WPS7616 Policy Research Working Paper 7616 MENA Export Performance and Specialization The Role of Financial Sector Development and Governance Christina A. Wood Judy Yang Middle East and North Africa Region Office of the Chief Economist March 2016 Policy Research Working Paper 7616 Abstract Industry and financial profiles of MENA firms may under- export data with country-level indicators, the results indi- pin the observation that MENA country exports are below cate that countries with more developed financial sectors potential and skewed toward low value-added goods that and stronger governance tend to have higher exports from are unable to spur rapid job creation and inclusive growth. sectors that are more reliant on finance external to the To assess this link, the paper combines analysis highlighting firm, and lower exports from sectors with higher shares external financing as a determinant of export performance, of tangible assets. Interestingly, financial sector develop- and analysis highlighting sector asset tangibility and gover- ment boosts exports less in MENA than in non-MENA nance. Why? Because high value-added sectors tend to have countries. To foster expansion of higher value exports, higher shares of intangible assets and to create innovative the results suggest a critical need for: (i) deeper financial products requiring substantial research and development sector development that strengthens market-based systems, or investments, thereby making these sectors more depen- such as asset registries and credit reporting agencies, and dent on external financing. Using sector- and firm-level (ii) strengthening of legal and governance frameworks. This paper is a product of the Office of the Chief Economist, Middle East and North Africa Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at cwood@worldbank.org and jyang4@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper 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. Produced by the Research Support Team MENA Export Performance and Specialization—The Role of Financial Sector Development and Governance Christina A. Wood Judy Yang The World Bank1 JEL Classification: F1, G0, K2, O14, O16 Key words: Exports, Financial development, Governance, External finance, Asset tangibility. 1 This analysis was commissioned as background paper for the Exporters in MENA regional study titled “Champions Wanted: Promoting Exports in the Middle East and North Africa” (published October 2015) supported by the MENA Chief Economist’s Office at the World Bank. We thank participants at the Exporters in MENA Workshop, held at the World Bank, for comments. This paper has benefited immensely from comments and input by Olivier Cadot, Ana Margarida Fernandes, Melise Jaud, Bob Rjikers and Murat Seker. We thank them for their time, comments, and for providing data. All errors are our own. MENA Export Performance and Specialization—The Role of Financial Sector Development and Governance 1. INTRODUCTION AND MOTIVATION The political events reshaping the Middle East and North Africa (MENA) region emphasize the importance of more inclusive development for political stability and accelerated growth. Research shows that exporting is vital for enabling more inclusive economic growth. Exporting firms tend to be larger, more productive, less financially constrained and more skill and capital intensive than their non-exporting counterparts (Melitz, 2003). Exporting firms tend to create more jobs, pay higher wages and offer better working conditions than non-exporting firms (e.g. Bernard and Jenssen, 1995, Brambilla et al, 2010, Frias, Kaplan and Verhoogen, 2008).2 Research also shows that countries with more sophisticated export baskets grow faster (Hausmann, Hwang and Rodrik, 2007), suggesting that export composition matters for a country’s growth and development prospects. Not exporting enough of the right products appears detrimental for sustained inclusive growth. MENA suffers from both insufficient exports and concentration of exports in the wrong products. Exports in the MENA region are concentrated in sectors that are low value-added and labor intensive, which may contribute to the absence so far of sustained export driven inclusive growth. Firms’ decisions on what to export, and where to, determine the evolution of aggregate levels and sector specialization of exports. Understanding the factors that influence firms’ export decisions is thus critical for ensuring that more appropriate policies aimed at fostering export growth and specialization are implemented. To understand trade specialization, the paper’s estimation strategy follows the line of research3 based on the theoretical Hecksher-Ohlin-Vanek framework, which predicts that a country will specialize in the production and export of goods more intensive in the country’s abundant factor endowments (Kletzer and Bardhan, 1987). This paper considers a country’s levels of financial sector development (FSD) and governance as among the country’s endowment basket. The paper then uses product and firm level data from the BACI International Trade Database4 and the World Bank’s Export Dynamics Database (EDD)5 to assess how FSD and governance matter for export levels and specialization profiles. Delving deeper into the existing literature, we find that export shares of sectors more dependent on external financing—funding external to the firm, i.e., from sources other than cash flow from operations—are higher in countries with more developed financial sectors (Beck, 2003; Svaleryd and Vlachos, 2005, Rajan and Zingales, 2001). Another strand of research finds that export shares 2 Andrew Bernard and Bradford Jensen “Exporters, Jobs and Wages in U.S. Manufacturing… 3 See Beck, 2003; Braun, 2003; Braun and Larrain, 2005; Claessens, Stilin, and Laeven, 2003; Demirguc-Kunt and Maksimovic, 1998; Fisman and Love, 2003a; Fisman and Love, 2003b; Hur, Raj, and Riyanto, 2006; Raddatz, 2006; Rajan and Zingales, 2001; Svaleryd and Vlachos, 2005. 4 BACI: International Trade Database at the Product Level, Gaulier and Zignago (2010). 5 EDD website: http://econ.worldbank.org/exporter-dynamics-database 2 from sectors comprised of more intangible assets 6 are higher when countries have stronger property rights (Claessens and Laeven, 2003; Hur, Raj, and Riyanto, 2006). These studies have utilized data with limited coverage of countries from the MENA region: Hur, Raj, and Riyanto’s (2006) study of finance and trade across 42 countries includes only one MENA country (Syria). Svaleryd and Vlachos (2005) examine only OECD countries and do not include any MENA countries. While Beck’s (2003) study of financial dependence and international trade does include five MENA countries (out of 56 countries total), it does not examine the aspect of asset tangibility. This paper combines aspects of the analysis focused on external financing as a determinant of export performance (Beck, 2003; Rajan and Zingales, 2001) and aspects of the literature focused on sector asset tangibility and governance (Braun, 2003), to assess the role of financial sector development and governance on export performance and specialization. Inclusion of more MENA countries in the analysis allows for MENA-specific characteristics to be discussed. Improving the understanding of what stimulates and inhibits exports in MENA is vital, as a number of empirical studies have shown that this region exports below its potential.7 Twenty-two percent of MENA firms report that customs and trade regulation are a major constraint to business operations (Figure 1, Enterprise Surveys8 ). Yet, the potential for trade is great. The region is strategically located, nested between European, African, and Asian markets. MENA is also endowed with human capital, experiencing a growing base of educated youth looking for work. In terms of the business climate, the Global Competitiveness Index ranks the MENA region better than upper-middle-income countries in the areas of infrastructure, health, primary education, and higher education. The literature finds that the presence of trade agreements between MENA countries and the rest of the world does not stimulate trade. On the other hand, domestic reforms are effective in stimulating trade by inducing a supply response from improved governance, regulations, openness, and labor markets (Freund, Portugal-Perez, 2013). The significance of institutional quality 9 towards promoting trade has also been documented by Francois and Manchin (2007) and Anderson and Marcoullier (2002). As regards financial sector development, MENA exhibits low levels of financial inclusion. Only 27.1 percent of MENA firms have a loan or line of credit, compared to 43 and 47 percent in the Eastern Europe and Central Asia and Latin America and Caribbean regions, respectively. 