Policy Research Working Paper 10637 Leveraging Trade for More and Better Job Opportunities in Developing Countries A Framework for Policy Ben Shepherd Deborah Winkler Macroeconomics, Trade and Investment Global Practice December 2023 Policy Research Working Paper 10637 Abstract Trade and labor markets are intimately connected. This government’s dual challenge in a sustainable way: policies connection presents governments with a dual economic for people, sectors, and places. The framework includes pol- challenge that cannot be resolved without social com- icies to mitigate losses and facilitate movement of workers, promise: maximizing aggregate gains but minimizing classical trade policies, and a broad set of complementary disaggregated costs, which can include losses to individuals policies that reduce trade costs. It also looks at the need and groups. This paper draws on recent research to develop for fiscal space to implement policies, and highlights the a framework for thinking rigorously about these linkages. It tension between tariff reductions and trade-related taxes, then examines aspects of policy design and implementation especially in countries where trade taxes account for a sig- that relate directly to labor market outcomes. It discusses nificant proportion of total government revenue. three sets of policies that are required to help resolve the This paper is a product of the Macroeconomics, Trade and Investment Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at dwinkler2@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Leveraging Trade for More and Better Job Opportunities in Developing Countries: A Framework for Policy Ben Shepherd 1 Deborah Winkler2 JEL Codes: F13; F15; J08. Keywords: Trade; Economic integration; Labor markets; Policy. 1 Principal, Developing Trade Consultants. Email: ben@developing-trade.com. 2 Senior Economist, Trade and Regional Integration, World Bank Group. Email: dwinkler2@worldbank.org. The authors would like to thank Sébastien Dessus, Thomas Farole, Mona Haddad, Mary Hallward-Driemeyer, Claire Hollweg, Hagen Kruse, Gladys Lopez-Acevedo, Maryla Maliszewska, Dino Merotto, Doug Nelson, Nithin Umapathi, as well as the participants of the World Bank authors’ workshops “Leveraging trade for more and better job opportunities in developing countries” in May and November 2022, for their helpful comments and suggestions. The work was carried out under the overall supervision of Sébastien Dessus and Antonio Nucifora. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the World Bank Group and its affiliated organizations, its Executive Directors or the governments they represent. 1 INTRODUCTION Trade and labor markets are intimately connected. While labor market functioning is an important factor in a country’s ability to respond to trade opportunities or shocks, there can also be potentially important labor market disruptions when trade integration changes. The focus of this paper is on the responses to shocks, but both realities stem from the fact that changes in trade costs—such as reductions in tariffs or non-tariff measures (NTMs)—produce relatively large resource reallocations and much smaller net gains in economic welfare. 3 “Resource reallocation” in this context means movements of factors of production: workers and capital. So the reallocation effects of trade cost changes can have important human consequences because they involve sometimes substantial movements of people from one task to another, one sector to another, and when activities cluster spatially, from one place to another. This paper aims to identify aspects of trade policy and complementary policies for people, places, and sectors, which can maximize the aggregate economic gains from increased trade integration, while minimizing disaggregated labor market losses. Talking about the problem in this way makes clear that any government contemplating bringing about a change in trade costs faces two challenges. The first is a familiar aggregate problem: how can policy maximize the gains from trade by allowing resources—workers and capital—to move so as to take advantage of specialization by comparative advantage? But no less important is the second disaggregated problem: how can policy minimize the costs for particular actors, especially workers, involved in that movement? It is important to state from the outset that there is no single, perfect answer to these twin challenges. Indeed, the answers to the two questions sometimes conflict in important ways. In addition to this two-fold policy challenge within countries, experiences regarding the labor market effects of trade vary between countries. To emphasize such differences in country experiences, it helps to look at one aspect of the literature examining the impacts of increased import competition from China following its 2001 World Trade Organization (WTO) accession. Estimates of the impact vary for the United States, but the key insight is that the biggest changes are at the local level: aggregate impacts are much smaller, though still substantial. A reduced form estimate from Acemoglu et al. (2016) gives the figure of 2.0 million to 2.4 million US manufacturing jobs lost, which translates into a substantial amount of labor market disruption, which in turn means lost jobs and impaired well- being. But a similar approach using data for Ethiopia by Ngoma (2022) finds a stunningly different result: increased imports from China translated into major economic gains, including more jobs for Ethiopian workers, with corresponding increases in incomes and well-being. Better understanding the mechanisms that give rise to these disparate labor market impacts from changing trade costs is a key