations as a moderate to major problem I 45 , I 40 35 30 25 20 15 x) 5 0 Source: World Bank staff estimates based on Doing nsiness, 2005, and BEEPS 2002. 6.15. However, the regulatory reform agenda i s far from complete. Figure 6.4 shows that notwithstanding the progress made to date the regulatory framework for businesses places producers inBulgaria at a substantial disadvantage compared to their peers in other countries in the region. The number o fprocedures requiredto open abusiness inBulgaria is more numerous 118 than in selected countries in the region-11 procedures in Bulgaria compared to 6 in Estonia and 4 in Ireland (see Figure 6.4: A). The procedures include the following: (1) check whether the proposedname o fthe business already exist and get a certificate for the registeredname; (2) notarize statutes; (3) deposit capital in a bank; (4) pay the court fee at the Treasury; (5) register with the Commercial Register at the District Court; (6) publishinthe State Gazette; (7) make a company seal; (8) register with BULSTAT; (9) register with the Regional Social Security Institute within 7 days o f court registration; (10) register with the territorial tax authorities within 14 days o f registration; and (11)register for VAT. 6.16 Bulgaria has the highest minimum capital requirement as a share of per capita income in Europe (see Figure 6.4: B). The Bulgarian Company Law has borrowed extensively from the German law, adopting, among other statutes, a minimum capital requirementfor limited liability companies to open a business. Minimumcapital has been cited in a number o f studies as an outdated legal concept,52and its problems are further amplified in Bulgaria, where the required amount (5,000 lev) i s the highest in Europe as a share o f average income (Figure 6.4: B). The high capital requirement for limited liability companies prevents many would-be entrepreneurs from starting up a business. Moreover, the ability o f the entrepreneur to withdraw the capital immediately after registration is completed overrides its designed purpose. Consequently, the high minimum capital requirement serves no useful purpose other than preventingpoorer would-be entrepreneurs from starting up new businesses. 6.17 These features of business regulations for opening a business hamper the competitiveness and productivityof Bulgarian firms. Regulatory reform has accelerated in several recent EU entrants, most notably in Hungary, Latvia, Lithuania, Poland and the Slovak Republic. In January 2004, the Polish Parliament adopted the Law on Economic Freedom, which reduces the number o f licensing regimes from 9 to 5, introduces a silent consent rule in business registration and other areas o f administrative regulation, and limits the number and duration o f various inspections. By comparison, 39 licensing regimes are still in place in Bulgaria,53the silent consent applies only to a handful o f licenses, and there is no legislation on inspections. 6.18 Measures simplifying insolvency procedures have improved costs of closing a business but Bulgaria remains behind several countries in the region (see Figure 6.4: C and D). Several amendments to the Commercial Code and Civil Procedure Code were introduced inthe 2003 to reduce the time taken by insolvency procedures, to introduce binding deadlines and more stringent criteria for the licensing and dismissal of trustees, to improve the powers o f creditors, and to limit opportunities for appeals. As a result, the number o f years to execute insolvency was reduced from 3.8 to 3.3, and the cost was reduced from 18 to 14 percent of debt between 2003 and 2004. Notwithstanding this progress, closing a business inBulgaria is far more expensive than in other countries in the region. The time taken to execute insolvency inEstonia is 3 years, inHungary it is 2 years, inLatvia it is 1.1years, and inIreland it is only 0.4 years. A PHAREproject supporting further amendments to the insolvency legislation which 52World Bank, Doing Business in 2004: "Understanding Regulation," Chapter 2. 53Some experts argue that de facto there are 60 licensing regimes, however several o f them were lumped together underbroader categories. 119 would help Bulgaria to catch up with other countries in the region has been at a preparatory stage for some time without making muchprogress. 6.19 Enforcinga contract is overly complex and presents a significant obstacle to doing business in Bulgaria (see Figure 6.4: E). The procedures for contract enforcement are significantly more complex than in other countries in the region. The number o f procedures requiredto enforce a contract inBulgaria is 34 compared to 25 in Estonia, 21 inHungary, and 16 in Ireland. Furthermore, Bulgarian courts place a heavy reliance on written over oral procedures, and impose extensive requirements for legal justification at different stages o f enforcement. Most notably, appeals can be introduced at any stage o f the process. On September 23' 2003, changes to the Constitution were introduced to expedite reforms. These are discussed inthe third section o f this chapter. 6.20 In addition, court inefficienciesmagnifythe difficulties of enforcing contracts. Both banks and businesses report that court inefficiency i s a significant obstacle to doing business in Bulgaria. Enforcing a simple debt contract takes 440 days on average in Bulgaria, compared with 252 days inEUcountries, 335 inRomania, 330 inRussia, and 184 inGermany. There are no out-of-court settlement options, no specialized commercial courts, no summary proceedings or default judgments for enforcement, and extensive opportunities exist for debtors to appeal and delay the process. Through Article 348 o f the Civil Procedure Code, it takes 19 months to enforce a mortgage. The process can also, in effect, be delayed indefinitely by the debtor; this can be done through persistent rescheduling simply through not showing up. (Judicial reform is examined inthe next section.) 6.21 Continuingimplementationof deregulationshould be madepart of the core reform agenda for Bulgaria. In2002 more than one-third o f businesses report regulations as a major obstacle to doing business in Bulgaria. Regulatory reform has taken the first steps towards improving the investment climate but, as shown above, the regulatory reform agenda is extensive. The progress achieved thus far must evolve into a sustained process. According to the results o f a recent enterprise survey conducted by the World Bank, the cost o f licensing in terms o f time and monetary concerns i s still high - in 2003 the enterprises surveyed spent, on average, 26 staff days, 37.5 calendar days, and 1,279 leva to obtain licenses and permits. 6.22 Bulgaria needs to choose its approach to creating an independent body that can championand lead deregulatoryreforms. The OECD report on regulatory reform concluded that the "specialized mechanisms for overseeing regulatory reform activities.. . seem most effective when responsibility for regulatory reform i s at the ministerial level or higher.. . Experience shows that such capacities are most effective ifthey are independent from regulators (not tied to specific regulatory missions), horizontal across government, expert (have the capacity for independent judgment), able to take the initiative in promoting reform, and linked to political authorities or existing centers of oversight authority (such as centers o f government and finance and trade rninistrie~)."~~There are different options for setting up an institutional framework for deregulation. In some countries this issue i s addressed by creating a Better RegulationUnit inthe government-as inthe UnitedKingdom, Australia and Canada; in others 54The OECD Report on RegulatoryReform, Synthesis, Organizationfor Economic Cooperation and Development, Paris, 1997. 120 by establishing an independent agency which screens all proposed legislation-as in Denmark and the nether land^.^^ C.Judicialreform 6.23 Bulgaria has made importantprogress in the area of judicial reform in the last two years. Well functioning judiciary institutions are necessary, to establish a rules-based system needed for markets to function, to protect economic and social rights, to provide efficient and reliable dispute resolution mechanisms, and to ensure accountability. To this end, Bulgaria has taken important first steps over the last two years. In September 2003, amendments and additions to the Constitution of the Republic of Bulgaria, central to judicial reform, were promulgated. InMarch 2004, the National Assembly enacted consequential amendments to the Judiciary Act to bringit into line with the amendments and additions to the Constitution. These are major steps towards an appropriate framework for the judiciary in terms o f immunity, mandate, and replaceability. The amendment to Article 123 o f the Constitution now limits the immunity o f members o f the judiciary to acts undertaken in the exercise o f their judiciary prerogatives. The addition o f Paragraph 5 to Article 129 o f the Constitution on mandates limits senior judicial appointments to five years with eligibility for a second five year period in office. For other members o f the judiciary, irreplaceability i s provided under an amendment to Paragraph 3 o f Article 129 o f the Constitution by decision o f the Supreme Judicial Council after five years o f service and taking into account a performance appraisal. Irreplaceability i s further limited in that dismissal will occur on completion o f 65 years o f age; upon resignation; upon entry into force o f a prison sentence; or following a serious violation or systematic failure to perform their official duties, as well as actions bringing disrepute to the judicial system. Moreover, the Supreme Judicial Council approved regulations on the terms and conditions for the holding o f competitions for the appointment o f magistrates. It also approved an Anti- Corruption Program for the Judiciary. The program envisions measures to strengthen the accountability o f magistrates and the introduction o f a random case assignment system to prevent corrupt practices within the judiciary, and also more transparent access to court documentation and files. The Attorney's Act was adopted inJune 2004. 6.24 Although a comprehensive legislative framework for a more independent and efficient judiciary has been put in place, significant efforts are still needed to implement the enacted changes. Regulations and management systems are needed to implement the enacted changes, particularly in the areas o f anti-corruption efforts, improvement o f the court organization, reduction of court delays, and enhancement o f the quality o fjustice. The draft regulations to support the implementation o f the Attorney's Act for qualification exams are being prepared by the Bulgarian Bar Association. The draft o f the Mediation Law passed its first reading in the Parliament in April 2004, but there was very little progress for most o f 2004. To the credit o f the Bulgarian authorities, the Mediation Law entered into force in December 2004. Now the Bulgarian authorities need to focus on its implementation. As a result, the drafting o f regulations for the Association o f National Mediators has been delayed. Similarly, there have been delays in the submission o f the Administrative Procedural 55 A number of strategic institutionaloptions for Bulgaria were considered inthe Pre-Feasibility Study "On the Establishment o f a Better Regulation UnitinBulgaria," prepared in2003 by the Institute for Market Economics. 121 Code (APC) to the Parliament and development o f a plan for the establishment o f the specialized administrative courts. The draft on the development o f performance standards for judges i s yet to be completed. 56 Figure6.5 CourtsPerformanceinTerms of Time A. Percent of firms viewingcourts as quick 2 v) 35 30 '5 20 E 25 15 E ;10 w 5 5 n 0 $ B. The time delay to collect a debt through courts I IOoo v) 2 800 U L m 600 U tE 400 0 200 0 Source: Judicial System in Transition Economies, forthcoming 2005, ECSPE of the World Bank. 6.25 Compared to other countries in the region, courts in Bulgaria are slow in general, and enforcing a contract is overly complex and time-consuming (see Figure 6.5). Businesses report that courts in Bulgaria are in general slow and, as discussed in the previous 56 Inaddition, the promulgation o f Rules for Court Administration in October 2004, introducing the concept o f a court administrator and defining the court clerical staffs performance appraisal mechanisms, ranks, etc., will bring additional professionalism and transparency to the courts while relieving judges and court presidents from time- consuming administrative tasks. 122 section, the time taken to put a debt through the courts i s considerable. Enforcing a simple debt contract takes 440 days on average in Bulgaria, compared with 252 in EU countries, 335 in Romania, 330 in Russia, and 184 in Germany. There are no out-of-court settlement options, no specialized commercial courts, no summary proceedings or default judgments for enforcement, and extensive opportunities exist for debtors to appeal and delay the process. The delays in adopting a Mediation Law place businesses in Bulgaria at a disadvantage compared to other countries inthe region. Figure 6.6 Legal Complexity and Contract Intensive Money A. Lepal extensiveness and effectiveness I Legal Ewtensiveness Legal Effectiveness U 70 60 50 40 30 20 10 0 B. Contract intensivemoney (non-cash transactionsto M2,in percent) O g 51 Eastern Europe 0O758 (Median) -Bulgaria ~~ 0 7 - \ J 065 0 6 ' I 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Iurce: Figure A: Judicial System in Transition Economies, forthcoming 2005, ECSPE oi the WorldBank; Figure B: World Bank staff estimates based on databases of the World Bank. 123 6.26 Bulgaria has the most complex system of legislation in the region and the gap between the system and its implementationis large (see Figure 6.6). This complexity is a serious problem and hampers economic performance in Bulgaria (see Figure 6.6: A). The excessive level of complexity substantially increases the costs o f doing business inBulgaria and encourages informality. The complexity also contributes to problems inenforcing contracts. As a result, Bulgarians rely more heavily on cash transactions than on more efficient, sophisticated, contract-intensive forms o f payment compared to other countries in the region (see Figure 6.6: B). Bulgaria shows low contract intensive money (CIM) or the ratio o f the non-cash fraction to M 2 over the last several years. In contrast in Estonia, which also has a Currency Board Arrangement, low inflation, and a solid fiscal position, the contract intensive money has been increasing rapidly over the last several years showing that Estonians rely more heavily on contracts than on cash to carry out transactions. Furthermore, Clague et al. (1999) argue that contract intensive money i s a relatively simple but effective measure o f the nature o f property rights enf~rcement.