89424 Trade and Shared Prosperity in the Caribbean Region June 2014 1 This policy note is based on the seven chapters of the Caribbean trade report, “The New Trade Environment and Shared Prosperity in the Caribbean”, prepared by the World Bank for the Caribbean Growth Forum (CGF). The Caribbean trade report is the result of collaborative efforts of a large team led by Calvin Zebaze Djiofack (LCSPE), and including Erwin H. R. Tiongson, Massimiliano Cali, Jose Daniel Reyes, Claire Honore Hollweg, Maros Ivanic, Julie Saty Lohi, Mathias Thoenig, Thierry Mayer, Sashana Whyte, Tania Valeria Diaz Bazan, Dominique Njinkeu, Charles Udomsaph, Filipe Sousa, Vanessa Uchiyama, and Liliana Foletti. Peer reviewers were Daniel Lederman, Leader Economist (LCRCE), Souleymane Coulibaly, Lead Economist (AFTP3), Cecilia M. Briceno-Garmendia, Lead Economist(LCSSD), Pavel Isa, Consultant (LCSPE), and Marcelo Olarreaga (University of Geneva). Miriam Beatriz, Patricia Holt, and Sashana Whyte provided invaluable support in all aspects of the production of this report. Additional support and comments were provided by Andrea Gallina (LCSSO), Francisco Galrao de Carneiro (LCSPR-LCC3C), Miguel Eduardo Sanchez Martin (LSPE), and Raju Singh (LCSPR-Haiti). Finally, we would like to thank J. Humberto Lopez (Director, Poverty Reduction and Economic Management, Latin America and the Caribbean Region), Auguste Tano Kouame (Sector Manager, LCSPE), Sophie Sirtaine (Country Director, LCC3C), Francisco Galrao de Carneiro (Lead Economist and Sector Leader, LCSPR-LCC3C), Andrea Gallina (CGF Coordinator) for overall supervision and guidance to the team. The Caribbean Knowledge Series is an occasional series that presents World Bank knowledge in an accessible format. It is meant to assist knowledge sharing across the region and trigger policy dialogue on topics relevant for the Caribbean. This note was prepared to support the participatory policy dialogue in the context of the Caribbean Growth Forum (CGF). The CGF is an initiative facilitated by the Compete Caribbean Program, the Inter-American Development Bank, the World Bank and the Caribbean Development Bank, with the support of the Department for Foreign Affairs, Trade and Development of Canada, the United Kingdom’s Agency for International Development, CARICOM Secretariat, the University of the West Indies, the European Union and Caribbean Export. It aims to facilitate a multi-stakeholder dialogue to identify practical solutions for the growth challenge in the Caribbean. To learn more about the CGF methodology and progress in each Caribbean country visit: http://caribgrowth.competecaribbean.org/ 2 Disclaimer: The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the Executive Directors of the International Bank for Reconstruction and Development / The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Visit the entire “Caribbean Knowledge Series” collection at: http://worldbank.org/lac Design & Concept by Room Grupo Creativo | www.room.com.do Cover Photo: Room Grupo Creativo Photo Library 3 Trade and Shared Prosperity in the Caribbean region Introduction Despite fairly respectable economic growth The purpose of this note is to provide over the decades and a degree of high trade background information on the role of trade in openness, unemployment rates remain very the unemployment and poverty reduction in the high in the Caribbean, averaging 10% for the Caribbean and, based on recent World Bank region between 2002 and 2009, and poverty analysis, to suggest areas where greater policy reduction has been slow (World Bank, 2009).1 attention could promote trade and thus reduce Over the past five years, most countries in the poverty. We begin with a profile of employment region (The Bahamas, Barbados, Belize, Jamaica, in the Caribbean, and discuss the impact of trade and Suriname) have exhibited on average double- on employment during the global financial crisis. digit or nearly double-digit unemployment rates. We also review evidence on the role of trade in Furthermore, progress toward reducing poverty has employment and development over the long term, been relatively slow and remains uneven in the region. and whether the poor in the Caribbean benefit from In the larger Caribbean countries, an estimated one- export activities. We then discuss how addressing quarter to one-third of the population lives below the constraints on exports and reducing tariff levels could poverty line (measured at US$1/day). The region’s enhance growth and reduce poverty. The conclusion three most-populous countries (excluding Cuba) summarizes the main issues. have poverty rates of 16.4% (Dominican Republic), 44.1% (Jamaica), and around 59% (Haiti). Profile of Employment in the Caribbean The Caribbean economies are dominated by sizes of the Caribbean countries make it difficult to the services sectors, which account for more support manufacturing plants that are large enough than 60 percent of employment in every country to achieve economies of scale level while serving except Haiti and St. Kitts and Nevis (Figure only the domestic economy. In addition, the small 1). The development of a services sector in the market size coupled with their long distances to other Caribbean stems from the region’s natural amenities markets imposes high trade costs on Caribbean firms that encourage the growth of tourism and further and impedes their manufacturing and agricultural enhances the importance of services. Moreover, like exports. other small developing economies, the small market Figure 1: The Services Sector Dominates Employment in the Caribbean Source: World Bank, 2013b. Note: Average shares in 2000-2008. 