Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Note 1 The World Bank MOLDOVA TRADE STUDY Analysis of Trade Competitiveness 1 Table of Contents 1. Executive Summary .................................................................................................................. 6 2. Introduction ............................................................................................................................... 9 3. Part I. Export Outcomes Analysis ........................................................................................... 10 3.1 Overall Trade Trends .............................................................................................................. 10 3.1.1 Trade growth and balance ............................................................................................ 10 3.2 Openness to Trade............................................................................................................... 14 3.2.1 Outcomes-based indicators of openness ...................................................................... 15 3.2.2 Policy-related determinants of trade openness ........................................................... 17 3.2.3 Foreign direct investment ............................................................................................ 19 3.2.4 Composition of exports ................................................................................................ 22 3.2.5 Growth orientation in products .................................................................................... 24 3.2.6 Services ....................................................................................................................... 25 3.2.7 Main export markets ................................................................................................... 26 3.2.8 Growth orientation in markets ..................................................................................... 28 3.3 Export Diversification ......................................................................................................... 29 3.3.1 Product dimension ....................................................................................................... 29 3.3.2 Market dimension ..................................................................................................... 31 3.3.3 Contribution of diversification to export growth ...................................................... 33 3.4 Quality and Sophistication ............................................................................................ 33 3.4.1 Technological classification...................................................................................... 34 3.4.2 Quality....................................................................................................................... 35 3.4.3 Quality in top export products .................................................................................. 36 3.4.4 Sophistication............................................................................................................ 39 3.5 Survival ......................................................................................................................... 41 4. Part II. Assessing Moldova’s Main Competitiveness Challenges .......................................... 43 4.1 What Factors Enhance and Constrain Moldovan Firms’ Productivity? ....................... 44 4.1.1 Productivity .................................................................................................................. 44 4.1.2 Firm-level determinants of productivity ................................................................... 44 4.1.3 Linking productivity and quality of backbone services ............................................ 45 4.2 Linking Productivity with the Business Environment and Governance ....................... 52 4.2.2 Business regulations: taxes, red tape, and trade and customs procedures ................ 52 4.2.3 Governance and institutional quality ........................................................................ 56 5. Policy Recommendations ....................................................................................................... 58 6. Appendix ................................................................................................................................. 61 List of Figures Figure 1. Evolution of Merchandise Trade, 2000-2013 ................................................................ 11 Figure 2. Relationship Between Re-exports and Exports of Transportation Services .................. 12 Figure 3. Merchandise Exports Performance, 2000-2013 ........................................................... 13 Figure 4. Evolution of Services Trade, 2000-2013 ....................................................................... 13 Figure 5. Share of Services in Total Exports ................................................................................ 14 Figure 6. Openness to Trade, 2000-2013 ...................................................................................... 15 2 Figure 7. Trade in Services as Percent of GDP vs. Income Levels .............................................. 16 Figure 8. Final Bound vs. MFN Applied Tariffs on Imports, 2013 .............................................. 17 Figure 9. Bound vs. MFN Applied Tariffs on Imports, 2013 ....................................................... 18 Figure 10. FDI Inflows as Percent of GDP, 1992-2013 ............................................................... 19 Figure 11. Benchmarking Moldova’s FDI Attraction Levels ....................................................... 20 Figure 12. FDI Inward Stock, by Sector ....................................................................................... 21 Figure 13. FDI Inward Stock by Origin, 2012 .............................................................................. 21 Figure 14. Evolution of RCA, 2000-2013 .................................................................................... 23 Figure 15. Growth Orientation in Products, 2003-2013 ............................................................... 25 Figure 16. Sectoral Composition of Services Exports .................................................................. 25 Figure 17. Main Export Destinations, 2000-2013 ........................................................................ 27 Figure 18. Export Composition by Region ................................................................................... 28 Figure 19. Growth Orientation in Markets ................................................................................... 28 Figure 20. Number of Exported Products ..................................................................................... 30 Figure 21. HHI for Products ......................................................................................................... 30 Figure 22. Share of Top Products in Total Exports, 2000-2013 ................................................... 31 Figure 23. Number of Markets, 2000-2013 .................................................................................. 31 Figure 24. HHI for Markets, 2000-2013 ....................................................................................... 31 Figure 25. Share of Top Markets in Total Exports, 2000-2013 .................................................... 32 Figure 26. Decomposition of Export Growth Along Extensive and Intensive Margins ............... 33 Figure 27. Technological Classification of Exports (Lall Classification, %) ............................... 34 Figure 28. Export Quality Index, 2000-2011 ................................................................................ 35 Figure 29. Relative Quality of Traditional Exports, 2003-2013 ................................................... 36 Figure 30. Relative Quality of New Exports, 2003-2013 ............................................................. 37 Figure 31. EXPY for Moldova and Comparators, 2000-2013 ...................................................... 40 Figure 32. Export Relationships Survival Rates (2000– 2013) .................................................... 42 Figure 33. Export Relationships Survival Rates by Sector, 1996–2012 ....................................... 42 Figure 34. Median TFP in Moldova by Region ............................................................................ 44 Figure 35. Productivity Premium of Exporters, Innovators, Foreign Firms in Moldova ............. 45 Figure 36. Median Productivity Across Sectors in Moldova ........................................................ 46 Figure 37. Perception-Based Quality of Services ......................................................................... 47 Figure 38. Perception-Based Quality of Services by Region ....................................................... 48 Figure 39. Objective Indicators of the Quality of Services .......................................................... 49 Figure 40. Objective Indicators of the Quality of Services in Moldova, by Region ................... 50 Figure 41. Government Regulations, Moldova and Comparators ................................................ 53 Figure 42. Trading Across Borders Indicators .............................................................................. 54 Figure 43. Logistics Performance Index (LPI) for Moldova and Comparators, 2014 .................. 54 Figure 44. Percentage of Firms Perceiving Customs, Trade Regulations as Obstacles ............... 55 Figure 45. Which Aspect of the Business Environment Represents the Biggest Obstacle? ......... 56 Figure 46. Productivity Losses Associated with Corruption ........................................................ 57 List of Tables Table 1. Main Destinations of Re-exports .................................................................................... 12 Table 2. Main Products Re-exported ............................................................................................ 12 Table 3. Change in Moldova’s Shares of Merchandise Exports ................................................... 23 3 Table 4. Top 10 Exported Products .............................................................................................. 24 Table 5. Top 10 Destinations ........................................................................................................ 26 Table 6. Moldova’s Top 20 Exports, 2003 and 2013 ................................................................... 34 Table 7. Financial Indicators for Moldova and Comparators ....................................................... 49 Table 8. LPI (Rank), 2007-2014 ................................................................................................... 54 List of Boxes Box 1. Re-exports in Moldova ...................................................................................................... 11 Box 2: Policy Options for the Wine Sector in Moldova ............................................................... 37 Box 3. Measuring Relative Quality of Exports Using Disaggregate Trade Data ......................... 39 Box 4. Measuring Export Sophistication ...................................................................................... 41 Box 5. Estimating the Impact of Services Inputs Quality on Firms’ Productivity ....................... 51 4 Acronyms ATP Autonomous Trade Preferences BEEPS Business Environment and Enterprise Performance Survey CIS Commonwealth of Independent States DCFTA Deep and Comprehensive Free Trade Agreement ECA Europe and Central Asia EU European Union FDI Foreign direct investment FTA Free trade agreements HHI Hirschman-Herfindahl Index MFN Most Favored Nation MIEPO Moldova Investment and Export Promotion Organization NZW New Zealand Winegrowers RCA Revealed comparative advantage TFP Total factor productivity WTO World Trade Organization 5 1. Executive Summary As a small economy, Moldova’s growth and development prospects are closely related to its performance in international and regional markets. With a strong reliance on remittances and the exports of a few commodities, Moldova remains one of the poorest and most vulnerable countries in the region. Despite strong economic growth over the last decade, Moldova’s export performance has been unremarkable. In recent years, Moldova’s economic performance has been crucially affected by developments in the Russian Federation and geopolitical tensions between the latter and the European Union (EU). The possibility of increasing economic linkages with the EU through a Deep and Comprehensive Free Trade Agreement (DCFTA) calls for a thorough assessment of trade performance. In this report we have looked at the export performance and competitiveness of the Moldovan economy. Specifically, we addressed two complementary issues: First, we assessed how exporters perform in the global marketplace by looking at four dimensions of competitiveness: growth, diversification, quality upgrading, and survival. We benchmarked Moldova’s performance against that of comparator or aspirational countries. Second, we investigated constraints on Moldova’s competitiveness focusing specifical ly on a series of supply-side factors, such as the role of backbone services, access to finance, the business environment (particularly government regulations affecting trade and governance), and institutional quality. Using firm-level data from the World Bank/EBRD Business Environment and Enterprise Performance Survey (BEEPS), we assessed Moldova’s performance in these areas and then investigated their effect on firm productivity and performance. Moldova is highly integrated into the global market place, mainly as a consequence of high import penetration. The country’s trade to gross domestic product (GDP) ratio of 125 percent is substantially higher than that of its regional comparators. In contrast to its peers, however, Moldova has failed to deepen its relative trade openness over the last decade. This is because, despite the sum of exports plus imports having expanded substantially, from US$1,250 million in 2000 to US$7,800 million in 2013, income per capita grew even more. Indeed, GDP per capita, which was US$370 in 2000, reached US$2,470 in 2013. Even when measured in real terms, income per capita growth was impressive, at an average rate of 5.8 percent. The recent expansion of merchandise trade in Moldova has been primarily driven by imports. Imports experienced a six-fold increase between 2000 and 2013, growing at an average annual rate of 16 percent. In comparison, exports have been less dynamic, growing at a slower average annual rate of 13 percent and experiencing a 20 percent decrease in 2009. Although export growth has recovered over the last year, the trade balance has remained negative during the 2000- 2013 period. The sizeable trade deficit has been financed mainly by inflows of remittances, and to a lesser extent, by foreign direct investment (FDI). Moldova’s export performance compares poorly with regional peers. 6 Trade in services has achieved greater dynamism than trade in goods. Indeed, during the last decade, exports of services have grown faster than imports. While the services export growth averaged 16.5 percent per year, imports grew at an annual average 13.9 percent. FDI Inflows to Moldova increased tenfold between 2002 and 2007, reaching 12 percent of GDP in 2007, Although FDI inflows were hit by the 2009 financial crisis, the share of FDI in GDP in 2013 (3.11 percent) was above the level expected given its income per capita. The composition of Moldova’s main export basket has shifted in the past decade, with exports of machinery and vegetables increasing in importance at the expense of foodstuffs and textiles. Wine, the top export product in 2003, fell to the fourth place in 2013, in part due to trade restrictions imposed by Moldova’s main wine importer, Russia, and in part due to the dynamism of other export products. While wine exports accounted for 21 percent of total exports in 2003, they represented less than 9 percent in 2013. Indeed, in 2013, the main export product was co-axial cables and other electric conductors, mainly explained by the production of a handful of foreign firms that transformed the production structure of the country. Moldova has diversified its export product base, although less than comparator countries in the region, with the number of varieties exported growing from 274 in 2000 to 393 in 2013. Export revenues remain concentrated in a few products. The top five products account for one- third of total exports, making the country significantly vulnerable to external shocks. Moldova has achieved greater diversification along the market dimension, with the count of export destinations reached by exporters growing from 63 to 103 from 2003–13. Moldova significantly decreased its reliance on markets in the Commonwealth of Independent States (CIS), shifting toward European countries. The share of Moldovan exports going to Russia decreased from 39.5 percent in 2003 to 14 percent in 2013. In recent years, new markets, such as China, Egypt, and Turkey, have gained salience among top export destinations. However, there is still significant scope for further diversification and for expanding trade ties with more distant markets. While most of Moldova’s export growth is explained by the expansion of sales of the same products to the same markets, diversification in markets has been a growing driver of export growth. The increase in market destinations explains more than half of the export growth observed in recent years. In addition, the new products introduced accounted for a sizeable 9 percent export growth in recent years. Moldova’s export basket has experienced a quality upgrade between 2003 and 2013 but remains limited when considering regional peers. However, the sophistication of Moldova’s exports has not changed substantially in the last 10 years. Indeed, the country’s export sophistication has lagged behind the country’s growth during the last decade. It is also low when compared with other countries in the region. Moldova’s export survival rates are significantly below those of comparators. The probability of a Moldovan export relationship surviving past the first year is less than 40 percent. Our findings show that the chances of survival of export relationships vary according to destination and export product. Exports to preferential trade partners—Russia, Kazakhstan and Belarus—have the highest 7 survival rates. Exports of foodstuffs and vegetables and other low-technology products have the highest survival rates. Moldova’s competitiveness remains hindered by a series of supply-side obstacles, including poor access to finance, weak infrastructure and backbone services, and an inefficient business environment. Almost 25 percent of surveyed firms viewed access to finance as a severe obstacle to their performance. These perceptions reflect objective conditions. Moldovan firms confront significantly high cost and limited access to external funds. Indeed, almost 70 percent of all investments in fixed assets are financed with internal funds. High reliance on internal funds and high collateral requirements are associated with lower total factor productivity levels. There is negative premium—or “bribe tax”—associated with corruption. Corruption has an adverse effect on firm performance and productivity. Moldovan firms that allegedly used informal payments and gifts to deal with customs procedures and with the courts had lower productivity levels than their counterparts. Specifically, firms that relied on bribes were between 6 and 7 percent less productive than their counterparts. 