6 Intangible assets are assets with lower ability to be transferred between signatories to a contract (e.g. from borrower to creditor) when the terms of the contract are broken. Examples of intangible assets are goodwill, R&D, associated human capital, organizational capital and even accounts receivables, cash, inventory or related investments. Examples of tangible assets, those amenable for use as collateral, are real estate, machinery, industry plants. “The higher the proportion of tangible assets a firm or entrepreneur has the higher the effective protection the outside investor has, and the more likely the relationship will be stable and become feasible in the first place” Braun, 2003. 7 See Behar and Freund (2011), Bhattacharya and Wolde (2010), and Iqbal and Nabli (2007). 8 The World Bank’s Enterprise Surveys is a firm-level survey, conducted in over 110 economies, that provides useful insights on firm perceptions and experiences. It is important to note that most of the MENA Enterprise Surveys do not have weights, and the survey statistics cannot be extrapolated to be representative at the country or sector level. The exceptions are Iraq 2011 and Yemen 2010 surveys, which do follow the Enterprise Surveys Global sampling methodology. Note also that the surveys in the MENA region were conducted in different years. 9 Institutional quality factors include weighted components of the size of government, legal system property rights, sound money, freedom to trade internationally, and regulation. 3 Moreover, the region exhibits the world’s highest loan concentration ratio, reflecting a bias towards lending to large firms and excluding young and small firms (World Bank, 2011). Thirty- seven percent of MENA firms report that access to finance is a major constraint to doing business, higher than all regions except for Sub-Saharan Africa (Figure 1). Stimulating trade in the MENA region is not only an economic challenge, but also a governance challenge. Many empirical works find a strong relationship between poor governance and stunted growth in MENA.10 World Governance Indicator (WGI) scores for countries in the MENA region are much lower relative to other countries, and along certain dimensions of governance, MENA scores are lower than even lower-middle-income economies. Firms in MENA report being significantly more constrained along regulatory factors than in any other region. For example, 25.9 percent of MENA firms reported that business licensing and permits are a major constraint to their business operations, which is nearly double the rate in any other region (Figure 1). In addition to low governance ratings, political issues impede the movement of goods in the region. For example, Jordan neighbors the West Bank, which is largely excluded from the global market. Figure 1. Share of Firms Identifying Business Enviroment Factor to be a Major Constraint 50 43.0 40 36.6 33.4 Percent of Firms 30.8 30 26.6 25.9 22.3 19.7 20 17.8 17.1 19.1 15.9 15.1 18.3 11.8 10 8.4 7.6 9.2 0 Customs and Trade Regulations Access to Finance Business Licensing and Permits East Asia & Pacific Eastern Europe & Central Asia Latin America & Caribbean Middle East & North Africa South Asia Sub-Saharan Africa Notes: Surveys were implemented following the GLOBAL sampling methodology. However, the MENA average includes non-GLOBAL surveys since region has not had a complete survey roll-out. MENA average includes Algeria 2007, Egypt 2008, Iraq 2011, Jordan 2006, Lebanon 2009, Morocco 2007, Syria 2009, West Bank and Gaza 2013, and Yemen 2010. Source: Enterprise Surveys High-value added sectors tend to have higher shares of intangible assets and create innovative products requiring substantial R&D or investments, making these sectors more dependent on external financing. The composition of MENA exports is currently skewed towards low value- added goods. Investment in intangible-asset-intensive goods relies on a strong legal framework where assets are protected, the legal system functions efficiently and individuals or firms do not face risk of expropriation of their assets. To promote more rapid export driven inclusive growth, 10 See Elbadawi, 2005; World Bank, 2003; Aysan, Nabli, and Veganzones-Varoudakis, 2007 4 exports will need to skew more towards high value-added sectors, and this may require policy measures to strengthen particular features of the legal and governance framework in addition to more financial sector development. The analysis in this paper contributes to the FSD and governance policy reform discussion. The next sections are organized as follows. Section 2 benchmarks trade dynamics and two areas of MENA’s business environment focused in this study, financial sector development and governance. Sections 3 and 4 discuss the estimation strategy and data employed. Results are discussed in Section 5, and Section 6 concludes. 2. EXPORTS, FINANCIAL SECTOR DEVELOPMENT AND GOVERNANCE IN MENA 2.1. EXPORT DYNAMICS MENA’s export-to-GDP ratio outpaced the world average until the mid-1970s, subsequently took a downward turn and has continued to decline since then. By 2008, only 1.8 percent of global exports originated from MENA. Empirical studies repeatedly find that exporters in the MENA region export at rates below their potential. Behar and Freund (2011) find the MENA region under- traded by 60-70 percent in the 2000s. Bhattacharya and Wolde (2010) estimate MENA export levels are 86 percent lower than expected given the characteristics of their economies. Iqbal and Nabli (2007) find that non-oil exports in MENA are only one-third of their expected levels based on country characteristics. Between 2005 and 2010, MENA countries exhibited lower mean exporter size, and exported fewer products and to fewer destinations than other countries11 (World Bank’s Export Dynamics Database (EDD)). Using data from the EDD for the 2005-2010 period, summary statistics corroborate previous empirical studies showing that MENA countries export less and are less diversified than comparable countries in the rest of the world.12 Table 1 summarizes the sample statistics for 6 MENA countries and 26 non-MENA countries. The principal differences in export dynamics between these MENA and non-MENA countries lie in the number of exporters and the value of exports per exporter. The number of exporters and mean exporter size is much lower in MENA countries than in non-MENA countries. The average number of exporters in a MENA country is 5,552 and 13,258 in non-MENA countries. A final point to illustrate using the EDD, is that MENA and non-MENA countries export different goods. In MENA, trade is concentrated in labor-intensive sectors; the goods sector with the largest number of exporters is food products (25 percent of all exporters are in this sector). However, less than 10 percent of exporters from non-MENA countries are in this sector. In the rest of the world, the sector with the most exporters is more industrialized (machinery and electrical goods). The predominance in MENA of exports in the non-technical industries is related to the low level of regional skills, human capital and technological sophistication. 11Other countries exclude low-income countries. 12The Exporter Dynamics Database is restricted to data from 2005 – 2010, we also exclude agriculture and extractive industries. Since no MENA country is low-income, we also exclude low-income countries from the rest of the world. 