objective of this paper. Clearly, the United States and Ethiopia differ along many dimensions, so it is important to make an effort to identify which ones give rise to such distinct labor market experiences from closely related changes in policy—in this case reductions in trade costs vis-à-vis China, combined with gains in Chinese competitiveness in a variety of manufacturing sectors. But understanding the underlying mechanisms also matters from a human and 3 Anderson and Van Wincoop (2004) define trade costs as the set of factors that drive a wedge between producer prices in the exporting country and consumer prices in the importing country. They therefore include “classical” trade policy factors like tariffs, but also behind-the-border regulations, institutional quality, and other factors that affect the costs of moving goods or services between countries. This paper uses the term “trade policy” broadly to refer to any policy that affects trade costs. It therefore includes standard tariffs and non-tariff measures, but also a range of other policies that affect trade but that do not come under the traditional umbrella of trade costs. The expanding scope of both the World Trade Organization (WTO) and regional trade agreements supports this broad use of the term. 2 economic perspective. It is about ensuring that trade reforms benefit as many people as possible (human dimension). But trade policy reform also requires social consensus, which can be undermined when disaggregated disruptions cause extensive pain that leads to calls for halting or even reversing the course of trade liberalization with sometimes major consequences for the wider economy including consumers and producers (economic dimension) as has been the experience of the US (Autor et al., 2020). The objective of a policy framework for understanding the linkages between trade and jobs is therefore not to provide a “silver bullet” answer that can easily be picked up and applied without accounting for various country settings. Instead, this paper aims to provide an assessment of the ways in which policy can shape the human and economic dimensions, and to identify the types of policy measures that have particular effects. Individual countries will need to engage in their own deliberative processes to find the appropriate balance both (a) among the different policy measures that appear in the policy framework, and (b) between maximizing the aggregate gains and minimizing disaggregated costs. An understanding of the mechanisms at play can help identify indicative policy priorities for particular groups of countries, looking across a range of variables that distinguish them in an approximate way. The starting point for this paper’s analysis is that trade costs are only one factor affecting labor market performance. In many cases, other policies—such as fiscal and monetary policy—play a greater role in generating labor market outcomes at a macroeconomic level, such as the economy-wide unemployment rate or average earnings. As noted above, the welfare gains from lowering trade costs are relatively small relative to the dislocation that can be caused at a disaggregated level, so understanding the forces behind that dislocation helps develop a policy framework that policy makers can use to find an appropriate balance between maximizing aggregate economic gains and minimizing disaggregated labor market disruptions and losses. The first contribution of the paper is to distill a comprehensive framework for thinking about the links between trade and labor market outcomes. It uses and expands the quantitative model of Caliendo et al. (2019) to motivate an intuitive presentation of the underlying mechanisms through which a trade costs change affects changes in disaggregated labor market conditions. The reason for choosing this model is that it is comprehensive and dynamic, and represents current best practice in the academic literature. The key insight is that changes in relative prices induced by changes in trade costs produce expansionary or contractionary pressures for individual sectors depending in part on the economy’s pattern of comparative advantage. Those pressures in turn give rise to a choice for workers: stay in their existing sector, migrate to a growing sector or another location, or exit the labor force. Rather than exiting, which may not be a viable option, workers in many low and middle-income countries can undertake self-employment or solo-entrepreneurship, or participation in the informal labor market. Understanding the dynamic factors that influence these choices when moving is not costless is a key step in unlocking policy insights that can help promote welfare-enhancing movements, while protecting people from the human consequences of exit or informality. Based on this analysis, the paper develops a policy framework that starts from an analysis of the ways in which policies that reduce trade costs can be designed and implemented with the dual challenge— aggregate gains and disaggregated costs—in mind. It then looks at specific policies in relation to people, places, and sectors, as part of the set of complementary policies that should accompany reductions in trade costs. Finally, it deals squarely with the fact that many complementary policies involve significant real resource costs, which means that governments need to raise revenue. Trade policy reform, such as reducing tariffs, further aggravates this problem in some low- and middle-income countries, where 3 trade taxes contribute a significant proportion of total government revenue. The analysis therefore looks at ways of mitigating this problem in the short to medium term, while countries shift to an alternative tax base (typically income and consumption). Against this background, the paper proceeds as follows. Section 2 provides a detailed framework for thinking about the links between trade integration and labor markets, drawing on and expanding Caliendo et al. (2019) but presenting it in a non-technical way. The section then considers aspects of designing trade cost reductions that have direct implications for managing labor market disruptions. Sections 3 to 5 then look in more detail at policies for people, sectors, and places, respectively. Section 6 concludes and distills the paper’s policy implications. 