108075 Tr S ea a pe n d ci d ResearchDigest al Gl Is ob su a World Bank e liz on at io n VOLUME 11 NUMBER 1 FALL 2016 Insights from World Bank Research on Trade and Globalization S uccessful international integra- opportunities but also the challenge IN THIS ISSUE tion, supported by sound nation- of meaningful participation. A paper al policy and effective interna- using firm-level data shows the rise in Insights from World Bank Research tional cooperation, has underpinned domestic content of China’s exports on Trade and Globalization … page 1 most experiences of rapid growth, and attributes it in part to the coun- shared prosperity, and reduced pov- try’s investment liberalization, which What Drives Differences in Exporter erty. Today, country-level trade in final led to a greater variety of domestic Behavior across Countries? … page 2 goods and conventional trade poli- materials becoming available at lower cies such as tariffs are reasonably well prices. Another paper shows that own The Rising Domestic Content of Chinese Exports … page 3 understood, but important questions trade liberalization can boost export remain: How do firms and workers performance. It finds that exporting to The Quality Effects of Exporting engage in trade, and how does trade richer countries led firms in Portugal to to High-Income Countries … page 4 affect productivity and poverty? How raise the average quality of the goods can national trade-related policies en- they produce by purchasing higher- Labor Market Frictions and Trade Adjustment hance growth and reduce poverty? And quality inputs. Costs in Developing Countries … page 5 how can international cooperation The labor market effect of trade help in achieving these objectives? reform depends on how easily work- How Product Standards Affect These are the questions driving the ers can move from sectors exposed Firms’ Export Decisions … page 6 World Bank’s research on trade and to greater competition to those with globalization, including the research greater opportunities. A recent paper How Services Reform Helped Revive papers summarized in this issue. proposes a method to estimate labor Indian Manufacturing … page 7 Understanding trade patterns and their mobility costs using readily obtainable impact. The focus of this research is data. Its estimates of labor mobility Are the Benefits of Export Support Durable? … page 8 increasingly on the microeconomic costs worldwide can be used to explore aspects of international integration the effect of trade shocks on employ- Resource Rents and Local Development as well as the implications of the frag- ment and wages. in Post-Apartheid South Africa … page 9 menting of global production. The re- Designing national policy reform. The search group’s new Exporter Dynamics focus of trade policy reform has Are Politically Connected Firms More Database, drawing together customs shifted from goods tariffs to nontariff Likely to Evade Import Tariffs and transaction-level data for more than 70 measures such as technical regula- Other Taxes? … page 10 countries, is helping us to understand tions, to services trade policies, and the elements of trade success. Using to the devising of cost-effective strate- Trade Agreements under these data, a recent paper finds that gies for trade facilitation and export Threat … page 11 developing countries lack the large, promotion. dominant firms that boost the exports Trade-affecting technical regula- Recent World Bank Research of richer countries—and also see more tions have become visible like rocks Publications and Papers on Trade exporter churning, with higher entry in the ebbing tide of tariffs. A recent and Globalization … page 12 and exit rates. Export success comes paper assesses their effects by combin- in an environment that encourages not ing data on all exporting firms in 42 just small and medium-size firms but developing countries with new data on also large ones. pesticide standards for 243 agricultural The emergence of global val- ue chains is creating new trade (continued on page 9) 2 World Bank ResearchDigest What Drives Differences in Exporter Behavior across Countries? Analysis using a new database finds direct link between firm size and pro- countries develop, gross rates of entry that both a country’s size and its ductivity. A stark prediction from this into and exit from exporting decrease stage of development matter for the model is that larger countries should while survival rates of entrants into export more because they have more export markets increase. These mea- behavior of exporters firms rather than because they have sures of exporter dynamics are not cor- C larger firms. related with country size. ountries do not export; firms do. To understand how variation in al- These novel stylized facts suggest Yet while trade at the country locative efficiency across countries that as countries develop and exports level is well understood, knowl- affects exporter behavior, the authors grow, the export expansion happens edge of trade at the firm level has consider the literature on efficiency through growth in both the number been incomplete. What determines gains from within-sector resource real- and size of exporting firms, and as which firms participate in export- location across firms, which attributes more resources flow to the largest ing—and which of these survive and variation in firm size to distortions in firms. Consistent with the standard grow? The new Exporter Dynamics this resource allocation. One implica- model of trade with heterogeneous Database has helped address such tion from this literature is that econo- firms, the authors find that larger questions through novel indicators on mies with fewer distortions (more de- countries export more in large part the micro structure of the export sec- veloped economies) should have more because they have more exporters. tor. Drawing on national customs data exporters and higher survival rates of Consistent with resource allocation at the exporter level for more than 70 entrants because the most produc- improving with stage of development, countries, the database provides the tive firms are able to grow and export. they find that more developed coun- number and average size of exporters, But implications for average exporter tries have both more exporters and exporter concentration, and rates of size and exporter concentration at the more resilient exporters. Overall, the entry, exit, and entrant survival among top of the size distribution depend findings that both the average size exporters for each country at different on which firms are most constrained. of exporters and the concentration levels of disaggregation. If only the most productive firms are of exports at the top of the firm size The data show great variation in able to overcome regulatory hurdles in distribution increase with a country’s exporter characteristics and dynam- economies with allocative inefficien- stage of development are consistent ics across countries. Chile has a third cies while mid-productivity firms are with models where firm growth is con- more exporters than Morocco, and its held back—so that the size distribu- strained in developing countries, espe- exporters are three times as large on tion has a “missing middle”—exporters cially among high-productivity firms— average. Cameroon and Malawi have should be relatively larger in develop- the situation with a “truncated top” of high entry rates into exporting, but ing countries and very concentrated at the size distribution. A key policy im- only 25 percent of entrants survive af- the top of the distribution. In contrast, plication is that developing countries ter the first year. if high-productivity firms are the most export less because their exporters are In a recent analysis using the da- constrained and fail to invest, export- both smaller and less dominant on tabase indicators for 38 developing ers should be relatively small in more average. and 7 developed countries, Fernandes, distorted economies and less concen- Freund, and Pierola identify novel trated at the top of the size distribu- stylized facts on how exporter behav- tion—so that the distribution has a ior varies with country size and stage “truncated top.” of development. They then examine The findings show that export sec- whether these facts are consistent with tor characteristics are systematically predictions from trade theory with correlated with both country size and heterogeneous firms and from models stage of development. Larger countries with allocative inefficiencies. and more developed ones have more Systematic variation in export sec- exporters, larger exporters on average, tor characteristics across countries of and a greater concentration of exports different sizes or income levels could among the top 5 percent of exporting result from differences in resource firms. Greater numbers of firms explain endowments or in the process of about two-thirds of the additional resource allocation. To explore how export value of larger countries, while Ana Fernandes, Caroline Freund, and Martha differences in resource endowments larger firm size explains the remain- Denisse Pierola. 