5 Figure 2 illustrates WDI export indicators for MENA and non-MENA countries in 2010, by economy size. The logistics performance index scores a country on a range of aspects related to trade: customs, trade related infrastructure, price, tracking, etc. (Panel A). With regard to this indicator, most MENA countries underperform or perform near the expected level given in their level of GDP per capita. Inefficiencies in exporting in MENA are more apparent in terms of the number of documents necessary to export (Panel C), while costs to export in MENA are low for the majority of countries, except Iraq. Table 1. EDD Summary Statistics (2005-2010) Non-MENA MENA Country-level averages Average number of exporters 13,258 5,552 Mean Value per Exporter (USD2000) $2,413,509 $1,321,472 Median Value per Exporter (USD2000) $40,625 $59,289 Mean # of products per exporter 6.6 5.3 Mean # of destinations per exporter 2.93 2.6 N(country) 26 6 Notes: Unweighted averages. Low-income countries are excluded. Averaged at the country-year level Sources: Exporter Dynamics Database 2.2. FINANCIAL SECTOR DEVELOPMENT The financial sector in the MENA region is characterized as undiversified, uncompetitive, underexposed, and concentrated (Al-Ississ, 2012). The region exhibits low levels of financial inclusion and the world’s highest loan concentration ratio, reflecting a bias towards lending to large established firms and excluding young and small firms (World Bank, 2011). The financial sector is comprised largely of traditional banking institutions with a low presence of non-bank financial institutions; the insurance sector is small, leasing, factoring and other methods of external financing—i.e., funding sourced outside the firm—are limited. Figure 3 illustrates GCI’s Financial Market Indicator score for all countries by income. It shows that the majority of MENA countries have lower scores than expected given the level of GDP per capita. Above average performers are Bahrain and Jordan, neither one a large high-income country nor ones characterized by significant Islamic financing activity. The state of collateralization in the MENA region provides some insight as to why the percent of firms with loans or lines of credit is so low. Collateral lending in MENA is not common and when it does occur, it is heavily based on immovable (physical or tangible) assets such as land and buildings. However, in the developing world, immovable assets comprise 22 percent of capital stock (Fleisig Safavian, and De La Pena, 2006). In some areas of MENA, lending is also influenced by practices of Islamic financing, much closer in spirit to relationship banking rather than the arms- length market-based financing having greater presence in more developed countries. In Islamic financing cases, lending becomes a risk-sharing venture between the bank and firm; hence lending is more likely to be approved on the basis of personal acquaintance of borrowers (Alvarez, 2011). Islamic financing includes principles such as no interest rates, which turns lending into a risk- 6 sharing venture rather than one of risk-transfer. However, this reason alone is not satisfying in explaining the access to finance gap among the MENA countries in this report. Figure 2. Export Indicators, 2012 A. Logistics performance index, other B. Delay to export, other 80 IRQ 4 ARE 3.5 60 QAT TUN MLT SAU MAR BHR 3 EGY p YEM OMN KWT 40 y JOR LBN 2.5 IRN DZA YEM g IRN LBN 20 IRQ DZA KWT QAT 2 EGY JOR TUN BHR SAU MAR OMNMLT ISR ARE 1.5 6 8 Log GDP per capita (2012,USD) 10 12 0 6 8 Log GDP per capita (2012,USD) 10 12 C. Documents to export, other D. Cost to export, other 15 8000 6000 10 IRQ p p EGY DZA 4000 IRN BHR OMN KWT IRQ YEM 5 MAR JOR MLT SAU QAT 2000 TUN LBN ISR ARE IRN DZA YEM KWT JOR MLTBHR QAT EGY LBN OMN SAU TUN ISR MAR ARE 0 0 6 8 10 12 6 8 10 12 Log GDP per capita (2012,USD) Log GDP per capita (2012,USD) Source: WDI At the micro-level in MENA, there is inequity in access to finance by firm size. The financial sector is dominated by banks that primarily serve large firms and exclude small and medium firms. Firms in MENA use very little external financing for investment and working capital. In MENA, 5.9 percent of investments and 4.4 percent of working capital is externally financed by banks (Enterprise Surveys). This proportion is lower only in the Sub-Saharan Africa region. The percent of firms who have a checking or savings account, and who have a bank loan or line of credit is also low in the MENA region. Sixty percent of MENA firms have checking or savings accounts 7 (Table 2). Only 12.1 percent of MENA firms have a loan or line of credit, compared to over 40 percent in the Eastern Europe and Central Asia and Latin America and Caribbean regions respectively. Figure 3. GCI Financial Markets Indicator (2005-2009), MENA vs Non-MENA 6 ISR BHR MLT QAT 5 gci8finmkt KWT ARE JOR OMN SAU 4 TUN MAR EGY SYR 3 DZA 7 8 9 10 11 Log GDP per capita (2005,USD) Source: GCI, WDI Table 2. Finance Characteristics, By Region % Firms with a checking % Firms with a bank Proportion of loans or savings account loan/line of credit requiring collateral (%) MENA 59.3 12.1 75.2 EAP 87.7 37.6 76.6 ECA 90.2 40.1 80.3 92.9 47.6 72.4 LAC SAR 78.9 34.8 84.3 AFR 88.1 23.8 79.7 Notes: Surveys were implemented following the GLOBAL sampling methodology. However, the MENA average includes non- GLOBAL surveys since region has not had a complete survey roll-out. MENA average includes Algeria 2007, Egypt 2008, Iraq 2011, Jordan 2006, Lebanon 2009, Morocco 2007, Syria 2009, West Bank and Gaza 2013, and Yemen 2010. Source: Enterprise Surveys 2.3. GOVERNANCE Good governance has been shown to have a causal link to growth at the micro and macro levels.13 Governance has also been significantly linked with an array of other economic outcomes including GDP per capita,14 FDI, private investment, and openness to trade. The definition of governance 13 See La Porta et al 1999; Kaufmann, Kraay, and Ziodo-Lobaton, 2002; etc. 14 See Acemoglu, Johnson, and Robinson, 2001; Halls and Jones, 1999; Easterly and Levine, 2002; Rodrik, Subramanian, and Trebbi, 2002 8 matters: most studies find significant effects using “rule of law” as a proxy for governance, however, robustness of the results is weaker when using corruption or democracy as proxies. The role of governance in handicapping MENA’s economic growth is not a new thesis (see World Bank, 2003a). Weak governance has been estimated to have slowed MENA’s GDP growth by 1 to 1.5 percentage points (World Bank, 2003a). Poor governance also discourages foreign investment that can be important for job creation, innovation and trade. Unlike other regions, trade has not been a driver of job creation in MENA. Political instability is a primary source of risk and uncertainty for the region, and FDI has not only decreased but has also skewed towards sectors with low job creation and that do not support exporting (World Bank, 2013). The low quality of governance in MENA is readily apparent in the World Governance Indicators (WGI).15 World Governance Indicator scores for MENA countries are much lower relative to other countries. In five out of the six areas, MENA countries have significantly lower scores than both upper middle and high-income countries. Two governance values identified to be very relevant for MENA are inclusion and accountability (World Bank, 2003a). The WGI Voice and Accountability indicator is very closely aligned to these two values. Indeed, the MENA region ranks lower than any other in this indicator. Moreover, among all size WGI indicators, the MENA region has the lowest scores in Voice and Accountability. Figure 6 (in the Appendix) illustrates WGI rankings given GDP per capita, averaged from 1995- 2009. In the instance of Voice and Accountability, MENA countries perform the worst in this ranking given their level of development. 3. ESTIMATION The literature offers many studies on complementarities between country-level characteristics and policies (or country-country level interactions) towards growth and economic outcomes. 