~' D.Acceleratingdevelopmentof a rules-basedsystem: A reformagenda 6.27 Reducing legislative and regulatory complexity. Regulatory reforms have improved the investment climate in Bulgaria. Nevertheless, Bulgaria's legislative and regulatory system remains excessively complex and places producers inBulgaria at a disadvantage relative to their peers in other countries in the region. To reduce legislative and regulatory complexity to support stronger competition in the domestic markets, the recommended reform include the following, Simplijjhg the procedures for contract enforcement. The procedures for contract enforcement are significantly more complex in Bulgaria than in other countries in the region. The number of procedures required to enforce a contract in Bulgaria is 34 compared to 25 inEstonia, 21 inHungary, and 16 inIreland. Completing streamlining of licensing regimes. Important progress was made in streamlining licensing regimes at the national level. The next step is to review and streamline sector-specific regimes and municipal regulations. The system governing construction permits needs to be modernized-businesses report that the current system represents a large burden on the costs of doing business. The reform should aim at simplifying the regimes by eliminating a multitude o f overlapping regulatory structures, including sector-specific and municipal level regulatory structures. Making registration an administrative process. The current organization o f Bulgaria's registration system is high in cost and low in efficiency. Its location in the courts i s contrary to good European practices. It also damages economic growth by reducing the capacities o f the commercial courts to deal with legal disputes. A more viable option would be the establishment o f an expert registration agency, as foreseen in the Bulstat law. 57Clague, C., P. Keefer, S. Knack, and M.Olson (1999): "Contract-Intensive Money: Contract Enforcement, PropertyRights, and Economic Performance."Journal of Economic Growth 4, 185-211. 124 a Eliminating or reducing the level of minimum capital for private limited companies. Minimumcapital has beencited ina number of studies as an outdated legal concept and its problems are further amplified in Bulgaria, where the required amount (5,000 lev) i s the highest in Europe as a share o f average income. The high capital requirement for limited liability companies prevents many would-be entrepreneurs from starting up. Moreover, the ability o f the entrepreneur to withdraw the capital immediately after registration i s completed overrides its designed purpose. As a result, minimum capital requirements serve no useful purpose other than preventing poorer would-be entrepreneurs from doing business. a Reducing the scope and complexity of the required documentation. There is a need to reduce the scope and complexity o f the documents required to establish new businesses. To this end, several measures can be implemented including the use o f standardized forms, single registration numbers, and notification and self-certification rather than authorization. Simplijjdng bankruptcy procedures. Notwithstanding reforms to date in this area, closing a business in Bulgaria i s far more expensive than inother countries inthe region. The number o f years to execute insolvency was reduced from 3.8 to 3.3, and the cost was reduced from 18 to 14 percent o f debt between 2003 and 2004. This places Bulgaria at a disadvantage relative to other countries in the region. The time taken to execute insolvency in Estonia is 3 years, inHungary it i s 2 years, in Latvia it is 1.1years, and in Ireland it i s only 0.4 years. The law needs to be amended and implemented to bring bankruptcy procedures in Bulgaria to a higher level o f efficiency and strengthen exit/entry of firms inthe economy. a Introducing statutory response times ("silence is consent" rules). As adopted, the Law on Administrative Regulation and Administrative Control on Economic Activities does not provide for the use o f the silent consent rule inbusiness registration and most types o f business licensing. a Strengthening policy making on regulations. Bulgaria adopted the Law on Administrative Regulation and Administrative Control on Economic Activities in 2004. The Law establishes principles for the introduction o f new regulatory regimes. Among other dimensions, the Law requires that proposals for new regulations be based on a clear rationale and to be subject to cost-benefit analysis. a Introducing the use of internet sites. Bulgaria should start introducing internet sites to provide information, improve the availability o f information for entrepreneurs: map all procedures and licenses; and consolidate information into a single electronic source. 6.28 Bulgarianeeds to choose of its approach to creating an independentbody that can champion and lead deregulatory reforms in a more systematic and sustained manner, There are different options for setting up an institutional framework for the effective and systematic simplification o f the legislative system and for deregulation. In some countries this issue is addressed by creating a Better Regulation Unit in the government-as in the United 125 Kingdom, Australia and Canada, inothers, by establishing an independent agency which screens all proposed legislation-as inDenmark andthe Netherlands. 6.29 Developing an efficient judicial system. Bulgaria has made important progress in the area o fjudicial reform in the last two years. Although a comprehensive legislative framework for a more efficient judiciary has been put in place, significant efforts are still needed to implementthe enacted changes. The most immediate andurgentreforms include the following: a Adopting the Mediation Law and drafting its corresponding regulations. a Establishing specialized administrative courts, provided by the new Administrative Procedure Code-the draft Code, however, is still pending the first reading in the Parliament. a Developing a comprehensive monitoring mechanism for evaluating judicial performance, including re-designing the court statistics system and providing an objective and transparent evaluation o fjudges' performance. a Fully implementing the Rulesfor Court Administration passed in October 2004. These rules introduce the concept of a court administrator and should help bringprofessionalism and transparency to court administration as well as relieve judges and court presidents from time-consuming administrative tasks. a Fully implementing a systematic approach to case management through automation, this should reduce the length of court proceedings and bring more transparency and accountability to the performance o fjudges. Upgrading judicial training-for both newly appointed magistrates and continuous education. a Strengthen the new National Institutefor Judiciary established inDecember 2003. 6.30 Building on achievements to date, moving towards a Lisbon-growth path o f higher productivity and growth by deepening integration with the EU and global markets i s within reach provided the core reform agenda i s supported by macroeconomic policies that address risks and vulnerabilities. Potential benefits are large but macroeconomic risks and vulnerabilities can not be underestimated. 126 7. MACROECONOMICRISKS AND VULNERABILITIES 7.1 Macroeconomic policies need to address risks and vulnerabilities and to support progress towards monetary integration. Bulgaria's fiscal consolidation strategy o f the last several years targeted gradual adjustment in the overall fiscal balance that was aimed at reaching a balanced budget and the external debt reduction needed to meet the Maastricht Treaty target o f a public debt to GDP ratio o f under 60 percent by 2005. The gradual adjustment in the overall balance has been accompanied by reductions in tax rates and significant reductions inpublic debt. Bulgaria reached the Maastricht Treaty target in2002 and achieved a balanced budget in 2003. As a result, Bulgaria's macroeconomic financial position has strengthened. Nevertheless, Bulgaria faces significant risks and vulnerabilities, which are manageable provided macroeconomic discipline i s maintained and the pace o f structural reforms is kept up. 7.2 Risks and vulnerabilities are involved in sustaining macroeconomic stabilization and reform achievements to date, namely: (i)the pressures towards the widening o f the external current account deficit and the rapid growth o f credit to the private sector; (ii) protracted progress in public expenditure policy reform, particularly the lack o f decisive measures that will increase the efficiency o f social expenditures and their fiscal sustainability, and further delays inoperationalizing the National Revenue Agency; and (iii) failure to maintain a flexible fiscal position to ensure gradual and sustained progress towards monetary integration. In the context of the domestic environment, there are risks of loss of momentum in the implementation o f structural reforms and weakening in macroeconomic discipline. The economy i s also vulnerable to external shocks, including further increases in oil prices and interest rate increases. There are also risks related to the adjustment from an accommodating to a more neutral monetary stance inhighincome economies and their growth prospects. A. Addressingrisks andvulnerabilities 7.3 Short-term and medium-termrisks weigh heavily on the fiscal stance. In the short run,there arerisks associatedwiththe rapidgrowth ofcredit to the private sector. Similarly, the fiscal stance should remain sufficiently flexible to respond to such external risks as greater capital flow volatility or the falling short o f FDIflows from levels needed to finance a relatively large external current account deficit. The mix o f macroeconomic policies i s further complicated by the rapid growth o f relatively short term capital flows resulting from interest rate differentials between Bulgaria and the EU-25. In particular, either sufficiently effective measures to reduce the expansion o f credit to the private are implementedor further tightening of the fiscal stance will be necessary to safeguard stability. This implies that if growth is to be sustained it will need to rely more heavily on productivity growth than on the continued expansion o f financial outlays. 7.4 In the short term, the challenge is to strengthen macroeconomic adjustment to ensure that the external current accountdeficitsremain at prudentlevels. Macroeconomic management could be further complicated if there i s greater capital flow volatility after capital controls are removed. Typically capital controls are removed well in advance o f the EU accession date. Bulgaria currently maintains controls on credit operations, real estate transactions, personal capital movements, money market instruments and other capital transactions. 58 The fiscal stance needs to remain sufficiently flexible to managepotential capital flow volatility. As mentioned inChapter 1, FDIflows have been financing the external current account deficit and making a significant contribution to rebuilding the capital stock in the Changein Credit and CAD Stock of Private Debt and Cumulative Quarterly (percentagepoints of GDP) CAD (percentof cumulative Quarterly.GDP) 7 - " 11"1- 14 A 6 4 2 12 .--0 - 5 3 5 f? n 10 P ue 4 2 B 9 a .r Q c 3 I t 6 0 5c% b) .-"2 0 4 -1 02 02 02 02 03 03 03 03 04 04 I I - 2 0 0 I 1-2-362 -1 1999 2000 2001 2002 2003 2004 Mar- Jun- Sep Dec-Mar- Jun- Sep Dec-Mar- Jun- Hsholdsand NPISHs -+ 1 -Curr, acct deficit Hsholds & NRSHs Curr. acct deficit ~ - ~~~~ ~~~ Credit to the Non-Government Sector, 1996- Credit to Private Companies by Maturity 2004 (at 1995 prices) (at 1995 prices) 300 - UCreditsto households n n *250 NCreditsto private CI BO OMorethan lyear 140 5200 cornpanies NBetween land5years me0 OCredits to SOEs 5m -._;no is0 -='1 e no QO 0 80 60 I 40 50 20 0 0 1996 1997 2998 1999 2000 2001 2002 2003 2004 1996 1997 1998 1999 2000 2001 2002 2003 2004 Source: World Bank staff estimates based on BNB and NSI lata; 2004 i s preliminary estimate. economy. With the privatization program coming to an end, stability and a solid business environment are becoming increasingly important for Bulgaria to remain competitive in attracting FDI flows. FDI flows are also affected by global economic developments and prospects. Hence, the risk o f FDIflows falling short from expected levels due to shocks can not be entirely ruled out. Furthermore, short capital inflows are growing fast-both in terms o f levels and as a share o f external balances. Short-term capital inflows more than doubled, from 649 to 1,633 million euros, between 2003 and 2004. Their share in the external current account '*There is a view that these controls are not excessive inthe case of Bulgaria. Nevertheless, given Bulgaria's vulnerability to a short fall inFDI and the rapid expansion o f short-term capital inflows, this report supports a cautious policy stance as that followed by Bulgaria over the last few years. 128 deficit increased from 27 to 55 percent during the same period (see Figure 7.1). Much o f these inflows represent money attracted to Bulgaria by the relatively high interest rates in relation to the returns on comparable securitieshvestments in the Eurozone. A sudden sharp reduction o f these rates would reduce the differentials, which could produce a rapid capital outflow causing a considerable shock to the economy. 7.5 The short-term capital inflows are resultingin the rapid expansion of private debt, underscoringnew policy challenges (see Figure 7.1). The foreign debt o f private firms and households more than quadrupled between 1999 and 2004. Households have been enjoying rapidly rising standards o f living since the crisis o f the late 199Os, and may anticipate further increases in the years to come, with EU accession rapidly approaching. Similarly, with improvements in the investment climate and in anticipation o f EU accession, private firms are undertaking investments at a rapid pace-to the extent that they have turned to the financial markets to finance productive investmentthat contributes to future productivity growth and the ability to repay obligations, both domestic and foreign." While more than a third of the foreign debt growth is FDI-related, the capacity of lenders to correctly assess the risk associated with certain consumer loans determines, to a large extent, the quality o f the rapid expansion o f credit to households and private firms. While statistics on non-performing loans currently indicate that bank portfolios remain solid, there are often time lags before bad credits manifest themselves, It i s important, therefore, to closely monitor the effectiveness o f recent measures to reduce the growth o f credit to the private sector. It is also important for banking supervision to continue to watch closely for signs o f inordinate loan portfolio deterioration. Figure 7.2 DifferenceinCPI and PPI Real EffectiveExchaneeRates (1997=100) Source: World Bank staffestimatesbasedon BNB and NSI data. 7.6 Furthermore, as part of the process of catching up in productivity and living standards, the trend towards the appreciation of the real exchange rate is likely to continue. Known as the Balassa-Samuelson effect, the relative productivity differentials betweenthe traded and non-traded goods sectors may lead to faster inflationinBulgaria than in 59Statisticsonthe composition o f consumercredit, unfortunately, are currentlyunavailable. 