1 The study covered the following 15 Caribbean countries: Antigua and Barbuda, Barbados, Belize, Jamaica, St. Lucia, The Bahamas, Dominica, the Dominican republic, Grenada, Guyana, St. Kitts and Nevis, St. Vincent and the Grenadines, Suriname, Haiti and Trinidad and Tobago. 4 The importance of services to employment may capita (Figure 2). On average, industrial employment be even greater than its impact on output, since in the Caribbean is in line with the region’s level of the Caribbean service sectors (e.g. tourism) income per capita, while agriculture accounts for the tend to be more labor intensive than industry or smallest share of employment of the three sectors. agriculture. Indeed, in most Caribbean countries, However, agriculture is particularly important for poor the share of services employment is greater than what households, and remains a large employer in Haiti, would be expected given their level of income per Belize, and Dominica. Figure 2: The Share of Employment in Services is Higher than Expected, Given Income per Capita Source: Authors’ calculations using World Bank (2013a) data. Labor markets in most Caribbean countries have the region (Haiti and Guyana). The global financial improved over the past decade, with steady crisis of 2007-09 did not greatly interrupt the steady increases in the ratio of those employed to the improvement in labor markets, with the exceptions of total population and in the labor participation Jamaica, where the employment ratio has declined rate among 15-64 year olds (Figure 3- panel a sharply since 2008; the small, tourism-dependent and b). However, progress has not been sufficient economies of Barbados and The Bahamas; and to a to significantly reduce the high unemployment lesser extent the Dominican Republic. and poverty rates of the low-income countries in Figure 3: Employment to Population Ratio (%), 2000-2011 (a) Caribbean countries and comparators Source: Authors’ calculations using World Bank (2013a) data. 5 (b) Small Caribbean countries and Haiti (individual series) Source: World Bank (2013a). The Role of Trade in Employment and Development Trade plays an important role in employment in domestic market, is 17 percent, again slightly above the small, open economies of the Caribbean. the average in other developing countries. Exporting Firms involved in export activities account for 34 firms account for 55 percent of manufacturing percent of formal employment in the region, slightly employment in the Caribbean, compared with 51 above the 32 percent average for other developing percent in other developing countries.2 The data countries covered by the World Bank Enterprise for manufacturing are probably more reliable than Survey (World Bank, 2013c) (Figure 4). The share for the total, as estimates of exports from the hotel of employment that is directly or indirectly generated and restaurant sector, one of main sources of export by exports, which excludes the workers of exporting revenues for several Caribbean countries, are firms who are involved in production destined for the probably understated.3 Figure 4: Share of Exporters and Share of Exports (Direct or Indirect) in Total Employment Source: Authors’ calculation using WBES data. The contribution of trade with new growth poles a tiny share of total employment (Figure 5, panel a) to employment in the Caribbean countries is still and even of manufacturing employment (Figure 5, very modest. The data reported in the WBES (World panel b). That is in part due to the small size of these Bank, 2013c) suggest that exporters and exports sectors in the economies but also to the particularly of metals and mineral fuels, that largely dominate small contribution of their exports to employment in Caribbean trade to new growth poles, account for the Caribbean relative to other developing countries.4 3 The WBES questionnaire does not explicitly state that sales of services to foreigners should be treated as exports, so many respondents may have failed to report export revenues properly. 4 The share of employment linked to exports in overall employment in these sectors is 9.3% in the Caribbean countries and 14.3% in the other developing countries. 6 Figure 5: Importance of Exports to the New Growth Poles: Metals and Mineral Fuels (a) Shares in total employment (b) Shares in manufacturing employment Source: Authors’ calculation using World Bank (2013c) data. Contrary to other developing countries, exporters in other developing countries do, is that the quality of employment in the Caribbean the well-educated workers are relatively more exporting firms, in terms of wage levels, gender abundant in the Caribbean. In Caribbean firms for participation, and skill of workers, does not which data is available (only one-third of the entire seem to be different to that of non-exporting sample), 71 percent of workers have secondary firms.5 Econometric analysis shows that, after school education, compared to 66 percent in other controlling for industry, foreign ownership, firm size, developing countries6. and age, as well as sector and country effects, the difference between wages paid by exporters The differences between the share of women and non-exporters is not statistically significant. By employees in the exporting and