16 For example in the MENA region, Nabli and Veganzones-Vatoudakis (2007) find that macroeconomic stability and structural reforms are complementary in boosting GDP per capita, and that structural reforms are ineffective without macroeconomic stability. The literature on country-sector complementarities is smaller, and none of this research focuses specifically on understanding factors for trade in the MENA region. In this respect as well, this paper is a contribution to the literature. To estimate how sector-level trade specialization is influenced by country-sector complementarities (see Box 1 for more detailed information), equations of the following basic form are estimated (Eq. 2). Variations of this basic specification are also explored and will be described in due course. 15The World Governance Indictors are developed by the World Bank. World Governance Indicators and are available for six areas; Rule of Law, Regulatory Quality, Political Stability & Absence of Violence/Terrorism, Control of Corruption, Government Effectiveness, and Voice and Accountability. World Governance Indicators values are normalized between -2.5 and 2.5 with a higher value being associated with better governance. 16 For an example related to trade, see Freund and Bolaky, 2008 9 ∗ ∑ ∑ (2) Observations are at the level of country (c) and sector17(s). Following the literature, we average all available data across several years to focus on the cross-sector aspects of trade specialization (Beck. 2003; Hur, Raj, and Riyanto, 2006; and Svaleryd and Vlachos, 2005). The dependent variable is the sector’s share in total exports, characterizing trade specialization. Full country and 2-digit ISIC industry controls are included to allow for an examination of within effects. The variable is the error term. The coefficient on the country-sector interaction term is of principle interest. Estimated coefficient estimates can only be interpreted as correlations, since no attempts have been made to control for simultaneity or endogeneity. Reverse causality cannot be ruled out; the possibility that governance and financial sector development can also be a consequence of trade cannot be dismissed. Using the proper econometric techniques, the role of governance has been found to have significant impact on growth and not the other way around. 18 However, the directionality between trade and governance is less clear although some research finds that openness to trade creates positive spillovers that promote better governance.19 Box 1. Country Endowments and Sector Dependencies, Predictions from Theory The impact of the interaction between external finance dependence on trade specialization onto trade is expected to be positive. Theoretically, Kletzer and Bardhan (1987) argue that countries with better developed financial sectors have higher exports from sectors that rely more on external financing, holding all else constant. If financial sector development reduces the costs of external finance to firms, then countries with more developed financial institutions will specialize in sectors that rely more on external finance. Comparative advantage through financing can be important for trade by facilitating risk diversification, providing finance to sectors in need, or allowing the acquisition of improved technology to enhance productivity. The impact of the interaction between asset tangibility and trade specialization is expected to be negative. The higher a sector’s level of asset tangibility the more it facilitates collateralization and liquidity in markets where default risks are high or when contract enforcement is weak (Braun, 2003). If higher asset tangibility facilitates lending when contract enforcement is weak or when default risk is high, then countries with a more (less) developed governance or financial sector institutions should have lower (higher) export shares in industries that use more tangible assets. Predicted Effect onto Trade Specialization Financial Sector Development External Finance Dependence + Financial Sector Development Asset Tangibility - Governance Asset Tangibility - 17 Specifically, industries are identified by the 2-digit harmonized systems product classification code (http://www.foreign- trade.com/reference/hscode.htm). 18 For example, see La Porta et al 1999; Rodrik, Subramanian, Trebbi, 2002 19 See Berg and Kruegter, 2003; Islam and Montenegro, 2002; Wei, 2000 10 11 4. DATA 4.1. EXPORT DATA The principal trade data source is the BACI World trade database developed by CEPII, created using COMTRADE. BACI data from 1995 to 2009 are used for this exercise. In total, 20 MENA countries and 136 countries from other regions are covered (see Table 6 for a list of MENA countries). One advantage of the BACI to the COMTRADE is that exporter and importer declarations have been checked for gross inconsistencies. The version used is aggregated at the country-sector level, where sectors are defined by the ISIC 2-digit level. It is important to be clear that there is no information on exporter characteristics in this database such as age, legal status, ownership, share of exports over total sales, or employment levels. Moreover, this study cannot answer why some firms export and others do not, or what are the determinants of becoming an exporter. The World Bank’s Export Dynamics Database (EDD) is also used as a second source of customs data. The EDD includes exporting data from 1997-2011, but analysis is limited to 2005-2010 since data are available for the majority of countries during this period only. All agricultural and extractive industries are excluded. The EDD’s country coverage is narrower than BACI’s, however EDD comprises additional information in terms of mean and median export value per exporter, which allows consideration of the intensive margin of trade. The full EDD data set comprises 43 countries, 7 of which are in the MENA region. Kuwait is the only high-income MENA country in the EDD. Egypt, Morocco, and Yemen are lower-middle-income, and Iran, Jordan, and Lebanon are upper-middle-income. Depending on our level of analysis and availability of control or business climate variables, MENA coverage ranges from three countries to seven countries. 4.2. DEPENDENT VARIABLES The share of total exports in a given sector is calculated per industry for every country and every sector using the BACI. Trade specialization at the sector-level in a given country ( ) is calculated as the value of total exports in sector s divided by total export value in country c (Table 7 in the Appendix). A sector’s export value is calculated as the number of exporters multiplied by the mean exporter size. EX = 100 ∗ (3) In addition to the share of exports, the mean and median value of exports per exporter is also used from the EDD to examine the intensive margin. 12 4.3. COUNTRY FINANCE AND GOVERNANCE INDICATORS Private credit to GDP ratio is used to proxy of financial sector development (Table 9 in the Appendix). This is in line with the literature which has generally utilized private credit to GDP ratios to proxy for financial sector development (Beck, 2003; Braun, 2003; Classens, Stijin, and Laeven, 2003; Fisman and Love, 2003a). Financial indicators are worse in MENA than in high- income countries, and are lower than expected given their income levels (Figure 4). For example, the private credit to GDP ratio in high-income countries is 117.2 compared to only 49.25 in MENA. In the area of governance, property rights index is used as a proxy, which has also been used by Claessens and Laeven (2005), and Hur, Raj, and Riyanto (2006). Global Competitiveness Reports offer a wide range of scores ranging from 1(worst) - 7(best) on a country’s development. These scores are strongly correlated with income levels. High-income countries score the best across all topics, in most cases at least one-point higher than any other income group. Low-middle-income countries also have the lowest GCI scores in the group. The GCI property rights index in MENA is higher in than non-MENA countries, but much lower than in high-income countries. A second proxy for governance is the set of World Governance Indicators (WGI), developed by the World Bank (See Kauffman, Kraay, and Mastruzzi, 2009). World Governance Indicators values are normalized between -2.5 and 2.5 with a higher value being associated with better governance. The WGI Voice and Accountability indicator captures the quality of interaction between citizens and government and is particularly relevant for the MENA region, as well as being instrumental towards promoting a stable environment for production and investment (World Bank, 2003a). Among all six WGI indicators, the MENA region averages the lowest in this measure. Voice refers to the voice of the citizens and their freedom of expression, choice in government, and ability to demand action from those in power. Accountability refers to the accountability of the government and refers to the presence of sanctions whether through voting, institutional oversight, or oversight from a free media. 