129 its key trade partners.60 The trend in real appreciation, o f course, need not result in currency misalignment if the competitiveness o f the economy is maintained, but in the future policymakers may feel compelled to maintain very restrictive fiscal policies in order to reach a lower inflation rate comparable to other countries inthe region, if the Balassa-Samuelson effect i s strong, putting at risk growth and employment.61 7.7 The relative productivity differentials between the traded and non-traded goods sectors indicate that the Balassa-Samuelson effect is relevant to inflationary dynamics in Bulgaria. Figure 7.2 shows the CPI-based and PPI-based real effective exchanges o f the lev. Between 2001 and the middle o f 2004, the lev appreciates in real terms by nearly 30 percent when nominal rates are deflated by CPI indices. At the same time, the PPI-based measure appreciated less than 10 percent. Since the CPI i s constructed from both traded and non-traded prices and the PPI i s thought to have a structure that i s more representative o f traded goods, the relative gap between the two indices indicates that the Balassa-Samuelson effect i s relevant to inflationary dynamics in Bulgaria. Econometric studies measuring the importance o f this effect in EU accession countries have estimated an inflationary impact ranging from 0.2 to 2.5 percentage points. However, the difference in annual growth o f the two indices exceeds 4 percent. Inthe case o f Bulgaria, at least half o f the accumulated real exchange appreciation i s related to the presence o f significant impediments to resource reallocation from lower to higher productivity activities. As shown in Chapters 2 through 6, these are mainly deficiencies that are due to limitedlabor market adjustment, poor conditions o f basic public infrastructure, relatively large skills mismatches, limited competition in the domestic markets, and the remaining deficiencies ininstitutional quality which contribute to a price bias infavor o f nontradables. B. Monetary integration 7.8 In addition to implementing reforms neededto move towards a Lisbon-growth path, enhanced macro-financial discipline is central to progressing gradually towards monetary integration. Bulgaria recently announced its decision to adopt the euro as early as possible upon EUaccession. The expected date o f adopting the Euro is mid-2009. Bulgaria's currency, the lev, was pegged to the Deutsche Mark at the exchange rate o f one new lev for one Deutsche Mark in 1999. Hence, the lev is pegged to the euro under the Currency Board Arrangement. As Bulgaria integrates with the EU and becomes a member o f the Eurozone, it will have greater opportunity to increase productivity through enhanced trade, greater competition, and transparency o f prices. Other benefits could include a lower risk premium on interest rates. However, greater monetary integration presents new policy challenges and is not without risks. The preparation for and eventual adoption o f the euro increases the need to implement reforms that will increase productivity and growth performance to accelerate real convergence, in 6o More concretely, it i s positedthat an increase inthe productivity o f tradables relative to non-tradables, if larger than inother countries, will cause an appreciation o f the real exchange rate. Higher productivity inthe tradable goods sector, where price increases are generally kept incheck through competition, will bidup wages inthat sector and, with labor being mobile, wages inthe entire economy will rise. Producers o f non-tradables will be able to pay the higher wages only ifthe relative price o f non-tradables rises, which will lead to an increase inthe overall price level inthe economy. For a discussion o f the implications for euro zone membership, see Buiter and Grafe (2002). 61 Chobanov and Sorsa (2004) examined the real exchange and found n o evidence o f significant misalignment. Preliminary analysis conducted for this report confirms this conclusion. 130 addition to enhanced financial discipline to ensure adherence to the conditions for participation inthe currency area. 7.9 There are three phases on the road to the adoption of the euro. The European Economic and Monetary Union (EMU) consists o f three stages coordinating economic policy which culminate with the adoption o f the euro, and hence with the participation inthe Stability and Growth Pact. All member states o f the European Union participate in the EMU. Twelve member states have reached the third stage and have adopted the euro. The United Kingdom and Denmark have opt-outs and need not adopt the euro. The other eleven member states are required to implement the reforms needed to move to the third stage and to eventually adopt the euro. The Copenhagen criteria require that entry members acceding to the EUbe able to meet the necessary conditions for monetary union within a period o f time-about 10 years. Prior to adopting the euro, member states participate in the European Exchange Rate Mechanism (ERM 11)for about two years. 7.10 Bulgaria aims to progress as fast as possible towards adopting the euro. Bulgaria i s currently in the EU pre-accession process whereby the candidate country makes irreversible progress towards a functioning market economy, competitiveness, and sustainable macroeconomic stability. By the time o f accession, Bulgaria will need to remove all capital controls. The second phase, which commences after membership inthe EU, entails entry inthe ERM 11. Bulgaria has taken the initiative o f announcing its official position o f keeping the Currency Board Arrangement for the minimumterm provided for by EU legislation-generally interpreted to mean two years-without devaluing the lev against the currency o f any other member state o f the Eurozone.62 At the same time, this must be accomplished without employing capital or exchange controls which are to be eliminated before accession. The final phase i s the full adoption o f the euro and the implementation o f the Stability and Growth Pact o f the EU. This limits the size o f the fiscal deficit to 3 percent o f GDP, and calls for a public debt ratio o f less than 60 percent o f GDP, low inflation, and interest rates close to the EU average.63 While Bulgaria currently meets the first two criteria, it has some distance to go on inflation, but a larger gap exists between the interest rates in Bulgaria and those o f the countries in the Eurozone. 7.11 The challenges posed by the objective of progressing towards monetary integration involve real and financial convergence. While these dimensions have been present for sometime, it i s timely to review them carefully to assess the impact o f reforms Bulgaria has implementedover the last several years interms o f asymmetries. This is because the structural differences between Bulgaria and countries in the Eurozone result in significantly different effects o f common external shocks-basically asymmetries, on both aggregate demand components and sectoral output. While they are declining, these asymmetries remain relatively large. 62The EFW requires the establishment o f a central parity and allows for a 15 percent fluctuation band. 63To be specific, the member state i s to maintain an average inflation rate of no more than 1.5 percentage points above the rates o f the three best performing members over a period o f one year. Similarly, the interest rate criterion calls for a one-year average nominal long-term rate not exceeding 2 percentage points above the three best performing members interms o f price stability. 131 7.12 A potentially important macroeconomic asymmetry vis-a-vis major trade and financial partners is suggested by simple correlations. Table 7.1 shows the simple correlations o f Bulgaria's business cycle with that o f the EU-15, the Eurozone, and selected EU member countries. Over time, the patterns o f consumption and imports in Bulgaria are becoming increasingly similar to those observed by important trade and financial partners. Business cycle correlations on consumption and on imports increased substantially in the last three years, reaching about 0.8 and 0.7, respectively. To some extent this reflects deeper integration with the EU, but it also reflects the rapid expansion o f consumption and imports, where consumer goods represent the lion's share in total imports, led by the rapid growth o f credit to the private sector. Hence, the foundation o f these trends i s less solid than it may at first appear. Indeed, the trend i s not as strong on growth rates or on sectoral growth patterns. The business cycle correlation on GDP only increased from about 0.3 to 0.4 in the last three years, and a correlation o f 0.4 reflects substantially large asymmetries inpatterns o f growth. Table 7.1 Business Cycle Correlations, 1997-2004 and 2002-2004 '' 19971 20021 19971 20021 19971 20021 19971 20021 19971 20021 2004 2004 2004 2004 2004 2004 2004 2004 2004 2004 Household Bulgaria with: GDP Final Cons. Consumption Exports Imports Bulgaria 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 EU--15 0.29 0.40 0.58 0.77 0.37 0.66 Eurozone 0.29 0.40 0.56 0.76 .... -0.24 -0.06 -0.28 -0.10 0.34 0.63 Germany 0.86 0.90 0.58 0.82 0.72 0.72 -0.12 0.07 0.59 0.69 Greece 0.37 -0.35 0.56 -0.12 .. 0.51 0.57 0.43 0.45 France -0.26 -0.17 0.29 0.48 0.16 0.36 -0.64 -0.61 0.04 0.18 Italy 0.40 0.47 0.57 0.51 0.42 0.30 0.19 0.37 0.14 0.52 Bulgaria with: Industry excl. Service, excl. Agriculture Construction Construction Services FS, RE, BS Bulgaria 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 EU-15 0.84 0.85 0.11 0.03 0.77 0.80 0.83 0.86 0.89 0.84 Eurozone 0.82 0.83 0.10 0.05 0.73 0.76 0.81 0.87 0.89 0.85 Germany 0.33 0.43 0.43 0.68 0.93 0.96 0.93 0.99 0.93 0.94 France -0.47 -0.41 -0.05 -0.30 -0.04 0.03 0.26 0.37 0.15 0.07 Italy 0.82 0.85 -0.12 -0.33 -0.03 -0.04 0.84 0.88 0.85 0.93 Note: a/ service sector excludespublic administration; FS= financial services, RE = real estate, BS = business services. Exports and imports include both merchandiseand services. Periods of coverage are: 1997:Q3 to 2004:Q3 for expenditures, and 199842to 200443; owing to data limitations, investment could not be examined. Source; World Bank staff estimates basedon data from Bulgarian National Institute of Statistics and Eurostat. 7.13 Progressing towards monetary integration implies that the core reform agenda should reduce structural differences, both real and financial, between Bulgaria and other Eurozone countries. Large structural differences can mean that macroeconomic policies aimed at promoting stability in the Eurozone as a whole need not correspond to short-run economic realities inBulgaria. A shift inmonetary policy, for example, inthe Eurozone aimed at limiting the inflationary impact o f higher energy costs, could result in a much more pronounced 132 contraction o f growth and employment in Bulgaria than in the broader currency area.64 This i s because structural differences across countries can mean significantly different effects o f common external shocks, which need not warrant the same direction or intensity o f policy action at the national level. 7.14 Asymmetries of shocks and responses to shocks seem to be declining, but the gap remains large (see Table 7.2). The statistical analysis conducted for this report indicates that synchronization is less clear in the data on inflation and the responses to shocks.65 There i s mainly a lack o f harmonization o f inflation rates and o f responsiveness to demand and supply disturbances, though inthe case o f demand the trend appears to be towards more co-movement. Table 7.2 Asymmetry of Shocks and Responsesto Shocks (correlations) a! 1997-2004 and 2000-2004 Shocks: 1997-2004 2002 - 2004 1997- 2004 2000 -2004 Bulgariawith: Growth Inflation Growth Inflation Demand Supply Demand Supply Bulgaria 1.oo 1.oo 1.oo 1.oo 1.oo 1.oo 1.oo 1.oo Euro Zone 0.29 -0.31 0.40 -0.29 0.03 -0.13 0.19 -0.24 EU-15 0.29 0.05 0.40 -0.18 0.05 0.19 0.18 0.02 France -0.26 0.10 -0.17 -0.52 -0.14 0.10 0.01 -0.21 Germany 0.86 0.24 0.90 0.15 -0.01 -0.15 0.07 0.13 Greece 0.37 0.36 -0.35 0.28 0.09 0.13 0.30 0.14 Italy 0.40 -0.32 0.47 -0.13 0.02 0.12 0.06 -0.39 Notes: a/ Demand and supply shocks are derived from the residuals of structural VAR model with growth and inflation as endogenous variables and long-run demand shocks being restricted to equal zero. Estimatesare based on quarterly data. Source: World Bank staff estimates based on data from Eurostat. 7.15 Bulgaria has a considerable distance to go in catching up in its growth path toward EUefficiency.66 This is suggested bythe differences inproductivity levels, utilization levels o f factors o f production, and the structure o f production discussed in Chapter 2. The differences between Bulgaria and its major trading partners indicate that, even if they were to face common external shocks, the disparities in the structure o f production could translate into significant asymmetric impacts on output and employment. For example, an adverse shock that affects agriculture more heavily than other sectors could lead to a relatively larger employment loss in Bulgaria than in the Eurozone, where agricultural output and employment are a fraction o f the size inBulgaria. 64 The relevance o f this example to current circumstances canbe seenfrom the recent experience inBulgaria with, large inflows o f relatively short-term capital. A strong tightening o f monetary policy inthe Eurozone could significantly reduce interest rate differentials, reverse large volumes o f capital inflow, and dampen growth in Bulgaria. 65 For a description o f the statistical procedure used to estimate the shocks, see T. Bayoumi and B.Eichengreen (1992).They extend the methodology developed by Blanchard and Quah (1989) to examine the impact o f demand and supply shocks on output and employment. Inthe long run, the cumulative impact o f demand shocks on the change inoutput i s assumed to be zero, while supply shocks are allowed to affect both prices and output inboth the short runand the long run. The speed o f harmonization could accelerate once Bulgaria has joined the EUand euro zone. Ithas been argued that membership ina currency union tends to lead to greater business cycle synchronization. Darvas and Szapary (2004) findthat inthe case of the Baltics synchronization has been weaker than inthe rest o f the new EU memberships, which may relate to initial conditions and trade patterns. 133 7.16 Greater adaptability of the economy, particularly in labor markets, is needed to minimize the negative impact on growth and employment caused by adverse circumstances. Given the lack of synchronization of economic developments in the medium term, the Currency Board Arrangement, and, later, a common monetary policy for a diverse set o f countries, the adverse consequences o f a macroeconomic shock on growth and employment can only be minimized through labor and product market flexibility. Greater flexibility is also needed to ensure that shocks do not tend to raise average unemployment over time. One way in which this could be done i s if skills are upgraded and labor market participation increases, so that, inaddition to making better use o f the available labor force and facilitating netjob creation by upgrading skills, labor market flexibility increases. 67 Similarly, reducing regulatory complexity and improving the quality o f the institutional framework to facilitate competition in the domestic markets would be needed to improve the adaptability o f product markets. Moreover, faster adaptability in labor and product markets would facilitate greater diversification o f the economy and would reduce output volatility. 