EntErprisE survEys 60353 EntErprisE notE sEriEs tradE Trade Performance in Eastern Europe and Central Asia 2010 Murat Seker I n the Eastern Europe and Central Asia (ECA) region, countries show great variations in their levels of openness and how intensively they trade. There is a strong and positive correlation in export and import market participation rates and a similar relationship between how intensively firms trade in each market. Using a firm-level data set, this note shows that the difference in export intensity between large and small firms is almost 30 percentage points; however, this difference is not seen among all countries. A comparison of industries shows that the garment industry is the most export and import intensive industry in the region. The note then focuses on customs efficiencies and EntErprisE notE no. 19 finds that countries with inefficient customs services export and import less intensively. Finally, this note analyzes how trade has evolved since 2005. The percentage of importers and import intensities have increased among the member countries in the European Union (EU); these statistics exclude the trade of EU countries with each other. In the region overall, import intensity has increased by 10 percentage points; on the other hand, there has not been a significant change in export intensity. International trade plays a crucial role in the development as exporting to improving a firm's performance. In his of countries (Freund and Bolaky 2008), making trade survey on technology diffusion, Keller (2004) summarizes performance central to the objectives of theoretical and empirical literature policy makers. Researchers have analyzed on how imports provide knowledge detailed firm-level data sets in order to In the ECA region, and technology transfer in a macro understand the micro foundations of almost 70 percent perspective. Using a firm-level data set, achieving high trade performance. This of exporting Seker (2010) shows how firms that import research has shown that trading firms are intermediate goods are more innovative intrinsically more productive and they firms also import and grow faster than non-trading firms. grow faster than non-trading firms (see intermediate The micro nature of the data used in Bernard et al. 2007 and Lopez 2005 for this analysis allows us to explore both reviews of the literature). Using detailed goods. exporting and importing patterns across firm-level data from the ECA region, this different firm characteristics. World Bank Group note evaluates the trade performances of countries in this The data for the analysis are collected through the World region. Moreover, the note relates trade performance to Bank's Enterprise Surveys (ES).1 A total of 11,306 firms customs efficiencies. Productive firms that want to increase were surveyed from 29 countries in the ECA region in 2008 their exposure to foreign markets might be constrained and 2009.2,3 In the surveys, a stratified random sample of by cumbersome customs clearing processes. Finally, this firms were selected that were representative of a country's note explores the change in trade patterns in the region manufacturing and service sectors. The surveys include between 2005 and 2008-2009. several questions related to international trade such as Firms engage with the foreign markets by exporting (1) what percentage of a firm's sales was due to direct or goods, importing materials or supplies, or by performing indirect exports (export intensity), (2) what percentage of both activities. Most of the existing studies on trade have material inputs or supplies were of foreign origin (import focused on exporting. Importing can be equally as crucial intensity) (asked only to firms in the manufacturing sector), and (3) what was the duration of time needed to Figure 1 Percentage of exporting and clear customs for imports and exports (time to import and importing firms time to export). The export and import data covered for the EU-10 countries is different from the other countries 60 · Slovenia in the region. For the EU-10 countries, trade measures · Serbia Percentage of Exporters exclude trade between EU member countries.4 · Czech · FYROM 40 · Bosnia · Turkey Export and import activities across the · Latvia · Estonia · · BelarusHungary ECA region · · Slovakia · Armenia Poland · Albania 20 Countries in the region show great variation in their levels · Bulgaria · Ukraine · Kosovo · Moldova · Kyrgyz · Romania · Tajikistan of openness and how intensively they trade (Table 1). The · Russia · ·Kazakhstan Montenegro · Azerbaijan · Uzbekistan differences between the most and the least open countries 10 20 30 40 measured as the percentage of exporters and importers Percentage of Importers are around 56 and 34 percentage points, respectively. The · ECA (non-EU) · EU-10 Fitted values countries with the highest trade participation rates for both importing and exporting are Slovenia, Turkey, and Source: Enterprise Surveys. FYR Macedonia. The percentage of exporters in Slovenia is three times more than the regional average. These firms export (which stands out as an outlier). Similarly, three countries are also among the most export intensive average export and import intensities across countries are countries. On the other hand, the Russian Federation, also positively and significantly correlated. Comparing the Kazakhstan, and Uzbekistan are among the least globally EU-10 with the rest of the region shows that among EU- integrated countries for both exporting and importing. 