8 2. Introduction As a small economy, Moldova’s growth and development prospects are closely related to its performance in international and regional markets. With a strong reliance on remittances and the exports of a few commodities, Moldova remains one of the poorest and most vulnerable countries in the region. Although the country grew at an annual average of 5.6 percent between 2001 and 2008, it was severely hit by the global economic crisis, experiencing a 6-percent contraction in 2009. After recovering in 2010, the economy contracted again in 2012 as the economy was hit by a drought-induced contraction in agriculture and weaker external demand due to the Eurozone crisis. In response, the government has launched the “Moldova 2020” national development strategy, which aims to move from “a remittance- and consumption-driven model of growth to an export-driven model in order to reduce the economy’s vulnerabilities and spur job creation.”1 The expansion of exports of goods and services, at the center of the new development model, will require attracting foreign investment to facilitate participation in regional value chains, and encouraging productivity upgrading and innovation to enhance efficiency and competitiveness. This report provides an overview of Moldova’s trade competitiveness. Its objectives are twofold: (i) to present a comprehensive analysis of Moldova’s recent trade performance and (ii) to identify policy measures and interventions that can enhance the competitiveness of Moldova’s export firms and the value added of their exports. Specifically, we address the following questions: How has Moldova’s trade in goods and services evolved in the last decade? How does this performance compare to that of peer countries and groups? What are the potential opportunities for trade expansion? What policies are needed to increase trade and investment integration and to gain from it? The framework of analysis of this report draws upon the World Bank’s Trade Competitiveness Diagnostics (Reis and Farole, 2012). The report is divided into two main parts. Part I contains an exports outcome analysis. It assesses export performance along four dimensions that contribute to form a comprehensive picture of the sustainable competitiveness of the export sector, including (i) the level, growth, and market share performance of existing exports (the “intensive margin”), (ii) diversification of products and markets (the “extensive margin”), (iii) the quality and sophistication of exports (the “quality” margin), and (iv) the survival of export flows (the “sustainability margin”). Part II investigates constraints on Moldova’s competitiveness, focusing specifically on a series of supply-side factors, such as the role of backbone and input services and utilities, and access to finance; and the business environment, particularly government regulations affecting trade and governance and institutional quality. Using firm-level data from the World Bank/EBRD Business Environment and Enterprise Performance Survey (BEEPS), we assess Moldova’s performance in these areas and then investigate their effect on firm productivity and performance. 1 World Bank, 2014, “Republic of Moldova: Policy Priorities for a Private Sector Development,” p. 1. 9 The rest of the report is structured as follows: Section 2 examines overall trends in trade flows, including the growth of exports and imports, the degree of trade openness, and the recent evolution in foreign direct investment flows. In Section 3, we concentrate on export outcomes, analyzing the sectoral composition, the growth orientation, and degree of diversification of Moldovan exports. We also analyze the evolution in the quality and sophistication of exports and the survival of export relationships in different markets and sectors. In the second part of the report, we look at productivity dynamics of Moldovan firms in comparative perspective, and then investigate the impact of access to finance, backbone services, trade and customs regulations, and corruption on firm productivity. 3. Part I. Export Outcomes Analysis 3.1 Overall Trade Trends 3.1.1 Trade growth and balance Trade in goods has expanded dramatically over the last decade, with imports being substantially more dynamic than exports. The remarkable growth in trade has been driven primarily by imports, resulting in a progressively deteriorating trade balance. Indeed, since 2000, exports expanded at an average rate of 13 percent per year, while import growth averaged 16 percent per year. Exports contracted slightly in 2006, as a result of trade sanctions imposed by one of its main partners, Russia, and more markedly in 2009, following the onset of the global financial crisis, and the decline in demand from its other main partner, the EU (Figure 1a). The underwhelming performance of exports in recent years, when compared to the more dynamic imports, have led to a progressive deepening of the trade deficit, which in 2013 was 10 times larger than in 2000 (Figure 1b). Exports decreased slightly, from US$1,619 million to US$1,529 million, between 2013 and 2014. 2 Re-exports have been increasing in importance since 2005. They accounted for 33 percent of total exports in 2013. There are different definitions for the concept of “re-exports.” In this report, re-exports measure the exports of previously imported goods without any transformation. In Moldova, this category grew by almost 150 percent between 2004 and 2005, increasing their share of total exports from 1.8 to 20 percent. Re-exports represented almost 40 percent of the value of exports in 2008, as Moldova took advantage of EU special preferences not enjoyed by other CIS members. While representing an expanding external inflow, re-exports are typically associated with little job creation value addition, relative to direct exports (see Box 1). 2 According to both Comtrade and UNCTAD, exports increased by 12.3 percent in 2013 and fell by 3 percent in 2014. 10 Figure 1. Evolution of Merchandise Trade, 2000-2013 a. Imports, exports, and trade balance b. Imports, exports, and re-exports 3,000 6,000 2,000 5,000 1,000 4,000 0 Millions USD -1,000 3,000 Millions USD -2,000 2,000 -3,000 1,000 -4,000 -5,000 0 -6,000 Exports Imports Trade Balance Exports Re-exports Imports Note: The value of exports in panel (a) includes re-exports. Source: Authors’ calculations based on data from UN Comtrade. Box 1. Re-exports in Moldova 11 Re-exports have become increasingly important in Moldova. Since 2000, they have been growing at a cumulative rate of 35 percent per year. Because they imply no domestic transformation of merchandise, it has been argued that re-exports have little impact on jobs or domestic value creation. Anecdotal evidence suggests, however, that the surge in this type of low is explained by firms that prefer to rely on Moldovan transport companies’ know-how in certain markets—mainly in Russia, rather than relying on their own. Under that hypothesis, the value addition associated with these flows happens in the services sector rather than in manufacturing. Figure 2. Relationship Between Re-exports and Exports of Transportation Services 1200 450 Re-exports and exports of transportation Exports of Transportation Services, in 400 services are correlated. Figure 2 shows the Re-exports, in Million USD 1000 evolution of exports of transportation services 350 800 300 and of re-exports, broken down into those to Million USD Russia (blue) and those to the rest (red). The 250 600 series show co-movement. For example, in 200 levels, the correlation between re-exports to 400 150 Russia and exports of transportation services is 100 86 percent, while when measured in growth rates, 200 50 it is still a sizable 30 percent. 0 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Re-exports to other countries Re-exports to Russia Exports of Transportation Services Source: Authors’ calculations based on UN Comtrade and UNCTAD. Russia accounts for almost half of re-exports of Moldova. Table 1 reports the top five destinations of re-exports for 2013. It reveals that although Russia accounted for half of them expressed in values, other destinations play an important role, including Romania and Italy. In terms of the main products being re-exported, the picture is more diversified, with the top five products accounting for only 22 percent of total re-exports (Table 2). Table 1. Main Destinations of Re-exports Table 2. Main Products Re-exported Top 5 Market Share of Share of Value in Top 5 products re- Value in destinations of re- re-exports re-exports USD Mln exported in 2013 USD Mln exports in 2013 in 2013 Russian Federation 399.7 49.4 Other Medicaments 90.81 11.23 Romania 99.7 12.3 Ignition wiring sets 36.42 4.50 Italy 71.3 8.8 Other mountings 24.83 3.07 United Kingdom 37.4 4.6 Shampoos 17.60 2.18 Unspecified 31.1 3.8 Petroleum oils 15.32 1.89 Source: Authors’ calculations based on UN Comtrade. Source: Authors’ calculations based on UN Comtrade. Moldova’s merchandise export growth has lagged behind regional comparators. Moldovan merchandise exports increased 3.5 times their dollar-value between 2000 and 2013. As Figure 3 reveals, this performance pales in comparison with the dynamism exhibited by other countries in Eastern Europe and Central Asia, such as Albania and Lithuania, which saw their exports grow 12 almost nine times between 2006 and 2013. Yet, Moldovan exports have experienced a sustained increase since 2009, growing at an average rate of 20 percent between 2009 and 2013. Indeed, exports doubled during this 4-year period. Figure 3. Merchandise Exports Performance, 2000-2013 1000 900 Moldova 800 Ukraine 700 Slovakia 600 2000=100 Belarus 500 Albania 400 300 Hungary 200 Lithuania 100 Georgia 0 Poland Source: Authors’ calculations based on UN Comtrade data. On the other hand, trade in services has achieved significant dynamism in the past decade. Exports and imports of services have expanded at comparable rates (Figure 4a). Since 2000, exports in services have grown at an average rate of 16.5 percent per year. During this period, imports have also grown but at a slower pace (average 13.9 percent per year). Indeed, Moldova’s trade balance in services has been slightly negative during the period under analysis. Figure 4b shows that although Moldova’s service trade balance is negative, it has grown over the last 10 years. As a consequence of their dynamism, services exports have increased importance as a share in total exports, accounting for one-third of total exports since 2008. The share of services in the total goods and services export basket reached 34 percent in 2008, and despite declining marginally after the global financial crisis, it recovered in 2012 (Figure 5a). The relative weight of services in total exports is larger for Moldova than for all of its comparators, except for Albania and Georgia (Figure 5b). Moldova’s export growth performance in services is also on par with that of its regional comparators. Services exports growth was only slightly below the growth rates of Albania (average 17.7 percent per year), Belarus (18 percent) and Georgia (21 percent). By contrast, Moldova outperformed other countries in the region, such as Hungary, Ukraine, the Slovak Republic, and Poland. Yet, a different picture emerges when one considers the services trade balance indicator. Figure 4. Evolution of Services Trade, 2000-2013 a. Imports, exports, and balance b. Trade balance in goods and services 13 10 900 0 -10 Percent of GDP 400 Millions USD -20 -30 -100 -40 -600 -50 -60 -1,100 Imports Exports Balance Services Goods Source: Authors’ calculations based on UNCTAD data. Figure 5. Share of Services in Total Exports a. Moldova, 2000-2013 b. Moldova and comparator countries, 2000, 2007, 2013 100% 70 25 Services exports as percent of total exports 90% 60 80% 20 70% 50 60% 15 40 Percent 50% 40% 30 10 30% 20 20% 5 10% 10 0% 0 0 ALB BLR GEO HUN LTU POL MDA SVK UKR Goods Services 2000 2007 2013 Services exports growth rate 2010-13 Source: Authors’ calculations based on UNCTAD data. Source: Authors’ calculations based on UNCTAD data. 3.2 Openness to Trade How open is the Moldovan economy to trade? There are two approaches to this question. The first focuses on outcomes. A typical indicator is the trade-to-GDP ratio. It weighs the combined importance of exports and imports of goods and services in an economy, and gives an indication 14 of trade integration in the global marketplace. The problem with this indicator is that some countries may trade more than others for reasons related to their geography (landlocked countries like Moldova, or islands typically present a handicap to trade) or their economic size (firms in large economies tend to trade internationally less than small economies because they have more scope to trade among themselves). For these reasons, outcome-based indicators of openness are typically complemented with policy-based indicators. The second approach then focuses on policies—one example is the tariff barriers that the country imposes on imports of goods or services, and the tariffs faced abroad for the products it exports. We follow both approaches to describe Moldova’s openness below. 3.2.1 Outcomes-based indicators of openness Empirically, there tends to be a concave relationship between trade openness and per capita income: trade increases with income per capita, at a decreasing rate. Figure 5 shows the location of each country in the world along these two dimensions for the periods 2000–03 and 2010–13. Moldova is highly integrated into the global marketplace, exceeding the levels of openness of other countries at comparable income per capita. Figure 6a and 6b show that Moldova is well above the predicted line, revealing that the country is more open and more integrated into the global economy than expected, given its development level. Indeed, with a trade-to-GDP ratio of 125 percent, Moldova’s trade openness surpasses that of regional peers of comparable small size, such as Georgia, Albania, and Ukraine. However, the comparison between the two panels suggests that Moldova has not considerably deepened its trade integration over the last decade. The trade- to-GDP ratio increased between 2000 and 2007, reaching a peak of 145 percent, but was negatively affected by the global financial crisis in 2008–09. Figure 6. Openness to Trade, 2000-2013 a. 2000-2003 b. 2010-2013 Note: The panels plot the relationship between trade openness and GDP per capita for all countries. Relevant comparators are labeled. The curve shows the expected trade openness for a given per capita income. The white band represents the 95 percent confidence interval. Countries above (below) the confidence interval are said to be more (less) open to international trade than what their economic development implies. Source: Authors’ calculations using data from WDI and UN Comtrade. 15 By contrast, some of Moldova’s comparators, most notably Georgia, and an aspirational country such as the Slovak Republic (also a member of the EU), have experienced important growth in their trade-to-GDP ratio and have moved further away from the average given their levels of development. Although Poland and Albania continue to trade less with the world than their income level would predict, the two countries have improved their positions. Only Ukraine, understandably given the political turmoil in recent years, has seen its trade openness and integration levels slip between the earlier and more recent periods. Moldova’s high openness to trade is mainly driven by imports. The faster pace at which imports grew when compared to exports is evident when considering the exports-to-GDP and imports-to-GDP ratios separately. As Figure A 1 in the Appendix shows, exports as a share of GDP have fallen between 2000-03 and 2010-13. Export growth has lagged behind the expansion in income per capita during the last decade. Still, Moldova remains more open in terms of exports than the average country with its level of development. The opposite trend is evident in imports. Imports as a share of GDP increased during the period under investigation. Moldova has moved further away from the line of prediction, as both its imports and its income per capita grew (see Figure A 2). Despite the dynamism of Moldovan services exports, the services trade-to-GDP ratio has slightly deteriorated in the past decade. Figure 7 benchmarks Moldova’s trade (exports plus imports) in services-to-GDP ratio against other countries in the world for 2005-07 and 2010-13. The comparison of the two panels shows that while Moldova is still above the predicted line, its position has shifted downward (toward the line of prediction) in the more recent period. Although Moldova still trades more in services than countries at the same level of development, its trade-to- GDP ratio in services has not expanded as fast as its income per capita over the last decade. This slight decrease has been driven by the reduction in imports of services, which have not expanded as fast as GDP (See Figure A 3 in the Appendix). Figure 7. Trade in Services as Percent of GDP vs. Income Levels a. 2005-2007 b. 2010-2013 Source: Authors’ calculations using data from WDI. 16 3.2.2 Policy-related determinants of trade openness Openness to trade is influenced by a variety of policy choices, in particular, decisions over a country’s trade regime. A country’s tariff profile crucially affects the efficiency and competitiveness of domestic producers and exporters, particularly as production processes are increasingly fragmented and firms participate in global value chains in which they need to import inputs to process domestically and export to the rest of the world. Tariff profiles also influence market access prospects for a country’s own export products. Moldova maintains a liberal tariff regime. The simple average most favored nation applied tariff rate is amongst the world’s lowest. According to the World Trade Organization, the simple average MFN applied tariff was 4.6 percent in 2013, with average non-agricultural tariff of 3.7 percent and agricultural tariff of 10.6 percent. Moldova’s final bound tariff level for 2013 was higher: 7 percent average (14 percent for agricultural goods and 5.9 for non-agricultural products). Moreover, Moldova has free trade agreements (FTAs) with more than 40 countries, including the EU-28 member states (DCFTA), CIS states, Eastern European countries, and Turkey. In addition, Moldova has signed a number of preferential trade agreements with other advanced industrial economies, including the United States, Canada, and Japan. There are different levels of protection across sectors. Animal products and foodstuffs enjoy relatively higher levels of protection. 