67See WesternEurope by 0.Blanchard and J. Wolfers (2000) for a discussion o f the case where the skills o f the long-term unemployed are so degraded that they become unemployable intheir careers and as a group cease to have an impact on wage rates. 134 8. CONCLUSIONS 8.1 In the last seven years, Bulgaria has made impressive progress in stabilization, growth, andpoverty reduction. Since 1997 Bulgaria has been implementinga comprehensive stabilization and structural reform program anchored in its process for EU accession. As a result, Bulgaria achieved, and has maintained macroeconomic stability since 1999; growth has ranged between 4 and 5 percent per year, contributing to increasing per capita income, at PPP, from US$5,508 in 1998 to US$8,260 in 2004; and unemployment has started to decline. The share o fthe private sector inthe economy had increased to around 75 percent by 2004, and trade in goods and services relative to GDP has expanded to above 100 percent since 2002. FDI flows have increased to an average o f 7 percent o f GDP per year since 1998 compared to an average o f 1 percent o f GDP in the first nine years o f transition. Investors' confidence has improved over time, both domestically and externally, with spreads o f Brady and Euro bonds declining since 2002. During the summer o f 2004, S&P and Fitch upgraded Bulgaria's long- term foreign currency debt to investment grade rating. Taken all together, Bulgaria has the unique opportunity o f the momentum o f its economic progress to deepen its economic transformation so as to accelerate real convergence. 8.2 The core reform agenda proposedin this report is centered on achievingsuccessful integrationwith the EU and the global markets. Bulgaria should build on its achievements to date to accelerate real convergence by expanding efficiency gains and competitiveness more broadly across the economy. To this end, Bulgaria needs to continue to act on several fronts as it has done since 1998 and to implement a core agenda o f reforms that will deepen trade integration with the EU and global markets. For Bulgaria to move towards potential output growth and become more competitive, policies and economic restructuring need to focus on achieving higher economic efficiency and productivity supported by a fiscally sustainable and well targeted social protection system. 8.3 The urgency of implementing a core reform agenda centered on productivity is morepressingfor Bulgaria than for other countries in the region where the demographic discountis not as large and is not increasingas rapidly. It is estimated that in20 years time the working age population will decline by 19 percent and the population older than 64 years o f age will increase by 17 percent relative to 2005 levels. This implies that labor productivity needs to increase by about 1percentage point per year just to compensate for the impact o f the decline in the labor force on per capita income. This increase would need to be larger if labor market participation where it remains low. 8.4 This report proposes three intertwined areas for the core reform agenda: (i) improving the quantity and quality o f physical and human capital accumulation, by focusing on upgrading skills and the transport network, and improving efficiency and transparency inpublic administration; (ii)enhancing labor market adjustment and the creation o f employment by implementing labor market reforms that better balance labor adjustment, the creation o f employment, and unemployment risks, and include the rationalization o f the social protection system to enable the reduction o f payroll taxes; and (iii)implementing institutional reforms in two areas-reducing legislative and regulatory complexity in order to improve competition in the domestic market, and moving toward a rules-based system by implementingjudicial reform to increase reliance of economic transactions on contracts and their efficient enforcement. This core agenda needs to be supported by macroeconomic policies that address risks and vulnerabilities and make progress towards monetary integration. 8.5 The first core area of reform is to upgrade the quantity and quality of human capital. To increase labor productivity, and hence improve prospects o fwage income levels, as well as to continue expanding export capacity beyond unskilled labor-intensive products, there is a need to upgrade the skills o f the labor force. This would call for a reform o f the education system, including universities and vocational schools, to: (a) strengthen the links between the skills acquired in the education system and those needed in the job market through establishment o f links and feedback mechanisms between the education system and labor markets; and (b) improve the efficiency and effectiveness o f public outlays in education by reallocating resources from underused facilities and excess staffing to modernization o f curriculum, textbooks, and teaching materials; implementinga composite per student resource allocation; and linking financing to educational results in addition to enrollment levels. 8.6 The second core area of reform is to upgrade the quantity and quality of physical capital. Bulgaria's transport network needs upgrading since maritime ports and close to two- thirds o froads are inpoor condition. Priority inresource allocationwould thus need to be given to maintenance rather than new investment. The sequencing o f reforms in this area would need to be based on cost-benefit analysis, and take into account liberalization o f international trucking in the context o f EU integration. Priority reform areas would be as follows: (a) upgrading and addressing bottlenecks along road segments that carry long distance international traffic; and (b) reform o f maritime ports, including introduction o f landlocked port, with the separation o f operational and commercial functions, privatization o f port services, and concessioning o f large container and bulk terminals to private operators. 8.7 The third core area of reform is to implement major public expenditure restructuring and promote efficient and transparent public sector management. The strategy o f having the public sector focus on the efficient delivery o f public goods, including an efficient and effective social safety net, while the private sector leads in the expansion o f investment, output, and employment, has served Bulgaria well. Bulgaria should remain on this path and should accelerate progress towards this end to continue to improve the standard of living and accelerate convergence. The reforms recommended in this report focus on the following: e Implementing public expenditure restructuring to reorient resources to upgrade the transport network and trade facilitation system and to upgrade the skills of the labor force is necessary. These measures are needed ifBulgaria i s to expand outward oriented sectors and export capacity beyond unskilled labor-intensive activities. To make room for such spendingand maintain it inthe face o f an aging population, Bulgaria will need to rein inthe growth o f social spending. Reforms inthe social protection system are needed to improve its efficiency and fiscal sustainability and to enable the reduction o f very high payroll taxes, which have created significant incentives for the informal economy and tax evasion, and which limit job creation. To this end, three areas are central: deepening the 136 reforms o f the social security system, reformingthe health care system, and reforming the social assistance system. Implementing public sector reform plays a determining role in the extent to whichpublic resources are translated into actual outcomes. Three areas o f reform are highlighted. Public expenditure management reform should strengthen the links between budget allocations and performance indicators; all expenditure accounts, including those o f the judicial system, should be incorporated into the budgetary system; and reforms should aim at eliminating the practice o f capping discretionary expenditures duringthe first three quarters o f the year. Public administration reform should focus on implementing fully the new merit criteria provided by the law; upon the completion o f functional reviews o f ministries and public entities, the government should implement a program to consolidate the public sector; and the scope and number o f administrative services should be increased subject to formal competitive processes. To improve public sector governance and financial accountability, the government should simplify an overly complex legislative and regulatory framework and implement fully the new financial accountability framework and the national and sectoral anti-corruption strategies. 8.8 The fourth core area of reform is to improve labor market adjustment and the creation of employment. The labor market in Bulgaria is performing poorly. Labor market regulations are a major constraint to creating o f employment and expanding productivity gains more broadly across the economy, to benefit more fully from economic integration with the EU. The current framework provides large benefits to those with labor contracts and seniority, leaving behind a large share o f the long-term unemployed with little job opportunities and even a larger share o f workers inthe informal sector with nearly no coverage o f social risks. This i s not equitable, efficient, or sustainable. The reforms recommended in this report include the following: Increasing employmentflexibility. Entry and exit from employment should be based on performance and on economic rationale, including allowing more flexible terms inhiring and firing due to fluctuations in production levels, performance, and absenteeism. In addition to allowing a more flexible use o f fixed contracts, reforms should include eliminating restrictions on working hours, holidays and overtime work. The links between wages and salaries and performance should be strengthened at the firm level rather than being mandated at sectoral or economy-wide levels. Disincentives to work beyond retirement age should be eliminated by allowing the flexible use o f temporary and part-time contracts and temporary work arrangements. To reincorporate the long- term unemployed, the government should consider relying on effective retraining programs, including on-the-job retraining; Eliminating disincentives to work. Two areas o f reform need to be implemented: (i) unemployment benefits and their duration should be adjusted to no more than OECD levels; (ii) number o f sick days should be limited, and taxes on wage income should the bethe samewhether wages are earned while sick or byworking. 137 Reducing disincentives to employment creation on the demand side. Very high payroll taxes are a major disincentive to employment creation. For type 3 jobs, payroll taxes are 43 percent o fwage income, o fwhich 75 percent i s paid by the employer. As discussed in the previous chapter, reducing payroll taxes is highly desirable not only for net job creation but also for reducing the size o f the informal economy and reducing tax evasion. Several reforms, however, need to precede payroll tax reduction, including the major rationalization o fpublic expenditures, and reforming social spendinginparticular. 8.9 The fifth area of reform is to implement institutional reform in two areas-the reductionof legislativeand regulatorycomplexity,andjudicial reform. Regulatory reforms have improved the investment climate in Bulgaria. Nevertheless, Bulgaria's legislative and regulatory system remains excessively complex and places producers in Bulgaria at a disadvantage relative to their peers inother countries inthe region. To reduce legislative and regulatory complexity, Bulgaria needs to undertake the following: to simplify regulatory regimes and municipality-implementedregulations; to simplify procedures for contract enforcement; to make registration into an administrative process (rather than one that i s managed by the courts); to eliminate the requirement for a minimumcapital for private limitedcompanies; andto introduce statutory response times ("silence is consent" rules). To this end, Bulgaria needs to choose its approach to creating an independent body that can champion and lead deregulatory reforms in a more systematic and sustained manner.. 0 To move to an eficient rules-based system, Bulgaria needs to implement judicial reform. Although a comprehensive legislative framework for a more efficient judiciary has been put in place, significant efforts are still needed to implement the enacted changes. Regulations and management systems are needed, particularly in the areas of anti- corruption efforts, the improvement o f the court organization, the reduction o f court delays, and the enhancement o fthe quality o fjustice. 8.10 Enhancedmacro-financialdisciplineand deeper economic restructuringare needed to address risks and vulnerabilities. There have been risks and vulnerabilities involved in sustaining macroeconomic stability and reform achievements to date, namely: (i) pressures the towards the widening o f the external current account deficit and the rapid growth o f credit to the private sector; (ii) protracted progress in public finance policy reform, particularly lack o f decisive measures to increase the efficiency o f social expenditures and their fiscal sustainability, and further delays in operationalizing the National Revenue Agency; and (iii) the failure to maintain a flexible fiscal position to ensure gradual and sustained progress towards monetary integration. In the context o f the domestic environment, there are the risks o f the loss o f momentum in the implementation o f structural reforms and the weakening o f macroeconomic discipline. The economy i s also vulnerable to external shocks, including further increases in oil prices and interest rate increases. There are also risks related to the adjustment from an accommodating monetary stance to a more neutral monetary stance in high income economies and their growth prospects. 138 8.11 Short-term and medium-term risks, and gradual progress towards monetary integration,weigh heavilyon the fiscal stance. Inthe short run,there are risks associated with the rapid growth of credit to the private sector and external risks such as greater capital flow volatility or the falling short o f FDI flows from the levels needed to finance the relatively large external current account deficit. In the medium term, pensions and the rapid growth o f expenditures in health care system (particularly the financing o f hospital care), and the social assistance system represent a serious risk to budgetary stability. Furthermore, the gradual progress towards monetary integration raises additional macroeconomic challenges. Greater monetary integration has potential benefits, but it is not without risks given the large structural differences between Bulgaria and countries in the Eurozone. The preparation for, and eventual adoption of, the euro will increase the need for deeper economic restructuring, in addition to enhanced macro-financial discipline, to ensure adherence to the conditions for participation in the currency area. 8.12 Going forward, a cohesive structural reform agenda supported by disciplined macroeconomic policies is central if Bulgaria is to translate EU accession into sustained improvementin standards of living for its population. Bulgaria has gained considerable expertise in recent years in managing its economy under a hard peg. Best o f all, it has demonstrated a willingness to subject its fiscal policy to the exigencies o f the currency board and a flexibility to do so that will hold in good stead as it integrates in European monetary arrangements. Still, these challenges underscore the centrality o f implementing the structural reforms outlined inthis report to get the Bulgarian economy insync with the rest o f the currency area and able to withstand common external shocks. Reforms to substantially enhance labor market adjustment, to improve competition in the domestic markets, to enhance the dependability o f contracts and to upgrade skills and transport network, will help achieve that, while accelerating real convergence. These are central for Bulgaria to translate EU accession into sustained improvement in standards o f living for its population. 