2 THINKING ABOUT TRADE, JOBS, AND POLICY: AN OVERVIEW This section presents a framework for analyzing the links between trade integration and labor market outcomes. To do so, it uses recent academic work that represents current best practice in that area. The presentation is non-technical, focusing on concepts and intuition. The purpose of delving into this material in detail is to motivate the development of specific policies in subsequent sections. 2.1 Part One: From Trade Costs to Relative Prices 2.1.1 The Consumer and Producer Channels The economic impact of a change in trade costs comes through changes in relative prices. An increase in the relative price of a sector’s output incentivizes producers to expand and export more, while a decrease has the opposite effect. Conversely, a decrease in the relative price of a sector’s inputs increases competitiveness and boosts exports downstream. But whether a sector expands or contracts, sectoral reallocations necessarily interact with the labor market. The linkages between trade and labor run deep, and are many-faceted. As a result, there is undoubted potential for these processes of price- induced changes in output, and therefore implied labor demand, stemming from changes in trade costs to affect specific types of workers performing tasks in particular sectors and locations (e.g., Grossman and Rossi-Hansberg, 2008). In terms of the current understanding of trade and labor market dynamics in the academic literature, the leading example of an analysis that incorporates the key aspects of trade—comparative advantage, general equilibrium thinking—and the most important facets of labor markets—imperfect mobility of labor, and geographical and sectoral segregation—is Caliendo et al. (2019). This section therefore presents a schematic version of the model, and derives a set of implications and policy questions surrounding the linkages between trade and labor markets. In doing so, it abstracts from the model’s mathematical foundations to focus on the core economic processes at work. The starting point for the model is a relationship between production and consumption decisions in the presence of trade costs as developed in Caliendo and Parro (2015). Figure 1 simplifies this set of relationships, by showing how a change in trade costs flows through to a change in national income by affecting the prices of intermediate and consumption goods (and services), which in turn affects the prices of traded goods and services, as well as the composition of each country’s trade. 4 Figure 1: Flowchart of key mechanisms in the trade, production, and consumption sections of the Caliendo et al. (2019) model. Source: Authors. 2.1.2 Trade Costs Changes in a GVC World To see how the flowchart works, it helps to take a concrete example. Consider a reduction in trade costs affecting electrical products, which country A both exports and imports. The policy change reduces the price of electrical products relative to other types of consumption where there is no policy change. But it also reduces the relative price of electrical products inputs into downstream production. So there is a twin effect of the policy change: competition from imports pushes prices down in electrical products, which makes competitive conditions harder for local producers, but that same effect spurs the competitiveness of producers using electrical inputs, either in another sector (e.g., motor vehicle companies) or within the sector (e.g., computer companies). Nonetheless, even before moving to the labor market, the implications for the electrical products sector are ambiguous. If the sector were a pure import-competing sector, with zero exports, then the impact at a sectoral level would be clear: prices for output would fall, imports would increase, and lower demand would lead to a flow of economic resources out of the sector as it contracts. But if the sector both exports and imports, the situation is more complex. Lowering tariffs allows resources to shift from pure import competition into the production of exports, including in the sector affected by trade cost reductions in this case. In addition, sectors typically use part of their own output as an intermediate input, so the positive downstream effect would also materialize to some extent for electrical products themselves. These effects are of particular importance in a GVC world, where at the firm level, importing of intermediates can take place simultaneously with exporting of intermediate or final goods further downstream. As even this simplified exposition makes clear, the implications of changing trade costs vary substantially according to the type of trade involved. Where reductions in trade costs focus on intermediate imports, they can act to boost exports downstream: this example is a classic case of the complex reality of protection in a GVC world. Moreover, the model puts a lower bound on the size of this effect, as it treats intermediates as substitutes, but some empirical work suggests that imported and locally produced intermediates may have some aspects of complementarity. As a result, displacement of local input production could be lower than suggested by the model. And there may also be cases where imported inputs facilitate new export production because alternative domestic inputs do not exist or not at sufficient scale, especially for smaller countries. This has been the case for Cambodia’s apparel exports, largely relying on imported textiles (Calabrese and Balchin, 2021). 