1. Figure 8 compares the MFN applied and final bound tariffs for different products in 2013, showing that sugars and confectionary products exhibit the highest bound tariff level (56 percent). No MFN applied tariffs surpass 25 percent. Moldova’s average tariff level is below that of the EU as well as that of some of its benchmark countries. With an average MFN applied tariff of 4.7 percent, Moldova is more open than its Western European trading partners. Its tariff structure is also more liberal than that of Belarus, Macedonia, and Ukraine (Figure 9). Moldova has lower bound tariffs than Albania and Armenia, but it has relatively higher applied tariffs than those countries. Interestingly, despite the overall liberal tariff schedule, Moldova maintains a relatively high tariff on glass containers—a crucial input into the production of one of its main export products: wine. Figure 8. Final Bound vs. MFN Applied Tariffs on Imports, 2013 17 60 Bound MFN applied 50 Percent 40 30 20 10 0 Note: *Simple Averages. Source: WTO Tariff Profile Data Base. Figure 9. Bound vs. MFN Applied Tariffs on Imports, 2013 10 9 8 7 6 Percent 5 4 3 2 1 0 Final bound MFN applied Source: WTO Tariff Profile Data Base. Moldova’s export expansion may be constrained by the relatively high tariff levels imposed by some of its main partners. The agricultural and non-agricultural tariffs imposed by some of Moldova’s main export destinations suggest Moldova has much to gain from completing the DCFTA with the EU. Since 2008, Moldova has enjoyed Autonomous Trade Preferences (ATPs) with the EU, which provide unlimited and duty-free access to the EU market for almost all products originating in Moldova. However, this agreement does not include certain agricultural products. 18 The most binding constraint to market access in the EU is related with standards. This is particularly relevant for exporters in agribusiness. For example, black caviar is certified to be shipped to the EU, but eggs for processing are in the process of getting certification, while eggs for consumption and meats are not yet compliant with EU standards. For apples, there are challenges associated with the varieties that are accepted in the EU, which are not the classic varieties produced in Moldova. Standards also apply to post-harvesting processes associated with sorting, calibration, packaging, and storage. Producers need to continue improving such processes (see Note 3 on Agribusiness for a thorough analysis of the challenges faced by that sector). 3.2.3 Foreign direct investment Foreign direct investment in Moldova grew strongly between 2003 and 2008, but was severely affected by the 2009 global crisis and has yet to recover to pre-crisis levels. FDI inflows grew from US$38 million in 1999 to US$711 million in 2008. Impressively, FDI increased tenfold in only five years, reaching 12 percent of GDP in 2007 and 2008 (Figure 10). Inflows plummeted in 2009 and, despite a slight recovery in 2011, fell again in 2012. In 2013, FDI as a percent of GDP (3.1 percent) was below its 1995 level (3.7 percent). The government of Moldova has used various instruments to attract FDI in the country, including investment promotion activities performed by the Moldova Investment and Export Promotion Organization (MIEPO), fiscal incentives, and special economic zones (see Note 4 on Special Economic Zones for a thorough analysis of how these operate). Figure 10. FDI Inflows as Percent of GDP, 1992-2013 800 14 700 12 600 10 500 8 Percent Million USD 400 6 300 4 200 100 2 0 0 FDI (Million USD) FDI (as percent of GDP) Source: Authors’ calculations based on UNCTAD data. Moldova’s performance in terms of FDI attraction is poor when benchmarked against other economies in its comparative group. In 2012-13, FDI stock in Moldova was among the lowest in the group. FDI inflows to Moldova in absolute terms were significantly lower than inflows to its 19 peers, particularly between 2005-08 and 2010-13. Indeed, in 2013 FDI inflows to Hungary and Ukraine were 13 and 16 times the value of inflows to Moldova. Moldova also underperformed relative to smaller economies in the region, with average inflows falling below those of Albania, Belarus, and Lithuania in the three periods analyzed. However, Moldova outperformed several of its peers, especially Belarus and Georgia (Table A 2). Moldova’s share of FDI in GDP is higher than expected given its level of development. Figure 11 benchmarks Moldova among other countries in terms of their level of openness to FDI. It shows that Moldova is considerably above the line, even after the decline in inflows triggered by the 2009 financial crisis. Moldova’s FDI-to-GDP ratio (3.5 percent in 2013) compared favorably to the level of FDI openness of other new and associated EU members, including Hungary, Poland, and Slovenia. The share of FDI in GDP income fell during the past decade while income per capita expanded. Yet Moldova’s performance in terms of FDI attraction is still above that of other new and potential EU entrants (see also Table A 3). Figure 11. Benchmarking Moldova’s FDI Attraction Levels a. 2000-2003 b. 2010-2013 Source: Authors’ calculations based on WDI data . Services account for almost two-thirds of inward FDI stock. In 2012, financial intermediation dominated, accounting for 28 percent of inward FDI stock (Figure 12b). The comparison of the sectoral composition of FDI stocks in 2005 and 2012 suggests an important expansion of this sector, which doubled its share in total stocks between these two periods (Figure 12a). This expansion took place at the expense of wholesale and retail trade services, which represented 26 percent in 2005 and decreased to 17 percent in 2012. FDI into the manufacturing sector, which includes industries such as automotive parts and components, textiles, and construction materials as well as food, tobacco and beverages, remained constant at around 23 percent of total inward stock. The relative low levels of FDI into transport and communications (6.8 percent) and the power sector (8.6 percent) points to untapped opportunities. In a landlocked country such as Moldova, 20 increasing competition in the transport and communications sectors, and improving the quality of services provided is crucial to stimulate firms’ competitiveness. FDI in these sectors can be a vehicle to achieve that. Figure 12. FDI Inward Stock, by Sector a. 2005 Transport & Manufacturing communications industry 7% 24% Electricity, gas & Hotels & restaurants water supply 2% Other 23% 2% Wholesale & retail trade Financial 26% intermediation Real estate & 10% business services 6% b. 2012 Other Transport & 3% communications Manufacturing Electricity, gas & 7% industry water supply 23% 9% Real estate & business services 10% Wholesale & retail trade Financial 17% intermediation 28% Healthcare & Agriculture, hunting, social assistance & forestry 2% 1% Source: Authors’ calculations based on data from the National Bank of Moldova. Russia was the largest investor in Moldova in 2012, accounting for 9 percent of total inward FDI stock. It was followed by Italy and Germany, each representing about 8 percent of total FDI (Figure 13). German firms, for example Dräxlmaier, have invested in the establishment of plants to produce cables and car parts to supply factories in the EU, while Italian investment has concentrated in the textile and clothing industries. Romania is the fourth main source of FDI in Moldova. The investments are most likely carried out by affiliates of foreign companies operating in Romania, which may enjoy greater knowledge of the Moldovan market than parent companies (UNCTAD, 2013). Other important sources of inward FDI include the United States, Switzerland, and Turkey.3 Figure 13. FDI Inward Stock by Origin, 2012 3 The fact that the Netherlands and Cyprus are among the top five investors in Moldova may reflect the fact that Moldova has a favorable double taxation treaty with them and hence offers advantages to firms based there, regardless of the ownership of the parent compa ny’s capital (UNCTAD, 2013). 21 Source: Authors’ calculations based on data from the National Bank of Moldova. 3.2.4 Composition of exports Goods Moldova’s four main export industries are vegetables, prepared foodstuffs and beverages, machinery and electronic equipment, and textiles. The relative weight of these industries in the Moldovan export basket has changed in the past decade. In the last decade, Moldova’s exports of vegetables and machinery increased, primarily at the expense of two sectors, namely, foodstuffs and textiles. Table 3 presents the evolution of export values by sector, their share in total exports, and the indicators of revealed comparative advantage (RCA), for two sub-periods: 2000-02 and 2011-13.4 Exports in vegetables increased at a compound annual growth rate of 17.6, reaching one-third of total exports in 2011-13. Indeed, driven by the expansion of sales in oil seeds, fruits and nuts, and cereals, the share of vegetables in total exports doubled between 2000-02 and 2010-13. Machinery and electronic equipment (medical or surgical instruments, vehicle/transport equipment, etc.) also experienced a dramatic expansion, almost tripling its share within the export basket. Moldova deepened its comparative advantage in both of these sectors, particularly in vegetables (Figure 14). Although losing considerable ground in recent years, foodstuffs remain Moldova’s second main export. The prepared foodstuffs and beverages industry accounted for almost half of the economy’s exports in 2000-02. As a result of a process of gradual diversification, its share in the export basket fell to 26.4 percent in 2011-13. While the RCA declined during these two periods, the sector continues to enjoy a relatively high comparative advantage, given the abundance of raw materials, such as fruits, vegetables, and cereals. Yet, exports of one of the top products in the sector, wine, were severely hit by the imposition of trade sanctions by main partner Russia in 2006. Wine exports, which reached US$556 million in 2005, were less than US$300 million in 2013. Moreover, the sector’s overall growth rate during this period has been unimpressive (Figure 14). 4 The RCA index is the ratio of a country’s export share in a specific sector to the world share of that sector in total world exports. An RCA index above 1 indicates that the country’s share of exports in a sector exceeds the global export share of that product, and is thus a measure of its competitiveness. 22 Table 3. Change in Moldova’s Shares of Merchandise Exports Average 2000-2002 Average 2011-2013 Exports Share Exports Share CAGR Sector RCA RCA (1000 USD) (%) (1000 USD) (%) (%) 01-05 Animal 18,772 3.6 1.7 37,643 2.7 1.5 7.2 06-15 Vegetable 90,416 16.5 6.0 459,344 33.0 10.5 17.6 16-24 Foodstuffs 238,391 44.0 14.9 369,669 26.4 8.5 4.5 25-27 Minerals 6,891 1.2 0.1 20,304 1.5 0.1 11.4 28-38 Chemicals 7,449 1.4 0.2 20,563 1.5 0.2 10.7 39-40 Plastic / Rubber 2,173 0.4 0.1 18,666 1.4 0.3 24.0 41-43 Hides, Skins 15,723 2.9 3.4 25,529 1.9 3.1 5.0 44-49 Wood 4,264 0.8 0.2 13,977 1.0 0.4 12.6 50-63 Textiles, Clothing 95,989 17.7 3.0 95,579 6.6 1.7 0.0 64-67 Footwear 6,382 1.1 1.2 11,616 0.8 1.0 6.2 68-71 Stone / Glass 14,157 2.7 0.9 44,665 3.2 0.9 12.2 72-83 Metals 5,963 1.2 0.2 63,202 4.6 0.6 26.6 84-85 Mach/Elec. 21,699 4.0 0.1 152,466 10.4 0.4 21.5 86-89 Transportation 5,363 1.0 0.1 12,138 0.9 0.1 8.5 90-97 Miscellaneous 8,913 1.6 0.3 62,695 4.3 0.8 21.5 Note: CAGR= compound annual growth rate. Source: Authors’ calculations using data from UN Comtrade. Textiles and apparel declined from the second to the fourth position. The sector’s RCA almost halved between 2000–02 and 2012–13. The erosion in the country’s comparative advantage in textiles, despite the availability of a skilled labor force, reflects in part the failure of domestic companies to convert acquired production knowledge into higher-value services. Indeed, Moldovan manufacturers tend to subcontract higher value-added services, such as brand development and marketing, concentrating on the lowest value-added activities. Figure 14. Evolution of RCA, 2000-2013 RCA 2000/02 RCA 2011/13 CAGR 16 30 15 14 25 13 12 20 11 10 9 15 RCA Percent 8 7 10 6 5 5 4 3 2 0 1 0 -5 Source: Authors’ calculations using data from UN Comtrade. 23 There has also been substantial churning at the product level when comparing the top 10 export products in 2003 and 2013. As Table 4 shows, wine in bottles, the top product in 2003, accounting for 21 percent of total exports, fell to the fourth place in 2013. Wine sold in bottles or containers holding less than 2 liters and in larger amounts, now jointly account for less than 9 percent of total exports. The main export in 2013 was coaxial cables and other electric conductors, reflecting the dramatic growth of the machinery industry in the last 10 years. While traditional agricultural exports such as seeds and apples continue occupying a top position in the external basket, bovine hides and skins have dropped from the top 10 ranking. Table 4. Top 10 Exported Products 2003 2013 Exports Share Exports Share Product (Millions USD) (%) Product (Millions USD) (%) Wine (in containers of< 2 liters) 162.7 21 Coaxial cable & other electric parts 150.6 9.30 Other grapes 34.3 4.45 Sunflower seeds 136.1 8.40 Spirits 25.02 3.23 Shelled walnuts 85.9 5.31 Apples 22.2 2.86 Wine (in containers of< 2 liters) 81.01 5.00 Shelled walnuts 21.6 2.78 Wheat seed, white, other 64.9 4.01 Bovine hides & skins (whole) 17.2 2.24 Other wine 62.0 3.83 Apple juice 16.98 2.19 Spirits 58.6 3.62 Other bovine hides & skins 16.84 2.17 Parts of seats 55.7 3.44 Sunflower seeds/safflower oil 16.4 2.11 Apple juice 48.8 3.01 Boneless bovine meat 13.5 1.73 Apples 47.01 2.9 Total Exports 775.9 100 Total Exports 1,619.8 100 Source: Authors’ calculations using data from UN Comtrade. 3.2.5 Growth orientation in products Moldova has gained international market share in some of its top industrial exports, particularly cables and other electric parts. Figure 15 plots the annualized growth rate of Moldova’s top exports between 2003 and 2013 against the world’s growth rate of exports of the same products during the same period. Products above the 45-degree line indicate that Moldova has gained market share in this product, as its exports have grown faster than everybody else’s. As Panel (a) shows, Moldova has clearly gained market share in coaxial cables and parts of seats. Yet, Panel (a) also highlights that demand for traditional export products from Moldova, such as sunflower seeds, other types of seeds, wines (in bulk), and apple juice, have also grown faster than world exports of these products. Moldova has been losing market share for some of its main export products, including wine (in bottles), apples, spirits, walnuts, and sunflower oil. The fall in demand for these products, in particular wine, reflects the imposition of trade sanctions by Russia in 2006 and the difficulties of the sector in reorienting exports to other markets. The restoration of a range of trade barriers to 24 Moldovan food and beverage products in 2014 will most likely deepen their relative position in international markets. Figure 15. Growth Orientation in Products, 2003-2013 a. All products b. Without coaxial cables & parts of seats Source: Authors’ calculations using data from UN Comtrade. 3.2.6 Services While Moldova’s services exports have been dominated by the transport and travel sub - sectors, ICT services have experienced a considerable expansion over the last decade. Transport services, including carriage of passengers and movement of goods, accounted for half of all services exports in 2000, followed by travel (24 percent) and communications services (10 percent). The share of transportation services has significantly declined since then, reaching 40 percent in 2013 ( Figure 16a). While the relevance of travel services has remained stable during this period, communication services as a percent of total exports have experienced a small increase (peaking at 17 percent in 2010 and staying at 13 percent in 2013). By contrast, information and computer services have expanded rapidly, with total exports growing tenfold between 2004 and 2008, and more than doubling between 2009 and 2013. Moreover, the subsector’s participation in the services export basket grew from less than 1 percent in 2005 to 6 percent in 2013. Building on this impressive export performance, the ICT services sector has also expanded its contribution to Moldova’s GDP, reaching 0.9 percent in 2013. Figure 16. Sectoral Composition of Services Exports a. Moldova, 2000-2013 b. Moldova and comparators, 2013 25 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% ALB BLR GEO HUN LTU POL MDA SVK UKR Transport Travel Transport Travel Communications Construction Communications Construction Insurance Financial Insurance Financial Computer and information Royalties Computer and information Royalties Other business Personal/cultural Other business Personal/cultural Other Other Source: Authors’ calculations using data from UNCTAD. The structure of Moldova’s services exports is relatively diversified when compared with regional peers. Figure 16b shows that other countries in the region, particularly Albania and Georgia, are strongly reliant on exports of travel services. Similarly, Lithuania and Belarus exhibit a pronounced concentration in transportation services. The communications sector, in turn, appears to be larger in Moldova than in comparators. Moreover, while several countries in this group have witnessed the dramatic expansion of ICT services, in none of these countries did the subsector grow at such a rapid pace as in Moldova (average 60 percent per year between 2002 and 2013). 