2016. “Exporter Behavior, Coun- affect exporter behavior, the authors ing third. In contrast, larger firm size is try Size and Stage of Development: Evidence from rely on a standard model of trade with relatively more important in explaining the Exporter Dynamics Database.” Journal of heterogeneous firms that draws a why richer countries export more. As Development Economics 119: 121–37. World Bank Research Digest 3 The Rising Domestic Content of Chinese Exports The domestic content of exports has China’s exports has been rising, show- firms with different levels of domestic been falling in most countries—but ing that it increased from 65 percent content—cannot explain the upward rising in China. What has caused in 2000 to 70 percent in 2007. China’s trend during the sample period. bilateral exports to major trading part- The authors also find that the sub- China to defy the trend? ners showed increases in domestic stitution of domestic for imported O content of a similar magnitude. These materials was induced by the country’s ver the past two decades, increases in domestic content are sta- liberalization of trade and foreign di- thanks to the growing fragmen- tistically significant. rect investment (FDI) since the early tation of global production, The authors’ results confirm the 2000s. To guide their empirical analy- exporting firms have come to rely upward trend found in an earlier study sis, they build a model featuring a less on domestic inputs for produc- using an approach based on input- translog cost function. This permits tion. Indeed, the domestic content of output tables (Robert Koopman, Zhi estimation of the time-varying elastic- exports has been declining in most Wang, and Shang-Jin Wei, “Estimating ity of substitution between domestic countries. China is an exception, de- Domestic Content in Exports When and foreign input varieties to study spite its deep engagement in global Processing Trade Is Pervasive,” Journal how different government policies may value chains. What has caused China of Development Economics 99 [2012]: affect a country’s domestic content. to defy the trend? 178–89). But their estimate is signifi- Results show that for China, increasing Kee and Tang explore this ques- cantly higher. This difference highlights FDI and declining input tariffs have led tion in a recent paper. Using customs how ignoring firm heterogeneity may to a greater variety of domestic materi- transaction-level data merged with lead to downward aggregation bias in als becoming available at lower prices manufacturing firm survey data, the the approach based on input-output during the sample period. For the authors measure and analyze the tables. entire processing sector and for most rising domestic content of China’s Samples used to construct input- industries within that sector, imported exports (measured as the ratio of do- output tables often consist mainly of and domestic materials are gross sub- mestic value added in exports to gross large firms. Because large firms tend stitutes, with the estimated elasticity exports). Their transaction-level data to have lower domestic content as a of substitution ranging between 1.9 cover the universe of Chinese export- result of their high import-to-sales and 6.6. These large elasticities explain ers during the period 2000–07, allow- ratios, oversampling large firms in the why lower prices of domestic materials ing them to construct firm, industry, construction of input-output tables can result in such significant increases and aggregate measures of domestic can lead to lower aggregate estimates in domestic content at the firm and content over time to study their evolu- of domestic content. To illustrate, the thus the aggregate level in China. tion. They then use the merged data authors conduct a decomposition ex- The results confirm existing find- sets to examine whether changes in ercise. They show that they can lower ings that the reduction in input tariffs export composition, firms’ produc- their estimate of domestic content to and greater presence of FDI in down- tion costs, or material shares are re- a level that does not differ statistically stream sectors can lead to an expan- sponsible for China’s rising domestic from that in the earlier study just by sion of domestic product variety. They content. using a sample that includes only the also show that the rising domestic The burgeoning literature on mea- large firms satisfying the sample selec- content of China’s exports is due to the suring domestic content at the indus- tion criteria applied in constructing substitution of domestic for imported try and aggregate levels relies on in- the Chinese input-output tables. This materials—indicating that the country put-output tables. While this method suggests that aggregation bias driven is relying less on imports and becom- has the advantage of capturing by firm heterogeneity alone is suffi- ing more competitive in intermediate input-output linkages within and cient to explain the wedge between the input sectors. This suggests that China across countries, the presence of firm estimates. has been moving up the value chains. heterogeneity may result in significant What has caused the rise in the Given its sheer size, this may have aggregation biases in estimates of domestic content of China’s exports? important implications for world trade domestic content. The authors use a The authors’ firm-level regressions re- and the global economy. ground-up approach that embraces veal that it has been driven mainly by firm heterogeneity by measuring do- individual processing exporters sub- mestic content at the industry and stituting domestic for imported ma- aggregate levels as the weighted aver- terials. Other factors—such as rising ages of the domestic content for the production costs due to higher wages, Hiau Looi Kee and Heiwai Tang. 2015. “Do- underlying firms. a changing composition of Chinese mestic Value Added in Exports: Theory and Firm Their estimates confirm existing exports toward industries with high Evidence from China.” American Economic findings that the domestic content of domestic content, or a churning of Review 106 (6): 1402–36. 