139 TechnicalAnnexes 140 ANNEX A: DETERMINANTSOFPRODUCTIVITY: FIRM- LEVEL DATA' The firm-level analysis o f growth i s based on the FIAS data (described below). This information i s used to provide alternative estimates o f total factor productivity (TFP). It i s also used to investigatewhat are the correlates o f TFP across firms and, in particular, what i s the role o f corruption and access to credit in determining TFP. EstimatingProductivity Firm productivity is an unobservable firm characteristic. However, estimates o f productivity can be recoveredas the difference between actual output and predicted output from input quantities. For example, ina regression o f the type TFP i s obtained asti=InY, -In . Firm productivity can affect input choices. This implies that the error and the regressors in (1) might be correlated and that coefficient estimates obtained with OLS might be biased. A number o f solutions have been proposed in the literature to overcome this problem. These include using firm-level fixed effects - that would deal with time-invariant individual effects - and instrumental variable strategy for input choices. Following Olley and Pakes (1996), Levinsohn and Petrin (2000) argued that using information on intermediate input choices such as demand for electricity - which tracks productivity shocks quite closely and cannot be stored - one can effectively control for productivity shocks and thus obtain consistent andunbiased estimates ofpk and PI(see discussion inHallward- Driemeier et al., 2002, and Gatti and Love, 2005). Estimates are obtained using pooled OLS across sectors and OLS estimated separately by sector. Value added TFP i s employed further on to perform robustness checks on the estimation o f the impact o f the credit variable. In order to correct for the possible simultaneity bias of OLS estimation, instrumental variable estimation is applied. This appears suitable to the structure o f the dataset and the information it contains. Instrumentsthat are correlated with the input choice but not directly with productivity are likely to perform well. Although in a cross-section setup it i s not easy to come across such variables, the survey reports information that might be suitable for this purpose. The amount spent on wages i s a natural choice o f instrument for (log) employment. Moreover, firms report the average amount of annual sales that they spent since 2001 for new buildings, machinery and equipment, the past - and as such are unlikely to be correlated with current productivity shocks - as well as for research and development. These variables were mainly determined in but are expected to predict the current level of capital. Statistically, they are good predictors o f capital and labor and the over-identifying restriction test cannot reject the hypothesis that they are not correlated with the error (productivity) in the main regression. 'This analysis was preparedby RobertaGatti. Data Description The FIAS survey contains information on 548 Bulgarian firms sampled according to a number o f criteria: (i)size, so as to be representative o f SMEs and to include a minimum of 20 percent o f large firms; (ii) sectors, so as to mirror the distributions o f Bulgarian firms across manufacturing, mining, and services; and (iii) location, so as to include firms for large cities (200), small towns (loo), and the capital, Sofia. The survey was carried out in March-April 2004. The survey reports detailed information on administrative and bureaucratic constraints to business and a limitedamount ofbalance sheet-type data.* Of the surveyed firms, 62 percent work in manufacturing, 7 percent in construction, and 28 percent in service activities. Access to selling markets i s fairly dichotomous: 63 percent o f the firms sell only domestically: these are mostly micro and small enterprises engaged in manufacturing and service^.^ Exporters sell on average more than 60 percent of their output to foreigners, indicating that there might be important costs to setting up production for export. About three-fourths o f the exporters sell to EUmarkets (to Germany, Italy, and Greece). Half o f these sell also to Eastern Europe and Central Asia markets, in particular to Macedonia, Russia, and Turkey.About 75 percent of the respondents believe that their products can withstand competition once Bulgaria will have entered the EUmarket. Foreign ownership i s highly concentrated. About 10 percent o f firms in the sample are foreign owned, with an average 71 percent o f their capital in foreign hands. Foreign owned firms are mostly large (41 percent) and medium (23 percent) and are concentrated in textile and food (for a total o f 77 percent) and services (22 percent). Firmsreport to be working at 70 percent of capacity, substantially below their full potential. Capacity differs across sectors, being higher in construction and services, but does not differ significantly between exporting and non-exporting firms. A simple regression indicates that micro firms seem to be working at substantially lower capacity than large firms and that, once size and sector are controlled for, previously publicly owned firms also work at lower than optimal capacity. Although observations within industry category are too few to allow meaningful comparisons, wood and wood products seem to be working at the lowest capacity (54 percent) while constructionand other manufacturing are working at the highest. Firmsreport to having on average two days o f inventory onhand, which, short o f a perfect on-time management of inventory, would indicate that they are ill- equipped to deal with any sudden increases in demand. This practice i s common across firms of different sizes and sectors (see also the discussion inIFC, 2004). * See "Investment Climate and Regulatory Cost Survey," IFC, for more details. Firm size is defined as follows: micro enterprises (upto 10 employees); small (between 11 and 50 employees); medium (between 51 and 100 employees); large (more than 100 employees). 142 Table A.l: DescriptiveStatistics of ComplianceVariable N o of Mean Standard Min Max obs. deviation All sectors 308 69.18 27.80 0 100 By sector: Miningand quarrying 4 68.75 14.36 50 80 Manufacturing 270 68.52 28.26 0 100 Construction 12 77.50 17.65 50 100 Services 22 72.86 28.56 0 100 Notwithstanding the recent explosion in credit growth, firms report "cost o f financing" as their main concern. Second on the list i s "competitive and informal practices." employees and wages, and o f costs as a percentage o f total sales. This information - Firmsreport the value o f total sales and fixed assets, as well as information on available for a subset o f about 190firms - is usedto obtain estimates o f TFP. The information on the administrative constraints faced by firms i s particularly rich, including data on licensing requirements, time spent with officials, unofficial payments, and the relationship with tax administration. Although a full-blown analysis o f these issues i s outside the scope of the present work, this information i s used to assess whether bribing and time spent with officials has an impact on productivity. 143 ANNEX B: DETERMINANTSOF LONGRUNGROWTH: A GROWTHACCOUNTINGEXERCISE4 Growth accounting makes it possible to decompose observed real GDP growth into the accumulation o f factors o f production and the growth rate o f total factor productivity. GDP growth can then be seen as a weighted average o f the rate o f growth o f capital and labor, and the so-called "Solow" residual that i s commonly interpreted as a measure o f technological ~hange.~A number o f studies have implemented growth accounting exercises for transition countries, including Dobrinsky (2001) and Boldrin and Canova (2003). The analysis for this CEM adds to these existing contributions along two dimensions: (i) by using updated data up to 2004; and (ii) providing a detailed account o f sources o f growth for relevant sub- by periods o f transition, notably before and after the financial crisis. This i s particularly important since the crisis years introduce substantial accounting distortions (for example, inthe measurement of investment). Total Factor ProductivityEstimation Total factor productivity (TFP) i s estimated using a standard Cobb-Douglas production function o f the type: Y,=At (KtaL )D (1) where Yt i s the aggregate output at time t,K i s the stock o fproductive capital, and Lt i s labor or productive human capital. a i s the share o f capital in GDP and p captures the returns to scale. Total factor productivity i s measured by A, and represents the unexplained portion o f growth after decomposing it into inputs o f production-labor and capital. Taking logs and differentiating (1) gives the growth in aggregate output Y as a function o f the dynamics o f input factors and TFP: Y A K The growth accounting was performed by Roberta Gatti. Venelin Boshnakov provided excellent assistance. Ana Otilia Nutu and Catalin Pauna provided useful discussions and initial computations. There are important caveats to interpreting growth accounting results for policy purposes. First, as we have mentioned, TFP is computed as a residual a growth once the accumulation o f capital and labor are accounted for. The accumulation o f both factors is likely to be measured with error. It i s difficult to quantify capture the quality o f the labor force (although we try to correct for this problem using a measure o f human capital). Moreover, we do not have information on capacity utilization over time or on the initial capital stock. Second, it is unwise to interpret the decomposition o f the results causally. This i s an accounting exercise that does not correct for the fact that increases in efficiency cause themselves certain choices o f inputs.Finally, TFP growth is estimated under the hypotheses o f constant returns to scale and perfect competition. However, if there are increasing returns to scale, increases in inputs translate into more than proportional increases in output. When estimating TFP using the standard assumptions, part o f this increase in output would erroneously be attributed to technological growth (Gosh and Kraay, 2000). where X =-dt . dX Output and employment data are available at constant prices for the period 1990-2004. To construct the capital stock series we employ the perpetual inventory model, by accumulating over time annual investment, net o f depreciation. Capital stock dynamics i s captured by K, =KO I,,,=KO (ID) + + - where KOand K1measure the capital stock inthe two periods, Ii s investment and D i s depreciation. There seem to be no major concerns regarding the accuracy o f the measurement of the flow variables, including investment, since gross domestic investment (GDI) and gross fixed capital formation (GFCF) are computed by the Statistical Office (NSI) using relatively reliable standardized methodological norms. However, measuringthe capital stock poses substantial challenges. This i s not unique to Bulgaria, or even to the transitional economies o f Central and Eastern Europe, as it i s well knownthat measures o f physical capital at book value, as reported traditionally for fiscal purposes, are poor proxies for the size of the physical capital stock o f a company. A number of alternative hypotheses are used to obtain a measure o f initial capital in order to assess the sensitivity o f our results to asses their sensitivity to initial conditions. For a Cobb-Douglas production function with constant returns to scale ratio (in 1980) o f 1.5 (baseline 1 - consistent with much of the existing work). Next, ( p 1 ) we first present the results o f the decomposition for an initial capital to output we use actual data for gross capital formation in 1980 (as reported by NSI) to backtrack the value o f the capital-to-output ratio.6 To do so, we first assume a depreciation rate o f 7.5 percent and o f 6 percent a year (baselines 2 and 3). Capital accumulation i s then computed with the perpetual inventory method, usingactual data for gross fixed capital formation and a constant depreciation rate o f 6 percent. Our assumption on depreciation rates i s consistent with Dobrinsky (2001) but higher than 4.0 percent used, for example, by Boldrin and Canova (2003). We believe that a higher depreciation rate i s more suitable for a country like Bulgaria where, especially inthe early 1990s, the slow accumulation ofnew productive capital and the quality of the capital stock inherited from the pre-transition period resulted in a higher depreciation rate than in other Eastern European countries, such as the Baltic states. For all of the exercises, we report estimates for alphas o f 0.3,0.4, and 0.5. Labor i s taken to equal the labor force (working age individuals between 15 and 64 years o f age). However, as TFP i s estimated as a residual, it i s important to adjust the contribution o f labor to growth to account for differences in quality over time. We use information on educational attainment for the labor force and existing estimates o f returns to education to compute a measure o f human capital (H>. We compute these adjustments usingtwo alternative methods: TFP is also estimated under the hypothesis o f increasing returns to scale (,f7=1.2) for the Cobb- Douglas production function. Results are available upon request. 145 (i) The methodology employed by Loayza et al. (2004) i s applied to compute human capital H as a weighted average o f the shares Ej o f the labor force with education levelj, where the weights W, are the returns to schooling at each educational level. As a result, H = CjWjEj. Inthis computation, we use data on the structure o f the labor force by education from N S I and annual returns to education from Jolliffe (2001), who estimated a Mincerian regression using 1995 household data. From these annual returns, we compute weights as follows. The coefficient i s 1 for basic or lower education. For higher levels, the weights are computed as Wj = (1 x)( + n , - n k ) where x i s the return to one extra year o f schooling at levelj and (nj-no) i s the difference between year o f schooling at completed levelj and basic education. (ii)We follow Gosh and Kraay (2000) to compute human capital as H = e(y*s)L, where the annual return to schooling, s, i s assumed to be 6 percent, and y represents the average number of years o f schooling for the labor force (baseline 6). These adjustments are applied to all three baselines. 