3.2.7 Main export markets In recent years, Moldova has progressively decreased its traditional export dependence on the Russian market. The share of Moldovan exports going to Russia decreased from 39.5 percent in 2003 to 14 percent in 2013 (Table 5). Russia is now the second main destination after Romania, which in 2013 received 19 percent of Moldovan exports. Indeed, exports to Romania tripled between 2010 and 2013. The third and fourth partners, Ukraine and Italy, did not substantially change their importance as markets for Moldovan products. In 2013, Moldovan exports to Turkey expanded significantly. In fact, Turkey received more than 6 percent of exports, becoming the fifth most important market (Figure 17b). Table 5. Top 10 Destinations 2003 2008 2013 Market Share (%) Market Share (%) Market Share (%) 26 Russian Russian Federation 39.5 Federation 21.7 Romania 19.2 Romania 11.4 Romania 14.9 Russian Federation 14.3 Italy 10.5 Ukraine 11.5 Ukraine 7.3 Germany 7.1 Belarus 9.2 Italy 7.0 Ukraine 7.1 Poland 4.3 Turkey 6.2 Belarus 5.2 Italy 4.0 Germany 5.8 United States 4.3 Switzerland 3.9 Belarus 5.3 Austria 1.3 Germany 3.4 Poland 4.7 Kazakhstan 1.2 Kazakhstan 3.0 United Kingdom 4.2 France 1.2 United Kingdom 2.3 Switzerland 2.8 Source: Authors’ calculations using data from UN Comtrade. Figure 17. Main Export Destinations, 2000-2013 a. Exports region, 2000-2013 b. Top 10 markets, 2003, 2008 and 2013 100% 45 90% 40 80% 35 70% 60% 30 50% Percent 25 40% 20 30% 15 20% 10% 10 0% 5 0 EU-27 RUS-BLR-KAZ 2003 2008 2013 Other Europe & Central Asia Other Advanced Economies Romania Russia Ukraine Italy Turkey East Asia Pacific Latin America & Caribbean Germany Belarus Poland UK Switzerland Middle East & North Africa South Asia USA France Kazakhstan Sub-Saharan Africa Source: Authors’ calculations using data from UN Comtrade. Moldova has also shifted significantly away from CIS and toward European markets. In 2000, more than 50 percent of Moldovan exports were to CIS countries, while 20 percent went to Western European countries. In 2013, the EU, including new members from Eastern Europe, accounted for half of Moldova’s exports. The CIS, in turn, reduced its share in Moldova’s total exports to less than 40 percent (Figure 17a). The shift toward European markets is more evident in vegetables and foodstuffs. As Figure 18 shows, Moldova exported a larger share of vegetables and foodstuffs to the EU during 2010– 13 than in 2005-07. The importance of that market for animal products and metals decreased. Russia continues to be a very important market for animal products and other subsectors. Although Moldova has decreased its dependence on Russia and other CIS countries in most sectors, these markets are still important destinations for Moldovan exports. In animal products, 27 CIS countries appear to have absorbed the share lost by European markets. The comparison between the two panels also shows an increase in Asia’s share as a destination for several sectors, especially transportation equipment, wood, plastic and rubber, and chemicals. This is probably driven by increased demand for raw materials by China, which in 2013 became one of the 25 most important destinations for Moldova. Figure 18. Export Composition by Region a. 2005-2007 b. 2010-2013 Source: Authors’ calculations using data from UN Comtrade. 3.2.8 Growth orientation in markets Moldova has gained market share in Western European countries while losing ground in Russia and other CIS countries. Figure 19 shows the growth concentration of Moldova’s exports in terms of markets. In this figure, observations above the 45-degree line correspond to destinations in which Moldova has gained market share relative to other countries during the period 2003-13. This data confirms that Moldova has diversified its export destinations away from CIS countries and toward the EU. The country has also expanded its presence in Turkey and to smaller degree, in Romania. Russia, Belarus, and Ukraine, by contrast, fall under the 45-degree line, suggesting that the expansion of Moldova’s exports to these destinations has lagged behind the growth of world exports to these markets. Figure 19. Growth Orientation in Markets 28 Source: Authors’ calculations using data from UN Comtrade. 3.3 Export Diversification Since the early 1990s, Moldova has been heavily dependent on exports of a few products to the Russian market. Diversifying exports across markets and products reduces the risk in the country’s export portfolio to partner-specific or product-specific shocks. The analysis above hints at Moldova’s progress in this respect. In this section, we look closer at the country’s performance in terms of diversification along the product and destination dimensions. 3.3.1 Product dimension Moldova’s export basket broadened in scope over the last decade, but remains considerably less diversified than those of benchmark countries. Figure 20 shows the number of product varieties exported by Moldova and comparators in 2010-13. The number of products exported increased from 274 in 2000 to 393 in 2013. Only Georgia, with 351 export lines in 2013, has a narrower product scope than Moldova. However, Moldova lagged behind Georgia and other comparators in terms of its expansion over time. While Moldova expanded the number of exported products by 43 percent in 13 years, in Georgia the number more than doubled during the same period. Albania, another small country with low levels of product diversification, also experienced a remarkable improvement during the last decade, increasing the number of varieties exported by 143 percent. The number of products exported is useful to assess the extent to which a country is diversified, but it tells us little about the degree of concentration of exports in each of these product lines. The Hirschman-Herfindahl Index (HHI) allows comparing export concentration of two or more countries that may be equal in terms of number of products but may vary in terms of 29 concentration.5 Figure 21 shows the HHI for products for Moldova and comparators over the period 2000-13. Figure 20. Number of Exported Products Figure 21. HHI for Products 4,000 0.14 3,500 0.12 3,000 0.10 2,500 0.08 2,000 0.06 1,500 0.04 1,000 500 0.02 0 0.00 MDA UKR SVK BLR ALB HUN LTU GEO POL MDA UKR SVK BLR ALB HUN LTU GEO POL 2000 2001 2002 2003 2004 2005 2006 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2007 2008 2009 2010 2011 2012 2013 Source: Authors’ calculations using data from UN Comtrade. The concentration of Moldova’s export basket decreased between 2000 and 2013. Figure 21 shows that the HHI for Moldova fell slightly, from 0.04 in 2000 to 0.03 in 2013. Moreover, Moldova outperforms several of its peer countries: its export basket is less concentrated than the baskets of Belarus, Albania, Lithuania, and Georgia. While the Slovak Republic and Ukraine exhibit less concentration than Moldova (both with HHI indexes of 0.01), these two countries experienced an increase in the product concentration over the period 2000-13. However, Moldova’s export revenues are substantially concentrated in the top five export products. An alternative way to measure concentration is to look at the share of export value accounted for by the top five or three products exported. Figure 22 shows that Moldova has decreased its dependence on its five main exports in recent years. While the top five products accounted for over 40 percent of exports in 2005, they only represented 32 percent in 2013. Yet, Moldova still shows relatively high levels of concentration when compared with most of its peers. It only outperforms Belarus, Albania, and Georgia, for which the top five products accounted for over 40 percent of export revenues in 2013. 5 The HHI is computed as the sum of squared shares of each product (market) in total export. A country with a perfectly diversified export portfolio will have an index close to zero, whereas a country exporting only one export (market) will have a value of 1 (least diversified). 30 Figure 22. Share of Top Products in Total Exports, 2000-2013 0.60 0.50 0.40 Proportion 0.30 0.20 0.10 0.00 MDA UKR SVK BLR ALB HUN LTU GEO POL 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Source: Authors’ calculations using data from UN Comtrade. Overall, Moldova has achieved a moderate diversification in its export basket, increasing the number of product varieties exported and reducing the degree of concentration, as evidenced by a falling HHI. Yet, the country still remains highly dependent on the exports of a few products. Moldova’s high product concentration becomes even clearer when considering the share of export revenues accounted for by the top three products (see Figure A 4). After decreasing in the first half of the 2000s, the top three products increased their share in total exports from 16.3 percent in 2007 to 23 percent in 2013. This trend reflected the effect of Russian sanctions on Moldovan food products, and the dramatic growth that cables and other mechanical parts experienced in recent years. 3.3.2 Market dimension Moldova has made significant progress in terms of diversification of export markets . The number of markets reached by Moldovan exports increased from 63 in 2000 to 103 in 2013 (Figure 23). Moldova was one the countries in its comparative group experiencing a larger expansion in the number of destinations served during this period. However, Moldova is still outperformed by most of its peers, with the exception of Albania and Georgia, which in 2013 reached only 87 and 98 markets, respectively. Moldova has also considerably decreased the concentration of its exports along the market dimension. The HHI for Moldova fell from 0.22 in 2010 to 0.07 in 2013, registering the greatest improvement in this indicator among countries in the comparative group. Indeed, in Moldova outperformed all of its peers except for Georgia, which achieved an HHI of 0.05 (Figure 24). Figure 23. Number of Markets, 2000-2013 Figure 24. HHI for Markets, 2000-2013 31 250 0.6 200 0.5 0.4 150 0.3 100 0.2 50 0.1 0 MDA UKR SVK BLR ALB HUN LTU GEO POL 0.0 MDA UKR SVK BLR ALB HUN LTU GEO POL 2000 2001 2002 2003 2004 2005 2006 2000 2001 2002 2003 2004 2007 2008 2009 2010 2011 2012 2013 2005 2006 2007 2008 2009 Source: Authors’ calculations using data from UN Comtrade. Moldova’s export revenues have become less concentrated in its top five markets over the last decade. Until 2003, the five main destinations accounted for 75 percent of the total value of exports. This share decreased progressively over the last decade, reaching 61 percent in 2008 and 54 percent in 2013. Moldova remains one of the most concentrated exporters in its comparative group, however. Only Belarus and Albania, which derive 75 percent of export revenues from their top five destinations, lag behind Moldova in this respect (Figure 25). Overall, Moldova has considerably diversified its export basket along the market dimension over the last 10 years. The number of markets reached increased significantly, and the degree of geographic concentration of the country’s export portfolio has fallen. Moldova also reduced its dependence on the top five and three markets (Figure A 5), although it lags behind comparators. Figure 25. Share of Top Markets in Total Exports, 2000-2013 1.00 0.90 0.80 0.70 2003 Proportion 0.60 2008 0.50 2013 0.40 0.30 0.20 0.10 0.00 MDA UKR SVK BLR ALB HUN LTU GEO POL Source: Authors’ calculations using data from UN Comtrade . 32 3.3.3 Contribution of diversification to export growth How has diversification contributed to export growth? Most of Moldova’s export growth is explained by the expansion of sales of the same products to the same markets. Still, diversification in markets has been a growing driver of export growth. Figure 26 shows Moldova’s export growth for two periods, 2005-08 and 2010-13, decomposed into the proportion explained by increased sales of the same products to the same market, and the proportion explained by increased sales of the same products to new markets, new products to the same markets, or new products to new markets. In 2005-08, 86 percent of total export growth was explained by more sales of the same products to the same destinations, while 13 percent of growth resulted from diversification along the market destination, that is, more exports of the same products to new markets. In 2010-13, the importance of diversification along the market dimension experienced a marked increase, accounting for over 60 percent of export growth. Diversification along the product dimension also increased in relevance during this period, explaining 9 percent of total export growth. Figure 26. Decomposition of Export Growth Along Extensive and Intensive Margins 90% 70% 50% 30% 10% -10% Net intensive Extinction Increase of new Increase of new Increase of old products in new products in old products in new -30% markets markets markets -50% 2003-2008 2010-2013 Source: Authors’ calculations based on UN Comtrade data. 3.4 Quality and Sophistication How has Moldova’s export basket evolved in terms of quality and sophistication? What goods countries produce and how they produce them both matter for development. All else equal, goods that embody greater value-added in terms of ingenuity, skills, and technology tend to fetch higher prices in world markets. Countries that produce goods that are more sophisticated than what their income levels would suggest tend to see higher rates of future economic growth. Upgrading product quality can be a source of both export and economic growth. This section looks at quality upgrading and sophistication of Moldovan exports. 33 3.4.1 Technological classification Moldova’s exports have increased in technological sophistication but remain concentrated in primary products. Figure 27 shows the evolution of exports by categories of diverse technological levels using the Lall classification. Resource-based goods, which accounted for more than half of Moldova’s total exports in 2005, rapidly lost ground in the second half of the 2000s. In 2013, less than 30 percent of the country’s exports were resource-based. By contrast, medium- technology products, such as coaxial cables and other car parts, doubled their share in Moldova’s exports, accounting for almost 20 percent of total exports in 2013. This increase more than offset the fall in the share of low-technology goods since 2004. While primary products saw their share grow only marginally (from 26 to 32 percent), they have accounted for the highest share of Moldova’s exports since 2009. This is clear when looking at the list of 20 top exports in 2013 and their specific technological classification ( Table 6). Figure 27. Technological Classification of Exports (Lall Classification, %) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% High technology Low technology Medium technology Primary products Resource based Source: Authors’ calculations based on UN Comtrade data. Table 6. Moldova’s Top 20 Exports, 2003 and 2013 Technological Technological Top 20 Products in 2003 Top 20 Products in 2013 Classification Classification Wine (in containers < 2 liters) RB1 Coaxial cable & other electric parts MT Other grape RB1 Sunflower seeds PP Spirits RB1 Shelled walnuts PP Apples PP Wine (in containers < 2 liters) RB1 Shelled walnuts PP Wheat seed, white, other PP Bovine hides and skins PP Other wine RB1 Apple juice RB1 Spirits RB1 Other bovine hides and skins PP Parts of seats MT Sunflower-seed oil (crude) PP Apple juice RB1 34 Boneless pork meat PP Apples PP Plasters RB1 Glass containers RB1 Sunflower seeds PP Sunflower-seed oil (crude) PP Footwear (other) LT Other waste and scrap RB2 Sunflower-seed oil (other) PP Barley PP Other wine RB1 Electricity meters MT Cotton T-shirts LT Other cane or beet sugar RB1 Tobacco RB1 Maize PP Maize PP Textile materials LT Glass containers RB2 Rape or colza seeds PP Other cane or beet sugar RB1 Ignition wiring sets MT Source: Authors’ calculations using UN Comtrade data. Technology classification based on Lall (2000). Note: PP = primary products; RB1 = agricultural-based manufactures; RB2 = other resource-based manufactures; LT = low-technology manufactures; MT = medium-technology manufactures. In terms of technological sophistication, Moldova’s exports lag behind most of the countries in its comparative group, including not only Poland, Hungary, and the Slovak Republic, but also Ukraine and Georgia (Figure A 8). Like Moldova, less than 2 percent of Georgia’s exports fall in the high technology classification. However, unlike Moldova, Georgia has significantly decreased its reliance on primary and resource-based exports, significantly expanding its exports in medium technology products (from 19 percent in 2000 to 51 percent in 2013, see Table A 4). 3.4.2 Quality Analyzing data on unit prices of important export products against key competitors can provide a valuable assessment of the trends in a country’s export quality (see Box 3). Moldova’s export quality is low relative to regional comparators. Figure 28 plots the export quality index for Moldova and other countries for the period 2000-11. Despite the noticeable upward trend in Moldova’s export quality index, it is one of the worst performers in the group. Only Georgia lags slightly behind Moldova. Figure 28. Export Quality Index, 2000-2011 35 Source: Authors’ calculations based on IMF data. 3.4.3 Quality in top export products Figure 29 and Figure 30 present quality ladders for four of Moldova’s top exports, wine in bottles (“bottles”), apple juice, coaxial cables and car seat parts (“parts of seats”). Apart from using the relative unit value of each of Moldova’s top exports, we look at the relative PRODY of the importers (See Box 3, second part). Moldova’s export basket has experienced a quality upgrade between 2003 and 2013. Most of the country’s traditional exports have experienced increases in their relative quality (see Figure 29).  In wine, the relative unit value increased slightly, and it reached 0.2 in 2013, suggesting that Moldovan wine exporters receive one-fifth of the export price per bottle perceived by the exporter at the 75th percentile of the distribution of export prices (see Box 2 for a discussion on policy options to upgrade the wine sector).  In apple juice, relative quality increased substantially. While the relative unit value in 2003 was 0.5, it grew to 0.8 in 2013, suggesting that Moldovan exporters received 80 percent of the price obtained by the producer at the 75th percentile of the distribution of export prices.  In walnuts, relative unit values increased during the period from 0.57 to 0.71.  In spirits, they increased from 0.09 to 0.13. Figure 29. Relative Quality of Traditional Exports, 2003-2013 36 Source: Authors’ calculations based on UN Comtrade data. Moreover, some of Moldova’s new top products, such as coaxial cables and parts of seats, rank high in terms of their relative quality. The quality of Moldovan parts of seats improved (Figure 30). Figure 30. Relative Quality of New Exports, 2003-2013 Source: Authors’ calculations based on UN Comtrade data. Box 2: Policy Options for the Wine Sector in Moldova The Russian trade restrictions on Moldovan wine highlighted the need for diversifying markets. There are two ways in which this goal can be advanced. First, by improving the quality of Moldovan wine, making it more attractive to Western consumers. Second, by actively promoting Moldovan wines, currently not well known in world markets. In fact, governments across the world have supported promotion activities of their wine industries. Such is the case of Chile, Argentina, New Zealand, and the EU. Here, we review the initiatives of governments of Chile, New Zealand, and the EU. 37 Chile. A strong public-private partnership has been key in raising the international profile of Chilean wines during the last decade.  Wines of Chile (WoC), an industry body, works closely with the government’s export promotion agency, Pro-Chile. With offices in Santiago, New York, and London, and representations in Canada, Brazil, Europe and Asia, WoC has been very active in raising Chilean wines’ image and reputation internationally, through online marketing campaigns, trade and consumer fairs, organization of tastings and seminars in key markets, innovation and quality awards, etc. More recently, WoC launched a US$1 billion online campaign aimed at promoting Chilean wine in China.6  In addition, the Chilean economic agency CORFO has a dedicated branch seeking to attract foreign firms to invest in different areas of the local industry. In 2006, CORFO’S ‘Innova’ program launched a public-private cooperation project to support R&D with the goal increasing the quality and competitiveness of Chilean wine in international markets (Vinnova & Tecnovid).7 New Zealand is another example of a rapidly transforming industry that has relatively recently penetrated world markets in the New Zealand sector. Government support has been important in aiding the success of the industry. Today New Zealand is, together with the United States, one of the New Wine Countries (NWC) that invest more heavily in international promotion of wines in world markets (EU, 2014). Like in other NWCs, the government also collaborates in and funds several private sector initiatives.  