4 World Bank ResearchDigest The Quality Effects of Exporting to High-Income Countries Increasing exports to richer markets Hungary, Portugal, and the United and economically significant relation- may require quality upgrading by States reveals that firms charge higher ship between average destination entire complexes of suppliers and prices for goods sold to richer destina- income and input prices within firms. tion markets within narrow product Alternative candidates for explain- downstream producers categories, controlling for other desti- ing the effects of exporting on firms’ A nation characteristics. This cross- behavior are difficult to reconcile with growing body of research sug- sectional evidence is not definitive, the observed patterns. gests that exporting has impor- however, for two reasons. First, firms Overall, the findings support the tant effects on firms’ behavior. may charge higher markups in richer hypothesis that firms choose to sell Although results for residual-based countries, even for homogeneous higher-quality products in richer coun- measures of productivity are mixed, goods. Second, the cross-sectional tries, that doing so requires purchas- recent studies have found causal evidence does not settle the issue of ing higher-quality inputs, and that this effects of exporting on a variety of causality: even if export prices do re- mechanism is part of the explanation directly observable outcomes, includ- flect product quality, shocks at the firm for the effects of exporting on firms’ ing technology investments, mea- level may affect both which products behavior that have been documented sures of product quality, and working a firm chooses to sell and where it is by a number of studies. Although conditions. able to sell them, leading to a positive product quality is not directly observed What explains these impacts? correlation between price and desti- and caution needs to be exercised in Perhaps the most common theoretical nation income even in the absence interpreting the results, the empirical explanation emphasizes scale effects: of a causal effect of exporting on firm patterns documented in the authors’ in the presence of fixed investment behavior. analysis add to the accumulation of costs—for example, for purchases of In a recent paper Bastos, Silva, evidence in the literature that quality technology or the screening of work- and Verhoogen develop an approach choice for both outputs and inputs is ers—increases in sales volume due to to estimating the importance of the an important component of firms’ be- exports reduce the fixed costs per unit income-based quality choice chan- havior in the international economy. and tend to induce firms to undertake nel—the idea that firms sell higher- While the analysis focuses on such investments. In this explanation, quality products to richer consumers Portugal, a high-income economy, therefore, the effects of exporting on and that doing so requires purchasing these findings have implications for firms’ behavior would depend on the higher-quality inputs—in shaping our understanding of the upgrading volume of exports, and not on the firms’ behavior in the international process in developing countries. In characteristics of particular export economy. Direct measures of product particular, the results reinforce the destinations. quality are not available, and follow- idea that increasing exports to high- A separate explanation focuses on ing a growing literature they seek to income destinations may require qual- quality choice: the varieties that firms draw inferences about product quality ity upgrading by entire complexes of sell on export markets may differ from from information about prices. But suppliers and downstream producers, those that they sell on domestic mar- such inferences are confounded by not just by particular exporters. The kets, and the different varieties may the well-known fact that prices may empirical setting has the advantage of require different technologies, skills, reflect markups as well as product making it possible to cleanly identify and other inputs in production. If rich- quality. Their proposed solution to a causal relationship between average er consumers are more willing to pay this problem is to focus on how the destination income and material input for product quality, firms may choose input prices paid by a firm respond prices, but the basic findings seem to sell higher-quality varieties in richer to exogenous variation in the income likely to apply more broadly. countries to appeal to them. level of the firm’s export destinations. Empirically, the relative importance While output prices may clearly reflect of these different explanations remains various forms of pricing-to-market, an open question. Plant-level data sets input prices arguably do not. typically do not provide information Using this insight, and detailed on the destination of exports, which customs and firm-product-level data makes it difficult to distinguish among from Portugal, the authors examine the different channels. Newly avail- the effects of firm-specific real ex- able customs data sets on firms’ inter- change rate shocks leading to variation national transactions have provided in exports to different destinations on Paulo Bastos, Joana Silva, and Eric Verhoogen. some support for the income-based the input prices paid by manufacturing 2014. “Export Destinations and Input Prices.” quality choice mechanism. Cross- firms. The results indicate that there Policy Research Working Paper 6914, World sectional evidence for China, France, is a positive, robust, and statistically Bank, Washington, DC. World Bank Research Digest 5 Labor Market Frictions and Trade Adjustment Costs in Developing Countries A global map of labor mobility costs world and use these estimates to ex- America, 2.6 in Western Europe, and sheds light on how long it might take plore labor market responses to trade 2.2 in North America). for different countries to fully adjust shocks. The authors While negatively cor- to trade shocks use the United Nations It typically takes six related with per capita Industrial Development years for a country GDP, labor mobility costs C Organization (UNIDO) are positively correlated hanges in trade policies—such database, which provides to fully adjust to a with poverty rates. They as those relating to quotas, information on labor al- tariffs, sanitary measures, anti- locations and wages in trade shock, and are also inversely cor- related with tertiary dumping laws, or customs agree- manufacturing, to esti- the higher the educational attainment ments—affect market conditions mate a map of the labor significantly and induce responses mobility costs for 25 de- mobility costs are, and schooling quality, but are uncorrelated with in wages and employment. But inef- veloped and 31 develop- the longer this primary and secondary ficiencies in labor markets impede the ing countries. enrollment. And they are process through which workers and The authors find that transition takes positively correlated with firms adjust to new market conditions. the costs of labor mo- other frictions, distor- Moving costs, firing and hiring costs, bility are large, where the costs are tions, and constraints in the economy. and differences in sector-specific skills defined as the lifetime welfare loss Most importantly, the size of the labor can create frictions and contribute faced by a representative worker who mobility costs matters. It typically to labor market inefficiencies. One is forced out of a specific industry. On takes six years for a country to fully consequence of such frictions is that average, the labor mobility costs in adjust to a trade shock, and the higher trade shocks induce only gradual wage developing countries are equivalent the mobility costs are, the longer this and employment responses. to 3.7 times the annual wage in manu- transition takes. The inability to adjust Nevertheless, relatively little is facturing. In developed countries the quickly is costly. known about how sluggish labor mar- mobility costs are 2.7 times the an- The authors estimate trade adjust- ket adjustment affects economic well- nual wage—much lower, as expected. ment costs, which vary widely across being. The assessment of these labor Indeed, there is a negative correlation countries. On average, the costs of market responses requires estimates between per capita GDP and mobility adjustment to a trade shock in the of the costs of labor mobility, but costs (figure 1). Adjustment costs vary food sector can be as high as the ac- these estimates are seldom available significantly across regions (4.0 times tual (static) gains from trade, while the in developing countries. In a recent the annual wage in Eastern Europe, costs of adjustment to a trade shock paper Artuc, Lederman, and Porto 4.0 in Sub-Saharan Africa, 3.9 in South in the textile sector can be equivalent construct a map of estimates of labor Asia, 3.6 in the Middle East and North to about 60 percent of the gains from mobility costs across the developing Africa, 3.5 in East Asia, 3.2 in Latin trade. These results confirm the im- portance of incorporating labor market imperfections into the assessment of Figure 1. Labor Mobility Costs and Per Capita GDP trade policy and trade shocks. Labor mobility costs as a multiple of average annual local wage 6 5 4 3 2 1 0 Erhan Artuc, Daniel Lederman, and Guido Porto. 