13 Figure 4. Private Credit to GDP Ratio (1995-2009), MENA vs Non-MENA 200 150 privatecredit MLT 100 ISR JOR LBN TUN KWT 50 BHR MAR EGY ARE OMN QAT DJI SAU IRN LBY SYR DZA YEM IRQ 0 7 8 9 10 11 Log GDP per capita (2005,USD) Source: WDI 4.4. SECTOR-LEVEL DEPENDENCIES Two measures of sector-level dependencies are used: external finance dependence, and asset tangibility. Measures of an industry’s need for external finance are taken from Rajan and Zingales (1998). External finance dependence is calculated as the fraction of capital expenditures not financed with cash flow from the firm’s own operations, using data from U.S. firms in the 1980s. The United States is a good example to isolate industrial financing needs since markets are relatively frictionless especially among large companies. For example, industries requiring a lot of R&D such as drugs and pharmaceuticals require a lot more financing than say pottery. By treating external finance dependence as similar across all countries, one also avoids the problem of entangling country financial sector effects into industry-specific measure. An industry’s level of asset tangibility is derived in Braun (2003) using U.S. Compustat data from 1986-1995. The most tangible industries include metals and chemicals, less tangible industries include apparel and chinaware. Figure 5 illustrates the average external finance dependency and asset tangibility by country averaged over the period 1995-2009. The two sector dependency indicators are trade weighted by sector using BACI export values from this same time period. Figure 5a shows that richer, more developed countries tend to have lower average dependence on external finance. Figure 5b shows that richer, more developed countries tend to be skewed toward sectors comprised of more intangible assets, and MENA countries are no exception. Interestingly, MENA countries are split in terms of which export more in sectors with higher asset tangibility or external finance dependence, yielding a high-level of heterogeneity for analysis. 14 Figure 5. Sector Level Dependencies (1995-2009 average), MENA vs Non-MENA 5a. Dependence on External Finance – Country Averages across sectors .45 rajanzingales .4 IRQ LBY DJI .35 YEM QAT DZA KWT MLTBHR OMN JOR LBN MARSYR TUN IRN EGY ISR .3 SAU ARE 7 8 9 10 11 Log GDP per capita (2005,USD) 5b. Asset tangibility –Country average across sectors .32 SAU ARE EGY IRN ISR JOR MAR TUN LBN SYR OMN .3 MLTBHR KWT DZA DJI QAT YEM LBY IRQ tangibility .28 .26 .24 7 8 9 10 11 Log GDP per capita (2005,USD) Notes: Average external finance dependency and asset tangibility are trade weighted by sector using BACI customs data. Source: WDI, BACI, Braun (2003), Rajan Zingales (1998). 15 5. RESULTS 5.1. TRADE SPECIALIZATION First, we examine the outcomes when only financial sector development, proxied by the private credit to GDP ratio, is considered. The association between financial sector development with both sector-level external finance dependence ( ) and asset tangibility ( ) are included in the regression. The interaction terms with these two sector-specific requirements capture different mechanisms. The interaction term with captures how industries highly reliant on external financing benefit more from financial sector development since financing from sources outside the firm becomes less costly relative to the firm’s own funds. The interaction term with reflects the agency problem and the preference towards lending to firms that have tangible (i.e,. physical) assets to offer as collateral. Table 3 display estimates for equation (2) using the share of exports per sector as the dependent variable. Results for trade specialization are shown using data from both the BACI and the EDD. The BACI data contains many more MENA countries than the EDD data set. Results in columns (1) and (3) yield similar results to Beck (2003), a significant correlation is found between sector- level external finance dependence and a country’s financial sector development. The interpretation being that for a sector with a given level of external finance dependence, the share of exports from that sector is higher in countries where the financial sector is more developed. This result holds irrespective of the export data source, though the effect is much smaller using BACI data than the EDD, possibly reflecting the different country and temporal coverage of the data sets. Results from columns (2) and (4) are also consistent with Hur, Raj, and Riyanto’s (2006) study, which examines both sector level external finance dependence and asset tangibility. The summary statistics in Table 3 (lower section) shows that sector-level dependencies and private credit to GDP ratios between the MENA and non-MENA country groups are similar using either the BACI or the EDD data. Non-MENA export sectors are more dependent on external finance and have lower shares of tangible assets compared with MENA export sectors. The private credit to GDP ratio is also higher in non-MENA countries, highlighting their higher levels of financial sector development in those countries. If MENA countries were to have similar levels of financial sector development as the rest of the world, we would expect MENA’s exports to reflect greater specialization in sectors more dependent on external finance and more intangible assets, a point worth noting since these sectors tend to have higher value-added. Next, we examine results of the regression incorporating both country-level financial sector development and governance indicators. Table 4 presents results when financial sector development is proxied by the private credit to GDP ratio, and various governance proxies. The dependent variable is the share of exports by sector. Comparative advantages from a sector’s external finance dependence are naturally best linked with financial sector development. Asset tangibility is related to liquidity, contracts, and ownership, and more related to legal and governance factors. This specification is most similar to Hur, Raj, and Riyanto’s study (HRR, 2006). 16 Table 3. Financial Sector Development and External Finance Dependence Dep. Var = Sector’s Share of Total (1) (2) (3) (4) Exports Export Share Export Share (BACI 1995-2009) (EDD 2005-2010) Private Credit /GDP * EXTFIN 0.000936*** 0.000708*** 0.00238** 0.00237** (0.000160) (0.000146) (0.00106) (0.00103) Private Credit /GDP *TAN -0.00255*** -3.04e-05 (0.000443) (0.00185) Constant 0.101*** 0.146*** 0.292** 0.292** (0.00309) (0.00700) (0.131) (0.137) Observations 3,832 3,832 2,394 2,394 R-squared 0.280 0.287 0.101 0.101 Number MENA countries 20 20 5 5 Number of Non-MENA countries 119 119 26 26 Summary statistics Average Private Credit/GDP (MENA) 43.57 49.25 Average Private Credit/GDP (non-MENA) 57.49 54.70 Average EXTFIN (MENA) 0.239 0.250 Average EXTFIN (non-MENA) 0.285 0.258 Average TAN (MENA) 0.382 0.32 Average TAN (non-MENA) 0.316 0.30 Notes: Estimation results from Equation 2. Panel dimension: country-sector. Average EXTFIN and TAN values for MENA and non-MENA regions are simple averages of trade weighted country values. Source: WB staff calculations, BACI, EDD. Similar to HRR (2006), a relationship is found between trade specialization and the interplay between property rights and asset tangibility (column 2, Table 4). Sectors with high shares of intangible assets are expected to benefit the most from secure property rights, the exact attribute that the GCI indicator captures. Significant correlations between