The Wine Institute of New Zealand and the Grape Growers Council, two institutions dating from the 1960s and 1970s, joined forces in 2002 and established the New Zealand Winegrowers (NZW), which is funded by public (levies on wine and grape sales) and private funds. The NZW is in charge of promoting the “New Zealand Wine” brand. This generic, national -level positioning has allowed wines from New Zealand to achieve significant price premiums in world markets.  NZW has recently begun promoting New Zealand wines as “sustainably produced,” emphasizing local wine growers’ commitment to preserving biodiversity and protecting heritage sites. EU. The EU is the world’s largest wine producer, accounting for around 60 percent of total wine exports. Since 1962, the sector was regulated through the Common Organization of the Market (COM) in wine, which included subsidies and other types of support for European wine producers.  In 2008, the Council of the EU introduced a comprehensive reform aimed at increasing competitiveness of EU wine producers, raising the reputation of European wine in world markets. The reforms included the establishment of national support programs, allowing member states to choose from among different policy measures (European Court of Auditors, 2012).8  Member states can choose the measure “Promotion on third country markets,” which includes funding for five types of promotional activities: o Public relations, promotional or advertising campaigns; o Information campaigns on EU’s regulations for DOCs, geographical indications, etc. o Participation in events, fairs, and exhibitions; o Studies and information on new/potential markets; o Studies evaluating the results of promotional or informational campaigns.  A new quality policy was also applied in 2009 in order to enhance “the consolidation of quality wines (…) and their protection against usurpation in Europe and in third countries”9 (EC, 2012). 6 http://www.winesofchile.org/2012/10/wines-of-chiles-largest-online-marketing-campaign-aimed-at-chinese- market/ 7 http://www.iei.uchile.cl/postgrado/magister-en-estrategia-internacional-y-politica-comercial/estudios-de- casos/84943/el-caso-de-consorcios-del-vino-vinnova-sa-y-tecnovid-sa 8 http://www.eca.europa.eu/Lists/ECADocuments/SR12_07/SR12_07_EN.PDF 9 http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2012:0737:FIN:EN:PDF 38  The Comité Européen des Entreprises Vins (CEEV), created in 1960, is the representative professional body of the EU wine industry. It engages in advocacy and lobbying, information and knowledge dissemination, and provides legal, technical, and commercial advice to members. Summary of policy measures to promote wine exports competitiveness Chile New Zealand EU Wine export promotion programs x x x Support to generic advertising (country brands) x x Support to R&D and innovation x x Oenological practices, labeling rules x Geographical indications, designation of origins x Source: Based on EU (2014). Source: Authors’ elaboration. Box 3. Measuring Relative Quality of Exports Using Disaggregate Trade Data Method 1: Relative Unit Values We rely on UN-COMTRADE database to characterize the relative unit values of exports. As in Schott (2004), unit values were calculated simply as the quotient of general import values and quantities. Within any product (6-digit Combined Nomenclature code) for any given year, we then have a distribution of unit values of imports from the different source countries. For each good and exporting country , in time year , we generate a measure of relative quality as: = 75 75 Where denotes de unit value of the good and denotes the value at the 75th percentile of the unit value distribution across countries for that product. denotes the relative quality of the country’s export of that good, i.e., quality relative to other countries exporting the same good. Method 2: Relative PRODYs of the Importers An alternative way of looking at product quality is by examining the level of development of the countries that import the products. To do this, we calculate a weighted average of the income per capita of Moldova’s importers of a given good, and take the ratio of that average to the weighted average of the income per capita of importers of that good from any country in the world. We call that indicator “PRODY” of the importer. Source: Authors’ elaboration. 3.4.4 Sophistication Another way to estimate the sophistication of a country’s exports is by assessing the income level of countries that typically produce and export a similar basket of goods. Calculating a corresponding indicator of export sophistication, denoted EXPY, is a two-stage process (Hausmann, Hwang, and Rodrik, 2006). In the first stage, the income level associated with every tradable product is determined. This PRODY is taken as the weighted average of the GDP per capita of countries that export that product. The second stage involves measuring the income associated with a country’s export basket as a whole. The EXPY results from weighing the 39 PRODY values of the country’s exports by the share that each good contributes to total exports (see Box 4). The sophistication of Moldova’s exports has not changed substantially in the last 10 years. Figure 31 shows the evolution of EXPY over the period 2000-13 for Moldova and comparators. Moldova’s EXPY increased between 2000 and 2008, more rapidly between 2003 and 2006, in part reflecting the increased significance of apples and wheat, as the wine sector declined. The EXPY fell after 2009 and, while experiencing a moderate increase in 2013, it remains below its 2000 level. Moldova’s export sophistication is low when compared with other countries in the region. In line with our previos findings that measured quality based on unit values, Moldova is again one of the worst performers (Figure 31). However, Moldova’s exports appear to be more sophisticated than Albania’s. Figure 31. EXPY for Moldova and Comparators, 2000-2013 Source: Authors’ calculations based on UN Comtrade data. Moldova’s export sophistication has lagged behind the country’s growth during the last decade. When looking at the evolution of the Sophistication of Exports (EXPY) index against per capita GDP for Moldova and benchmark countries for 2000-13, we observe that Moldova’s income per capita growth was accompanied by an increase in export sophistication between 2003 and 2008 (Figure A 7). Yet, after 2009, the export basket became less sophisticated, despite Moldova’s continued economic growth. Other countries in the comparative group, in contrast, have performed better in this regard. The cases of Albania and Lithuania, both of which considerably saw their EXPY increase as fast as their GDP per capita, stand out. Much like Moldova, Belarus and the Slovak Republic have experienced a horizontal movement in the graph. 40 Box 4. Measuring Export Sophistication Calculating export sophistication, denoted by EXPY, is a two-stage process. The first stage is to measure the income level associated with each product in the world, termed “PRODY.” The PRODY of a particular product is the GDP per capita of the typical country that exports that good. Typical GDP is calculated by weighting the GDP per capita of all countries exporting the good. The weight given to each country is based on “revealed comparative advantage,” defined as the share of its exports that comes from that good relative to the “average” country. The PRODY for a single product is calculated by weighting the GDP per capita of all countries exporting that product. Therefore, a product that typically makes up a large percentage of a poor country’s export basket will have stronger weights towards poor countries’ GDP per capi ta. This will be less the case for a product that makes up a small percentage of a poor country’s exports but is a significant component of many rich countries’ export baskets. The second stage is to measure the income associated with a country’s export basket as a whole; this is its EXPY. From the first stage, each product that a country exports will have a PRODY. The EXPY is calculated by weighting these PRODY by the share that each good contributes to total exports. If butter makes up 15 per cent of a country’s exports, its PRODY will be given a weight of 0.15. Countries whose export baskets are made up of “rich-country goods” will have a higher EXPY, while export baskets made up of “poor -country goods” will have a lower EXPY.  x jk   X   x  PRODYk    j  Y and EXPYi    ik PRODYk x jk j k  Xi  j Xj j Source: Authors’ elaboration. 3.5 Survival For countries to achieve fast export growth and diversification, both successful entry into export markets and survival of export flows are crucial. The majority of export relationships (at the product-country level) forged by developing countries do not survive more than a few years. Assessing the dynamics of export participation and survival is valuable for diagnosing the export competitiveness of a country. From a policy perspective, understanding what the main challenges to export survival are is key to promote growth and ensure diversification. This section focuses on how Moldovan exports flows have performed along the sustainability margin. For this purpose, we use product-level data (HS-6 level) over the period 2000-13. The product level data provide us with an approximation to the issue of export survival, but allows international comparisons. Moldova’s export survival rates are significantly below that of comparators. Figure 32a shows that the probability of a Moldovan export relationship surviving past the first year is 40 percent, and the probability of maintaining that relationship for another year falls to 25 percent. In comparison, benchmark countries exhibit a stronger performance in terms of export survival. Poland’s 1-year survival rate, for example, is almost 20 percentage points above Moldova’s. Only Georgia is outperformed by Moldova. 41 Exports to preferential trade partners Russia, Kazakhstan, and Belarus have the highest survival rates. Figure 32b compares the survival rates of Moldovan exports by destination. The probability of survival after one year for exports to CIS CU partners is above 65 percent. Survival rates to the EU are also relatively high. Flows to the European market have almost 60 percent of lasting at least one year and 45 percent chance of surviving for two. By contrast, export flows to the Middle East and North Africa, South Asia, and Sub-Saharan Africa exhibit the lowest probability of one-year survival, at around 40 percent. Figure 32. Export Relationships Survival Rates (2000– 2013) a. Moldova vs. comparators b. By destination Source: Authors’ calculations based on data from UN Comtrade . Figure 33. Export Relationships Survival Rates by Sector, 1996–2012 Source: Authors’ calculations based on data from UN Comtrade . 42 Exports of foodstuffs and vegetables and other low-technology products have the highest survival rates. As Figure 33 shows, while vegetable exports have 50 percent chance of surviving at least one year, foodstuffs products exhibit a slightly higher chance. Textile and clothing exports also have survival rates above 40 percent. By contrast, higher-technology exports such as machinery and transportation equipment have lower chances of surviving at least one year. 4. Part II. Assessing Moldova’s Main Competitiveness Challenges Part I of this report provided an analysis of Moldova’s relative performance on four aspects of trade: the level and growth of exports, the diversification of products and markets, the quality and sophistication of exports, and the entry and survival of new exporters. This is only half of the story. In order to improve export competitiveness and performance, we need to understand the factors driving the observed outcomes, facilitating and constraining domestic firms’ productivity and competitiveness. What are the constraints that prevent Moldova from exploiting trade potential for long-term economic gain? How can Moldova take better advantage of the opportunities offered by access to the European market? The second part of this report contributes to shedding light on some of these constraints, by focusing specifically on supply-side factors affecting competitiveness. Two types of supply-side factors are particularly important determinants of a country’s competitiveness: factor conditions, which affect the cost and quality of production; and the incentive framework, including factors that establish the broad environment that influences private sector investment and participation in exports. In the analysis below, we focus on some of these supply-side factors and assess the extent to which they have constrained Moldova’s productivity and competitiveness. Among factor conditions, we focus on access to finance and on the quality of backbone services, including energy, transport, electricity, telecommunications, other input services. We consider the incentive framework, concentrating on two aspects of the business environment: (i) business regulations, and in particular, customs and trade policy regulations, and (ii) governance and institutional quality. How has Moldova performed in these areas? Are these supply-side factors barriers to firm productivity and competitiveness? We use firm-level data from the World Bank’s BEEPS to examine the productivity performance of businesses in Moldova and other countries in Europe and Central Asia (ECA).10 We are particularly interested in the links between the quality of services and other aspects of the business environment and firms’ productivity. In fact, enhancing firms’ productivity is a sine-qua-non condition for maximizing the gains from trade integration. The rest of this section is structured as follows. We begin with a discussion of Moldova’s productivity performance in a comparative perspective, investigating the determinants of productivity differentials across firms and countries in the ECA region. We then expand our 10 Our dataset contains 42,228 observations, comprising firms from 32 countries surveyed over 8 years (2002, 2005, 2007, 2009, 2011, 2012, 2013, and 2014). There are 1,245 observations from Moldova. Of these, 174 correspond to 2002, 350 to 2005, 363 to 2009, 51 to 2012, and 309 to 2013. 43 analysis to explore empirically the extent to which the supply-side factors listed above affect productivity dynamics in firms in Moldova and for the broader ECA region. Our goal is to uncover obstacles to the expansion of productivity and competitiveness in these countries. 4.1 What Factors Enhance and Constrain Moldovan Firms’ Productivity? 4.1.1 Productivity Productivity has been identified as one of the most important factors underlying the concept of “competitiveness” (Krugman, 1994; Porter, 1990). A dynamic and productive private sector promotes growth and expands opportunities for international trade. The process by which productivity expands and the economy upgrades works through a series of enablers, which also serve as intermediate indicators of competitiveness. True competitiveness is thus measured by productivity. How have Moldovan firms performed in this respect? Within Moldova, firms’ productivity levels vary across regions. Firms in the northern city of Balti, the third largest in the country after Chisinau and Tiraspol, exhibit the highest median levels of total factor productivity (TFP), followed by firms in other cities within the Northern region. By contrast, firms in the South appear to be the least productive, with the lowest median TFP (Figure 34). Although median productivity levels in Moldova’s capital city Chisinau are lower than in Balti and in the Northern region, they compare favorably with TFP performance of firms in the South and Centre regions. Figure 34. Median TFP in Moldova by Region Source: Authors’ calculations based on data from BEEPS. 4.1.2 Firm-level determinants of productivity Firms that are more integrated in the global economy enjoy a productivity premium. The results in Table A 5 in the Appendix show productivity performance varies across types of firms. In line with the literature, firms that are export-oriented and that are foreign owned show productivity premia when compared to others in the same country, same sector, and of the same size. When considering a sample of 32 countries in the ECA region, exporters are 16 percent more 44 productive than non-exporters. The productivity premium is even higher for foreign firms: controlling for other factors, foreign-owned establishment have a 17 percent higher TFP than domestic firms (Figure 35). Size matters too, with larger firms enjoying a premium over smaller ones. Focusing specifically on Moldova, younger firms appear to be more efficient than older ones, probably reflecting technological differentials. However, exporters in Moldova are equally more productive, relative to non-exporters, as in the average of ECA. Innovators are systematically more productive. Our results show that those firms that reportedly introduced new products or services in the three years previous to the year of the survey enjoyed higher levels of productivity. More precisely, innovation entails a 4 percent productivity premium (Figure 35). These findings are consistent with previous empirical evidence, showing that R&D and product innovation, by allowing the development and adoption of more efficient production technologies, fosters productivity. Productivity is also heterogeneous across sectors. For the sample as a whole, manufacturing firms are almost 7 percent less productive than firms in the services sector. Firms in transport and IT services are more productive than their counterparts in other sectors (Table A 5). The advantage for transportation services, however, is smaller (11 percent) but still significant. Focusing specifically on Moldova, manufacturing firms appear to be less productive than firms in the services sectors. The median level of TFP for manufacturing is lower than the median productivity level for construction, retail, and other services but higher than in transport services. However, manufacturing firms exhibit higher levels of productivity than companies in the transport sector (Figure 36). Figure 35. Productivity Premium of Exporters, Innovators, Foreign Firms in Moldova 0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 Foreign Exporter Innovator Note: The bar represents the regression coefficients from Model 3, Table A 5 Appendix The benchmark for comparison is that of domestic, non-exporter and non-innovator firms. Source: Authors’ calculations based on data from BEEPS. 