0 10,000 20,000 30,000 40,000 50,000 2015. “A Mapping of Labor Mobility Costs in the Per capita GDP (US$) Developing World.” Journal of International Source: Calculations based on UNIDO data. Economics 95 (1): 28–41. 6 World Bank ResearchDigest How Product Standards Affect Firms’ Export Decisions Firms are less likely to export to and obtaining certifications) and vari- These findings are consistent with markets where product standards able ones (such as from inspection recent trade models with heteroge- are more stringent than those they procedures). For firms in developing neous firms that predict that only the countries these additional costs can most productive firms are able to over- face at home be high enough so that SPS standards come the fixed costs of exporting. The E act as barriers to entry. need to obtain information on foreign vidence shows that the substan- Knowing how such standards af- standards and adjust production pro- tial decline in tariffs over recent fect firms’ export decisions and export cesses to comply with those standards decades has fostered growth in success is critical for both economists increases the fixed costs of export- world trade. Less well understood are and policy makers. In a new paper us- ing. This helps explain why relatively the consequences of the growing use ing firm-level data, Fernandes, Ferro, stringent standards have a stronger of nontariff measures by both devel- and Wilson provide econometric effect on the entry and exit decisions oped and developing countries. The evidence on the effect of SPS stan- of smaller exporters. These findings successful conclusion of the Trans- dards on trade. The authors focus on have clear implications for developing Pacific Partnership negotiations has pesticide standards set by exporting countries trying to reduce poverty by revived interest in the role of nontariff and importing countries on food and expanding agricultural trade. measures, particularly the regulatory agricultural products, determining Differences in SPS standards across standards on food and agricultural the maximum levels of residues for countries are likely to persist. Full products known as sanitary and phyto- pesticides legally permitted on unpro- global harmonization is unlikely—the sanitary (SPS) standards. cessed food. They estimate the effect World Trade Organization’s Agreement There are legitimate health and of these standards on firms’ decisions on Sanitary and Phytosanitary safety reasons for countries to impose to export, enter, or exit a product- Measures has created no more than a SPS standards. But while these stan- destination market as well as their ef- presumption in favor of the interna- dards may address market failures and fect on export values, quantities, and tional Codex Alimentarius standards. spillovers, they may also create trade unit prices. To do so, they combine Moreover, the development of new frictions and serve protectionist mo- two novel data sets, one covering all and deeper trade agreements that in- tives. Moreover, domestic food and ag- exporting firms in 42 developing coun- clude nontariff measures (such as the ricultural standards often deviate from tries and one covering pesticide stan- Trans-Pacific Partnership) will have a international ones. Higher-income dards for 243 agricultural products in great impact on third-country firms countries generally have greater soci- 80 importing countries over the period in developing countries that may be etal awareness of and concerns about 2006–12. unable to meet the new agreed-upon the standards of the food they con- The evidence shows that standards standards. sume and thus tend to impose stricter have significant effects on individual What can developing country gov- SPS standards. firms’ decisions to enter foreign mar- ernments do to support local firms’ What are the consequences of kets and their eventual success in entry into and success in foreign mar- SPS standards for trade? The effect those markets. An increase in the kets? The best strategy is to provide is ambiguous, particularly for devel- stringency of standards in the destina- the necessary “quality infrastructure” oping countries. SPS standards can tion country relative to those in the so that firms have all the information facilitate trade by signaling that prod- exporting country significantly lowers they need on standards in foreign des- ucts allowed into the market are safe firms’ probability of exporting, deters tinations, supply the necessary test- for consumers. And complying with exporting firms from entering new ing facilities, and maintain efficient such standards may require quality markets, and fosters exit from existing customs clearance procedures that improvements that enhance consum- markets. Relatively stringent standards will reduce the overall costs to export. ers’ demand for imports. This creates also cut exporters’ values and the Developing countries also need to incentives for developing country quantities sold in foreign markets. The revise their SPS standards to ensure exporters to modernize their supply effects on market entry and exit deci- that they reduce the relative cost of chain structure, enabling them to sions are greater for smaller exporters exporting for local firms while also improve their competitiveness while than for larger ones. Network effects— protecting the health and safety of strengthening domestic standards. which are captured by the presence of their citizens. But SPS standards can also firms from the same country that are increase the costs of exporting. all exporting the same product to a Complying with stringent standards given destination—reduce the nega- Ana Fernandes, Esteban Ferro, and John S. Wil- involves both fixed costs (such as for tive effect of relatively stringent stan- son. 2015. “Product Standards and Firms’ Export upgrading production systems, imple- dards on firms’ decisions to enter new Decisions.” Policy Research Working Paper 7315, menting quality control procedures, markets. World Bank, Washington, DC. World Bank Research Digest 7 How Services Reform Helped Revive Indian Manufacturing A key factor in the success of Indian suppliers, now enjoy world-class com- transport, insurance, and telecom- manufacturing in recent years lies munications services. munications boosted the productivity outside manufacturing—in the Transport has seen an improvement of manufacturing firms. The reforms but not yet a revolution. The average benefited both foreign and locally services sector turnaround time for a container at owned manufacturing firms. A one- O major Indian ports has declined from standard-deviation increase in the ag- ne element of India’s rapid about eight days in 1990 to four today. gregate index of services liberalization economic growth since the This is still a long time by international resulted in a productivity increase of early 1990s has been the mod- standards, but the improvement has 11.7 percent for domestic firms and est resurgence of manufacturing in made a difference for firms that com- 13.2 percent for foreign enterprises. the country. Conventional explana- pete in highly variable markets such The additional effect of transport sec- tions have focused on policy reforms as textiles and electronics, where the tor reforms was greatest, followed by in manufacturing industries, notably ability to respond quickly to changes that of reforms in telecommunications trade liberalization and the disman- in demand is crucial. and banking. tling of the “license raj.” Few have rec- Banking reforms seem to have Despite significant improvements, ognized that a key factor lies outside helped too. Manufacturing firms in services reforms in India remain in- manufacturing itself, in the services India saw an improvement in their ac- complete and barriers to domestic sector. cess to and cost of finance as a result and foreign competition persist. Even The neglect of services is surpris- of the banking sector liberalization. outside India, today most of the policy ing because it should be obvious how Two World Bank Investment Climate barriers to competition, and to foreign much manufacturing firms depend on Surveys point to a positive trend: in direct investment, are not in goods services—for credit and insurance, for 