non-exporting contrast, the same estimation procedure finds that firms in the Caribbean is not statistically exporting firms in other developing countries pay, on significant.7 However, the share of women in the average, 17 percent more than do non-exporters. This workforce is 4.5 percentage points higher in exporters difference between Caribbean and other developing of manufactures than in other manufacturing firms. countries holds for the manufacturing sector as well. This is a large difference, considering that the This difference seems to be driven at least in part average female share of employment in the sample by quality certifications and investment in training. of Caribbean manufacturing firms in the WBES is 35 The results suggest that these investments yield a percent. considerably higher effect on wages in the rest of the world than in the Caribbean. As exporters tend to be Exporters to the new growth poles also do not more highly represented than non-exporters in areas appear to provide a higher quality of employment like industry, foreign ownership, and firm sizes, wage than other firms. That is, econometric analysis premiums drop significantly once these investments suggests that their wage levels, the share of are controlled for. employees with a secondary school education, and the share of female employees, are not significantly Similarly, it does not appear that the share of different from other firms. skilled workers (those with secondary school In short, the contribution of trade to Caribbean education) is significantly larger in Caribbean development is through increasing employment, exporting than non-exporting firms, while improving efficiency, and providing opportunities exporting firms in other developing countries for greater economic activity. In contrast to many employ 2.8 percent more secondary-educated other developing economies, export activity does workers than non-exporters. One reason why not make a significantly greater contribution to Caribbean exporters do not tend to employ more development than production for the domestic market educated workers than non-exporters, while in the Caribbean. 5 See, for example, Isgut (2001) and Kaplan (2006) for wage premium in exporting sectors. See Schank, Shnebel and Wagner (2004) for a contrary view based on data for Germany. On exporting and the share of skilled workers, see Iacovone and Javorcick (2013) and Verhoogen (2008). On exporting and gender employment, see Korinek (2005). 6 It is worth noting however that the availability of skilled labor varies significantly within the Caribbean. 7 Boosting the employment of women is important because women are more economically excluded than men on virtually every global measure (World Bank, 2013a), and there is a strong, positive correlation across countries and time between the relative position of women in society and the level of economic development (Duflo, 2012). 7 The Impact of Trade on Employment during the Global Financial Crisis Exporting firms generally had lower rates of before the 2009 survey. Measuring employment employment growth during the financial crisis only in firms that survived the crisis may overstate than non-exporters. The difference in some cases the overall growth rate in employment. In addition, was substantial. In Trinidad and Tobago, employment it is likely that firms’ exit during the global crisis was growth among exporters was 4.7 percentage points more common among exporters than among firms lower than among non-exporters (Table 1). Moreover, producing for the domestic market. Thus the results these calculations are only based on reports from may understate the difference between the growth those firms that existed in 2007 and survived until at rates of employment in firms serving the domestic least 2009, and thus do not include firms that folded market versus in exporters.8 Table 1: Employment Growth in Caribbean Countries, 2007-09 (Annual geometric growth rate) Employment (Geometric) Growth Rate 2007-2009 [SD Outliers Dropped] Source: Authors’ calculation using World Bank (2013c) data. Observations with values greater (less) than three times the standard deviation plus (minus) the mean of their respective group defined by country, sector, and size are dropped. The impact of the crisis on employment was the growth rate of employment was 13 percentage diverse across sectors, with the manufacturing points lower than for non-exporters. The growth rate sector showing stronger resilience than tourism of employment in the hotel and restaurant sector was sector (hotel and restaurant). During the crisis, also negatively, and significantly, related to the share the difference between employment growth among of exports in total sales. The impact of the crisis also Caribbean direct exporters of manufactures and differed significantly among countries. For example, that of non-exporting manufacturing firms was not in Jamaica the tourism sector demonstrated significantly different from zero. By contrast, the considerable resilience, while tourism revenues were growth rate of employment among exporters in the hard hit in most other Caribbean economies. hotel and restaurant businesses was significantly worse during the crisis than among non-exporters. This was particularly true for direct exporters, where 8 There also may be spillover effects from exporters to non-exporters (and vice-versa), as exporters may generate demand for goods and services produced by non-exporters. So if exporting firms failed more frequently than non-exporters, these comparisons would fail to capture the full extent to which the export channel reduced Caribbean employment during the crisis. 