governance and asset tangibility level of export sectors are also observed when governance is proxied by the GCI Legal Rights index and World Governance indicators (columns 3-8). Considering the legal framework or governance factors of a country to be endowments, it can thus be expected that in countries with well-developed legal frameworks, trade is specialized in sectors having lower shares of tangible assets. When the legal framework is strong, banks are more likely to enter into contracts and less prone to worry about risks associated with default. The quality of a country’s legal framework influences financial sector development as well as the ease of access to external financing.20 20 La Porta, Lopez-de-Silanes, Shleifer, and Vishny, 1998; Jensen and Meckling, 1976; Williamson, 1988; Grossman and Hart, 1986; Hart and Moore, 1990; Hart, 1995 17 Table 4. Trade Specialization (Dep. Var = Sector’s Share of Total Exports) (1) (2) (3) (4) (5) (6) (7) (8) Private Credit / GDP * EXTFIN 0.000708*** 0.000945*** 0.00107*** 0.000774*** 0.000817*** 0.000770*** 0.000809*** 0.000818*** (0.000146) (0.000182) (0.000149) (0.000160) (0.000158) (0.000159) (0.000162) (0.000168) Private Credit / GDP * TAN -0.00255*** (0.000443) Property Rights * TAN -0.0500** (0.0220) Legal Rights * TAN -0.0343** (0.0143) WGI Regulatory Quality * TAN -0.122*** (0.0263) WGI Govt Efficiency* TAN -0.132*** (0.0291) WGI Political Stability* TAN -0.111*** (0.0237) WGI Corruption* TAN -0.0863*** (0.0243) WGI Voice* TAN -0.101*** (0.0271) Constant 0.146*** 0.178*** 0.222*** 0.136*** 0.151*** 0.146*** 0.134*** 0.133*** Observations 3,832 2,770 2,685 3,832 3,832 3,832 3,832 3,832 R-squared 0.287 0.311 0.316 0.286 0.288 0.286 0.284 0.285 Number of MENA countries 20 14 14 20 20 20 20 20 Average Stats WGI*TAN interaction (MENA) 12.82 1.504 1.066 -0.0483 -0.114 -0.0314 -0.00605 -0.238 WGI*TAN interaction (non- 17.05 1.397 1.595 0.0803 0.0521 0.0831 0.0610 0.0805 MENA) Notes: Estimation results from Equation 2. Panel dimension: country-sector. Data is averaged over 1995-2009. Source: WB staff calculations, BACI 5.2. INTENSIVE MARGINS Export specialization is not the complete story regarding export dynamics in MENA. While specialization can offer insights to efficiencies and comparative advantages gained from sector- level needs and country endowments, it is also useful to examine the average volumes of trade per exporter. The estimation of exporter size is very similar to the trade specialization specification. Here, the data series is expanded to include a time dimension (4). The data set used in this case is the EDD. ∗ ∗ ∗ ∗ log (4) tariff TradeOpen The dependent variable, “SIZE” (exporter size), is the mean or median export value per exporter. A few country characteristics are included such as the log of GDP per capita (PPP), weighted tariff rates, trade openness, as well as full country and 2-digit industry fixed effects. In addition to the complementarity effects previously discussed, the main effect of the country-level endowment is also expected to be positively related to mean and median exporter size. Exporter size should be higher in countries with better financial sector development. Another motivation to include the time dimension is to capture the variation in the macroeconomic variables during the Arab Spring. 18 MENA interaction terms are added to the baseline equation (4) to capture differential regional effects. Since the estimation of exporter size includes country-controls, some countries are excluded due to lack of a complete data profile. In estimation of equation (4), country coverage ranges from three countries with full country controls, to all countries without any country controls. In regressions with complete controls, only Morocco, Jordan and Egypt are represented. For the analysis of exporter size, data is at the country-sector-year level and additional variables are included in the regression to control for country level factors: the log of GDP per capita (PPP), trade openness and a tariff rate21 (Table 7 in the Appendix). GDP can benchmark for market size. Country-level determinants included in the regression are the log of GDP per capita (PPP), trade openness, and a tariff rate. As a measure of trade openness, the total export and import value in manufacturing and services is divided by a country’s GDP (TradeOpen) using data from the WDI.22 The share of trade relative to GDP is similar in MENA and non-MENA regions, 0.79 and 0.78 respectively. Results are examined when financial sector development is proxied as previously by the private credit to GDP ratio. Table 5 shows results where the complementarity between external finance dependence and private credit-GDP ratio is significantly and positively associated with both mean and median exporter size. Comparing regions, financial impacts on export size are smaller in MENA when associated with mean exporter size (Private Credit / GDP * MENA). For a given sector, an identical increase in the private credit-GDP ratio yields a smaller increase in mean exporter size in MENA than in non- MENA countries. However, large firms in MENA may be benefiting more from financial sector development than smaller firms. The interaction term Private Credit/GDP*EXTFIN*MENA is positive in the case of mean exports but not median exports. This may illustrate a bias in access to finance depending on firm size. This is reasonable considering that large firms are favored by the financial sector in MENA but not significant to the same degree as in the rest of the world. MENA has the world’s highest loan concentration ratio, illustrating a bias in lending to large firms versus young and small firms (World Bank, 2011). Two more points on the size distribution are also examined, the 25th and 75th percentiles. Financial country-sector complementarities predict significantly larger increases in the 75th percentile of exporter size in MENA than non-MENA countries. There are no regional differences at the 25th percentile of exporter size. This is further evidence of gains from access to finance being reaped primarily by large firms in MENA. Further firm-level analysis would be interesting here to examine differences in access to finance by firm size. 21 Both average and weighted tariff rates were tested and yield very similar results. 22 This measure is constructed using constant 2000 USD and includes both manufacturing and service goods. 19 Table 5. Mean and Median Exporter Size23 log(Mean Exporter Size) log(Median Exporter Size) (1a) (2a) (1b) (2b) Private Credit / GDP 0.00185 0.00237 0.00150 0.00247 (0.00229) (0.00255) (0.00243) (0.00289) Private Credit / GDP * EXTFIN 0.00656*** 0.00692*** 0.00428*** 0.00450*** (0.00133) (0.00133) (0.00110) (0.00111) Private Credit / GDP * MENA -0.0200*** 0.00823 (0.00583) (0.00864) Private Credit / GDP * EXTFIN * MENA 0.00441*** -3.78e-05 (0.00150) (0.00220) Property Rights -0.0725 0.0634 -0.132 0.0238 (0.141) (0.140) (0.143) (0.151) Property Rights *TAN -0.102 -0.109 -0.254 -0.229 (0.327) (0.316) (0.244) (0.241) Property Rights *MENA 0.558*** 0.229* (0.0896) (0.123) Property Rights *TAN *MENA -0.202** -0.191*** (0.0957) (0.0641) Constant 7.571*** 8.695*** 10.71*** 12.47*** (2.263) (2.322) (2.193) (2.259) Observations 9,816 9,816 9,816 9,816 R-squared 0.451 0.430 0.451 0.430 Number of MENA countries 3 3 Source: EDD (2005-2010) Notes: Three MENA countries include Morocco, Egypt, and Jordan. Results without country controls to allow for the inclusion of more MENA countries yield similar results and are available upon request. Full industry, country, and year controls are included. Errors are clustered by country-year. Standard errors are in parentheses. Significance level markers: *0.10 ** 0.05 *** 0.01 Next, the Property Rights*TAN interaction term is discussed. In developing countries, banks prefer immovable assets (i.e., a high degree of asset tangibility) such as land for collateral, which highlights an agency problem. The influence of asset tangibility on trade specialization can be magnified or diminished by the development of legal framework and other governance-related factors such as property rights and credit