5 Nonetheless, this set of effects is lost when reductions in trade costs focus on final consumption goods and services: there is no boost to downstream production from lower input costs. There can still be a general equilibrium reallocation in favor of exports, but the effect will be more muted than in the example above. It is therefore more likely that consumer gains from lower prices will be associated with producer losses, as the formerly protected industry contracts. 2.2 Part 2: From Relative Prices to Labor Market Impacts When it comes to the labor market and its relationship to trade (Figure 2), Caliendo et al. (2019) provide a much richer set of mechanisms than models that focus on the relationship between changing trade costs and a country’s trade composition only (e.g., Caliendo and Parro, 2015). The starting point is that labor markets are segmented by sector and location. These distinct sector-location pairs (or markets) differ from each other by sector, location or both. When relative prices change, there is an induced change in relative demand for labor: sectors with a higher relative price have an incentive to pull in extra workers, while sectors with a lower relative price tend to engage in some combination of shedding workers or lowering wages. However, due to mobility frictions there are adjustment costs to moving from one sector-location to another sector-location. Such frictions can arise from a variety of sources, including policies in areas like housing and land use. Workers will only undertake this internal migration if the expected benefits from higher future wages outweigh those costs. If they do not, they will either stay put, or exit the labor market entirely. As noted above, exit may not be a feasible option for workers in low and middle-income countries, so moving beyond the strict confines of the model, their choices also include self-employment and informal labor. Combining segmentation of markets with these kinds of frictions makes it possible to run a complex causal chain all the way from changes in trade costs in Figure 1 to migration outcomes in Figure 2. Figure 2: Flowchart of key mechanisms in the labor market section of the Caliendo et al. (2019) model. Source: Authors. Note: dotted lines indicate additional mechanisms of importance in low and middle-income countries that are not included in the Caliendo et al. (2019) model. 2.2.1 Relative Price Changes with Segmented Labor Markets To see how this process works, it helps to take a concrete example. The previous section looked at a reduction in trade costs affecting electrical products, and showed that the consumer and producer channels combine to translate that change into a new set of relative prices (Figure 1). The starting point for Figure 2 is that shift in relative prices, in this case a lowering of the price of electrical products relative to other goods or services. Exporters using electrical inputs will become more competitive 6 and so will be able to expand production as they sell more overseas. Import-competing firms will find the marketplace tougher, and so will contract (see discussion in section 2.1.2). Depending on the relative strength of these underlying channels, the overall impact on sectoral labor demand is ambiguous. If the sector is predominantly export-oriented, that overall impact will likely be positive, as producers in the sector are, on average, eager to hire more workers. But to do so, they need to convince workers to move from some other part of the economy. They will be willing to pay higher wages than prior to the policy change to incentivize workers to move. Importantly, the change in wages needs to be high enough to compensate for the adjustment costs involved in moving. As a result, not all workers who could profitably move in a frictionless world will in fact move. Some will move to other sectors in the same location, perhaps at lower wages, while others will exit the labor market as their home sectors contract. At the same time, movement across sectors implies geographical movement: locations with a strong concentration of electrical products exporters will see in-migration, while areas specialized in other sectors will see out-migration. What does the model mean for the welfare of workers, who are also consumers? First, relative changes in real wages can differ by sector and location, which means that the labor market impacts of changing trade costs are, in the first instance, highly localized. These impacts are directly driven by derived demand: changes in relative prices for output affect the relative demand for labor. But disaggregated labor market effects can also be aggregated over a full national economy, so a complex pattern of narrowly defined gains and losses can translate into an overall labor market impact for a country. Second, the presence of frictions affecting migration means that workers do not necessarily move even if wages are higher in an expanding sector. This feature of the model coincides with the important stylized fact that workers are imperfectly mobile, both in terms of sectors and geographically. While not included in this model, one could also consider imperfect mobility of workers across tasks. It also creates the possibility of workers remaining in a sector even though doing so makes them worse off: a