4.1.3 Linking productivity and quality of backbone services Poor factor conditions, such as limited or costly access to finance and deficient backbone and input services are significant barriers to firms’ productivity and competitiveness. For example, lack of 45 access to finance to fund working capital may be a barrier to participation in export markets, given the greater risks and longer payments terms involved in exporting. In addition, limited access to affordable finance prevents producers from undertaking investments that will improve productivity. Figure 36. Median Productivity across Sectors in Moldova Source: Authors’ calculations based on data from BEEPS. Similarly, limited access to high quality, efficiently priced inputs and backbone services can also increase the costs of production and discourage exports. In particular, poor quality of utilities, such as an inadequate supply of power, water, and telecommunication services is a significant constraint on firm productivity. As a result of the unreliability of the power supply in many developing countries, for example, firms tend to rely on generators, which significantly increase the costs of production. Access to quality and competitively priced water, in turn, is crucial for firms in the agricultural sector. In Moldova, firms perceive access to finance as the main obstacle to their operations. Figure 37a shows the percentage of firms that perceive access to finance, and the quality of electricity, transport and telecommunications, as at least a moderate obstacle. Figure 37b presents the proportion of firms, for Moldova and for all countries in the ECA region that perceive each of these as major or severe constraints to their operations. The data shows that Moldovan firms, as their counterparts in other countries in the region, complain primarily about limited access to finance and about the quality of electricity services. These results have been validated by fieldwork. When we interviewed firms in Chisinau, these factors were also identified as problematic. While more than 40 percent of respondents said access to finance was at least a moderate obstacle, 30 percent said poor quality of electricity was a problem. Moreover, 23.5 percent of Moldovan firms considered access to finance to be a major or severe obstacle. A lower, but still considerable, proportion of firms identified electricity and telecommunications as a problem. 46 Moldova’s perception-based performance indicators are on par with those of regional peers. As Figure 37a illustrates, concerns over access to finance are not exclusive to Moldova, with firms across benchmark countries also identifying this as the main obstacle. Half of Polish firms viewed access to finance as an obstacle, and almost one third of respondents indeed identified this as a major or severe problem. Only in Albania and in Georgia, the quality of electricity seems to pose a greater constraint on productivity than finance. When compared to the regional average, however, perceptions of Moldovan firms are slightly more negative than those of their peers. Figure 37. Perception-Based Quality of Services a. Moldova and Comparators b. Moldova and Regional Average 70 ECA MDA 60 Telecommunications 50 40 Electricity Percent 30 20 Transport 10 Finance 0 MDA ALB BLR GEO HUN LTU POL SVK UKR 0 5 10 15 20 25 Finance Electricity Transport Telecommunications Percent Note: Percentage of firms that identified each category of services as at Note: Percentage of firms that identified each category of services as a least a moderate obstacle to their operations. major or a severe obstacle to their operations. ECA regional average. Source: Authors’ calculations based on BEEPS data. Within Moldova, there are substantial regional differences in perceptions regarding the quality of services. In larger cities, such as Chisinau and Balti, firms tend to perceive the provision of services as less of an obstacle than in other cities and regions in the country. Among firms in the capital city, access to finance seems to be the main perceived constraint to productivity ( Figure 38). Almost 18 percent of firms in Chisinau and 17 percent of firms in Balti said finance was at least a major or severe obstacle to their operations. Access to finance appears to be the most severe obstacle in other regions as well, except in the South. In the southern region of the country, firms tend to see the quality of telecommunications, transportation, and electricity services as more severe obstacles than access to finance. More specifically, about 35 percent of southern firms said poor electricity, transport, and telecommunications services were major or severe obstacles to their operations. Perceptions of services’ quality vary according to firm characteristics. The results in Table A 6 in the Appendix suggest that firms that are more integrated into the global marketplace are more constrained by poor infrastructure services than other firms. For example, all else equal, exporters are more likely than firms that focus on the domestic market to view finance, access to electricity, transport, and telecommunications as major obstacles to their operations. Exporters seem to be 47 particularly concerned about transport services. All things equal, exporting firms are 10 percent more affected by poor quality of transport than firms that focus on the domestic market. Figure 38. Perception-Based Quality of Services by Region 40 35 30 Balti 25 Percent Chisinau 20 North 15 Centre 10 South 5 0 Finance Electricity Transport Telecommunications Source: Authors’ calculations based on BEEPS data. Foreign-owned firms are more likely than domestic companies to perceive poor quality of transportation services as a major obstacle. Compared to domestic firms, foreign-owned companies value transport services 3 percent more. However, foreign firms appear to be less constrained than their national counterparts by access to finance, electricity, and telecommunications services. Larger and older firms also tend to perceive service quality as less of an obstacle to their performance. Finally, compared to exporters and investors based in other countries, Moldovan exporters are less constrained by the quality of electricity and telecommunications, while foreign-owned firms operating in this country tend to consider transportation services as a less important problem. Services valuations also vary according to firms’ sector of operation. All else equal, firms in the manufacturing sector are more concerned about access to finance and the quality of electricity services than firms in services (Table A 6). Conversely, manufacturing firms tend to care less about the quality of telecommunications. Manufacturing firms appear to value telecommunications services almost 6 percent less than firms in other sectors. By contrast, providers of IT services perceive poor telecommunications and electricity services as important obstacles to their production process but value transport services substantially less (40 percent) than firms in other sectors. Are these perceptions consistent with actual conditions on the ground? It seems they are. Moldovan firms appear to face real constraints in terms of access to finance. The evidence from the survey suggests that firms rely heavily on internal funds for investments in fixed assets. On average, 66 percent of fixed assets purchases are financed with internal funds (Table 7). Firms in Moldova confront a significantly high cost of external financing. One component of it, for example, the collateral-to-loan ratio for Moldova (137 percent) is above the regional average, although slightly. 48 Table 7. Financial Indicators for Moldova and Comparators Required collateral Reliance on internal funds for to loan ratio investment in fixed assets Slovak Republic 117.9 60.7 Belarus 118.1 71.8 Lithuania 122.1 60.5 Poland 130.6 70.8 Moldova 137.0 66.9 Albania 140.2 77.5 Hungary 158.3 57.9 Ukraine 171.3 71.7 Georgia 192.5 67.5 ECA (avg) 134.3 68.11 Source: Authors’ calculations based on data from BEEPS. Moldovan firms also confront electricity and water services bottlenecks. Data from the BEEPS suggests that, on average, firms in Moldova wait for 17 days to obtain an electricity connection and 11.4 days to get a phone line for a new establishment. This is largely in line with the regional average—an average wait time of 15 days for a phone line and 17 for an electricity connection (Figure 39a). Although incidents of power outage appear to be relatively rare in Moldova, when these happen they last on average a week, significantly longer than the average for the ECA region (4.8 days). Water services appear to be an even more serious constraint. Moldovan firms must wait almost 2 months to obtain a water connection, and almost 20 days longer than the average firm in the region. The average length of water outage incidents is also higher in Moldova (11.4 days) than for the ECA region and for most of its comparators (Figure 39b). The quality of electricity and water services also varies across regions within Moldova. Electricity services in the northern region of the country, and particularly around cities like Edinet, Glodeni, and Balti, appear to be more efficient than in southern and center regions ( Figure 40a). While it takes 53 days to establish an electric connection in the South, it takes only 7.5 in Glodeni. Electricity services in the capital are not particularly efficient either, with firms reportedly waiting on average 28 days for power and 15 days for a phone line. Despite limited data availability for water services at the subnational level, Figure 40b suggests that their quality is weaker in the central region. Figure 39. Objective Indicators of the Quality of Services a. Electricity Services b. Water Services 49 Source: Authors’ calculations based on BEEPS data. Figure 40. Objective Indicators of the Quality of Services in Moldova, by Region a. Electricity Services b. Water Services Source: Authors’ calculations based on BEEPS data How does the quality of services’ provision affect firms’ performance? We use firm-level data on the ECA region to examine the effects of different types of services —access to finance, electricity, transport, and telecommunications—on the productivity of firms (See Box 5). Our findings indicate that poor quality of services is associated with significant productivity costs for firms in Moldova, as well as in other countries in the region. Firms’ productivity is adversely affected by poor provision of backbone services. Evidence suggests that, in particular, limited access to finance significantly undermines firm productivity. As Table A 7 in the Appendix shows, a unit increase in the perceived valuation of the quality of access to finance is associated with a 2.2 percent decrease in total factor productivity. While negative, the effects of electricity, transport, and telecommunications services on productivity are not statistically significant. 50 The productivity losses associated with limited access to finance are also evident when focusing on more objective indicators. Models 1-5 in Table A 8 in the Appendix consider the effects of actual indicators of the quality of services on firms’ productivity. The findings uncover a statistically significant impact of the cost of borrowing on productivity. The higher the collateral- to-loan ratio that a firm is required when borrowing funds, the lower its productivity. More precisely, a 1 percent increase in the required collateral decreases productivity by 0.5 percent. Weak electricity services also undermine productivity. In addition, our results confirm the effects of the poor quality of electricity on firms’ productivity performance. Power outages, irrespective of their duration, are associated with lower productivity levels. In particular, all else equal, the productivity of those firms that reportedly experienced power outages in the year of the survey was almost 3 percent lower than the productivity of those that did not suffer interruptions in their power supplies. Column 5 in Table A 8 also suggests that he longer a firm has to wait to obtain an electric connection, the lower its productivity. Deficient water services are also a constraint to productivity, as our results in column 5 show. The longer a firm has to wait to establish a water connection in a new facility, the greater its productivity loss. Specifically, for each additional day that a firm has to wait to get water supply, its productivity falls by 0.11 percent. The coefficient for water connection services in Moldova is significant and positive, which is surprising, and suggests that it is capturing the effect of an omitted variable that is both associated with water connection and with productivity (Table A 8). Box 5. Estimating the Impact of Services Inputs Quality on Firms’ Productivity How do access to quality services inputs affect firms’ performance? Answering this question requires access to a dataset on the basis of which we can obtain comparable measures of quality services input provision and of firms’ performance. Then, it is necessary to test whether a systematic relationship exists between the two. Our approach follows that of Arnold, Mattoo and Narciso (2006). The measure of firms’ performance chosen is productivity. We use three alternative measures: (i) labor productivity (t he ratio of output to total labor costs), (ii) total factor productivity (TFP) estimated in two as a residual of a Cobb-Douglas production function, with output as a function of the capital stock, labor and intermediate inputs; and (iii) TFP estimated as a residual from a translog specification in which output is expressed as a function of the capital stock, labor, intermediate inputs and their squared terms, and their cross-products. The performance of services sectors is also obtained from the Enterprise Surveys. We used subjective measures of local services performance, which are firms’ valuations as to how much of a constraint they consider electricity, telecommunications, transport, and access to finance for their businesses. Firms are asked to select, on a scale from 0 to 4, whether they consider each of these dimensions to be not an obstacle for their operations (0), a minor obstacle (1), a moderate obstacle (2), major obstacle (3) and severe obstacle (4). The empirical strategy consists in regressing the measure of productivity on measures of the performance of services, controlling for factors that are typically identified in the literature as relevant for firms’ performance, which include firm’s export status, firm’s size, and firm’s age. In add ition, we control for country-year fixed effects, to eliminate the potential of distortions due to changes in the relative values of the different currencies in which output, wages, intermediates and capital stock are expressed and to eliminate the effect of country-year unobservables that may affect both productivity and the perception of services’ quality, as well as sector-fixed effects to control for time-invariant and sector-specific unobservables. Concerns about endogeneity arise because it is possible that poor performance affects firms’ perceptions about the obstacles that services input provision represent. This would imply a bias upwards in the coefficient linking services performance with 51 productivity. This makes a specification that links firm-level perceptions of services quality with firm-level productivity inappropriate. Our strategy, following Arnold et al (2006) consists in aggregating the individual firm’s responses to the services-related questions on the right hand side at the regional level, within each country. This reduces the influence that an individual firm’s performance has on the regressor. In addition, it is likely to better summarize the quality provision of se rvices in a given region. The chosen specification is as follows: = + + + + (1) where is the indicator of productivity (labor productivity, residual from Cobb Douglas or residual from translog), is a country-year fixed effect, is a sector fixed effect, ServPerformance is a vector of perception based indicators of obstacles represented by access to finance, electricity, transport, and telecommunications, that vary at the regional level, X is a vector of controls varying at the firm level, and is an error term assumed orthogonal to the regressors. Source: Authors’ elaboration. 4.2 Linking Productivity with the Business Environment and Governance Business environment and governance issues have a fundamental impact on competitiveness well beyond export markets. The business environment plays an important role in firm-level productivity, acting as an enabler or an obstacle for their growth. An effective business environment should promote firm behavior that is allocatively efficient on a macro basis over the long term. This requires: (1) a regulatory regime that supports private sector activities without creating unnecessary obstacles to running a business, (2) a non- discriminatory tax environment, (3) a legal framework that promotes market competition, and (4) sound governance and capable institutions that minimize the wedge between policy and practice. In the remainder of this paper, we focus on the first and the fourth of these aspects of the business environment. 4.2.2 Business regulations: taxes, red tape, and trade and customs procedures Extensive compliance requirements associated with government regulatory procedures, such as paying taxes, getting licenses, or dealing with customs procedures for trading across borders, can be detrimental to firms’ competiveness in international markets. Indeed, excess regulations may add extra costs for regular firms in terms of time and money. Heavier regulation is generally associated with greater inefficiency of public institutions and more corruption. When looking at objective indicators of red tape and the cost of government and tax regulations, Moldova’s performance is on par with that of regional comparators. As Figure 41 shows, the average time spent by businesses’ management dealing with requirements imposed by government regulations is lower in Moldova than in Poland, Ukraine, Belarus, and Albania. By contrast, the Slovak Republic and Lithuania appear to have less time-consuming government regulations. The time spent by firm executives filing or paying taxes is high across the region (average 35 per year). 