2002, 61 percent of Indian firms report- but in services. For example, countries transport and telecommunications. ed that access to finance was an obsta- in Southeast Asia that have reaped Moreover, reforms in the 1990s visibly cle to their business, but in 2006 only huge benefits from the liberalization transformed these services sectors, 41 percent had the same complaint. of trade and investment in goods, such with greater openness and improved These improvements have en- as Malaysia and Thailand, continue to regulation leading to a dramatic hanced firms’ ability to invest in new maintain restrictions on foreign pres- growth in domestic and foreign invest- business opportunities and better ence in services ranging from transport ment. Indian manufacturing firms are production technology, to exploit to telecommunications. The productiv- no longer at the mercy of inefficient economies of scale by concentrating ity improvements observed in manu- public monopolies but can now source production in fewer locations, to ef- facturing industries in India following services from a wide range of domestic ficiently manage inventories, and to services liberalization suggest that, in and foreign providers operating in an make coordinated decisions with their addition to holding back the develop- increasingly competitive environment. suppliers and customers. ment of the services sectors, these As a result, they have access to better, In a recent article Arnold, Javorcik, barriers also penalize manufacturing newer, more reliable, and more diverse Lipscomb, and Mattoo analyze the link firms. business services. between services reforms and manu- Wider appreciation of this link The telecommunications market, facturing productivity in India. The between services sectors and manu- of course, exemplifies the transforma- authors collected detailed information facturing industries may help cre- tive power of liberalization interacting on the pace of reform across Indian ate broader political support for with technological change. It is easy services sectors, with a particular focus pro-competitive regulatory reform in to forget that in the 1980s India’s com- on entry and operational restrictions. services. It may also provide greater munications minister, C. M. Stephen, To make this information amenable to perspective for international trade declared in parliament that telephones econometric analysis, they aggregated negotiations, which continue to focus were a luxury, not a right, and that it into time-varying reform indexes. on goods—agricultural and manufac- any customers dissatisfied with their They then related the total factor pro- turing products—and only notionally service were welcome to return their ductivity of about 4,000 manufacturing address impediments to services trade phone—because there was an eight- firms to the state of liberalization in and investment. year waiting list for service. Today services sectors, taking into account India has the world’s second-largest other aspects of openness, such as tar- telecommunications network and the iffs on output and intermediate inputs third-largest Internet user base, with as well as foreign direct investment in Jens Arnold, Beata Javorcik, Molly Lipscomb, one of the lowest call tariffs. Indian final and intermediate goods sectors. and Aaditya Mattoo. 2016. “Services Reform and businesses, once severely handicapped Their results suggest that pro- Manufacturing Performance: Evidence from In- in communicating with customers and competitive reforms in banking, dia.” Economic Journal 126 (590): 1–39. 8 World Bank ResearchDigest Are the Benefits of Export Support Durable? An export promotion program in half the cost of their export business levels and diversification no longer Tunisia benefited participants—but plans, which focused on one of three differed significantly from those of the only in the short term. Why did the possible objectives: to become a sub- control group. Moreover, the effects stantive exporter (if the firm had little varied across firms: small and large benefits not last? or no export experience), to diversify firms saw no positive impact on export T destination markets, or to develop levels, and even the positive impact for rade liberalization has not al- new export products. Grants were used medium-size firms was temporary. ways led to better export perfor- mostly to cofinance the cost of techni- The results show no evidence that mance. The focus of trade policy cal assistance and marketing services the temporariness of the impact re- has therefore shifted in recent years provided by local and foreign experts. flected spillovers to nonbeneficiary toward trade facilitation and export The analysis combines firm-level firms that helped them to catch up or promotion, with substantial resources data from several sources—the FAMEX greater exposure of beneficiary firms going to export processing zones, program, the National Statistical to crisis-affected economies. On the export assistance programs, and the Institute and Investment Promotion contrary, the impact may have been modernization of border management Agency, and the customs agency—into temporary because it was hard for a and customs procedures. Yet there is a unique data set on Tunisian export- short-lived, arm’s-length interven- not much evidence on the impact of ers. Results are estimated using a tion to durably enhance competitive- such initiatives. propensity-score-weighted regression ness. Indeed, the program led to no To add to this evidence, a recent method that controls for firm fixed ef- improvements in product quality as paper by Cadot, Fernandes, Gourdon, fects and whose weights are obtained would be reflected in higher unit prices and Mattoo assesses the effects of from a probit regression for selection or greater product sophistication. Tunisia’s FAMEX export promotion into the FAMEX program that accounts Instead, it primarily benefited firms program on participating firms. While for, among other things, past firm ex- that initially had no internal export evaluations of such programs typically port performance. Several firm-level unit, suggesting that assistance was focus on their contemporaneous or export outcome variables are consid- rudimentary. FAMEX may have placed short-term impact, the authors also ered, ranging from firms’ total exports too much emphasis on “low-hanging consider the longer-term impact. to their numbers of products and des- fruit” (helping domestic exporters The FAMEX program was aimed at tinations as well as such dimensions attend or set up representation at helping Tunisian firms overcome bar- as export unit values, product sophisti- foreign fairs) rather than on more riers to selling in foreign markets and cation, and distance to destinations. complex activities aimed at improving enhance their competitiveness. Its The estimates suggest that com- products and processes, which would rationale was that Tunisian firms were pared with a control group, FAMEX enhance longer-term competitiveness. poorly informed about export markets beneficiaries initially enjoyed a boost Despite the fairly transient impact, and had difficulty identifying the right in total exports along with greater di- this relatively low-cost intervention target markets, product segments, and versification of destinations and prod- still generated about 2 Tunisian dinars sales channels. The program provided ucts (figure 1). But three years after of additional private profit per dinar firms with matching grants cofinancing the intervention, beneficiaries’ export of program expenditure. The firms’ benefit-cost ratio was 3.57 while the Figure 1. Effect of Tunisia’s FAMEX Program on Export Outcomes for Participating Firms government merely broke even, as the Change in export outcome for FAMEX firms relative to control firms (%) additional corporate tax revenue just 70 covered the public cost of the program. 60 These estimates raise the question of why the firms did not undertake such 50 investments on their own. Of course, 40 firms’ inability to borrow against future 30 profits, or lack of information about the benefits of investing in export pro- 20 motion, could still provide a rationale 10 for the program. 