8 Do the Poor Benefit from Exporting? On the face of it, poor Caribbean economies richer households. Moreover, in every country, poor benefit significantly from employment in workers account for a larger share of employment in exporting firms. In all countries with adequate data, exporting firms than in non-exporting firms. The fact the poorest 40 percent of households have a higher that the share of educated workers does not differ share of their workers employed in exporting sectors greatly between exporting and non-exporting sectors (Figure 6). In Jamaica, for example, about a third of (see above) is a further indication of the participation workers from the poorest 40 percent are employed in in exporting sectors by the poor. exporting sectors, compared to only about a fifth in Figure 6: The distribution of Employment by Exporting and Non-Exporting Sectors Dominican Republic Jamaica Haiti Source: Household surveys and World Bank staff calculations. However, in some cases the job opportunities average. In the case of Jamaica, for example, the created by trade may be less important than exporting sector employs two-thirds of its workers suggested by these numbers (Figure 7). That from the poorest 40 percent of households, but the is because many of the poor are employed in metal sector, representing a significant share of all exporting sectors (e.g., agriculture) that account for trade (35 percent or 46 percent, if the mining of metal a much smaller share of exports than other exporting ores is included) does not employ workers from poor sectors (e.g., manufacturing). The weighted average households at all. As a consequence, weighted by of employment opportunities created by trade— each industry’s share of all merchandise exports, that is, the poorest 40 percent’s average share of poor workers account for only a fifth of export employment, weighted by a sector’s share of total employment. trade—may differ substantially from the non-weighted 9 Figure 7: The Distribution of Employment by Income Group: Main Exporting Industries (In percent of total employment by industry) Dominican Republic Haiti Jamaica Source: Household surveys and World Bank staff calculations. Meanwhile, employment opportunities provided average, the earnings of the poorest 40 percent by trade with the new growth poles have been are not significantly different between exporting limited, reflecting the concentration of trade in and non-exporting sectors, though there are some a few industries that employ few poor workers. exceptions. The resulting disparities in income or Many of the industries that trade with the new growth consumption between poorer households and richer poles employ mostly skilled workers from richer households are also similar across sectors. In the households. In part, this has been due to the capital- Dominican Republic where relevant data exist, the intensive production methods in the sectors, such average earnings of workers from among the poorest as mining, that account for a large share of exports 40 percent of households are about the same in to the new growth poles, while more labor-intensive exporting and non-exporting sectors alike.9 sectors that employ more poor workers account for a small share of this trade. However, the Dominican According to the broad literature on the impact Republic has been an important exception due to of changes in earnings on poverty, there are the large share of poor workers in mining, which several reasons why the income gains from dominates exports to the new growth poles. employment in the exporting sector may be insufficient to significantly raise the incomes In addition, the income gains from employment of the poor in the Caribbean. First, poor workers in the exporting sectors appear limited. On may earn low wages, reflecting low levels of human 9 By construction, of course, workers from the bottom 40 percent of the distribution have much lower per capita consumption than workers from all other households. Consumption levels, however, reflect all income sources available to a household, including both private and public transfers. Thus, it may still be useful to investigate the relative income gains from employment in exporting sectors. 10 capital. Second, poor workers may be paid adequate of schooling, compared to 12 years in the richest unit wages but are not able to work enough hours. households. They may work only part-time during the week or hold temporary contracts during the year, leading In addition, in sectors where average years of to seasonal unemployment. Third, a large number education are more or less the same among of dependents (e.g. household members in school workers, workers from the poorest households or retired, or working-age household members that are employed in informal jobs or working on their are unemployed or out of the labor force) may drive own account, and thus unable to benefit from down the income of poor households. economies of scale and unable to participate in value chains. These workers are therefore likely to Indeed, the data generally suggest low levels be less productive, on average, than richer workers, of human capital among the poor in Caribbean despite possessing the same level of human capital. countries, constraining