fall in their sector’s relative price lowers labor demand, but migration costs mean that not all workers move to other sectors, so those who remain receive a lower wage. Finally, there is the possibility of exit from the labor market, or alternative choices in economic environments where that option is constrained, which is driven both by market circumstances and the availability of fiscal transfers that replace lost wages in the absence of employment. In the model, welfare—either for each segmented region or for the aggregate economy—is based on a consumption measure. This approach, combined with the mechanisms just mentioned, means that everyone in an economy benefits from lower import prices when trade costs fall, but for some, that benefit can be offset by lost labor income, either because their wages are lower, or they exit the labor market. So the aggregate economic effect of a trade costs change depends on summing all of these impacts across a full country, and can be positive or negative depending on variables like the ease of movement across sectors and geographical location, the pattern of trade costs changes, and the economy’s pattern of comparative advantage. Caliendo et al. (2019) consider a shock to the US labor market, namely an increase in imports from China following that country’s accession to the WTO. How might the analysis differ in the context of low and middle-income countries? The flowcharts in Figures 1 and 2, which summarize the linkages running from a change in trade costs to labor market outcomes, capture general features of any economy in a dynamic general equilibrium framework. As such, they are just as applicable to low and middle-income economy settings as to high-income economy ones. But the way in which the forces play out at a disaggregated level can be different in both types of countries, in particular because factors related to the ease of migration, as well as the extent of wage differences, may differ depending on an 7 economy’s structure. In addition, Figure 2 identified a three-way choice facing workers: migrate, stay, or exit. But in many low and middle-income countries, there are also other choices: enter the informal labor market, or consider self-employment. These choices—indicated by dotted lines in Figure 2— can have important social implications because working conditions can differ substantially between the formal and informal sectors. 2.2.2 Contrasting Country Experiences More concretely, there is evidence that the local labor market effects of increased import competition are not uniform across countries, and that local conditions—labor market institutions, patterns of specialization in sectors and tasks, and available margins of adjustment—lead to substantially different outcomes. However, the local labor market impacts of trade integration remain relatively understudied in low and middle-income environments, in part due to difficulties collecting the necessary data. To highlight how different outcomes can be even when the shock considered is similar in historical terms, it is useful to compare estimates of the impact of the “China Shock” from the US, Peru, and Ethiopia. These three countries importantly differ in their type of trade integration, with Peru specializing in commodity exports, Ethiopia in limited manufacturing GVCs, and the US in innovative GVC activities (World Bank 2020a). The discussion focuses on comparable reduced form estimates and leaves Caliendo et al. (2019) to one side, as it is a structural model. For the US, Acemoglu et al. (2016) estimate that US job losses from Chinese import competition over the 1999 through 2011 period were of the order of 2.0 million to 2.4 million people. 4 By contrast, Mansour et al. (2022) show that in the case of Peru, increased import competition from China did not lead overall to decreases in employment or informality for men, but did have adverse impacts for women because they tended to move into the non-tradable sector, where wages were lower, or leave the labor force entirely. 5 Even more strikingly, Ngoma (2022) finds that a one percentage point increase in Chinese import penetration in Ethiopia leads to a 15.2% increase in industry employment, because the dominant effect is to boost industrial performance by facilitating access to intermediates. The latter finding is in line with Ethiopia’s participation in limited manufacturing GVCs. Similarly, several studies for low- and middle-income countries have examined the local labor market effects of exports rather than import competition using a comparable specification. Recent studies for China and Vietnam find a positive effect of exports on labor market outcomes at the subnational level (Erten and Leight 2019; McCaig 2011, McCaig and Pavcnik 2018). Other papers find that increasing exports per worker result in higher wages for workers generally and in declining informality in India and Sri Lanka (Artuc et al. 2019) and Bangladesh (Robertson et al. 2020). Vazquez and Winkler (2023) study the employment, income, migration and informality channels across Mexican municipalities and find that expanding exports per worker increased labor force participation, but not employment rates. Exports also raised total labor incomes, but not average labor incomes. These results imply a growing labor supply which they attribute to increased immigration and movement from the informal to the formal labor market following the export expansion. In contrast, a recent study on the Arab Republic 4 Even within high income countries, there is evidence of substantially different effects from the “China shock”: Dauth et al. (2017) find no significant impact on manufacturing employment in Germany, while De Lyon and Pessoa (2021) find negative effects for the UK, and Citino and Linarello (2002) find evidence of negative effects at a local level, but only small aggregate impacts. In the case of Japan, Hayakawa et al. (2021) find that the type of trade matters: there are negative effects of import penetration in the case of product-level competition, but positive effects of import penetration in upstream (input producing) industries. 