52 Tax procedures are particularly cumbersome and inefficient in countries like Ukraine, which exhibits the worst performance among the countries in the sample.11 In addition, excessive regulations associated with getting import licenses or dealing with customs procedures also have adverse effects on firm performance and competitiveness. In particular, some of the conditions faced by developing countries, such as excessively high fees for customs and other documentation, high port and handling charges, and lengthy and complex customs clearance procedures, significantly increase the costs of exporting. Moreover, in some cases, cumbersome and costly regulations may create incentives for firms to rely on bribes to facilitate movement of goods. Figure 41. Government Regulations, Moldova and Comparators Source: Authors’ calculations based on BEEPS data . How has Moldova performed in terms of business regulations concerning trading across borders, trade facilitation and logistics? In this section, we examine two types of data: aggregate data derived from the World Bank’s Doing Business and Trading Across Borders indicators; and firm-level data derived from the BEEPS. Both point to significantly high costs faced by Moldovan when engaging in international trade. Moldova’s performance on trading across borders is weak. In Doing Business (DB) 2014, Moldova ranked 150 out of 185 economies and slipped eight places compared to DB 2013. The DB2014 does not indicate any major reform in the section of trading across borders (Figure 42). The average cost to export one container is $1,545 and the average cost to import one container is $1,870. At the same time, it still takes 32 days to export goods from Moldova compared to 25 on average in ECA and 11 days in OECD countries. For imports, the number of days is respectively 35, 26 and 10. 11 The time that business managers spend complying and dealing with government regulations can be interpreted as a “time tax” imposed on firms (De Rosa, Gooroochurn, and Gorg, 2010). It can be viewed as an opportunity cost, which may hinder the firms’ performance. However , our results show the opposite. 53 Figure 42. Trading Across Borders Indicators a. Time to Export and Import b. Cost of Exporting and Importing 35 3,000 30 2,500 25 2,000 20 1,500 15 1,000 10 500 5 0 0 ALB BLR GEO HUN LTU MDA POL SVK UKR ALB BLR GEO HUN LTU MDA POL SVK UKR Time to export (days) Time to import (days) Cost to export (USD per container) Cost to import (USD per container) Source: Authors’ elaboration using data from the World Bank’s Doing Business Database. Poor logistics are another constraint on export expansion. Several recent surveys of importers and exporters (including potential exporters) yielded strong indications that logistics and customs are considered key obstacles and constraints to growth in Moldova’s foreign trade, especially for exports. Anecdotal evidence from discussions with firms in Chisinau confirmed that logistics is a problem for their operations. Moldova ranks 90 in 2014 Logistics Performance Index, scoring 2.65 with the 64.3 percent of the highest performer (Table 8 and Figure 43). Table 8. LPI (Rank), 2007-2014 2007 2010 2012 2014 LPI Rank (Overall) 106 104 132 94 Customs 110 129 129 98 Infrastructure 128 123 98 85 Logistics Services 112 132 142 118 Timeliness 111 97 126 109 Source: Authors’ calculations based on LPI surveys data . Figure 43. Logistics Performance Index (LPI) for Moldova and Comparators, 2014 54 4.00 140 SCORE RANK 3.50 120 3.00 100 2.50 80 2.00 60 1.50 40 1.00 0.50 20 0.00 0 POL HUN SVK LTU UKR MDA BLR GEO Source: Authors’ calculations based on LPI surveys data . Data from the BEEPS survey confirms that trade regulations and logistics are a constraint to business performance in Moldova. Almost 35 percent of Moldovan firms said customs and trade regulations were at least a moderate obstacle to their operations. This is slightly higher than the average for the ECA region (24 percent) and for other countries in the comparative group (Figure 44). Not surprisingly, all else equal, exporters and foreign firms, which participate in international trade, are more likely to perceive customs and trade regulations as serious constraints (Table A 9 in the Appendix). Figure 44. Percentage of Firms Perceiving Customs, Trade Regulations as Obstacles Moderate Obstacle Major/Severe Obstacle 25 20 15 10 5 0 MDA UKR GEO BLR ALB HUN LTU POL SVK Source: Authors’ calculations based on BEEPS data. However, Moldova performs better than other countries in the region in terms of the expediency of customs procedures. The average number of days it takes for a firm’s goods to clear customs after arriving at a point of exit, 2.8 days, is lower than the regional average (3.9). Only Belarus, and Lithuania seem to have more expedient customs services. Obtaining import licenses, in turn, takes slightly longer in Moldova (12 days) than in other countries in its 55 comparative group (except for Belarus). However, Moldova’s average wait for an import license is below the regional average for ECA firms (14 days). Lengthy customs and import licensing procedures appear to have an adverse impact on firms’ productivity. We estimated a model of TFP for exporters only that includes a variable measuring the average number of days that a firm takes to clear customs. Our results, reported in Table A 10 in the Appendix) confirms that customs and trade regulations are constraining the performance of Moldovan exporters. Each day that companies have to wait to clear customs is associated with a productivity decline of 0.5 percent. Delays in obtaining an import license also have a negative effect on TFP, but this effect is not statistically significant. The interactive term of this variable (days to obtain import license) and the dummy for Moldova, on the other hand, has a statistically significant and positive coefficient, suggesting that Moldovan firms face no productivity loss associated with delays in import licenses application process. 4.2.3 Governance and institutional quality Moldovan firms identified corruption as the most constraining aspect of the business environment. Indeed, corruption and political risk appear to be perceived as important barriers to business performance across the region (Figure 45). However, when considering the whole sample of ECA firms, tax regulations and administration emerges as the most severe obstacle associated with the business environment. Political instability is also more constraining for Moldovan firms than for others in the region. Figure 45. Which Aspect of the Business Environment Represents the Biggest Obstacle? ECA MDA Corruption Political instability Customs and trade regulations Business licensing and permits Tax regulations and administration Labor regulations Crime, theft and disorder Courts 0 5 10 15 20 25 Percent Source: Authors’ calculations based on BEEPS data. An extensive body of literature has examined the links between corruption, productivity, and economic growth. Corruption results in inefficiency in resource allocation and decreases firms’ incentives to invest, thereby undermining firm productivity. On the other hand, some have argued that corruption may “grease the wheels” of production by helping firms overcome red tape and excessive government regulations. We use data from the BEEPS to examine empirically the links between corruption and productivity in the ECA region. In addition, we follow De Rosa, 56 Gooroochurn, and Gorg (2010) in looking at the ways in which corruption and government regulations interact to influence productivity levels. Moldovan firms face a “corruption tax” on productivity. The results in Table A 11 in the Appendix show that those firms that perceive corruption as an obstacle to their operations are 5.4 percent less productive than those that don’t see corruption as a problem. Moreover, Moldovan firms that reported having relied on “informal payments or gifts” to deal with customs officials or with courts are less productive than those that did not reveal such payments. Specifically, firms that engaged in bribing were between 6 and 7 percent less productive than firms of similar characteristics that did not (Figure 46). The cost of using bribes to obtain business licenses, on the other hand, is very high. Firms that allegedly relied on bribes in their application for business license were 50 percent less productive than counterparts. Figure 46. Productivity Losses Associated with Corruption Bribes courts Bribes customs Corruption perceptions 0 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 -0.08 Source: Authors’ calculations based on BEEPS data. Other governance issues, such as the functioning of courts and political instability, also influence firms’ performance in Moldova. Those Moldovan firms that perceive the functioning of courts as a major obstacle are 15 percent less productive than other counterparts. By contrast, perceptions of political instability are associated with higher, not lower, productivity levels (Table A 11). 2. There are differences across firms in terms of the extent to which they view obstacle. Exporters and foreign firms are less likely to perceive corruption as an obstacle. These firms are also less likely to engage in “informal payments or gifts” to secure government contracts. Exporters, however, are more likely to rely on informal payments to deal with customs and import licenses ( Table A 12). 57 5. Policy Recommendations What do we learn from these results in terms of policy actions? How can Moldova improve its export competitiveness and increase the product and market scope, quality, and sophistication of its export basket? Short Medium to Longer Term 1. Trade and investment facilitation 1.1. Increase market access  Implement DCFTA provisions in all sectors.  Upgrade quality and standards of exports. Replace national standards with for Moldovan exporters  Increase communication activities related to technical aspects of EU and international standards. the DCFTA, with emphasis on tariff-rate quota management and  Promote capacity-building and strengthening of the National Food Safety other customs issues. Agency.  Negotiate other preferential trade agreements at the country level.  Develop and accredit national SPS laboratories.  Raise awareness of the implications of the DCFTA and the requirements of EU markets in specific sectors and regions.  Promote capacity building of the Ministry of Economy as the main institution responsible for DCFTA implementation. 1.2. Maintain liberal trade  Eliminate remaining tariff peaks in some products, such as wine and investment policies bottles.  Eliminate restrictions to trade in services, in the form of barriers to mobility of high-skilled workers, which could be hindering technological and managerial upgrading.  Push for urgent adoption by the Parliament of pending trade-related legislation. 1.3. Actively promote FDI  Take advantage and better promote Moldova’s key assets in terms of  Improve the business environment FDI attractiveness—extremely fertile soils, a low-cost labor force and o Implement anti-corruption/corruption initiatives. trading access to both the European and CIS markets. o Prosecute openly high-ranking officials charged with committing  Identify policy reforms to remove obstacles to investment retention and corrupt practices. attraction, including a reform of MIEPO’s institutional set-up. o Continue the comprehensive civil service reform effort, with  Strengthen investment protection policies. emphasis on the judiciary and the prosecutor’s service. 58 o Reduce the cost of doing business by streamlining procedures for construction and obtaining licenses. 1.4. Ensure that gains from  Encourage FDI technology and knowledge spillovers to the rest of the  Undertake efforts to increase the absorptive capacity of domestic firms), FDI materialize economy, for example, by which is a crucial condition for these spillovers to materialize. These may o Promoting FDI in upstream sectors, which results in more include: varieties of inputs, better prices, and often-higher quality o Educate and train labor force products, with economy-wide gains. o Invest in R&D o Allowing FDI to compete with domestic firms (horizontally), o Implement supplier development programs which can induce competition-driven or imitation-driven productivity gains in domestic firms. o Encouraging interaction of foreign and domestic firms, either through client or suppliers’ relationships that can also lead to productivity spillovers. 1.6. Improve logistics, Customs streamlining and modernization transport, and infrastructure  Implement simplified customs procedures and electronic (paperless)  Sign on to the Revised Kyoto Convention (RKC) and adopt a new declarations for all customs regimes, to facilitate exports and Customs Code aligned with it. imports.  Connect to the EU New Computerized Transit System transit system.  Reduce customs controls at the point of clearance, including additional controls targeting valuation.  Ease the VAT regulation related to the investment in modern industrial equipment. Defer the VAT payment for the imported raw materials and accessories.  Improving quality and reliability of transport and logistics services  Implement Transport and Logistics Strategy, including allocation  Align the trade logistics framework and procedures with international for road maintenance. good practices. This may include:  Develop national public investment program to improve roads, o Introducing professional and performance standards for service railways, and airports, which is fundamental to facilitate the providers, movement of goods across borders, o Allowing increased scale of logistics service providers through  Promote private (national and foreign) investment to complement mergers and acquisitions, and these public efforts. o Promoting the introduction of new technologies for tracking and  Establish an independent road safety government champion and security. provision of budget to implement the National Strategy for Road  Adopt appropriate legislation and regulations to modernize the railways Safety. sector in accordance with the commitments under the EBRD financing agreements, and as part of the requirements for EU accession set forth in the AA.  Establish a Naval Agency by restructuring the Giurgiulesti Harbor Master 59 Service and introduce effective procedures for supervision of the merchant fleet registered under the flag of Moldova to ensure compliance with international conventions.  Develop a medium term action plan and strategy for the development of the transport sector, to fulfill Moldova’s commitments under the recently signed AA. 2. Support for firms’ internationalization 2.1. Address informational  Adopt a short-term strategy for MIEPO, outlining programs and  Ensure full implementation of the SME Strategy and coordination failures funding for export promotion, for example, through marketing and  Monitor impact of investment policy interventions. information campaigns.  Adopt institutional strategies for the SME and export/investment agencies, giving each adequate institutional set-ups, programs, and budget to implement their mandates.  Support the creation of business excellence networks.  Engage actively the Moldovan diaspora. 2.2. Improve access to  Address informational asymmetries between borrowers and lenders  Establish a system of incentives to build a dynamic funding ecosystem for finance and other market failures exporting start-ups and high-growth services firms.  Promote training on financial skills among micro and small firms so  Encourage angel investors to network and to promote that type of as to improve the quality of financial information that firms provide investment in the growth of early stage businesses. to banks.  Encourage the regional clustering of firms that operate in a certain sector, so as to increase the stock of information about the returns of relevant activities, and provide useful signals to the banking sector.  Encourage angel and venture capital investors, for example, through the provision of tax incentives. 2.3. Promote firms’  Promote product and process innovation among private sector firms productivity upgrading,  Establish incentives for firms to offer management training and invest in human  Ensure that a well-functioning and deep financial market exists to support firm investment in these areas.  Ensure that a well-functioning labor market allows for productivity gains to translate into better jobs. 60 6. Appendix Table A 1. Selected Evidence on Export Promotion Interventions and Its Estimated Impact Country Intervention Effect Documented by Grants to encourage investment in technology, training or physical Ireland capital, feasibility studies, technology acquisition, typically not Positive effect of grants on export intensity (among already Gorg et al (2008) exceeding 45%-60% of capital cost, and paid in installments subject exporters) when grants were large enough. No significant to periodic reviews effect on turning non-exporters into exporters. USA Export promotion expenditures at the state level (no information on Bernard and Jensen (2004) the exact type of instrument) Weak evidence of increased participation in global markets Czech Republic Public export credit guarantees against political and commercial Janda et al (2013) risks, no thresholds on size or legal form of the exporter Evidence of increased export flows in the short and longer run Increased the value of exports and expanded the extensive Tunisia margin (helped diversify). They were found useful to Cadot et al (2013) Matching grants for export development (new products, new encourage first-time exporters. Three years after receiving the markets or export skills for first time exporters). Eligible activities grant, however, export performance of recipients was again on needed to address informational constraints to enter export markets. par with that of non-recipients of the grant Network of Export Promotion Agency Offices abroad (78 offices Korea, Rep. with the mandate to provide information and bolstering the trade- Kang, K. (2011) investment infrastructure—business matchmaking, international exhibitions and marketing of IT and cultural industries) Positive effect on export values at the macro level Canada Trade missions co-financing (macro level) No effect on exports Head and Ries (2006) Chile No effect on exports from trade missions, trade shows. Alvarez (2004) Trade missions, trade shows, and exporter committees Positive effect from participation in exporter committees Positive effect of the combination of all interventions relative Counseling (training on export process, information on to participation in only one of them. Effect is concentrated Colombia Volpe and Carballo (2010) opportunities and target markets), participation in international trade mainly on the extensive margin, and within it, on market fairs, support in setting up an agenda of commercial meetings diversification. Random assignment of export opportunities to handloomers Egypt Atkin et al (2014) producing rugs Strong evidence of increased efficiency and quality Source: Authors’ elaboration 61 Figure A 1. Exports as Percent of GDP vs. Income Level a. 2000-2003 b. 2010-2013 Source: Authors’ calculations using data from WDI and UN Comtrade. Figure A 2. Imports as Percent of GDP vs. Income Level a. 2000-2003 b. 2010-2013 Source: Authors’ calculations using data from WDI and UN Comtrade . 