their ability to be more The evidence in Jamaica suggests that indeed, the productive. The disparities are very large in some majority of principal earners among the poorest 40 countries. In Haiti, for example, the principal earner percent of households are working as self-employed among the poorest households has only 1 year of workers. schooling, on average, compared to 9 years of schooling in the richest households. In the case of the Dominican Republic, the principal earners in the poorest households average about 5 years Trade reforms and shared prosperity: CGE-micro simulation To assess the effect of alleviating trade constraints category of scenario is considered where the on poverty, we relied on a sequential CGE micro- Caribbean countries reinforce their specialization on simulation approach that allowed us to capture tourism as their traditional domestic market. not only the macroeconomic effects of the reforms but also the corresponding poverty and inequality Results for the Dominican Republic show that implications. The model used is a multi-country changes in tourism and improvements in innovation static model relying on the Global Trade Analysis would bring proportionally more benefits for urban project (GTAP) framework (Hertel, 1997), using households, while alleviating constraints on access Social accounting Matrices of 2007, and including to key services would bring a faster decrease in all fifteen Caribbean countries of interest for this extreme poverty (table 2). Even though reforms to report (the focus countries): Antigua and Barbuda, reduce trade costs and improve access to emerging Bahamas, Barbados, Belize, Dominica, Grenada, markets have a much lower impact, they are still Guyana, Haiti, Jamaica, Saint Lucia, St. Kitts and poverty reducing. Extreme poverty changes little Nevis, St. Vincent and the Grenadines, Suriname, when expanding access to emerging markets. As Trinidad and Tobago and the Dominican Republic. underlined previously, trade to new growth pole relies The micro-simulation module, which uses household on metals and minerals fuels that are very capital- income and expenditures surveys to evaluate the intensive and use little labor. A reduction in trade implications of CGE results for households, focuses costs would have a larger impact on extreme rural on Haiti, the Dominican Republic, and Jamaica. poverty compared to urban. Tariff reforms (adoption of the CAROCIM CET) are associated with a slight Three sets of scenarios are considered that rise in poverty, as tariff levels are already relatively correspond to three areas of policy interventions: low in the DR compared to the CET. trade policy measures, consisting of deeper regional integration among Caribbean countries, and better market access to new growth pole countries; trade facilitation measures consisting of reductions in trade costs; and measures to eliminate supply- side constraints, such as enhancing productivity through improvements in innovation, and reduction in the infrastructures gap. As a benchmark, a fourth 11 Table 2: Dominican Republic. Percentage change in poverty by urban rural divide Source: Authors’ construction. In the case of Haiti, our model shows that easing to innovation would reduce poverty reduction by the access to innovation would reduce poverty reduction greatest amount, followed by reforms to reduce trade by the greatest amount, followed by reforms to reduce costs (see table 3), reflecting the gap between Haiti trade costs (see table 3), reflecting the gap between and the best Caribbean performers in these areas Haiti and the best Caribbean performers in these (see Policy Note 1). The poorest households living in areas (see Policy Note 1). The poorest households urban areas would gain the most if these reforms are living in urban areas would gain the most if these undertaken. Extreme urban poverty falls almost 34 reforms are undertaken. Extreme urban poverty falls per cent in the innovation scenario and 21 per cent in almost 34 per cent in the innovation scenario and the trade cost scenario. Tourism has the third largest 21 per cent in the trade cost scenario. Tourism has impact on extreme urban poverty reduction. Rural the third largest impact on extreme urban poverty poverty also falls in all scenarios except the scenario reduction. Rural poverty also falls in all scenarios where access to emerging markets is improved, with except the scenario where access to emerging extreme poverty declining faster than overall poverty. markets is improved, with extreme poverty declining An increase in tourism would be poverty reducing, faster than overall poverty. An increase in tourism even though most of the gains accrue to urban areas. would be poverty reducing, even though most of the Tariff reduction will also benefit the urban poor most, gains accrue to urban areas. Tariff reduction will also reflecting their higher dependence on imported food benefit the urban poor most, reflecting their higher products. dependence on imported food products. In the case of Haiti, our model shows that easing access Table 3: Haiti. Changes in poverty by urban rural divide Source: Authors’ construction. 