protection. The interaction of property rights and asset tangibility with exporter size (Property Rights *TAN) is insignificant. Strong property rights are not correlated with larger mean or median exporter size in industries characterized by higher asset intangible. However, despite the insignificance, the effects are significantly different in MENA and non-MENA regions (Property Rights*MENA). One explanation is that the GCI Property Rights Index is higher in MENA than other regions, and offers more effective protection than in non-MENA. Property Rights in MENA also appear to be 23 We estimate regressions with average or weighted tariffs. Trade regulations are generally viewed to be much more restrictive in MENA countries. Weighted tariff rates in MENA average 17.4 percent, and 8.7 percent in non-MENA countries. Tariff data from the WDI is unavailable for Lebanon or Kuwait. Egypt, Iran, and Yemen have only two years of tariff data. The availability of tariff data does not always coincide with availability of EDD or business climate data. 20 promoting exports high in intangible assets at both the mean and median levels (Property Rights*TAN*MENA). Median size can be seen as a more equitable measure of the size of exporters since a few very large firms will not pull up the average values. For MENA, governance (proxied here by property rights)24 appears to be helping a wider range of exporters, while financial access is benefiting mostly large firms. 6. CONCLUSION Domestic policies and the investment climate are critical for promoting growth in MENA. A number of studies conclude that trade positively affects growth and living standards. 25 In predicting export flows, institutional quality and domestic policies can be even more relevant than international trade agreements. In this paper, trade is examined through the interaction between financial sector development and governance on the one hand and relevant sector-level dependencies on the other hand. This paper finds that the quality of a country’s financial sector development and governance framework has a positive correlation with sector-level dependencies. In MENA, financial sector development appears to be linked to mean exporter size, but not median export size, i.e., the contributions of exports by large firms matter. On the other hand, improved governance in MENA has a positive correlation with both median and mean exporter size, with beneficial impact on both large and small exporters. These results are related to other research, which has also found that governance effects dominate economic effects in some cases (Rodrik, Subramanina, and Trebbi, 2004). For the MENA region, the literature has also found that good governance predicts significant improvements in growth and private investment. Political instability has not only decreased the levels of FDI in MENA in the 2000’s, but has also skewed it away from tradable sectors (World Bank, 2013). For countries seeking to increase the size and composition of their exports in a manner that fosters more rapid job creation and inclusive growth, the analysis and results of this paper point to consideration in the following two policy areas. First, a critical need for more financial sector development to create or strengthen market-based systems, such as asset registries and credit reporting agencies (in addition to the more fundamental objectives of access to, and efficiency/stability of, the financial system that may be needed). Second, a strengthening of legal and governance frameworks, particularly as regards property rights and credit protection. 24 The authors are open to sharing results where other governance indicators are used, but only show property rights here for brevity. 25 See Dollar and Kraay, (2003); Freund and Bolaky (2008); Chang, Kaltani, and Loayza (2009). 21 REFERENCES Al-Ississ, Mohammad. 2012. “The Access to Credit Dichotomy: Arab Financial Markets Caught Between the Pull of the Disenfranchised and the Push of the Entrenched,” in Addressing the 100 Million Youth Challenge: Perspectives on Youth Employment in the Arab World in 2012, Geneva: The World Economic Forum. Alvarez, Alejandro de la Campa. 2011. “Increasing Access to Credit through Reforming Secured Transactions in the MENA Region,” World Bank Policy Research Working Paper 5613. Anderson, James E., and Douglas Marcoullier. 2002. “Insecurity and the pattern of trade: An empirical investigation,” Review of Economics and Statistics, 84(2): 342-352. Aysan, Ahmet Faruk, Mustapha Kamel Nabli, and Marie-Ange Veganzones-Vatoudakis. 2007. “Governance, Institutions, and Private Investment: An Application to the Middle East and North Africa”, in Breaking the Barriers to Higher Economic Growth: Better Governance and Deeper Reforms in the Middle East and North Africa, ed. by Mustapha K. Nabli (Washington D.C: The World Bank). Baldwin, Richard E. 1989. “Exporting the Capital Markets: Comparative Advantage and Capital Market Imperfections,” in David B. Audretsch, Leo Sleuwaegen, and Hideki Yamawaki (eds.), The Convergence of International and Domestic Markets, Amsterdam: North-Holland , pp135- 152. Beck, Thorsten. 2003. “Financial Dependence and International Trade”, Review of International Economics, 11(2), 296-316. Behar, Alberto and Caroline Freund. 2011. “The Trade Performance of the Middle East and North Africa,” Working Paper. Besley, Timothy. 1995. “Property Rights and Investment Incentives: Theory and Evidence from Ghana,” Journal of Political Economy 103(5):903-937. Bhattacharya, Rina, and Hirut Wolde. 2009. Constraints to Trade in the MENA Region. IMF Working Paper. Braun, Matias. 2003. “Financial Contractibility and Assets’ Hardness: Industrial Composition and Growth”, Mimeo. Braun, Matias, and Borja Larrain. 2005. “Finance and the Business Cycle: International, Inter- Industry Evidence”, The Journal of Finance 60(3): 1097-1128. Cebeci, Tolga, Ana M. Fernandes, Caroline Freund, and Martha Denisse Pierola. 2012. “Exporter Dynamics Database,” World Bank Policy Research Working Paper 6629. Chang, Roberto, Linda Kaltani, and Norman V. Loayza. 2009. “Openness can be good for growth: The role of policy complementarities”, Journal of Development Economics 90: 33-49. Claessens, Stijin, and Luc Laeven. 2003. “Financial Development, Property Rights, and Growth”, The Journal of Finance 58(6): 2401-2436. 22 Dasgupta, Dipak, Mustapha Kamel Nabli, Christopher Pissarides, and Aristomene Varoudakis. (2007). “Making Trade Work for Jobs – International Evidence and Lessons for MENA”, in Breaking the Barriers to Higher Economic Growth: Better Governance and Deeper Reforms in the Middle East and North Africa, ed. by Mustapha K. Nabli (Washington D.C: The World Bank). Demirguc-Kunt, Asli, and Vojislav Maksimovic. 1998. “Law, Finance, and Firm”, Journal of Finance 53(6): 2107-2137. Dollar, David, and Aart Kraay. 2003. “Institutions, trade, and growth”, Journal of Monetary Economics 50:133-162. Djankov, Simeon, Caroline Freund, Cong S. Pham. 2010. “Trading on Time”, The Review of Economics and Statistics 92(1): 133-162. Elbadawi, Ibrahim A. 2005. “Reviving Growth in the Arab World,” Economic Development and Cultural Change 53(2): 293 Fisman, Raymond, and Inessa Love. 2003a. “Trade Credit, Financial Intermediary Development, and Industry Growth”, The Journal of Finance 58(1): 353-374. Fisman, Raymond, and Inessa Love. 2003b. “Financial Dependence and Growth Revisited”, NBER Working Paper Series 9582 Fleisig, Heywood, Mehnaz Safavia, and Nuria de la Pena. 2006. “Reforming Collateral Laws to Expand Access to Finance,” The World Bank, Washington D.C. Francois, Joseph, and Miriam Manchin. 2007. “Institutions, infrastructure, and trade,” World Bank Policy Working Paper 4152. Freund, Caroline, and Alberto Portugal-Perez. 2013. “Assessing MENA’s trade agreements”, in The Arab Spring Implications for Economic Integration, edited by Michael Gasiorek. Freund, Caroline, and Bineswaree Bolaky. 2008. “Trade, regulations, and income”, Journal of Development Economics 87:309-321. Gaulier, Guillaume and Soledad Zignago. 