5 This evidence squares up well with the findings of Ben Yahmed and Bombarda (2020) for Mexico, who show that tariff cuts increased the probability of working formally for both men and women in manufacturing, but led to a decrease in formality for women working in service sectors. 8 of Egypt finds no significant effect of rising exports on wages, informality, and female labor force participation at the local level (Robertson et al. 2021). Again, the country sample varies in terms of its type of trade integration, with countries specialized in commodities faring worse (Egypt) than countries participating in limited manufacturing (Bangladesh, Sri Lanka, Vietnam) or advanced manufacturing and services GVCs (China, India, Mexico). Clearly, when analyzing the causal pathways from increased trade integration to labor market effects, the type of trade integration matters as do labor market institutions and occupational composition of labor. These examples clearly demonstrate that while high-income countries are more studied because of the availability of data, there is significant potential for outcomes in low and middle-income countries to diverge from that benchmark in important ways. A key objective of the remainder of this paper is to provide a framework for thinking about this disparate set of results, within the overall set of economic mechanisms identified above. 2.3 Bringing It All Together: Thinking about Impacts and Policies The two previous sections have described the sometimes complex economic mechanisms that link trade cost changes and labor markets, and have highlighted the potential for disparate outcomes both at a disaggregated level within countries, and between countries, due to differences in a range of economic and social factors. How can that information translate into an awareness of the types of policies that can help promote aggregate economic welfare, while paying due attention to the kinds of disaggregated economic losses governments need to mitigate? A first guiding principle is that trade policy is typically only one factor, and potentially not the most important one, when it comes to understanding overall labor market performance. A well-known principle in trade policy analysis is that policy responses should be targeted as closely as possible to the locus of any dysfunction (e.g., Corden, 1997). Thus, if labor market slack is due to a country’s macroeconomic conditions, the appropriate policy response is likely some mix of fiscal and monetary policy, not, in the first instance, trade policy. Similarly, if disaggregated labor market losses can be attributed to difficulties in moving across sectors or places, then the best policy response is to take action to support those movements and make them easier, which again is not usually part of trade policy as such. Second, the key issue for managing the disaggregated economics of trade and jobs is a human one: when people lose jobs, or see reduced incomes, they suffer losses that may be significant and durable. Those losses are all the more important in low-income environments, where they could translate into changes in poverty rates, or have adverse implications for broad development efforts under the UN Sustainable Development Goals (SDGs). As such, the first part of a broader agenda for managing the intersection between trade and jobs is to focus on the human dimension: that is, policies for people. Section 3 looks at the sorts of policies that can support people through labor market transitions, as well as ensure that trade costs changes do not cause undue hardship. It distinguishes between policies mitigating losses for workers and polices facilitating movement of workers. Third, the analysis in the previous section makes clear that the effects of trade costs are primarily felt, in an economic sense, at the level of sectors and locations. Aggregate impacts are necessarily net, the result of adding up disaggregated gains and losses. But realities in affected sectors and locations are just as much about gross effects, that is to say, the sizes of both sides of the ledger, and what they imply for movement and displacement. As such, there is a role for discussing policies that are specific to sectors and places, as part of a broad-based response to the trade and jobs linkages discussed here. Sections 4 and 5 discuss these areas. 9 In particular, there are implications for trade policy itself. While trade costs changes can be disruptive for labor markets at a disaggregated level, there may be ways in which those changes can be designed so as to proactively manage anticipated disruptions. Such management could either involve the use of particular types of policy instruments, or the phasing of changes in time so as to facilitate less painful adjustment. While trade liberalization lowers trade costs, its fundamental role in the context of trade and labor market linkages is to initiate a sectoral reallocation process via changes in relative prices (as illustrated in section 2.2.1). This process of sectoral reallocation can be supported by complementary policies. However, deploying policies for people, sectors, and places often implies real resource costs. So it is important to think about where those resources come from. A related concern relates to the link