0 Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 −10 Olivier Cadot, Ana Fernandes, Julien Gourdon, Total exports Number of destinations Number of products and Aaditya Mattoo. 2015. “Are the Benefits of Note: For each year, the change in a firm’s export outcome is calculated as the difference between Export Support Durable? Evidence from Tunisia.” that year’s outcome and the corresponding outcome before treatment. The figure shows the difference between these changes for FAMEX firms and the corresponding changes for control Journal of International Economics 97 (2): firms based on propensity-score-weighted regression estimates. Year 1 is the year of treatment. 310–24. World Bank Research Digest 9 Resource Rents and Local Development in Post-Apartheid South Africa on services reforms and panel data for about 4,000 Indian firms, it finds that banking, transport, insurance, In post-apartheid South Africa the collectively bargain for a larger share and telecommunications reforms all dismantling of coercive institutions of industry profits. The bill of rights of had significant positive effects on the affected the distribution of rents the 1996 constitution not only prohib- productivity of both foreign and locally ited the state from discrimination on owned manufacturing firms. from natural resource exports any grounds but also included explicit The limited success of trade liber- A reforms to labor laws—notably by im- alization in spurring exports in some large body of evidence has posing fair labor practices—and the countries has renewed interest in pro- documented how bad economic right to strike. The mining industry had active trade policies. A paper assessing institutions can stunt develop- the highest unionization rate of all sec- Tunisia’s FAMEX export promotion ment. But can a bad legacy be turned tors, exceeding 70 percent in both 1996 program finds that beneficiaries ini- into a relatively more prosperous and 2011, and thus was in a particu- tially saw both higher export levels future? In a recent paper Bastos and larly strong position to benefit from and greater diversification in products Bottan study how reforms at the end these reforms. and destinations. But these effects of apartheid in South Africa spurred Examining relative changes be- lasted for only about three years. development for marginalized com- tween 1996 and 2011, the authors find Enhancements in product quality or munities. In October 1996, just after that incomes converged faster among sophistication could have durably the end of apartheid, communities marginalized communities with higher strengthened competitiveness. located just inside the former self- initial employment in the mining in- The political support for policy governing territories set aside for dustry. The results accord with stan- reform depends less on its aggregate black inhabitants were considerably dard bargaining theory in which the benefits than on its distributional ef- poorer than communities located dismantling of coercive institutions fects. One study examines how the just outside these areas. Fifteen years improves the negotiating position of dismantling of coercive institutions later the gaps remained sizable. But unionized workers in the sector. Other after the end of apartheid in South incomes converged at a different pace possible explanations for the income Africa affected the distribution of rents across local communities of the for- convergence of marginalized commu- from natural resource exports. Another mer homelands. nities in the post-apartheid period are presents evidence suggesting that What explains this difference? The difficult to reconcile with the observed during Tunisian president Ben Ali’s authors argue that the different rates patterns. reign, firms owned by the president of income growth for black commu- and his family were more likely to nities in the former homelands may Paulo Bastos and Nicolas Bottan. 2016. “Re- evade import tariffs and other taxes. be explained in part by the degree of source Rents, Coercion, and Local Development: Understanding the political forces exposure of marginalized communi- Evidence from Post-Apartheid South Africa.” Pol- that shape existing policies and their ties to rents from natural resource icy Research Working Paper 7572, World Bank, implementation is a precondition for Washington, DC. exports—as measured by the initial successful reform. share of people in each community Pursuing international trade coopera- who were employed by the mining (continued from page 1) tion. International trade cooperation industry. Although South Africa is a has stalled, multilaterally at the World leading mineral producer and exporter, Insights from World Bank Trade Organization (WTO) and now the industry has traditionally been Research on Trade and even in the preferential mega-regional controlled by a few privately owned trade initiatives. Recent research finds Globalization mining investment houses. Moreover, that the WTO is structured to encour- in the apartheid era the mining in- products in 80 importing countries for age policy-efficiency-enhancing out- dustry was characterized by a highly 2006–12. It finds that where standards comes through a process of shallow uneven distribution of income: profits in the importing country are more integration—that is, reciprocal liber- were high, but wages for black workers restrictive than those in the export- alization of border barriers and rules were low. In the gold mining industry ing country, firms are less likely to ex- against discrimination. The “deep” in- in 1972, for example, monthly salaries port—with smaller exporters the worst tegration envisaged by mega-regional were about 18 rand for African miners affected. agreements focuses on regulations but 400 rand for white miners. Black Conventional explanations for the on consumer product safety, the en- workers were not included in the legal post-1991 growth of India’s manufac- vironment, and labor standards. The definition of employee and so could turing sector focus on trade liberaliza- implied tradeoffs—between efficiency, not become members of legally regis- tion and industrial delicensing. But local preferences, and sovereignty— tered labor unions. recent research shows that a key factor need to be much better understood After the end of apartheid, re- lies outside manufacturing—in the and explained to a skeptical public. forms created scope for workers to services sector. Using a new data set 10World Bank ResearchDigest Are Politically Connected Firms More Likely to Evade Import Tariffs and Other Taxes? In Tunisia firms owned by former Underreporting was also common. prices and other firms’ prices widening president Ben Ali and his family For the economically active firms that with the tariff rate. were more likely than others to submitted tax declarations, the au- Why does this matter? Tax evasion thors checked whether the sales these is not only unfair as well as damaging evade import tariffs and other taxes firms declared to the tax authorities to government revenues. It is also inef- during his reign were consistent with their reports to ficient because it gives the perpetra- D the social security and customs ad- tors a cost advantage over those who uring the protests of the Arab ministrations. Tax declarations were are compliant that is not based on Spring, one of the chief de- considered anomalous if the sales performance. mands of those who took to the reported were lower than the reporting What has happened since the streets was ending abuse of power firm’s wage bill, total imports, or total Jasmine Revolution? The ousting of by the ruling elites. Two recent pa- exports. No less than 15.3 percent of Ben Ali during the revolution led to a pers unveil the scale of such abuse nonconnected firms submitted a tax reduction in tax evasion and in the un- in Tunisia. They document evidence declaration that was anomalous in derreporting of unit prices in product suggesting that firms owned by former this way. All else being held equal, lines in which the Ben Ali family had president Zine al-Abidine Ben Ali and connected firms were 8.4 percentage been active. But it sparked an increase his family were more likely to evade points more likely to submit such an in overall tariff evasion. Despite