62 Figure A 3. Services Imports as Percent of GDP vs. Income Levels a. 2005-2007 b. 2010-2013 Authors’ calculations using data from WDI. 63 Table A 2. Share of FDI in GDP for Moldova and Comparators 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 ALB 3.96 5.08 3.04 3.13 4.73 3.24 3.61 6.15 7.5 8.24 8.93 6.89 7.1 9.69 BLR 1.14 0.78 1.69 0.96 0.71 1.01 0.97 3.99 3.6 3.81 2.52 6.7 2.31 3.09 GEO 4.29 3.41 4.72 8.38 9.6 7.06 15.11 17.2 12.23 6.12 6.99 7.26 5.76 6.15 HUN 5.96 7.47 4.51 2.56 4.19 6.99 6.06 2.9 4.1 1.57 1.73 4.58 11.22 2.38 LTU 3.29 3.65 5.09 0.96 3.41 3.94 6.01 5.12 4.14 -0.04 2.18 3.36 1.65 1.16 POL 5.51 2.99 2.08 2.12 5.09 3.39 5.74 5.54 2.8 3 2.95 4 1.24 -1.17 MDA 9.9 6.99 5.06 3.72 5.63 6.38 7.58 12.3 11.75 3.83 3.58 4.11 2.41 3.11 SVK 13.33 10.78 23.97 8.94 9.55 6.49 10.4 5.36 5.16 -0.01 2.03 3.64 3.09 0.62 UKR 1.9 2.08 1.63 2.84 2.64 9.06 5.2 6.93 6.06 4.11 4.76 4.41 4.44 2.05 Source: Authors’ calculations based on data from UNCTAD. Table A 3. FDI to GDP Ratio and Income per Capita, 2000-2003 and 2010-2013 (avg.) 2000-2003 2010-2013 Share of FDI GDP per Share of GDP per in GDP capita (log) FDI in GDP capita (log) Turkey 0.78 9.45 Hungary -0.87 10.02 Belarus 1.09 8.98 Slovenia 0.39 10.24 Iceland 1.94 10.44 Iceland 1.24 10.59 Slovenia 2.78 10.07 Turkey 1.64 9.79 Latvia 3.07 9.46 Croatia 1.70 9.91 Serbia 3.10 9.04 Poland 1.84 10.01 Poland 3.16 9.61 Lithuania 2.15 10.03 Lithuania 3.24 9.51 Bosnia & Herzegovina 2.24 9.13 Bosnia and Herzegovina 3.33 8.78 Slovak Republic 2.49 10.13 Albania 3.78 8.67 Czech Republic 2.55 10.25 Hungary 5.07 9.85 Latvia 3.43 9.91 Croatia 5.47 9.72 Moldova 3.47 8.34 Moldova 5.59 7.85 Bulgaria 3.53 9.64 Macedonia 6.07 9.05 Belarus 3.66 9.72 Slovak Republic 6.64 9.70 Macedonia, FYR 3.91 9.33 Bulgaria 7.16 9.20 Serbia 4.12 9.44 Estonia 7.19 9.71 Cyprus 4.21 10.29 Czech Republic 7.23 10.00 Estonia 5.92 10.08 Cyprus 9.05 10.26 Albania 8.61 9.18 Montenegro n/a 9.20 Montenegro 14.05 9.54 Source: Authors’ calculations based on data from WDI. 64 Figure A 4. Top 5 and 3 Products, Number of Products, and HHI 0.45 450 0.40 400 0.35 350 Number of products 0.30 300 Share 0.25 250 0.20 200 0.15 150 0.10 100 0.05 50 0.00 0 Share Top 5 Products Share Top 3 Products HHI Products Number of Products Source: Authors’ calculations using data from UN Comtrade. Figure A 5. Top 5 and 3 Markets, Number of Markets, and HHI 0.90 120 0.80 100 0.70 Number of markets 0.60 80 0.50 Share 60 0.40 0.30 40 0.20 20 0.10 0.00 0 Share Top 5 Markets Share Top 3 Markets HHI Markets Number of Markets Source: Authors’ calculations using data from UN Comtrade . 65 Figure A 6. Moldovan Wine Against Competitors, Growth of Unit Values, 2000-2013 (Index with base=100 in 2006) 250.00 Argentina 200.00 Australia Chile 150.00 France Georgia 100.00 Italy Moldova 50.00 New Zealand South Africa 0.00 United States Uruguay Source: Authors’ calculations based on UN Comtrade data . Figure A 7. EXPY vs. GDP per Capita, 2000-2013 Source: Authors’ calculations based on UN Comtrade data. 66 Table A 4. Technological Classification of Exports for Moldova and Comparators, 2000 and 2013 % of Exports ReporterISO3 Technological Classification 2000 2013 ALB High Tech 0.01 0.02 ALB Low Tech 0.7 0.39 ALB Medium Tech 0.05 0.05 ALB Primary Products 0.1 0.39 ALB Resource Based 0.14 0.15 BLR High Tech 0.04 0.03 BLR Low Tech 0.2 0.13 BLR Medium Tech 0.37 0.27 BLR Primary Products 0.04 0.12 BLR Resource Based 0.34 0.45 GEO High Tech 0.05 0.02 GEO Low Tech 0.04 0.08 GEO Medium Tech 0.19 0.51 GEO Primary Prods 0.2 0.15 GEO Resource Based 0.52 0.25 HUN High Tech 0.29 0.25 HUN Low Tech 0.15 0.12 HUN Medium Tech 0.39 0.41 HUN Primary Products 0.06 0.07 HUN Resource Based 0.11 0.15 LTU High Tech 0.07 0.05 LTU Low Tech 0.25 0.18 LTU Medium Tech 0.21 0.23 LTU Primary Products 0.11 0.13 LTU Resource Based 0.36 0.41 MDA High Tech 0.02 0.02 MDA Low Tech 0.23 0.29 MDA Medium Tech 0.06 0.2 MDA Primary Products 0.29 0.27 MDA Resource Based 0.39 0.22 POL High Tech 0.09 0.12 POL Low Tech 0.31 0.22 POL Medium Tech 0.34 0.36 POL Primary Products 0.09 0.1 POL Resource Based 0.17 0.2 SVK High Tech 0.07 0.21 SVK Low Tech 0.28 0.18 SVK Medium Tech 0.39 0.42 67 SVK Primary Products 0.05 0.05 SVK Resource Based 0.21 0.14 UKR High Tech 0.05 0.04 UKR Low Tech 0.31 0.22 UKR Medium Tech 0.3 0.31 UKR Primary Products 0.12 0.21 UKR Resource Based 0.22 0.21 Source: Authors’ calculations with data from WITS. Figure A 8. Technological Classification of Exports (Lall Classification, %), 2013 100% 90% 80% 70% ResourceBased 60% PrimaryProds 50% MediumTech 40% LowTech 30% HighTech 20% 10% 0% ALB BLR GEO HUN LTU MDA POL SVK UKR Source: Authors’ calculations based on data from UN Comtrade. 68 Table A 5. Determinants of Productivity (1) (2) (3) VARIABLES TFP TFP TFP Capital 0.0522*** 0.0569*** 0.0516*** (0.00913) (0.00888) (0.00927) Capital*MDA -0.0216 -0.0211 (0.0402) (0.0396) Materials 0.481*** 0.473*** 0.473*** (0.0111) (0.0110) (0.0124) Materials*MDA -0.0335 -0.0469 (0.0618) (0.0616) Labor 0.367*** 0.385*** 0.369*** (0.0172) (0.0165) (0.0179) Labor*MDA 0.0723 0.102 (0.0657) (0.0661) Firm size 0.000256*** 0.000220*** 0.000254*** (6.92e-05) (6.34e-05) (7.05e-05) Firm size*MDA -4.31e-05 -3.27e-06 (0.000323) (0.000338) Exporter 0.134*** 0.116*** 0.161*** (0.0216) (0.0205) (0.0224) Exporter*MDA -0.0478 -0.0588 (0.114) (0.111) Foreign 0.176*** 0.166*** 0.171*** (0.0314) (0.0313) (0.0318) Foreign*MDA -0.0747 -0.0681 (0.145) (0.146) Innovator 0.0403** (0.0159) Innovator*MDA 0.0927 (0.0971) Firm age 0.000528 0.000279 0.000956 (0.000642) (0.000622) (0.000649) Firm age*MDA -0.00784** -0.00824** (0.00385) (0.00412) Manufacturing -0.0680*** -0.0707*** 0.000956 (0.0201) (0.0199) Manufacturing*MDA 0.189** (0.0931) Construction -0.0282 -0.0485** (0.0219) (0.0207) Construction*MDA 0.108 (0.126) Transport 0.113*** 0.115*** (0.0279) (0.0292) 69 Transport*MDA -0.239** (0.102) IT 0.345** 0.318** (0.135) (0.140) Constant -182.2 -113.5*** -101.9 (536.2) (37.84) (560.5) Country, year, country-year dummies YES YES YES Sector dummies NO NO YES Observations 8,121 8,121 8,121 R-squared 0.930 0.928 0.932 Robust standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1 70 Table A 6. Valuation of Services’ Quality by Type of Firm (1) (2) (3) (4) VARIABLES Transport obstacle Finance obstacle Electricity obstacle Telecomm obstacle Firm Size 1.85e-05 -7.34e-05*** 1.99e-05 -2.00e-05 (1.90e-05) (2.45e-05) (1.50e-05) (1.89e-05) Firm Size*MDA 0.000187 -0.000251 0.000283 0.000326 (0.000215) (0.000201) (0.000212) (0.000224) Exporter 0.103*** 0.0363** 0.0318** 0.0761*** (0.0145) (0.0165) (0.0161) (0.0154) Exporter*MDA -0.0693 0.0210 -0.168* -0.187** (0.0981) (0.103) (0.0975) (0.0845) Foreign 0.0499*** -0.273*** -0.0715*** 0.0193 (0.0186) (0.0202) (0.0205) (0.0196) Foreign*MDA -0.280** 0.0237 -0.0648 0.00667 (0.117) (0.126) (0.126) (0.115) Firm age -0.00119*** -0.000115 -0.00176*** -0.00212*** (0.000421) (0.000507) (0.000457) (0.000423) Firm age 0.00720* 0.000185 0.00483 0.00346 (0.00377) (0.00416) (0.00366) (0.00316) Manufacturing 0.0129 0.150*** 0.109*** -0.0565*** (0.0135) (0.0157) (0.0153) (0.0142) Manufacturing*MDA -0.0761 -0.0701 0.0155 -0.0461 (0.0786) (0.0841) (0.0850) (0.0687) Construction -0.0210 0.162*** -0.208*** -0.0716*** (0.0189) (0.0232) (0.0196) (0.0197) Construction*MDA -0.189 -0.249 -0.509*** -0.271* (0.152) (0.152) (0.146) (0.139) Transport -0.0810*** 0.0353 -0.162*** -0.0270 (0.0276) (0.0301) (0.0266) (0.0273) Transport*MDA 0.611*** 0.225 -0.122 0.213 (0.209) (0.169) (0.165) (0.189) IT -0.390*** -0.101* 0.141** 0.323*** (0.0409) (0.0593) (0.0596) (0.0653) IT*MDA 1.279 0.0823 0.140 -0.451 (0.929) (0.569) (1.106) (0.534) Constant 170.7 347.5 177.8 -3.881 (352.1) (522.3) (468.5) (334.1) Country, year, country-year dummies YES YES YES YES Observations 41,017 36,347 41,471 33,819 R-squared 0.095 0.081 0.165 0.108 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 71 Table A 7. Determinants of Productivity Based on Perceptions of Service Quality (1) (2) (3) (4) VARIABLES TFP TFP TFP TFP Capital 0.0526*** 0.0514*** 0.0515*** 0.0342*** (0.00930) (0.00968) (0.00935) (0.0130) Capital*MDA -0.0241 -0.0229 -0.0207 0.0370 (0.0410) (0.0407) (0.0394) (0.0262) Materials 0.473*** 0.480*** 0.474*** 0.496*** (0.0124) (0.0131) (0.0125) (0.0165) Materials*MDA -0.0447 -0.0617 -0.0435 0.152*** (0.0636) (0.0638) (0.0628) (0.0481) Labor 0.369*** 0.359*** 0.369*** 0.368*** (0.0180) (0.0190) (0.0180) (0.0240) Labor*MDA 0.102 0.123* 0.0970 -0.139*** (0.0675) (0.0709) (0.0673) (0.0482) Exporter 0.165*** 0.158*** 0.167*** 0.142*** (0.0227) (0.0231) (0.0228) (0.0301) Exporter*MDA -0.0205 -0.0377 -0.0305 -0.0992 (0.112) (0.115) (0.113) (0.0702) Foreign 0.170*** 0.163*** 0.171*** 0.181*** (0.0318) (0.0329) (0.0320) (0.0422) Foreign*MDA -0.0753 -0.0835 -0.0976 -0.0679 (0.145) (0.146) (0.144) (0.138) Firm size 0.000254*** 0.000305*** 0.000252*** 0.000211*** (7.13e-05) (8.02e-05) (7.06e-05) (7.83e-05) Firm size*MDA -3.97e-05 -5.67e-05 -1.26e-06 0.000168 (0.000328) (0.000355) (0.000339) (0.000209) Firm age 0.000925 0.000929 0.000931 0.00128 (0.000646) (0.000648) (0.000647) (0.000868) Firm age*MDA -0.00804** -0.00902** -0.00810** 0.00200 (0.00401) (0.00436) (0.00409) (0.00298) Electricity obstacle -0.00820 (0.00680) Electricity obstacle*MDA 0.0360 (0.0396) Finance obstacle -0.0225*** (0.00705) Finance obstacle*MDA 0.0261 (0.0402) Transport obstacle -0.00417 (0.00803) Transport obstacle*MDA -0.0226 (0.0420) Telecomm obstacle -0.0102 72 (0.0101) Telecomm obstacle*MDA 0.0317 (0.0502) Constant -101.8 -84.86 -83.49 -293.4 (559.5) (567.2) (559.9) (601.9) Country, year, country-year dummies YES YES YES YES Observations 8,090 7,617 8,056 5,648 R-squared 0.932 0.929 0.931 0.945 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A 8. Determinants of Productivity Based on Objective Indicators of Services’ Quality (1) (2) (3) (4) (5) VARIABLES TFP TFP TFP TFP TFP Capital 0.0465*** 0.0524*** 0.0676*** 0.0361 0.0678*** (0.0106) (0.00934) (0.0131) (0.0623) (0.0168) Capital*MDA -0.108 -0.0205 -0.0837* -0.426*** -0.152 (0.0737) (0.0399) (0.0471) (0.0833) (0.101) Materials 0.529*** 0.472*** 0.517*** 0.344*** 0.538*** (0.0225) (0.0125) (0.0176) (0.0793) (0.0268) Materials*MDA 0.0525 -0.0433 0.00212 0.143 0.0116 (0.0923) (0.0625) (0.0446) (0.0868) (0.185) Labor 0.365*** 0.370*** 0.324*** 0.389*** 0.281*** (0.0228) (0.0180) (0.0276) (0.111) (0.0390) Labor*MDA 1.06e-05 0.103 -0.0238 3.164*** 0.162 (0.0705) (0.0668) (0.0989) (0.281) (0.203) Exporter 0.0849*** 0.168*** 0.144*** 0.376 0.130*** (0.0234) (0.0229) (0.0317) (0.238) (0.0441) Exporter*MDA -0.0734 -0.0409 0.0523 -2.595*** 0.0916 (0.158) (0.110) (0.155) (0.488) (0.220) Foreign 0.0876*** 0.171*** 0.158*** 0.224 0.156** (0.0339) (0.0322) (0.0439) (0.295) (0.0655) Foreign*MDA -0.325** -0.0948 -0.439 3.732*** -0.721 (0.145) (0.145) (0.269) (0.507) (0.567) Firm size 0.000145** 0.000254*** 0.000216* 0.000103 0.000444*** (6.19e-05) (7.08e-05) (0.000115) (0.000124) (0.000116) Firm size*MDA 9.01e-05 -6.35e-05 0.00281 -0.0310*** 0.000731 (0.000681) (0.000347) (0.00210) (0.000993) (0.000797) Firm age 0.000700 0.000817 0.000688 0.0106* 0.000814 (0.000764) (0.000650) (0.000939) (0.00538) (0.00118) Firm age*MDA -0.00905 -0.00817** -0.0426 -0.757*** -0.0149 (0.00937) (0.00402) (0.0370) (0.0771) (0.0157) Collateral -0.000529*** (0.000126) Collateral*MDA 0.000125 73 (0.00103) Power outages dummy*MDA -0.0354 (0.0959) Power outages dummy -0.0283* (0.0165) Power outages length 7.41e-05 (0.00195) Power outages length*MDA 0.0303 (0.0204) Water Connection -0.00117** (0.000520) Water connection *MDA 0.00476*** (0.000533) Electric connection -0.000270 (0.000250) Electric connection*MDA -0.00115 (0.00191) Constant -554.9** -82.29 -946.0 3,372*** 897.1 (272.5) (557.9) (751.7) (328.9) (1,299) Country, year, country-year, sector, dummies YES YES YES YES YES Observations 2,755 8,017 3,053 244 1,308 R-squared 0.938 0.931 0.940 0.942 0.957 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A 9. Perceptions of Trade Regulations as Obstacles (1) (2) Customs regulations Business lic. and permits VARIABLES as obstacles as obstacles Firm size -4.74e-05* -3.80e-05*** (2.55e-05) (1.45e-05) Firm size*MDA 6.46e-05 -0.000137 (0.000153) (0.000162) Foreign 0.0304 0.000698 (0.0205) (0.0204) Foreign*MDA 0.0792 0.00625 (0.106) (0.102) Innovator 0.0543** 0.0484** (0.0223) (0.0202) Innovator*MDA 0.0710 0.0844 (0.102) (0.0866) Firm age -0.000950* -7.52e-05 (0.000540) (0.000508) Firm age*MDA 0.000440 -0.00754** (0.00449) (0.00313) 74 Exporter 0.0330* -0.0318* (0.0172) (0.0166) Exporter*MDA -0.0337 0.0308 (0.0862) (0.0846) Constant 1,477 -241.6 (1,494) (611.8) Country, year, country-year, sector dummies YES YES Observations 15,447 18,439 R-squared 0.088 0.065 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A 10. Effect of trade and customs regulations on productivity (1) (2) (3) (4) VARIABLES TFP TFP TFP TFP Capital 0.0617*** 0.0573 0.0573 0.0667*** (0.0148) (0.0831) (0.0829) (0.0201) Capital*MDA -0.00679 -0.453** -0.495** -0.0882* (0.0797) (0.208) (0.201) (0.0510) Materials 0.425*** 0.388*** 0.388*** 0.471*** (0.0219) (0.0520) (0.0519) (0.0271) Materials*MDA -0.185* -0.271*** -0.314*** 0.0118 (0.102) (0.0705) (0.0685) (0.0595) Labor 0.390*** 0.0427 0.0428 0.344*** (0.0274) (0.0977) (0.0975) (0.0368) Labor*MDA 0.278** 0.924*** 1.161*** 0.0850 (0.130) (0.279) (0.271) (0.0928) Foreign 0.126** 0.190 0.191 0.165*** (0.0507) (0.166) (0.165) (0.0506) Foreign*MDA 0.197 1.173*** 1.002*** -0.102 (0.230) (0.408) (0.372) (0.134) Firm size 0.000251** 0.00236*** 0.00236*** 0.000260** (0.000107) (0.000692) (0.000690) (0.000119) Firm size*MDA -0.000652 -0.00225 -0.00207 0.000428 (0.000792) (0.00147) (0.00158) (0.000609) Firm age 0.00145 0.00211 0.00212 0.00221** (0.000888) (0.00418) (0.00417) (0.00108) Firm age*MDA -0.00587 0.0213 0.0430 -0.00988** (0.00512) (0.0344) (0.0289) (0.00482) Innovator -0.0321 0.124 0.119 0.0241 (0.0299) (0.228) (0.227) (0.0346) Innovator*MDA 0.256* -0.650** -0.569* 0.0382 (0.153) (0.323) (0.311) (0.107) Customs time -0.000520* (0.000300) Customs time*MDA 0.00417 75 (0.00541) Import license time -0.00307 -0.00295 (0.00246) (0.00245) Import license time*MDA 0.0183* (0.0107) Customs reg. obstacle 0.0299 (0.0229) Customs reg. obstacle*MDA -0.155** (0.0658) Constant 3.322*** 6.139** 6.155** 2.357*** (0.404) (2.439) (2.434) (0.389) Country, year, country-year, sector dummies YES YES YES YES Observations 2,113 316 316 1,635 R-squared 0.912 0.870 0.870 0.921 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A 11. Corruption and Productivity (1) (2) (3) (4) (5) (7) (8) (9) VARIABLES TFP TFP TFP TFP TFP TFP TFP TFP Capital 0.0568*** 0.0607*** 0.0574*** 0.0568*** 0.0414*** 0.0754*** 0.0568*** 0.0568*** -0.00876 -0.00929 -0.00912 -0.00876 -0.00761 -0.0165 -0.00875 -0.00876 Capital*MDA -0.0161 -0.0252 -0.0199 -0.0226 -0.0309 -0.103** -0.0199 -0.0226 -0.0373 -0.0404 -0.0461 -0.0381 -0.071 -0.0491 -0.0383 -0.0381 Materials 0.466*** 0.476*** 0.480*** 0.466*** 0.584*** 0.394*** 0.466*** 0.466*** -0.0121 -0.0128 -0.0126 -0.012 -0.014 -0.0186 -0.0121 -0.012 Materials*MDA -0.0416 -0.0509 -0.0577 -0.0378 -0.0528 -0.0176 -0.0432 -0.0378 -0.0607 -0.0648 -0.0644 -0.062 -0.0932 -0.0707 -0.0617 -0.062 Labor 0.388*** 0.377*** 0.374*** 0.388*** 0.336*** 0.348*** 0.388*** 0.388*** -0.017 -0.0189 -0.0188 -0.017 -0.0168 -0.0281 -0.017 -0.017 Labor*MDA 0.082 0.0792 0.076 0.0858 0.0558 0.180* 0.0917 0.0858 -0.0614 -0.0675 -0.0691 -0.0636 -0.0808 -0.0923 -0.0629 -0.0636 Exporter 0.143*** 0.138*** 0.138*** 0.141*** 0.0693*** 0.260*** 0.142*** 0.141*** -0.0207 -0.0229 -0.0227 -0.0207 -0.0174 -0.043 -0.0207 -0.0207 Exporter*MDA -0.0434 -0.0213 -0.0605 -0.0501 0.049 -0.426** -0.0463 -0.0501 -0.108 -0.121 -0.133 -0.109 -0.138 -0.2 -0.11 -0.109 Foreign 0.158*** 0.159*** 0.158*** 0.158*** 0.0720** 0.274*** 0.156*** 0.158*** -0.0311 -0.032 -0.0318 -0.031 -0.0293 -0.0724 -0.0305 -0.031 Foreign*MDA -0.0369 -0.0594 0.0422 -0.0506 -0.0483 -0.188 -0.0557 -0.0506 -0.143 -0.155 -0.177 -0.144 -0.0809 -0.204 -0.143 -0.144 Innovator 0.0354** 0.0350** 0.0347** 0.0344** -0.00713 0.0421 0.0352** 0.0344** -0.016 -0.0176 -0.0173 -0.016 -0.0139 -0.0536 -0.016 -0.016 76 Innovator*MDA 0.0872 0.147 0.154 0.108 0.1 0.126 0.113 0.108 -0.0969 -0.107 -0.117 -0.0982 -0.112 -0.179 -0.0987 -0.0982 Firm size 0.000216*** 0.000199*** 0.000201*** 0.000217*** 7.47e-05** 0.000701*** 0.000217*** 0.000217*** -6.37E-05 -6.33E-05 -6.32E-05 -6.35E-05 -3.36E-05 -0.000187 -6.37E-05 -6.35E-05 Firm size*MDA -5.96E-05 6.39E-05 -3.68E-05 -1.74E-06 -1.21E-05 0.00089 -5.76E-05 -1.74E-06 -0.000307 -0.000346 -0.000359 -0.000309 -0.000296 -0.00104 -0.000307 -0.000309 Firm age 0.000589 0.000524 0.000433 0.000616 -0.000175 -0.00102 0.000584 0.000616 -0.000617 -0.000658 -0.000642 -0.000617 -0.000492 -0.00124 -0.000617 -0.000617 Firm age*MDA -0.00731* -0.00886** -0.00806* -0.00797** -0.0015 -0.0382** -0.00702* -0.00797** -0.00404 -0.00414 -0.00446 -0.00403 -0.00376 -0.0166 -0.00397 -0.00403 Courts -0.0225 -0.0181 Courts*MDA -0.151* -0.0835 Bribes customs 0.00717 -0.00722 Bribes customs*MDA -0.0686** -0.0607* -0.0321 -0.0365 Bribes courts 0.00903 -0.00742 Corruption dummy -0.0545*** -0.0193 Corr. dummy*MDA 0.128 -0.084 Corruption major obst. -0.0311* -0.0162 Bribes percent 0.00266 -0.00418 Bribes percent*MDA -2.98E-05 -0.00473 Corruption obstacle -0.0545*** -0.0193 Pol. instability -0.0204 -0.0158 Pol. Instability*MDA 0.187** -0.0887 Bribes lic. 0.371 -0.352 Bribes lic.*MDA -0.904** -0.39 Corruption obst.*MDA 0.128 -0.084 77 Constant 2.66*** 2.46*** 2.664*** 2.71*** 1.14*** 3.66*** 2.66*** 2.71*** -0.21 -0.266 -0.23 -0.21 -0.303 -0.367 -0.21 -0.21 Country, year, country- year, sector dummies YES YES YES YES YES YES YES YES Observations 8,121 6,576 6,631 8,121 4,534 2,486 8,121 8,121 R-squared 0.93 0.931 0.933 0.93 0.962 0.895 0.93 0.93 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table A 12 (1) (2) (3) (4) (5) Customs Corruption Bribes Bribes Bribes per reg. obstacle obstacle customs percent sales Firm size -4.27e-05* -6.88e-05*** -2.75e-05** -0.000341*** -0.000523*** (2.48e-05) (2.31e-05) (1.12e-05) (7.66e-05) (0.000134) Firm size*MDA 0.000106 0.000138 -0.000149 -0.00122* -0.000715 (0.000152) (0.000182) (0.000289) (0.000631) (0.000701) Exporter 0.0403** -0.0492*** 0.264*** 0.249** -0.284* (0.0170) (0.0162) (0.0167) (0.0990) (0.167) Exporter*MDA -0.0293 0.0440 0.312** 0.213 -0.260 (0.0894) (0.0940) (0.126) (0.421) (0.561) Foreign 0.0289 -0.0389* 0.132*** -0.216* -0.427** (0.0205) (0.0206) (0.0222) (0.114) (0.176) Foreign*MDA 0.0874 -0.0838 0.0303 0.997* 0.454 (0.105) (0.115) (0.151) (0.575) (0.791) Firm age -0.000946* 0.000284 -0.00318*** -0.0148*** -0.0171*** (0.000537) (0.000494) (0.000465) (0.00211) (0.00447) Firm age*MDA 0.000690 -0.00684* -0.00931** -0.00308 -0.0152 (0.00449) (0.00350) (0.00444) (0.0156) (0.0222) Manufacturing -0.0601*** 0.00709 0.0270* 0.0175 -0.114 (0.0173) (0.0155) (0.0150) (0.0961) (0.159) Manufacturing*MDA -0.0411 -0.153* -0.194** -0.237 0.923* (0.0878) (0.0815) (0.0966) (0.335) (0.559) Construction -0.0636** 0.0503** -0.00476 0.821*** 0.763*** (0.0278) (0.0219) (0.0210) (0.129) (0.190) Construction*MDA -0.207 0.163 -0.338*** 1.511 2.389 (0.198) (0.137) (0.116) (2.145) (2.869) Transport 0.00356 0.0304 0.0562* 0.0647 -0.0343 (0.0336) (0.0304) (0.0325) (0.168) (0.231) Transport*MDA 0.143 0.214 0.186 -0.362 0.344 (0.172) (0.138) (0.220) (0.621) (1.610) IT -0.161** 0.0310 -0.120*** 0.354 1.473* (0.0628) (0.0533) (0.0452) (0.404) (0.858) IT*MDA 0.670*** -0.697*** 0.245 -0.523 -5.460*** (0.154) (0.143) (0.884) (0.878) (1.711) Constant 1,366 636.3 362.6 1,616 -16,604** 78 (1,514) (610.2) (458.4) (3,754) (7,371) Country, year, country- year dummies YES YES YES YES YES Observations 15,447 23,328 34,237 18,690 7,705 R-squared 0.085 0.102 0.086 0.062 0.148 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 79