146 ANNEX C:A MODEL OF BULGARIA'S GROWTH CYCLE T h i s section describes the model that generated the simulations shown in Figure X o f the main text, which replicates the main aspects o f the transition. The objective o f this stylized version o f transition i s to focus on a few aspects o f the economy that are crucial in explaining the phases o f Bulgaria's growth cycle and, in particular, the speed o f the current recovery. The starting point i s a version of the standard Solow (1956) growth model. At the beginningo f the transition total output may be produced using old capital K, and old technology, A,, and new capital, K, ,and new methods o f production, A,. Total GDP is given by: (1) r,=A,K,: +A,K,q where: 0 Ai i s an index o f total factor productivity. When technology i s new, i = n ;when it is old i = 0 . New technology is more productive than old technology, i.e., A,, 2 A,, Kj, is the capital stock. If j =o capital is "old", if j =n capital is "new". At the beginning most capital is old and the dominant exchange technology is A,. The initial output collapse i s the result o f breaks in the planning system, which governs exchange transactions among economic units. This i s the phenomenon called "disorganization" by Blanchard and Kremer (1997), and i s represented by a fall in productivity from A to A,. Disorganization occurs because A i s not immediately replaced by A,, which i s linkedto new investment. ' The remaining part o f the output loss i s the result o f the immediate obsolescence o f part o f the capital stock caused by contact with higher quality goods and services. Ina nutshell, part o f Bulgaria's stock o f capital was equivalent to `black and white TVs, which could project an image and were useful while there were no "plasma TVs"0 but became obsolete the moment the consumer opportunity set expanded to include more advanced, internationally available, technologies. This implies that there will be no new investment in old capital inputs and the stock will decline over time at a pace that depends on depreciation and the ability to transform old into new capital. This i s captured inequation (2): where: 8, is the rate of depreciation ofthe old capital stock. 'Before the breakdown of planningthe technology parameter is A 2 A, The cumulative output loss occurs because there is not an instantaneous adjustment of the capital stock. Initially, most o f the output i s produced usingthe old technology and because A, i s lower than before, there is a marked fall in per capita GDP. As time goes by, GDP continues to shrink as the old stock depreciates at an annual rate of (1-6,) and the absolute output loss i s not offset by the parallel increase inoutput produced usingthe new technology. Transition occurs as new capital replaces old capital, and new capital evolves as a result of investment. New capital evolves over time according to: (3) K,, = (1- )K,,-, +I,, 4 and (4) I,,I FS, DS, + where: 6, is the rate of depreciationofthe new capital stock; I,,isinvestmentinnewcapitalinperiodt; DS, is domestic savings inperiodt; FS, i s foreign savings inperiod t. Until 1998, Bulgaria's transition was characterized by stop and go policies and political uncertainty. As a consequence, the economy suffered a long period o f output decline, with an initial collapse in output followed by a "second dip" in the middle o f the decade. The protracted recession duringthe transition would have been shorter if market rules had replaced planning rules overnight, but that was impossible. Political uncertainties delayed reform thus aggravating the negative effects caused by the existing institutional vacuum. Slow and partial reform increased the output losses, a phenomenon that Murphy, Shleifer and Vishny (1992) have termed the cost o f "partial reform." The speed of growth since the trough is determined by new investment. It would be optimal to instantaneously adjust the capital stock to the new equilibrium level but new investment i s constrained by domestic savings, the capacity to attract foreign savings and the capacity to transform old capital into new capital (see equation (4)) Inthe model, domestic savings is a constant fraction, s, of GDP, a fraction that was probably lower at the bottom of the output trough. Foreign savings are a function of expectations held by foreign investors about possible future profits. Again, the more uncertain was the political and economic environment, the lower was the capacity to attract foreign direct investment. The capacity to transform old capital into new capital in the economy is low because a large share o f investment i s irreversible and institutional barriers and regulatory constraints interfere with the reallocation o f existing factors o f production 148 to new uses. Still, we assume that there i s a fraction E o f the old capital stock that can be transformed annually into new capital. The model in equations (1)-(3) is calibrated using some simplifying assumptions and simulated. The simulated time-paths depicted in Figure X of the maintext were generated, from lowest to highest, respectively, usinginvestment rates of 15 percent, 17 percent (the 1990-2003 average for gross capital formation as a percent o f GDP in Bulgaria), and 19 percent (the 1998-2003 average for gross capital formation as a percent o f GDP inBulgaria). The path predictedby the model reproduces the U-shaped output cycle that characterizedBulgaria and other transition countries. The trough and the speed o f recovery following the initial fall o f output depend on technological parameters and on the rate of investment inthe economy. Note that with the 19 percent savings rate, the trough becomes negligible. Matlab Codefor the Simulations inFigure C.l: % This is a simple growthmodel for Bulgaria. % It intends to describe the U-shaped behavior o f GDP % duringthe transition. % GDP (y) i s producedwith old (0) and new (n)technologies % that coexist for awhile % y(t)=An*(kn(t))"alpha+Ao* (ko(t))"alpha % Old capital, ko(t), evolves according to ko(t)=( 1-deltao)*ko(t-1) % New capital, kn(t), evolves according to kn(t)=(1-deltan)*kn(t-l)+in(t) % Investment innew machinery i(t) i s determinedby % i(t)=s*y(t) YOFirstthe parameters o f the model are specified clear Yo New and Old TFP (An) and (Ao) are An=l.3; Ao= 1; Yo Capital's Income Share is alpha alpha=. 8; % New capital depreciates at a rate that is higher than old capital deltan=.04; deltao=. 1; % The savings rate is initially assumed to be 17% ,the 1990-2003 average s=O. 15; % The initial old capital stock is assumed to be 10times greater % than the initial new capital stock koO= 1000; knO=lOO; YoThe simulation is runfor 50periods T=50; % The equation we want to solve is: % kn(t)=(1-deltan)*kn(t-l)+s*(An*kn(t)"alpha+Ao*ko(t)"alpha) % This equation is valid for t=l onward. % Given t=l we have: 149 YOkn(1)=( 1-deltan)*kn(O)+s*(An*kn(l)"alpha+Ao*ko( 1)"alpha) % New initial output, F(kn(l)), F(kn(l))=kn(1)-B1-C1*kn(1)"alpha equals % where B1=(1-deltan)*kn(O)+s*Ao*ko( 1)"alpha) % and C1= s*An % once Kn(1) is solved, the problem canbe iterated to solve for the next % period's capital % Kn(2)=( 1-deltan)*kn( l)+s*(An*kn(2)"alpha+Ao*k0(2)~alpha) YOand repeated. The for t=l knl(t)=(1-deltan)*knO+s*(An*knOAalpha+Ao*koOAalpha) k o l(t)=(1-deltao)*koO end for t=2:T knl(t)=(1-deltan)*knl (t-l)+s*(An*knl(t-l)"alpha+. .. Ao *k o1(t-1)"alpha) k o l(t)=(1-deltao)*kol (t-1) end k n l = [ k n O knl(2:T)I ko1=[koO ko1(2:T)l yl=An*knl ."alpha+Ao*kol ."alpha il=s*y 1 s=O. 17; koO=lOOO; knO=lOO; T=50; for t=l kn2(t)=( 1-deltan)*knO+s*(An*knOAa1pha+Ao*koOAalpha) ko2(t)=( 1-deltao)*koO end for t=2:T kn2(t)=( 1-deltan)*kn2(t-l)+s*(An*kn2(t-l)"alpha+. .. Ao *ko2(t- 1)"alpha) ko2(t)=(1-deltao) ko2(t-1) * end kn2=[knO kn2(2:T)] ko2=[ koO ko2(2:T)] y2=An*kn2 ."alpha+Ao*ko2 ."alpha i2=s*y2 s=0.19 for t=l kn3(t)=(1-deltan)*knO+s* (An*knO"alpha+Ao *koO"alpha) ko3(t)=( 1-deltao)*koO end for t=2:T kn3(t)=(1-deltan)*kn3 (t-l)+s*(An*kn3 (t-1)"alpha+. .. Ao*ko3(t- l)/'alpha) ko3(t)=( 1-deltao)*ko3(t-l) 150 end kn3=[knO kn3(2:T)] ko3=[ koO ko3(2:T)] y3=An*kn3."alpha+Ao *ko3."alpha i3=s*y3 kn= [knl'kn2' kn3'1 k o = [kol' ko2' ko3'1 y = [yl' y2' y3'] i= [illi2' i3'] subplot(2,2,1) plot(kn) title('new capital') subplot(2,2,3) P W Y ) title('gdp') subplot(2,2,2) plot(ko) title('o1d capital') subplot(2,2,4) plot(i) title('investment') References: Blanchard, 0.and M.Kremer (1997). "Disorganization," Quarterly Journal of Economics, November, pp. 1091-1126. Murphy,K.,A. Shleifer and R.Vishny (1992). "The Transition to a Market Economy: Pitfalls o f Partial Reform," Quarterly Journal of Economics, August, 1992. Solow, R.M.(1956). "A Contribution to the Theory o f Economic Growth," QuarterlyJournal of Economics, February, pp. 65-94. 151 ANNEX D: CROSS-COUNTRYCONVERGENCEMODEL This annex presents a cross-country convergence model to examine per capita income growth in Bulgaria in the context o f transition countries and to assess the extent to which Bulgaria's growth has exceeded or fallen short o f the predictions o f a conventional cross-country convergence framework. Table D.l reports growth rates for the 28 transition countries starting from 1991. Overall, growth rates proved to be highly variable, with a standard deviation o f annual growth o f about 8.7 percentage age points, more than 10 times the average growth in the whole period. High growth variability reflects, among other things, the output effects o f the wars inthe Balkans, and o f the financial crisis and high inflation rates that at different times have plagued the region. Thus, it i s not surprisingto find a negative correlation between growth variability and average growth rates (Figure D.1).Growth variability inBulgaria was overall substantially lower than average (5.6 % age points), beingpredictably higheraroundthe years o f financial crisis. Although the initial output fall (and negative growth rates) characterized the transition in all o f the countries, marked differences in growth patterned by country groups and regions became evident early on the process. It i s standard to group transition countries according to a mix o f criteria relating to a number o f geographic, performance-based, and political characteristics. Although these groupings are certainly useful for summarizing styled facts, finding suitable comparators to assess a single country's performance i s not an easy task. For example, using geographical criteria can be misleading as regions encompass diverse experiences that are not necessarily meaningfully comparable. When compared to the performance o f other Balkan countries, Bulgaria's growth i s substantially above average. However, this reflects diverse effects, among which were wars (directly affecting Bosnia, Croatia, and Serbia) and deep financial crises (Bulgaria and Albania). Similarly, comparisons o f the performance o f EU accession countries versus non-EU accession countries, selects samples according to ex-post performance, making the drawing o f inferences problematic. Cross-country regressions are useful tools for assessing a country's growth performance without encountering these problems. Inparticular, the analysis makes it possible to compare a country's actual growth rate with its predicted values from a set o f variables that proxy for that country's structural and policy characteristics. These estimates are particularly intuitive in the context o f a simple conditional convergence framework where subsequent per capita income growth rates are predicted to be negatively correlated with initial income once the position o f a country's steady state i s suitably controlled fore8 An alternative approach compares cumulative growth rates across countries since the lowest point o f recession with the level o f income at the trough o f recession (see EBRD, 2004). However, making inferences from this approach i s problematic because the breakpoints are endogenous to the growth process. *** *** FigureD.l: Growth Variability versus. Growth Rates during the Period 1991-2003 -5 -4 -3 -2 -1 0 1 2 3 4 Annual average real growth rate of GDP per capita (%) Figure D.2: Average Annual Real Growth Rate of GDP per Capita (%),1991-2003 -*-3t 1 -. KGZ 4 - UKR v 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 Level of GDP per capita (1995 USD) in the start of transition -years 1991 to 1994 A simple correlation between initial GDP per capita in 1991 (for most countries) and subsequent growth up to 2003 does not provide any evidence of absolute (or unconditional) convergence over the whole period. It i s noted that, in this simplified 154 context, Bulgaria's performance is in line with countries that had similar initial income levels. In contrast, recent data suggest a fairly marked convergence pattern (Figure D.3) highlightingthat, over time, countries' rank inper capita income changed and, overall, a closing o f the gaps across steady state positions might have occured. Figure D.3: Average Annual Real Growth Rate of GDP per Capita (%), 2000-2003 18 TKM 16 14 12 lo-+ME ARM +KAz TJK UKR LVA +MKD 0 1 2 3 4 5 6 7 a 9 10 11 12 Year 2000 GDP per capita (thousand 1995 USD) It should be noted, however, that simple correlations between initial income and growth can be misleading as, among other factors, suitable proxies for the position o f the steady state are unaccounted for. Table D.2 reports the results o f some basic cross- country regressions for transition countries which include initial income as well as a basic set o f variables that are often used as proxies for the steady state (for example, the quality o f institutions) and for policy proxies. In view o f the wealth o f existing literature on cross-country growth determinants - world-wide and for transition countries only - the objective o f this work i s not to obtain new and exhaustive estimates for standard cross- country regressions for transition countries, but instead to use the resulting estimates to assess Bulgaria's growth performance in the context o f the other transition countries. Columns 1 and 2 report results from a small panel o f 3-year and 5-year periods, respectively, per country, while column 3 reports benchmark results for the same two periods panel for a world sample. 155 Table D.2: Cross Country Growth Regressions Average annual Average annual Average annual growthrate for the growthrate for the growth rate for the period period period Sample Transitioncountries Transition countries Whole world WPPC -0.113 -0.977 -0.761 (0.14) (0.77) (3.OO) ** open -0.009 -0.004 0.004 (0.41) (0.18) (1.28) Inflation, consumerprices -0.016 -0.01% 0.000 (annual % ) (2.05) (2.16)* (0.20) School enrollment, -0.023 0.015 -0.001 secondary(% gross) (0.50) (0.26) (0.10) Resources 0.766 1.103 (0.85) (1.02) ye1995 5.975 (3.52)** ye2000 8.113 1.763 0.305 (3.63)** (1.53) (1.04) (control of )corruption 0.408 0.880 (0.27) (2.72) ** Continent dummy: EAP -0.116 (0.20) ECA 1.572 (2.53)* SSA -1.947 (2.92)** SA -0.666 (0.85) MENA -1.284 (2.15)* LAC -1.669 (3.08)** Constant 1.419 10.320 8.156 (0.30) (1.10) (4.24)** Observations 49 41 254 R-squared 0.70 0.36 0.24 Notes: Dependent variable: GDP per capita growth (Source, WDI (2004). Regressors: (log) GDP in 1995 constant dollars (WDI, 2004); Open=import+export shares in GDP (WDI, 2004), Secondary school enrolment, gross (WDI, 2004), Resources=O, if country is resourcepoor, =1, if country has moderate level of corruption (Kaufmann et al., 2004). Robust t statistics in parentheses. * significant at 5 percent; of natural resources,=2, if the country i s richinnatural resources(Falcetti et al., 2003); Corruption=control ** significant at 1percent. 