2010, “BACI: International Trade Database at the Product-Level. The 1994-2007 Version,” CEPII Working Paper, N°2010-23, October. See more at: http://www.cepii.fr/CEPII/en/bdd_modele/presentation.asp?id=1#sthash.NLPdG5W7.dpuf Gianetti, Mariassunta. 2003. “Do Better Institutions Mitigate Agency Problems? Evidence from Corporate Finance Choices,” Journal of Financial and Quantitative Analysis 38(1): 185-212. Grossman, Sanford, and Oliver Hart. 1986. “The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration”, Journal of Political Economy 94(4): 691-719. Haber, Stephen, Armando Razo and Noel Maurer. 2003. The Politics of Property Rights: Political Instability, Credible Commitments, and Economic Growth in Mexico, 1876-1929. Cambridge, U.K.: Cambridge University Press. Hart, Oliver D. 1995. Firms, Contracts, and Financial Structure. London: Oxford University Press. 23 Hart, Oliver and John Moore. 1990. “Property Rights and the Nature of the Firm”, The Journal of Political Economy 98(6): 1119-1158. Hur, Jung, Manoj Raj, Yohanes E. Riyanto. (2006). “Finance and Trade: A Cross-Country Empirical Analysis on the Impact of Financial Development and Asset Tangibility on International Trade”, World Development 34(10): 1728-1741. Iqbal, Farrukh, and Mustapha Kamel Nabli. 2007. “Trade, Foreign Direct Investment, and Development in the Middle East and North Africa,” Chapter 12 in Breaking the Barriers to Higher Economic Growth: Better Governance and Deeper Reforms in the Middle East and North Africa, ed. by Mustapha K. Nabli (Washington: The World Bank). Jensen, Michael C., and William H. Meckling. 1976. “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure”, Journal of Financial Economics 3: 305-360. Karshenas, Massoud, 2001, “Structural Obstacles to Economic Adjustment in the MENA Region: The International Trade Aspects,” in The State and Global Change: The Political Economy of Transition in the Middle East and North Africa, ed. by Hassan Hakimian and Ziba Moshaver. Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi. 2009. “Governance Matters VIII Aggregate and Individual Governance Indicators 1996-2008”, World Bank Working Paper 4978. Kletzer, Kenneth, and Pranab Bardhan. 1987. “Credit Markets and Patterns of International Trade”, Journal of Development Economics 27: 57-70. La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert W. Vishny. 1998. “Law and Finance”, Journal of Political Economy 106(6): 1113-1155. Nabli, Mustapha Kamel, and Marie-Ange Veganzones-Vatoudakis. (2007). “Reform Complementarities and Economic Growth in the Middle East and North Africa”, in Breaking the Barriers to Higher Economic Growth: Better Governance and Deeper Reforms in the Middle East and North Africa, ed. by Mustapha K. Nabli (Washington D.C: The World Bank). Raddatz, Claudio. (2006). “Liquidity and vulnerability to financial underdevelopment”, Journal of Financial Economics 80: 677-722. Rajan, Raghuram G., and Luigi Zingales. 1998. “Financial Dependence and Growth”, The American Economic Review, Vol. 88, No. 3 (Jun., 1998), pp. 559-586 Rajan, Raghuram, and Luigi Zingales. 2001. “Financial Systems, Industrial Structure, and Growth”, Oxford Review of Economic Policy 17(4):467-482 Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi. 2004. “Institutions Rule: The Primacy of Institutions Over Geography and Integration in Economic Development,” Journal of Economic Growth 9: 131-165. Sabbagh, Karim, Chadi N. Moujaes, Leila Hoteit, Jussi Hitunen, and Hatem Sammam. 2012. “Education for Employment – Commiting Enterprises to Education Development in the Middle East and North Africa”, in Addressing the 100 Million Youth Challenge: Perspectives on Youth Employment in the Arab World in 2012, Geneva: The World Economic Forum. 24 Svaleryd, Helena, and Jonas Vlachos. 2005. “Financial markets, the pattern of industrial specialization and comparative advantage: Evidence from OECD countries,” European Economic Review 49: 113-144. Williamson, Oliver E. 1988. “Corporate Finance and Corporate Governance”, Journal of Finance, 43(3): 567-592. World Bank. 2000. “Pilot Investment Climate Assessment: Moroccan Manufacturing Sector at the Turn of the Century: Results of the Firm Analysis and Competitiveness Survey, FACS-Morocco.” Washington, D.C. --- 2003a. MENA Development Report: Better Governance for Development in the Middle East and North Africa, Enhancing Inclusiveness and Accountability, Washington DC: The World Bank. --- 2003b. MENA Development Report: Trade, Investment, and Development in the Middle East and North Africa: Engaging with the World, Washington DC: The World Bank. --- 2011. MENA Development Report: Financial Access and Stability: A Road Map for the Middle East and North Africa, Washington DC: The World Bank. --- 2013. World Bank Middle East and North Africa Economic Developments and Prospects, October 2013, Washington DC: The World Bank. World Economic Forum. 2012. Addressing the 100 Million Youth Challenge: Perspectives on Youth Employment in the Arab World in 2012, Geneva: The World Economic Forum. 25 APPENDIX FIGURES Figure 6. World Governance Indicators (1995-2009), MENA vs Non-MENA Figure 6a: Voice and Accountability 2 MLT 1 ISR WGI_voice_mean 0 LBN KWT MARJOR QAT OMNBHR ARE DJI -1 YEM EGY TUN DZA IRN IRQSYR SAU LBY -2 7 8 9 10 11 Log GDP per capita (2005,USD) Figure 6b: Corruption 2 1 ISR MLT QAT KWT ARE WGI_corr_mean OMNBHR JOR TUN SAU 0 MAR EGY DJI LBN DZA IRN YEM SYR LBY -1 IRQ -2 7 8 9 10 11 Log GDP per capita (2005,USD) 26 Figure 6c: Political Stability 2 MLT 1 OMN QAT ARE TUN KWT WGI_poli_mean 0 LBY BHR DJI MARJOR SYR SAU EGY -1 IRN LBN ISR YEM DZA -2 IRQ -3 7 8 9 10 11 Log GDP per capita (2005,USD) Figure 6d: Government Efficiency 2 ISR 1 MLT WGI_govteff_mean ARE TUN OMNBHR QAT KWT JOR 0 MAR SAU EGY LBN DZA IRN SYR DJIYEM LBY -1 IRQ -2 7 8 9 10 11 Log GDP per capita (2005,USD) 27 Figure 6e: Regulatory Quality 2 MLT 1 ISR BHR ARE WGI_regqual_mean OMN JOR QAT KWT TUN 0 MAR SAU LBN EGY DJIYEM DZA -1 SYR IRN LBY IRQ -2 7 8 9 10 11 Log GDP per capita (2005,USD) Figure 6f: Rule of Law 2 MLT 1 ISR OMN KWT ARE WGI_rulelaw_mean BHR QAT JOR SAU TUN 0 MAR EGY SYR LBN DJI IRN DZA LBY -1 YEM IRQ -2 7 8 9 10 11 Log GDP per capita (2005,USD) 28 DATA APPENDIX Table 6. BACI - MENA Country List Country code Country Name SAU Saudi Arabia LBY Libya BHR Bahrain DZA Algeria OMN Oman IRN Iran, Islamic Rep. SYR Syrian Arab Republic KWT Kuwait ISR Israel TUN Tunisia LBN Lebanon YEM Yemen, Rep. IRQ Iraq QAT Qatar JOR Jordan DJI Djibouti ARE United Arab Emirates EGY Egypt, Arab Rep. MAR Morocco MLT Malta Source: BACI Table 7. Measures of Exporter Size and Trade Specialization ( ) Description Source Sector Share of Total Exports in industry s and country c / Total BACI, EDD Exports Export Value of Goods and Services in country c. log(Exports) log(Total Exports) BACI log(Mean Exports) log(Mean Export Value per Exporter) USD 2000 EDD log(Median Exports) log(Mean Export Value per Exporter) USD 2000 EDD Source: Exporter Dynamics Database, World Development Indicators, BACI 29 Table 8. Measures of Sectoral Variation (s) Description EXTFIN External Finance Dependence (Rajan & Zingales, 1998) TAN Asset Tangibility (Braun, 2003) Table 9. Country-level factors and endowments ( ) Description Source FINANCIAL SECTOR DEVELOPMENT Domestic credit Domestic credit to private sector refers to FS.AST.PRVT.GD.ZS, World to private sector financial resources provided to the Development Indicators, (% of GDP) private sector, such as through loans, World Bank purchases of non-equity securities, and trade credits and other accounts receivable, that establish a claim for repayment. GOVERNANCE Property rights How would you rate the protection of Global Competitiveness property rights, including financial Index, World Economic assets, in your country? [1 = very weak; Forum 7 = very strong] Legal Rights How efficient is the legal framework in Index your country for private businesses in settling disputes? [1 = extremely inefficient; 7 = highly efficient] Regulatory (normalized mean zero) Quality Government (normalized mean zero) Efficiency World Governance Indicators, Political Stability (normalized mean zero) World Bank Corruption (normalized mean zero) Voice and Voice and Accountability (normalized Accountability mean zero) 30