12 Conclusion: Conclusion: Addressing Constraints on Exports and Promoting Broad-based Benefits of Trade Our analysis shows that international trade plays 40 percent of households. In the case of Jamaica, a major role in terms of job creation and poverty this requires strengthening the quality of education. reduction in the Caribbean, more so, on average, than in the other developing countries. That is 2. Strengthening links to the value chain particularly true for manufacturing, where over one among small enterprises: Where workers from in four jobs is related to exports and where exporters poorer households are unable to take advantage of account for more than half of formal employment. scale and network effects, addressing the constraints may require a deeper understanding of the complex However, in general, poor Caribbean households dimensions of the business environment. Part of have not benefited fully from the employment the problem could be a burdensome regulatory opportunities created by trade. Although poor environment that encourages informality or that limits households in most Caribbean countries have scale. Another part of the problem could be the some access to employment opportunities in the existence of a large number of relatively unproductive exporting sectors, it appears that workers from the micro firms that are unable to participate fully in the poorest households generally do not earn enough value chain. In some economies, there are emerging from their jobs in exporting sectors. In some cases, models for aggregating micro firms or for creating differences in earnings and productivity are driven by networks of micro firms to enable smaller firms to commensurate disparities in levels of human capital. take advantage of scale. In a few cases, however, despite having comparable levels of human capital, workers from the poorest 3. Addressing key impediments to trade 40 percent of households earn significantly less performances: our analysis shows that acting than other households, for at least two possible to remove some key trade impediments may reasons: First, there is some qualitative evidence also directly help the poor: i) Lowering the cost to suggest that though some workers have similar of transport and strengthening trade services could numbers of years of education on paper, the quality reduce food costs and thus benefit the poorest of education actually received may not be the same. consumers, particularly in Haiti, the poorest country Second, where both the quality and quantity of of the sub-region which depends heavily on food education received by households are comparable, imports; ii) Improving access to transport and poorer household seem unable to participate fully in infrastructure services, as well as devoting more productive activities boosted by scale and network resources to innovation (e.g. investment in research effects. Instead many of them are self-employed, and development) would boost productivity and possibly in informal enterprises. thus reduce prices on both domestic production and imports, which also would reduce poverty; There is a role for policy in alleviating poverty iii) Strengthening regional integration through the by helping promote the shared benefits of trade. implementation of the CARICOM common external Considering the variety of issues involved in this tariff (CET) would reduce poverty and inequality in area, it will likely require a multi-pronged approach most countries in the region, by lowering the cost involving the following measures: of imported inputs, increasing economies of scale. However, the CET would not significantly reduce 1. Promoting quality education for all: poverty in countries such as the Dominican Republic, Where human capital is an important constraint, where tariff schedules are already lower than the the requirements are consistent with an overall CET for key items. development strategy that promotes quality education for all. In the case of the Dominican Republic, for example, this requires promoting access to education, particularly among the poorest 13 References Duflo, E. (2012). Women Empowerment and Economic Development. Journal of Economic Literature 50 (4): 1051–1079. Hertel, T. (1997). Global Trade Analysis: Modeling and Applications. Cambridge University Press. Hollweg, C. H. & Reyes, J.D. (forthcoming). Recent Export Dynamics of Merchandise and Services Trade in the Caribbean. World Bank. Mimeo. Iacovone, L. & Javorcick B.S. (2012). Getting ready: Preparation for exporting, CEPR Discussion Paper No. DP8926. Isgut, A.E. (2001). What’s Different About Exporters: Evidence from Columbian Manufacturing. Journal of Development Studies. Vol. 37(5):57-82. Kaplan, D. S. & Verhoogen, J. (2006). Exporting and Individual Wage Premia: Evidence from Mexican Employer-Employee Data. Columbia University. http://www.columbia.edu/~ev2124/research/KaplanVerhoogenJan06.pdf Korinek, J. (2005). Trade and Gender: Issues and Interactions. OECD Trade Policy Working Paper No. 24. http://www.oecd.org/trade Schank, T., Schnebel, C., & Wagner, J. (2004). Exporting Firms do not Pay Higher Wages, Ceteris Paribus. First Evidence from Linked Employer-Employee Data. IZA Discussion Paper No. 1185. Institute for the Study of Labor. Bonn. Verhoogen, E.A. (2008). Trade, quality upgrading, and wage inequality in the Mexican manufacturing sector, Quarterly Journal of Economics 123 (2): 489-530.doi: 10.1162/qjec.2008.123.2.489 World Bank. (2013a). Gender at work, Washington DC: World Bank publications. World Bank. (2013b). World Development Indicators, Washington D.C. World Bank publications. World Bank. (2013c). World Bank Enterprise Survey. http://data.worldbank.org/data-catalog/enterprise-surveys 14 15 worldbank.org/lac 16