156 The panel specifications contain a parsimonious set o f regressors to capture conditional convergence.' These include: initial (log) per capita GDP; structural and institutional policies (control o f corruption from Kaufmann et al., 2004); openness (export plus import shares in GDP); stabilization policies (inflation rates), and o f human capital (as proxied by the level o f secondary education). Because o f selected data availability w e end up with a small unbalanced panel for a total o f 21 countries. It is first noted that there is a highdispersion o f growth due, among other factors, to the effects o f the oil shock in resource rich countries." Unsurprisingly, Bulgaria's growth i s below that predicted in the period 1995-1999, reflecting the disruption following the financial crisis. As a consequence o f low growth, Bulgaria fell in the country ranking of per capita income at the beginningo f the following period (2000). In 2000-03 Bulgaria's growth i s instead slightly above what is predicted on the basis o f its initial income controlling for openness level, control o f corruption, inflation and time dummies." An alternative approachis to estimate growth ina cross-sectional setup. Usingour short panel, however, allows us to better visualize how country rankings interms o f per capita income at the beginning of the period changed over time. loNote that including a variable indicating the extent to which countries are resource-rich does not substantially change the estimates. A vast literature discusses the merits of alternative estimation methods for cross-country regressions (see Caselli et al., 1996). 157 ANNEX E:THE INFORMAL ECONOMY'~ The key elements in the estimation o f the size o f the informal economy in this CEM are time series data on currency holdings and an error correction model. Information on tax compliance from the FIAS firm survey i s used to analyze the determinants of informality. Methodology Inthis study, the basis for estimating the volume ofunrecorded economic activity i s the demand-for-currency approach. This approach i s based on the pioneering work o f Cagan (1958) who correlated demand for currency and tax pressure for the UnitedStates over the period 1919-1955. Cagan's approach was further developed by Tanzi (1980, 1983) who econometrically estimated a demand for currency function for the United States for the period 1929-1980. The underlying assumption i s that transactions related to the shadow economy are undertaken in the form o f cash payments in domestic currency. Inturn, individuals and firms underreport economic activity to avoid taxation. This approach relies on the unique characteristic o fbanknotes when they are used to settle transactions. Anonymity o f payments makes banknotes the best vehicle for the settlement o f unrecorded transactions (Goodhart, 2000). Economic agents who want to hide their business activity can choose to use domestic or foreign banknotes. If the domestic currency i s sufficiently stable and agents that are involved in hidden economy transactions are mainly domestic, the demand for domestic currency can serve as a good proxy for the scope o f the unrecorded economic activity. Instandard theories o f money demand, money may be demanded for at least two reasons: as an inventory to smooth differences between income and expenditure streams, and as one among several assets in a portfolio. Since we assume that all transactions in the shadow economy are settled in the form o f cash payments, demands for currency functions have to control for conventional factors such as: income, payments habits, interest rates, exchange rates, and also the major factors causing people to work in the shadow economy. The basic regression equation for currency demand proposed by Tanzi (1983) is: = a, +a,InT, +a, In - +a3InR, +a4In (y), l2This work was prepared by Roberta Gatti, World Bank, and Kalin Hristov, Bulgarian National Bank. where In denotes natural logarithms; C/M2 is the ratio o f cash holding to current and deposit accounts; Tis a weighted average tax rate (to proxy incentives to do business in the unofficial economy); WS/Y i s a proportion o f wages and salaries in national income which i s used to capture changing money holding patterns; R i s the interest rate paid on savings deposits and, thus, a proxy for the opportunity cost o f holding cash; and Y/N is a per capita income that captures the development o f the economy. The expected sign for both per capita income and the interest rate is negative, while the expected sign for both taxes and ratio o f wages and salaries in national income i s positive. From equation (l), the estimates ofthe undergroundeconomy canbe derived as follows: 1. The predicted level o f the currency ratio C/M2can be derived from the estimated regression. Given the actual figure o f M2 for each observation within the sample, the A predictedlevel o f currency holdings can be calculated (C). 2. The next step i s to solve equation (1) assuming that the tax variable i s zero while the coefficients o f other variables remain unchanged. The estimated demand for currency A A A i s defined as (C). The difference between C and C provides information for the A A A accuracy o f the fit o f the estimated equation. The difference between C and C gives the estimation o f how much currency i s held due to tax reasons. The difference between these two estimations provides an estimation o f the "illegal currency" incirculation (eI). The difference between actual total currency incirculation (C)plus demand deposits (D), which i s money incirculation (MI), and estimated "illegal currency" yields the estimated "legal currency" in circulation for transaction purposes ( e,). Dividing GDP by legal currency gives an estimate o f the income velocity o f legal currency (VI). Assuming that illegal and legal currency have the same velocity, an estimate o f unrecorded economy can be derivedby multiplying illegal currency by the velocity o f currency. Inshort, C D M - C, =C, . Money velocity is vL = + CL GDP, - =vI so that the estimate o f the ~ informal economy i s recovered as H =7V.I Cl Descriptionof the Data Data on currency in circulation and its structure by denominations i s available from the official statistics o f the central bank. The estimation o f the equation of the demand for currency can provide useful information for inferring the volume of unrecorded economic activity. The basic underlying assumption o f this approach is that agents choose not report economic activity to avoid taxation. In this context, informal activity i s seen as a direct consequence o f hightaxes. Standard demands for currency equations based on the Tanzi model use a logarithmic specification o f a single equation in levels. Since most o f the variables 159 included in the specification are non-stationary, consistent estimates can be obtained by the single-equation error correction model (ECM). Cointegration analysis helps to establish long-run relationships among integrated variables. Single equation E C M requires weak exogeneity o f independent variables, which implies that the cointegrating vector and the feedback coefficients enter only the currency incirculation equation. Assuming one cointegration relationship among the variables o f interest and weak exogeneity o f taxes, wage income to nominal GDP ratio, real GDP per capita, interest rate, and exchange rate allows us to use a single-equation ECM. The aim of the modeling exercise is to identify the model with the best forecasting properties and to derive an estimation o f unrecorded economic activity using the methodology described in the previous section. We use quarterly data covering the period from the first quarter o f 1998 to the thirdquarter of 2003. Because of the high level o f asset and currency substitution, we estimate two separate specifications where different variables approximate the opportunity cost o f currency holdings. Inthe first equation we use the weighted interest rate on deposits. In the second equation we use the BGN/U.S.dollar exchange rate, which i s measured as the quarterly average BGN/U.S. dollar exchange rate. This variable captures the opportunity costs o f domestic currency holdings. In addition the exchange rate variable controls also for the variation inbroad money (M3) that i s caused by exchange rate volatility. Since 20 percent o f deposits are denominated in U.S. dollars, exchange rate volatility affects the volume o f broad money and currency to broadmoney ratio. The wage income variable i s taken from national accounts estimates. The variable comprises only labor income generated within the economy. Social security and health insurance contributions are accounted for inthe tax variable. Inorder to measure the real tax burdenwithin the economy, we deviate from the Tanzi (1983) specification. The tax variable that we use includes revenues from direct taxes (corporate and personal income tax), indirect taxes (value added tax, excises and customs duties) and social security and health insurance contributions. Tanzi's model uses a tax variable that measures only income tax, thus implicitly assuming that there are no incentives to hide labor or avoid indirect taxes. To capture the incentives to underreport activity in order to generate an unofficial cash flow that can be used to pay bribes, we include inthe regression a proxy for the control o f corruption inthe economy (from Kraay et al., 2003). Estimatingthe following equation, we derive a range estimation for the volume o f unrecorded economy. 160 = c o n s t a n t + a , * ~- + a , * ~ (L3)-i -+a,*~ -+a4*(yr)+a5*(x)- - p --6, *--6 t *--8 wi *yr-J4 * i +seasonals+coc+E [:3 yn 1, 2 Yn 3 Where small letters refer to natural logarithm and A i s a first difference. C-is currency incirculation A43 - i s broad money T-is the total amount o ftaxes collectedby the Government Yn -is nominal GDP Wi - i s the wage income inthe economy Yr -i s real GDP per capita -i s BNG/USD exchange rate x=i, the weightedinterest rate on deposits indomestic currency inmodel (1) and x= e, BNG/USD exchange rate inmodel (2) coc - is a variable which measures the control o f corruption inthe economy. Firm-levelEvidenceon Compliance:HypothesesandMethodology As with other illegal business practices, it is not easy to obtain data on the extent o f activity in the informal sector by firms. However, the FIAS survey contains the following question: "Recognizing the difficulties many enterprises face in fully complying with taxes and regulations, what percentage age of total sales would you estimate the typical establishment in your area o f activity reports for tax purposes?" (henceforth COMPLIANCE). Answers to questions about "typical firms in your area o f activity" are routinely interpreted as answers about one's own firm, with the advantage that respondents are willing to answer truthfully. This survey approach is common practice in work on corruption (see for example, the work o f Svensson, 2002). In the discussion that follows, we first examine the incentives to underreport economic activity and then discuss the results from an estimation o f the correlates o f the extent of informal activity. Estimation is conductedby OLS, correcting for possible heteroskedasticity inthe error structure. Inthe FIAS survey, 356 firms report informationon COMPLIANCE. On average, 88 percent o f the output appears to be reported for tax purposes. As predicted, compliance i s lowest in micro firms (81 percent ) and highest inmedium-large firms (93 percent ). Across sectors, compliance is lowest in services (86 percent ) and highest in construction (94 percent ). Inthis context, regression analysis i s useful to disentangle the individual contributions o f the different determinants o f informality. In particular, we investigate three competing hypotheses: (i) ratehypothesis.Consistentwiththediscussionintheprevioussection,we Tax investigate whether being subject to a higher tax rate amounts to an incentive to avoid taxation. In the data we are able to discriminate between sole proprietor firms, which pay personal taxation with a rate up to 29 percent (the highest 161 bra~ket),'~and incorporated firms, which in 2004 paid the 19.5 percent corporate tax rate. Inour sample, 76 percent o f firms are in~orporated.'~ (ii)Enforcement hypothesis. The differential enforcement of tax laws can constitute a further incentive to avoid tax compliance. Anecdotal evidence suggests that it might be easier for small firms to hide their output, as outside control over their activities i s minimal. Note that the overlap o f micro and small firms with firms with sole proprietorship is far from perfect. Fifty-seven percent o f micro firms are not incorporated, while 16 percent o f small firms are not incorporated. As predicted, most o f the medium forms (94 percent) and virtually all o f the large firms are incorporated. The lack o f collinearity among incorporation (our proxy for the incentive to avoid higher tax rates) and size (especially among micro and small firms) will allow us to disentangle the two hypotheses. (iii) hypothesis.Therearemanyanglestothelinkbetweentheinformal Bribe economy and corruption (see Johnston et al., 1999). For example, firms might have to pay tax authorities in order to be able to avoid the tax burden. However, in an economy where corruption is systemic, firms might have to generate unofficial output to be able to pay bribes to conduct their ordinary business. To assess whether this is the case, we use information on bribe paying to tax authorities as well as information about the overall level o f corruption. Respondents report how regularly they must make unofficial payments "to deal with taxes and tax collection" (TAXOBST3, Q. C51g). 9 percrent of firms inthe relevant sample report paying tax authorities from time to time or more often. In separate sections, firms report unofficial payments related to registration, connection to utilities, application for permits and licenses, etc. Firms also independentlyreport the share o f output that they have to pay inbribes to government officials. Thanks to such detailed information we are able to build different measures o f corruption, including total bribe paying built as the sum o f all unofficial payments made for registration procedures (section c25), licenses and permits (section c27), land acquisition and cadastre registration (section c30), construction permits(section c32), connection to utilities (section c34) and customs (section c42). About 13percent o f firms report bribe amounts, averaging 30,000 BGN (circa US$ 12,000) for those who made payments. To smooth outlier observation, we will use the log o f this sum (LNBRIBES-ALL). In a separate section, firms report the percentage o f output they spend on bribes - an average o f 1.1 percent overall, 6 percent for those who report positive values. Finally we can combine the two sources o f information into bribe paying dummies - DBRIBEl, taking a value o f 1 if firms report payment amounts or a l3 A 2,000 BGNflat tax is imposedon small business with turnover below 50,000 BGN.Inthe sample, 35 companies reported sales below 50,000 BGN, 20 of which are sole proprietors. Note that the highest tax bracket applied in2004 to annual incomes above 7,000 BGN. All o f the non-incorporated firms reporting positive profits, reported profits above this threshold. l4 This percentage is indistinguishable from the percentage of incorporated firms inthe full sample of 548 f i r m s (77 percent). 