10 countries, the percentage of exporters is 10 percentage Countries with a high percentage of exporters also points higher than in the rest of the region (28 percent have high percentage of importers (figure 1). This vs. 18 percent) and that the EU-10 countries export more shows the complementarity between these two activities. intensively (12 percent vs. 7 percent). These differences In the region, almost 70 percent of exporting firms in exporter and export intensity percentages would likely in manufacturing sectors also import intermediate be higher if within-EU trade (for EU-10 countries) were goods. Given the fact that most global trade takes place included in the analysis. among intermediate goods, this complementarity can be Macro literature on trade and development shows that explained by the high integration of value chains across trade has a strong positive relationship with wealth (Freund the globe. The graph shows that in Albania, despite of and Bolaky 2008). Data from ES confirms this relationship. the high percentage of importers, roughly 20 percent of Countries that export more intensively have higher per capita incomes (figure 2). On average, EU-10 countries are almost Table 1 Countries with high and low trade performance Percentage of exporters Percentage of importers Export intensity Import intensity High Levels Slovenia 58 Albania 38 FYR Macedonia 21 Albania 79 Serbia 47 Slovenia 29 Slovenia 20 Estonia 63 Czech Rep. 38 FYR Macedonia 29 Turkey 17 Armenia 59 FYR Macedonia 38 Turkey 27 Estonia 15 Montenegro 52 Turkey 37 Estonia 23 Bosnia & Herz. 14 FYR Macedonia 51 Low Levels Tajikistan 9 Montenegro 10 Kazakhstan 2 Azerbaijan 21 Russian Fed. 7 Kazakhstan 9 Montenegro 2 Poland 20 Kazakhstan 5 Uzbekistan 8 Azerbaijan 2 Russian Fed. 19 Azerbaijan 4 Ukraine 5 Russian Fed. 2 Uzbekistan 17 Uzbekistan 2 Russian Fed. 4 Uzbekistan 1 Ukraine 16 Average 22 16 9 38 Source: Enterprise Surveys. 2 Figure 2 Export intensity vs. income Figure 3 Trade intensity and firm size · FYROM · Slovenia 20 Large (100 and over) · Turkey · Bosnia · Estonia 15 · Lithuania · Czech Export Intensity · Serbia · Albania · · Croatia · Latvia Hungary Medium (20-99) ··Belarus 10 Bulgaria · · Moldova · Poland Kyrgyz · Ukraine · Romania · Slovakia · Tajikistan 5 · Georgia Small (<20) · Montenegro · Azerbaijan · Russia · Uzbekistan 0 0 5,000 10,000 15,000 20,000 0 10 20 30 40 Per Capita Income Export Intensity Import Intensity · ECA (non-EU) · EU-10 Fitted values Source: Enterprise Surveys. Source: Enterprise Surveys. Note: The measure of income is gross national income per capita (in constant U.S. dollars) from World Bank Development Indicators (WDI). twice as rich as the rest of the countries in the region. Slovenia relationship between the probabilities of exporting is the richest and the second most export intensive country. and investment in R&D. For import intensity, the results are similar. In general, Trade and firm performance large firms are almost 20 percentage points more intensive Existing studies show that exporting firms are likely in importing than small firms. However, in Estonia, to be more productive and larger than non-exporting Tajikistan, and Kosovo, small firms have a significantly firms (Bernard et al. 2007). The firm-level data from the higher usage of foreign inputs than the large firms ES allows us to investigate how firms of different sizes (76 percent vs. 43 percent for Estonia, 48 percent vs. differ in trade intensity and whether this difference varies 15 percent for Kosovo, and 62 percent vs. 30 percent for across countries. To explore this relationship, firms were Tajikistan). divided into three size groups: small (5 to 19 workers), Some industries are more export oriented than the medium (20 to 99 workers), and large (100 workers or others. Hence export intensities are likely to vary across more). Small firms comprise 62 percent of the firms in the industries. Comparing trade intensities across seven major region. Medium and large firms comprise 29 percent and 9 manufacturing industries shows considerable variation percent, respectively. The data show that export intensity across industries (figure 4).5 The garment industry is the increases significantly with size (figure 3). The difference in export intensity between large and small firms is almost