cur- taxes, circumventing at least $1.2 bil- anomalous declaration. tailed capture by the Ben Ali clan, tariff lion worth of import tariffs on account In addition to assessing domestic evasion gaps have risen by 5 percent of their political connections between tax evasion, the authors examined on average, in part reflecting the rise in 2002 and 2009. While evasion by firms evasion of import tariffs by compar- informal trade with Algeria and Libya. previously owned by the Ben Ali fam- ing import transaction records with Thus, if anything, tax evasion seems ily seems to have subsided since the counterpart declarations by export- to have escalated since the Jasmine Jasmine Revolution, overall evasion in ers, who have limited incentives to Revolution. Tunisia has escalated. lie about how much they have sold How is it possible to know this? abroad. If imports are reported cor- Verifying tax declarations is difficult, rectly, they must be very close to since objective information on what exports reported in countries that firms should declare is typically not send goods to Tunisia. But if imports available. And in Tunisia information are declared incorrectly—or not at sharing between (and within) govern- all—then “evasion gaps” may arise. ment agencies was extremely limited These are measured as the difference during the Ben Ali era. But data made between exports reported in countries available by Tunisian authorities after selling goods to Tunisia and imports of the revolution made it possible to those same goods reported in Tunisia; detect evasion by identifying discrep- the greater this difference, the more ancies between declarations made imports are “missing” and the less rev- to different government agencies. By enue is collected by Tunisian customs. comparing reports made to the social Such evasion gaps are typically larg- security and customs administrations est for goods subject to high tariffs, with tax records, the authors were able where evasion is most lucrative. In this to assess the prevalence of nonreport- context they have become a standard ing and underreporting. proxy for tax evasion. Nonreporting was widespread. In Tunisia evasion gaps were great- Among firms not connected with the est for products imported by firms Bob Rijkers, Leila Baghdadi, and Hassen Arouri. Ben Ali family, 9 percent did not sub- owned by the Ben Ali family—and es- 2016. “Are Politically Connected Firms More mit a tax declaration despite being pecially large when such products were Likely to Evade Taxes? Evidence from Tunisia.” economically active. Nonreporting was subject to high tariffs. These gaps were World Bank Economic Review, ABCDE even more widespread among active due to the underreporting of prices. Supplement. Published electronically June 9, firms owned by Ben Ali or his family; Firms owned by the Ben Ali family re- 2016. doi:10.1093/wber/lhw018. with other conditions remaining the ported significantly lower prices than Bob Rijkers, Leila Baghdadi, and Gael Raballand. same, these connected firms were 4.6 their competitors for the same goods 2015. “Political Connections and Tariff Evasion: percentage points more likely than imported from the same country (and Evidence from Tunisia.” World Bank Economic nonconnected firms to have a missing consequently paid less tax for those Review. Published electronically November 17, tax declaration despite being active. goods), with the gap between Ben Ali 2015. doi:10.1093/wber/lhv061. World Bank Research Digest11 Trade Agreements under Threat Is 2016 the year that antiglobalization border barriers, such as import tariffs. governments apply behind the border, forces will halt what has been One important result is that countries such as regulations on consumer prod- a seven-decade push on global around the world today apply import uct safety, the environment, and labor tariffs under the GATT/WTO agree- standards. economic cooperation? ments at historically low rates. Are direct negotiations over such O Recent economic research has shed policies necessary? What are the n June 23, 2016, the United new light on the role that basic prin- tradeoffs—for efficiency, for local Kingdom held a referendum in ciples and formal rules play in making preferences, and for sovereignty—in which a majority of voters opted the GATT/WTO agreements work. The the approach of these mega-regionals to leave the European Union, after 43 first is reciprocity, or the mutual ex- to international cooperation as years of membership for the country. change of market access. Reciprocity compared with the shallow approach The unprecedented Brexit vote opens helps countries move out of prisoner’s- of the GATT/WTO? up questions about the future of U.K.- dilemma-like, tariff-ridden outcomes. Guidance from economic research EU relations, of further EU integration, A growing body of empirical evidence on this last question is still at an early and of the United Kingdom’s trade re- shows that these trade agreements stage. Relatively little is known about lationships with the rest of the world. play a critical role in galvanizing gov- the very newest negotiations taking In the United States the 2016 elec- ernments across countries to jointly place under the mega-regionals. Is tion season has featured presidential lower their import tariffs and then it possible that they are pushing too candidates from both major political maintain those tariffs at low rates. deep? Or is there something more parties actively speaking out against The GATT/WTO agreements also fundamental about international trade international trade. Particularly caught include two rules of nondiscrimina- that has changed and that requires up in the rhetoric have been interna- tion. Economic research has shown this additional depth? tional trade agreements. The target is that these rules affect how countries What is increasingly clear is that not just the North American Free Trade negotiate and that they sustain more forces emerging in 2016 are putting Agreement (NAFTA). One proposal has globally efficient levels of trade than if all forms of trade agreements under the United States potentially exiting governments were interacting in a non- the microscope. That these threats from the World Trade Organization cooperative manner and without trade are arising from voters in democra- (WTO). agreements. cies makes them fundamentally dif- The public debates taking place The first such rule is national treat- ferent from other recent threats to in the United Kingdom, the United ment. This requires that once the ex- cooperation. Consider, for example, States, and other countries bring into porter of a good has paid the import the acute economic shock of the Great sharp relief fundamental questions tariff necessary to cross the border to Recession, the most recent of these about international cooperation over sell into a foreign market, the govern- other threats suggesting the possibility trade policy. Why do countries sign on ment may not use additional taxation of an imminent return to 1930s-type to these trade agreements? What role or regulation that would affect that protectionism. do trade agreements play? product differently than a nationally What role for trade agreements? In a lengthy survey of recent re- produced good. The second is most- Today’s researchers must continue to search Bagwell, Bown, and Staiger favored-nation (MFN) treatment. With provide theoretical and empirical an- provide some initial answers and only well-codified exceptions, WTO swers. Before it is too late, they must perspectives. Their study is motivated members are supposed to apply the also come up with more effective ways primarily by a need to synthesize ma- same policy across all trading partners. of explaining the benefits of interna- jor intellectual advances in economic Negotiators of the mega-regional tional cooperation to the voting public. research over the past decade. trade agreements—the Trans-Pacific A main focus is the WTO. This Partnership and the Transatlantic multilateral agreement involves more Trade and Investment Partnership— than 160 countries that have collec- are pushing in a different direction tively established the basic rules for than the WTO. This direction is more nondiscriminatory trade in the global than the obvious fact that treating economy. The approach under the agreement insiders differently from WTO and its predecessor—the General outsiders is a major exception to the Agreement on Tariffs and Trade MFN rule. These new types of arrange- Kyle Bagwell, Chad P. Bown, and Robert W. (GATT)—has been to engage in a grad- ments involve governments negotiat- Staiger. Forthcoming. “Is the WTO Passé?” Jour- ual process of “shallow” integration. ing over new and different issue areas. nal of Economic Literature. (Also issued as The GATT/WTO has routinely convened “Deep” integration negotiations are Policy Research Working Paper 7304, World countries to voluntarily reduce their taking place directly over policies that Bank, Washington, DC, 2015.) 