162 non-zero percentage age o f output spent on unofficial payments; and DBRIBE2, taking a value o f 1ifthey report making unofficial payments (but not necessarily their amount) or a non-zero percentage age o f output spent on unofficial payments. All o f these measures are highly correlated and help in reducing the substantial measurement error associated with answers on corruption andbribepaying (see Table E.4). Table E.l: Summary Statistics for the Question "What Percent of Total Sales Would a FirmLikeYours Report for Tax Purposes?((32)'' by Size of the Firm and Sector of Activity Observ. Mean Std.Dev. Min Max Whole sample 356 88.7 20.4 2 100 Micro 98 81.55 24.09 4 100 Small 120 88.91 20.41 3 100 Medium 67 93.58 14.59 30 100 Large 67 93.79 17.19 2 100 Mining and quarrying 4 97.5 5 90 100 Manufacturing 221 88.71 20.15 2 100 Construction 31 94.84 15.25 20 100 Services 100 86.65 22.44 3 100 Table E.2:Cross-tabulationof the Firmsby Size andIncorporationStatus Inc. statussize Incorporated Sole proprietorship Total Micro 59 39 98 % row 60.2 39.8 100.0 % column 71.1 14.5 27.9 Small 20 100 120 % row 16.7 83.3 100.0 % column 24.I 37.2 34.I Medium 4 63 67 % row 6.0 94.0 100.0 % column 4.8 23.4 19.0 Large 0 67 67 % row 0.0 100.0 100.0 % column 0.0 24.9 19.0 Total 83 269 352 YOrow 23.6 76.4 100.0 % column 100.0 100.0 100.0 163 Table E.3: Pairwise Correlations between Bribery Variables COMPLIANCE LNBRIBES ALL DBRIBESl DBRIBES2 BRIBE1 COMPLIANCE 1.000 356 LNBRIBES-ALL -0.321* 1.000 350 537 DBRIBES1 -0.417* 0.840* 1.ooo 282 400 407 DBRIBES2 -0.354* 0.652* 0.869* 1.ooo 297 420 405 430 BRIBEl -0.260* 0.489* 0.556* 0.493* 1.ooo 273 388 395 395 395 Note: Indicates 5 percent level of significance. * Table E.4: Pairwise Correlations between the Corruption, in Corporation Status, Size of the Firm and Tax Obstacles COMPLIANCE INCORP SIZE TAXOBST ADM TAXOBST2 COMPLIANCE 1.ooo 356 INCORP -0.243* 1.000 356 548 SIZE -0.222* 0.474* 1.ooo 352 541 541 TAXOBST-ADM -0.058 0.043 0.069 1.ooo 356 548 541 548 TAXOBST2 -0.073 0.002 0.1oo* 0.213* 1.ooo 356 548 541 548 548 Note: *Indicates 5 percent level of significance. TAXOBST-ADM=1 is tax administration i s considered an obstacle; TAXOBST2=1 if treatment by tax authorities, filing for taxes, audits, etc., are an obstacle to business. Table E.5: Pairwise Correlations between the Variables for Corruption, FDI, Share of Domestic Sales and Export C52 FDI10 SELLDOMPCT EXPORT10 C52 1.ooo 356 FDIlO 0.097 1.ooo 355 545 SELLDOM-PCT -0.093 -0.210* 1.ooo 348 531 534 EXPORT10 0.090 0.187* -0.849* 1.000 348 531 534 534 Note: Indicates 5 percent level of significance. * Estimation begins with a basic specification o f the type: where the relevant correlates are included one by one. 164 ANNEX F: STATISTICAL TESTS OF LABORMARKET FLEXIBILITY This annex describes the data and three different econometric models used to test for labor market flexibility. The regression results and some additional sensitivity checks are briefly noted. Two different Vector Autoregression (VAR) models (unrestricted and restricted) were used to`test wage flexibility through the observation o f real wage and unemployment dynamics. The third model, an equation o f demand for employment, was used to examine the responsiveness o f employment to changes inthe real wage. In general, VARs provide relatively simple representations of relationships, but have proven to be useful tools for forecasting without the need for structural m0de1ing.l~ Ina VAR each of the variables is related separately to past values of itself and all other variables. The impulse responses derived from the parameter estimates can provide a useful summary o f the dynamic interactions among the variables. The restricted version of the VAR, known as Vector Error Correction (VEC), i s implemented when cointegration tests suggest the presence o f one or more long-run relations among non- stationary variables. In each case the appropriate lag structure is determined using statistical tests to ensure that the residuals o f the regression are reasonably close to being "white noise." In a VEC model, the number o f long-run relationships in the regression depends on the number o f "cointegrating" equations identified by the cointegration tests. Ifthereislabormarketflexibility,itispositedherethatthereisasignificantpositive relation between the real wage and unemployment. The employment equation examined for this report relates the number o f employed persons to factors that in principle should strongly affect hiring and dismissal. These include the level o f real GDP, the wage rate, seasonal patterns in economic activity, and the supply o f labor. Findings other than a significantly negative estimate o f the parameter relating employment and the wage, after controlling for the effect o f the other variables, would lead to the rejection o f the hypothesis that there i s labor market flexibility. Quarterly data are used in all regressions. The source o f the data i s the National Institute o f Statistics. Wages pertain to employees with contracts. Labor force surveys provide the unemployment statistics. Variables are expressed inlogs. Seasonal variations are captured by quarterly dummies. For the VARs the sample covers the third quarter o f 1997 to the last quarter o f 2004. The sample for the employment equation begins in the first quarter o f 2000, owing to the lack o f quarterly labor force numbers. ModelI:UnrestrictedVAR The regression was run in first-differences with four lags to purge the autocorrelation o f the residuals. As indicated by the regression output and the impulse `j For a review of VAR modeling see Enders(1995). responses, there appears to be no significant relation betweenunemployment and the real wage. Furthermore, over a 10-quarter time horizon, a large (2 standard deviations) "shock" to the unemployment rate would explain at most 20 percent of the variation in the real wage. Vector Autoregression Estimates Date: 02/28/05 Time: 12:02 Sample(adjusted): 1997:32004:3 Includedobservations:29 after adjusting Endpoints Standard errors in ( ) & t-statistics in [ ] D(LWAGE) D(LUNEMP) D(LWAGE( 1)) - 0.124762 -0.270553 (0.12064) (0.14079) [ 1.034171 [-I.92162] D(LWAGE(-2)) 0.206475 0.082130 (0.11972) (0.13972) [ 1.724681 [ 0.587821 D(LWAGE(-3)) 0.234125 0.176057 (0.09970) (0.11635) [ 2.348371 [ 1.513131 D(LWAGE(-4)) -0.115112 0.017149 (0.09900) (0.11554) [-I.I62741 [ 0.148421 D(LUNEMP(-1)) -0.010459 0.880624 (0.20128) (0.23491) [-0.051961 [ 3.748731 D(LUNEMP(-2)) 0.069792 -0.295847 (0.26153) (0.30523) [ 0.266861 [-0.969271 D(LUNEMP(-3)) -0.175530 0.235167 (0.24615) (0.28727) [-0.7131I ] [ 0.818631 D(LUNEMP(-4)) 0.307802 -0.208650 (0.17242) (0.20123) [ 1.785181 [-I.03689] C 0.004902 -0.000967 (0.00844) (0.00985) [ 0.580731 [-0.098131 D1 -0.077789 0.004244 (0.04611) (0.05381) [-I,686991 [ 0.078861 D2 0.039640 -0.154112 (0.03874) (0.04522) [ 1.02310] [-3.408241 166 D3 0.001962 -0.018331 (0.04733) (0.05524) [ 0.041461 [-0.331851 R-squared 0.583892 0.803325 Adj. R-squared 0.314646 0.676064 Sum sq. resids 0.023616 0.032166 S.E. equation 0.037272 0.043498 F-statistic 2.168618 6.312447 Log likelihood 61.99121 57.51087 Akaike AIC -3.447669 -3.138680 Schwarz SC -2.881892 -2.572903 Mean dependent 0.016073 -0.007520 S.D. dependent 0.045021 0.076427 Determinant Residual Covariance 2.60E-06 Log Likelihood(d.f. adjusted) 104.1687 Akaike InformationCriteria -5.528875 Schwarz Criteria -4.397320 Response to Generalized One S.D. Innovations f 2 S.E. Response of D(LWAGE) to D(LWAGE) Response of D(LWAGE) to D(LUNEMP) -,02i - 1 I , , , -.03( , , , , , , , , 1 2 3 4 5 6 7 8 9 1 0 1 2 3 4 5 6 7 8 9 10 Response of D(LUNEMP) to D(LWAGE) Response of D(LUNEMP) to D(LUNEMP) .06- '. .04- .02--.. -.___.,-------_-________ --- .oo -' ___-. 1.' ---------____ __----______- - -.02- *'.. ,_.- __--- ,_________------ \______ - 0 4 ( , I , , 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 167 Variance Decomposition Percent D(LWAGE) variance due to D(LWAGE) Percent D(LWAGE) variance due to D(LUNEMP) l 2 O I ' 0 ° ~ 80- 60- 40 - 18 2 8 3 , 4 58 , 6 7 8 9 10 Percent D(LUNEMP) variance due to D(LWAGE) Percent D(LUNEMP)variance due to D(LUNEMP) 80- 60- 40 - 20- 0 1 2 3 4 5 6 7 8 9 0 M o d e l 11. Vector E r r o r Correction Model The VEC model examined for this report brings together the concepts o f the Phillips Curve and the macro wage curve.16 Johansen tests indicated the existence o f two cointegrating relations with a trend and constant term among the real wage, unemployment, inflation, and productivity. *'The regression output suggests that wage flexibility is not statistically significant. The variation inthe wage due to the variation of the unemployment rate is relatively weak, particularly inthe early periods o f the shock, in comparison to the past variation o f the wage, but stronger than the effect o f productivity or inflation." The substitution o f GDP for productivity did not materially change the results. l6Among others, this model has been suggested by Blanchard and Katz (1997) and tested o n OECD countries by the OECD (1997). For an example of an application to a developing country, see Verner (1999). "Theinflationratecapturestheeffectofexpectations. Unlike VARs, statistical confidence bands for the impulse responses o f a VEC model are not defined. They can be simulated usingMonte Carlo techniques. 168 Vector Error Correction Estimates Date: 02/08/05 Time: 16:28 Sample(adjusted): 1997:32004:3 Included observations:29 after adjusting endpoints Standard errors in ( ) & t-statistics in [ 3 Cointegrating Eq: CointEql CointEq2 LWAGE(-1) 1.oooooo 0.000000 LCPI(-1) 0.000000 1.oooooo LUNEMP(-1) -0.027266 0.057634 (0.10192) (0.03954) [-0.267521 [ 1.457791 LPROD(-1) -0.262748 -0.383720 (0.21149) (0.08204) [-I.24238] [-4.677381 @TREND(95:1) -0.005720 -0.011402 (0.00161) (0.00062) [-3.560081 [-I8.29441 C -3.453062 -2.689757 Error Correction: D(LWAGE) D(LCPI) D(LUNEMP) D(LPR0D) CointEql -0.557101 -0.033022 0.153057 0.240624 (0.08305) (0.02843) (0.15129) (0.10090) [-6.707771 [-1.16161] [ 1.011711 [ 2.384721 CointEq2 0.297501 -0.719157 0.007597 0.107035 (0.33910) (0.11607) (0.61769) (0.41198) [ 0.877321 [-6.195991 [ 0.012301 [ 0.259811 D(LWAGE(-1)) 0.254870 -0.040146 -0.219943 -0.310037 (0.11295) (0.03866) (0.20575) (0.13723) [ 2.256391 [-I.03839] [-I.06897] [-2.259241 D(LCPI(-1)) -0.833998 0.129149 0.004483 0.573363 (0.19211) (0.06576) (0.34994) (0.23340) [-4.341241 [ 1.964071 [ 0.01281] [ 2.456591 D(LUNEMP(-1)) 0.069159 0.066727 0.526686 0.204126 (0.10398) (0.03559) (0.18940) (0.12632) [ 0.665141 [ 1.874911 [ 2.780821 [ 1.615891 D(LPROD(-1)) 0.182614 -0.284381 0.243326 -0.155292 (0.20146) (0.06896) (0.36698) (0.24476) [ 0.906441 [-4.124041 [ 0.663061 [-0.634461 C 0.031473 0.016504 -0.001934 -0.000905 (0.00635) (0.00217) (0.01157) (0.00772) [ 4.953891 [ 7.589531 [-0.16710] [-0.117291 D1 0.039521 -0.078907 0.087169 -0.367243 (0.06421) (0.02198) (0.11696) (0.07801) [ 0.615521 [-3.590471 [ 0.745321 [-4.707891 169 D2 0.139159 -0.094335 -0.007033 -0.056787 (0.10809) (0.03700) (0.19690) (0.13133) [ 1.287391 [-2.549701 [-0.035721 [-0.432411 D3 0.025956 0.009338 -0.016479 0.281945 (0.03860) (0.01321) (0.07032) (0.04690) [ 0.672361 [ 0.706721 [-0.234341 [ 6.011541 R-squared 0.778673 0.923531 0.745160 0.987659 Adj. R-squared 0.673834 0.887309 0.624446 0.981814 Sum sq. resids 0.012561 0.001472 0.041679 0.018541 S.E. equation 0.025712 0.008801 0.046836 0.031238 F-statistic 7.427317 25.49632 6.172948 168.9572 Log likelihood 71.I4511 102.2367 53.75416 65.49949 Akaike AIC -4.216904 -6.361154 -3.017529 -3.827551 Schwarz SC -3.745423 -5.889673 -2.546047 -3.356070 Mean dependent 0.016073 0.017030 -0.007520 0.015201 S.D. dependent 0.045021 0.026217 0.076427 0.231640 Determinant Residual Covariance 6.84E-14 Log Likelihood 299.4800 Log Likelihood (d.f. adjusted) 274.9543 Akaike Information Criteria -15.51409 Schwarz Criteria -13.15668 Model 111: EmploymentEquation The employment equation was estimated using labor force data to proxy for the supplyo f labor, which is requiredfor purposes o f identification. The demand for labor is captured by the level o f GDP. Autocorrelation o f the residuals i s purged by including lagged employment as an explanatory variable. In principle, if the labor market is flexible the regression results should indicate a significant inverse relationship between the real wage and employment. The regression results indicate that the level o f employment has no significant relation to the real wage, though the sign i s in the direction predicted by theory." The level o f GDP has the strongest impact on employment, and, as expected, has a positive sign. Past levels o f employment also have a positive and statistically significant impact on current employment. Seasonal factors are important. Inthis regression, labor supply has a negative and significant impact on employment, though the small coefficient indicates that this link can be ignored. Sensitivity checks using lagged supply indicate no significant relation between labor supply and employment. The findinghere i s consistent with the results o f a panel data analysis conducted by the IMF (2000), which examined a simpler employment equation for the Czech Republic, Hungary, Poland, the Slovak Republic, and Slovenia. 170 -Dependent Variable: LEMPL Method: Least Squares Date: 04/04/05 Time: 1154 Sample(adjusted): 2000:1 2004:3 Included observations: 19 after adjusting endpoints VariabIe Coefficient Std. Error t-Statistic Prob. C -2.143850 1.621972 -1.321755 0.2131 LEMPL(-1) 0.531036 0.108470 4.895712 0.0005 LWAGE -0.343847 0.358561 -0.958966 0.3582 LGDP 0.686750 0.192589 3.565884 0.0044 D1 0.232871 0.048656 4.786081 0.0006 D2 0.217246 0.042496 5.112194 0.0003 D3 0.007332 0.027362 0.267965 0.7937 LSUPPLY -0.000322 0.000145 -2.217308 0.0486 R-squared 0.971721 Mean dependent var 7.544676 Adjusted R-squared 0.953725 S.D. dependent var 0.086939 S.E. of regression 0.018702 Akaike info criterion -4.824802 Sum squared resid 0.003847 Schwarz criterion -4.427143 Log likelihood 53.83562 F-statistic 53.99660 Durbin-Watson stat 2.334297 Prob(F-statistic) 0.000000 References: Blanchard, O., and L. Katz (1997). "What We Know and Do Not Know About the NaturalRate o fUnemployment," Journal of Economic Perspectives, (11) 1, pp. 51-72. Enders, W. (1995). Applied Econometric Time Series. New York: John Wiley and Sons, Inc. IMF (2000). "Chapter IV: Accession of Transition Economies to the European Union: Prospects andPressures," WorldEconomic Outlook, October, pp. 138-174. OECD (1997). "Chapter 1: Recent Labour Market Developments and Prospects," Employment Outlook,pp. 1-25. Verner, D. (1999). "The Macro Wage Curve and Labor Market Flexibility inZimbabwe," The WorldBank, Policy Research Working Paper No. 2052. 171 MAP SECTION