Figure 4 Trade intensity across 30 percentage points. This difference is more pronounced manufacturing industries among EU-10 countries compared to the rest of the region: 42 percentage points vs. 18 percentage points Garment in respective order. In FYR Macedonia, large firms are 45 percentage points more export intensive than small Textiles firms, which is the largest difference in the region. Such Electronics differences are consistent with common trade theories Chemicals and (see Melitz 2003). Productive firms are more likely to be pharmaceuticals large and are also more likely to compensate the sunk Metals and machinery costs of trading. However, in Armenia, Kazakhstan, Non-metal and plastics Serbia, and Montenegro, there is no significant difference Food in export intensities between large and small firms. Another performance measure where exporters differ 0 20 40 60 fromnon-exporters is in investment in research and Export Intensity Import Intensity development (R&D). In 14 countries there is a positive Source: Enterprise Surveys. 3 most export and import intensive industry among EU- still higher than exporting time. One explanation for this 10 countries and in the rest of the region. Textiles is difference could be congestion in the ports. In 22 out of the second most export intensive industry in the EU-10 29 countries, the percentage of importers is higher than group, whereas the chemicals industry is the most export the percentage of exporters. Hence, customs processes intensive in the rest of the region. In both groups of are likely to last longer for importers than exporters. countries, the food industry is the least intensive in both The time it takes to clear customs varies across firms in exporting and importing. different size groups. Large firms spend roughly twice as much time clearing exports through customs than do small Effects of customs efficiency on trade firms. For importing, among EU-10 countries--although not performance the most import intensive group--medium-sized firms spend Trade facilitation is essential to a the most time clearing customs. For the rest country's trading success. One important of the region, the duration increases as the aspect of trade facilitation is the efficiency Among EU-10 size decreases. Small firms spend twice as of customs in handling traded goods. much time as large firms spend to clear countries, the custom. Firms may be discouraged from trading if they find it too costly and cumbersome proportion of The customs clearing times from the ES to clear goods through customs. Evidence importing firms data includes only the duration between from ES supports this hypothesis. the arrival of goods to the main point Countries with high customs clearing increased by 10 of exit and the time these goods clear times export less intensively (figure 5). The percentage points customs. We can compare the ES data average time to clear customs for export with the Doing Business (DB) database, purposes is around 4 days. The country 6 since 2005. which measures customs clearing times with the lowest duration of time to clear and incorporates the duration of all other customs is Montenegro with 1.1 days, followed by Bosnia procedures from the contractual agreement between and Herzegovina, and Albania with around 1.5 days. Times the parties to the delivery of goods.7 The data in DB is for the five countries with the highest duration of time to collected for the transportation of a standardized cargo clear customs range from 6 to 20 days. The relationship of goods by only ocean transport where ES data includes between importing and the time it takes to clear customs all means of transportation. Moreover, the DB data is quite similar to the one for exporting. For importing, considers only the most populous city in the country, the average customs clearing time is 9 days. However, whereas ES data covers firms from other cities. Despite two outlier countries, Uzbekistan and Armenia, have the differences in their definition and coverage, there is a durations of 52 and 28 days, respectively. Excluding these strong positive correlation between the variables in ES and two countries reduces the average time to 7 days, which is DB that measure the total time to trade. This relationship shows the representativeness of the data from the ES in how customs efficiency varies across countries. Figure 5 Export intensity and time to export Over time comparison Among the firms surveyed in 2008-2009, 2,342 of · FYROM · Slovenia them were also surveyed in 2005. From 2005, importing 20 · Turkey in the ECA region increased significantly. Among EU- · Estonia · Bosnia · Czech 10 countries, the proportion of importer firms increased 15 Export Intensity · Serbia from 60 to 70 percent. In the region, import intensity also · Croatia · Latvia · Armenia · Hungary ·· Belarus · Bulgaria increased by roughly 10 percentage points from a level of 10 Hungary · Slovakia · Poland·Moldova · Poland · Albania 55 percent (figure 6). On the other hand, there has not been · Romania · Kosovo · Ukraine · Georgia a substantial change in either the percentage of exporters 5 · Azerbaijan · Tajikistan Uzbekistan Russia · Montenegro · · Kazakhstan or the export intensities. On customs clearance times, the · duration for importing increased by 3 percentage points, 0 0 2 3 6 8 Time to Export (Days) which could be due to increased importing in the region, · ECA (non-EU) · EU-10 Fitted values whereas the time it took for exporters to clear customs increased only slightly. Source: Enterprise Surveys. In the region, the Russian Federation was the only Note: The graph on the left excludes Kyrgyz Republic and Tajikistan due to their especially high customs clearance times. country that had a significant decrease in the percentage 4 Figure 6 Change in trade intensity References over time Bernard, A. B., J. Branford Jensen, S. J. Redding, and P. K. Schott. 2007. "Firms in International Trade," Journal of Economic Perspectives 21(3): 105­30. 80 Freund, C. and B. Bolaky. 2008. "Trade, Regulations, and Income," 70 Journal of Development Economics 87: 309­321. 60 Keller, W. 2004. "International Technology Diffusion," Journal of 50 Economic Literature 42(3): 752­782. 40 Lopez, R. A. 2005. "Trade and Growth: Reconciling the Macroeconomic and Microeconomic Evidence," Journal of Economic Surveys 19(4): 30 623­648. 20 Melitz, M.J. 2003. "The Impact of Trade on Intra-Industry Reallocations 10 and Aggregate Industry Productivity," Econometrica 71: 1695­1725. 0 2005 2008 Seker, M. 2010. "Importing, Exporting, and Innovation in Developing Countries," 2009, World Bank Policy Research Working Paper No. Export Intensity Import Intensity 5156. Source: Enterprise Surveys. Notes 1 See www.enterprisesurveys.org for a detailed description of the data of exporters since 2005 (by 19 percentage points). In and methodology used for data collection. The Enterprise Surveys, export intensity, the Russian Federation and Uzbekistan implemented in Eastern Europe and Central Asia countries, are also known as Business Environment and Enterprise Performance are the two countries that had a significant decline of 6 Surveys (BEEPS) and are jointly conducted by the World Bank and and 5 percentage points, respectively. On the import side, the European Bank for Reconstruction and Development for this in several countries, including Armenia, Croatia, Estonia, geographic region. Lithuania, Moldova, Poland, Slovenia, Tajikistan, Turkey, 2 These countries are Albania, Armenia, Azerbaijan, Belarus, Bosnia and Uzbekistan, the percentage of importers increased and Herzegovina, Croatia, FYR Macedonia (FYROM), Georgia, Kazakhstan, Kyrgyz Republic, Kosovo, Moldova, Montenegro, around 10 to 30 percentage points. The only country Russian Federation, Serbia, Tajikistan, Turkey, Ukraine, and that had a decrease in the percentage of importers was Uzbekistan and 10 recent European Union members: Bulgaria, Uzbekistan. The countries that had significant increases Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, in import intensity were Albania, Armenia, Estonia, Romania, Slovak Republic, and Slovenia. 3 Although countries were surveyed in either 2008 or 2009, the Lithuania, and Slovenia. survey questions refer to fiscal year 2007, and similarly, for the over This note presents evidence on the import and export time comparison, the 2005 survey presents data from fiscal year performance of countries in the ECA region. It shows 2004. that countries that perform one activity extensively are 4 In the graphs presented in this note, variables labeled as ECA (non-EU) excludes EU-10 countries. likely to perform similarly in the other activity. In almost 5 These are two-digit manufacturing industries that are classified all countries, exporting firms are larger and more likely to according to ISIC rev 3.1. invest in R&D. This note also presents evidence on how 6 The exact question on customs clearing time in the survey is "When the time to clear customs can be related to lower trade this establishment exported (imported) goods directly, how many performance. Finally, this note shows that there have been days did it take on average from the time this establishment's goods arrived at their main point of exit (e.g., port, airport) until the time significant increases in imports in the region since 2005. these goods cleared customs?" 7 See www.doingbusiness.org for a detailed description of the data and methodology used. The Enterprise Note Series presents short research reports to encourage the exchange of ideas on business environment issues. The notes present evidence on the relationship between government policies and the ability of businesses to create wealth. The notes carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this note are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. 5