12World Bank ResearchDigest Recent World Bank Research Publications and Papers on Trade and Globalization Books and Reports Bastos, Paulo, Lucio Castro, Julian Cristia, and Carlos Mattoo, Aaditya, and Batshur Gootiiz. Forthcoming. Scartascini. 2015. “Does Energy Consumption “Regionalism in Services: A Study of ASEAN.” World Ahsan, Ahmad, Aaditya Mattoo, Batshur Gootiiz, Juan Respond to Price Shocks? Evidence from a Regression- Economy. Sebastian Saez, Martin Molineuvo, and Peter Discontinuity Design.” Journal of Industrial Economics Walkenhorst. 2015. ASEAN Services Integration Report: Özden, Çaglar, and Christopher Parsons. 2016. “On the 63 (2): 249–78. A Joint Report by the ASEAN Secretariat and the World Economic Geography of International Migration.” World Bank. Washington, DC: World Bank. Bastos, Paulo, Joana Silva, and Rafael Proenca. 2016. Economy 39 (4)w: 478–95. “Exports and Job Training.”Review of International Pierola Castro, Martha Denisse. 2015. Trade in Zimbabwe: Rijkers, Bob, Hassen Arouri, Caroline Freund, and Antonio Economics 24 (4): 737–56. Changing Incentives to Enhance Competitiveness. World Nucifora. 2014. “Which Firms Create the Most Jobs in Bank Development Studies series. Washington, DC: Beine, Michel A. R., Frederic Docquier, and Çaglar Özden. Developing Countries? Evidence from Tunisia.” Labour World Bank. 2015. “Dissecting Network Externalities in International Economics 31: 84–102. Migration.” Journal of Demographic Economics 81 (4): Chapters in Books 379–408. Working Papers Bown, Chad P., and Kara M. Reynolds. 2015. “Trade Anderson, James E., Ingo Borchert, Aaditya Mattoo, Carletto, Calogero, Jennica Larrison, and Çaglar Özden. Flows and Trade Disputes.” Review of International 2015. “Informing Migration Policies: A Data Primer.” and Yoto V. Yotov. 2015. “Dark Costs, Missing Data: Organizations 10 (2): 145–77. Shedding Some Light on Services Trade.” NBER Working In International Handbook on Migration and Economic Development, edited by Robert Lucas. Cheltenham, UK: Cherkashin, Ivan, Svetlana Demidova, Hiau Looi Kee, Paper 21546, National Bureau of Economic Research, Edward Elgar. and Kala Krishna. 2015. “Firm Heterogeneity and Cambridge, MA; and Policy Research Working Paper Costly Trade: A New Estimation Strategy and Policy 7465, World Bank, Washington, DC. Constantinescu, Cristina, Aaditya Mattoo, and Michele Ruta. Experiments.” Journal of International Economics 96 (1): 2015. “The Global Trade Slowdown.” In The Global Trade Blanchard, Emily J., Chad P. Bown, and Robert C. Johnson. 18–36. 2016. “Global Supply Chains and Trade Policy.” Policy Slowdown: A New Normal? edited by Bernard Hoekman. VoxEU.org eBook. London: CEPR Press. Cristea, Anca, Russell Hillberry, and Aaditya Mattoo. 2015. Research Working Paper 7536, World Bank, Washington, “Open Skies over the Middle East.” World Economy 38 DC. Fernandes, Ana Margarida, and Ricardo Lopez. 2015. (11): 1650–81. “Imported Technology and Firm Exporting: The Case of Bown, Chad P. 2015. “What’s Left for the WTO?” Policy Chile.” In Handbook on Trade and Development, edited Fernandes, Ana Margarida, and Tolga Cebeci. 2015. “Micro Research Working Paper 7502, World Bank, Washington, by Oliver Morrissey, Ricardo Lopez, and Kishor Sharma. Dynamics of Turkey’s Export Boom in the 2000s.” World DC. Cheltenham, UK: Edward Elgar. Economy 38 (5): 825–55. Constantinescu, Ileana Cristina, Aaditya Mattoo, and Kee, Hiau Looi, Cristina Neagu, and Alessandro Nicita. Fernandes, Ana Margarida, and Alberto E. Isgut. 2015. Michele Ruta. 2016. “Does the Global Trade Slowdown 2015. “Is Protectionism on the Rise? Assessing National “Learning-by-Exporting Effects: Are They for Real?” Matter?” Policy Research Working Paper 7673, World Trade Policies during the Crisis of 2008.” Reprint. In Emerging Markets Finance and Trade 51 (1): 65–89. Bank, Washington, DC. International Trade and the New Global Economy, edited Fernandes, Ana Margarida, and Caroline Paunov. 2015. “The Fernandes, Ana, Russell Hillberry, and Claudia Berg. 2016. by Edward D. Mansfield. Cheltenham, UK: Edward Elgar. Risks of Innovation: Are Innovating Firms Less Likely to “Expediting Trade: Impact Evaluation of an In-House Mattoo, Aaditya. 2015. “An EU-US Trade Deal: Good or Bad Die?” Review of Economics and Statistics 97 (3): 638–53. Clearance Program.” Policy Research Working Paper for the Rest of the World?” In Catalyst? TTIP’s Impact 7708, World Bank, Washington, DC. Freund, Caroline, and Bob Rijkers. 2014. “Episodes of on the Rest, edited by M. Sait Akman, Simon J. Evenett, Unemployment Reduction in Rich, Middle-Income Pierola, Martha Denisse, Ana Margarida Fernandes, and and Patrick Low. VoxEU.org eBook. London: CEPR Press. and Transition Economies.” Journal of Comparative Thomas Farole. 2015. “The Role of Imports for Exporter Özden, Çaglar, and Zovanga Louis Kone. Forthcoming. Economics 42 (4): 907–23. Performance in Peru.” Policy Research Working Paper “Brain Drain, Gain and Circulation.” In Handbook of 7492, World Bank, Washington, DC. Hillberry, Russell, and Phillip McCalman. Forthcoming. Globalization and Development, edited by Kenneth “Import Dynamics and Demands for Protection.” Reinert. Cheltenham, UK: Edward Elgar. Canadian Journal of Economics. Kee, Hiau Looi. 2015. “Local Intermediate Inputs and the Trade research at the World Bank, while funded Journal Articles Shared Supplier Spillovers of Foreign Direct Investment.” in large part by the World Bank budget, has also Artuc, Erhan, Frederic Docquier, Çaglar Özden, and Journal of Development Economics 112: 56–71. Christopher Parsons. 2015. “A Global Assessment received valuable support from the UK-funded of Human Capital Mobility: The Role of Non-OECD MacLaren, Donald, Russell Hillberry, and Suphanit Strategic Research Program, the Knowledge for Destinations.” World Development 65: 6–26. Piyapromdee. 2014. “‘Fair Trade’ Coffee and the Mitigation of Local Oligopsony Power.” European Review Change Program, and the Multi-Donor Trust Artuc, Erhan, and John McLaren. 2015. “Trade Policy of Agricultural Economics 4 (4): 537–59. Fund for Trade and Development. and Wage Inequality: A Structural Analysis with Occupational and Sectoral Mobility.” Journal of Mattoo, Aaditya, Fabien Bertho, and Ingo Borchert. 2016. “The Trade-Reducing Effects of Restrictions on Liner Working Papers can be downloaded at http://econ.worldbank.org International Economics 97 (2): 278–94. To download the World Bank Research E-Newsletter, Shipping.” Journal of Comparative Economics 44 (2): 231–42. go to http://econ.worldbank.org/research_newsletter The World Bank Research Digest is a quarterly publica- The Research Digest is financed by the Bank’s Editorial Committee: Chorching Goh (managing editor), tion disseminating findings of World Bank research. Research Committee and managed by DECDP, the Aslı Demirgüç-Kunt, and Shiva S. Makki. Editor: Alison The views and interpretations in the articles are those research support unit of the Development Economics Strong; production: Roula Yazigi. For information or of the authors and do not necessarily represent the Senior Vice Presidency (DEC). The Research Digest is free subscriptions, send email to research@worldbank. views of the World Bank, its Executive Directors, or the not copyrighted and may be reproduced with appropri- org or visit http://econ.worldbank.org/research_digest. countries they represent. ate source attribution. The World Bank 1818 H Street, NW Washington, DC 20433, USA Printed on Recycled Paper