between raising government revenue in some low- and middle-income environments, and the role of trade taxes as part of that effort. Indeed, there can be an important tension: while there are important economic gains from lowering trade taxes, for example, doing so creates an enhanced need for potentially costly complementary policies; but if a country relies to a substantial extent on trade taxes to finance government spending, then the trade tax reform may undercut its ability to provide much- needed services. So it is important to look at the links between trade costs reform, jobs, and fiscal policy space. Both issues are discussed in detail as part of policies for sectors in section 4. Policies for places emphasize a two-fold role of places when linking trade cost changes to labor market outcomes. First, the institutional quality of a location determines its competitiveness, including its ability to attract domestic and foreign firms that export as part of the sector reallocation process. Even when all regions in a country have expanding export sectors, there is spatial disparity across regions in terms of their attractiveness to exporting firms. Higher institutional quality not only attracts capital, but also workers seeking jobs in these firms. Second, a location’s connectivity to other marketplaces – domestic and abroad – determines how smoothly goods and services can be traded. Both types of policy also lower trade costs. In summary, policies discussed in this framework focus on movement – of traded goods and services (connectivity), of workers (labor market flexibility) and of capital (foreign direct investment) – but also supporting sector reallocation by lowering trade costs, putting in place complementary policies and ensuring a high quality of institutions. From a worker perspective, however, the most immediate type of policy is the labor support extended to workers to mitigate the losses they are facing. Before moving to details in the following sections, Figure 3 presents a schematic of a general policy framework based on these considerations. In working through the various policy dimensions, however, it is important to recognize two important limitations of existing work. First, the mechanisms of trade are well understood, but quantification of labor market linkages has focused on high income countries, particularly the US. The reason for this focus is largely pragmatic and related to access to data. The policy questions raised for developing countries are no less important, but deployment of newly developed tools has lagged. This limitation has consequences for understanding which policies to prioritize. While for some policy measures there is some evidence to differentiate by types of countries, including their level of development, world region, or type of trade specialization, in other cases reasoning necessarily proceeds from first principles. This, however, helps identify key areas where further research in developing countries is required. Second, and related to the first point, academic work on trade and labor tends to focus on relatively straightforward trade policy changes, such as liberalization of tariffs. This approach helps make mechanisms clear, but does not provide direct evidence on the labor market effects of complementary policy changes that reduce trade costs, such as rationalizing NTMs, FDI or competition policy at the 10 sector level, or reducing trade costs using other types of policy interventions such as institutional quality and connectivity at the local level. Figure 3: A policy framework for leveraging trade for more and better job opportunities. Source: Authors. The approach in the following sections is therefore to identify parameters and mechanisms that are important determinants of outcomes in Figures 1 and 2 above, and to consider policies that could conceivably have an impact on them. Evidence on the effectiveness of complementary policies is limited, in particular in developing countries, because it is challenging from a technical point of view to include them in sophisticated structural models. This is a point that future work urgently needs to address. 3 POLICIES FOR PEOPLE: LABOR SUPPORT AND LABOR MARKET FLEXIBILITY 3.1 Labor Support to Mitigate Losses for Workers Changes in the labor market affect people: their livelihoods, well-being, location, and occupation. While trade can have positive and negative effects both at the local and aggregate levels, it always causes dislocation because the power of trade integration as an economic mechanism stems exactly from its ability to reallocate economic resources—including workers. Since trade can have human consequences, and they can sometimes be negative, it should follow automatically that trade reforms should be accompanied by policies that aim to both facilitate beneficial movements of people and avoid the most negative human consequences of lost employment or lower wages. It is uncontroversial that trade generates gains and losses at an individual level. Indeed the result is so widely believed that it is referred to as a theorem in economics, stating that under certain assumptions – such as perfect labor mobility –changes in relative prices lead to stronger benefit in the real wages of those workers in production (or factors of production more broadly) who are more intensively used in the expanding sector, while other workers (factors of production) who are less intensively used see their real wages fall (Stolper and Samuelson, 1941). Under the more realistic scenario of imperfect labor mobility, such adjustment costs can also translate into job rather than income losses or – in the 11 case of low- and middle-income countries –increased informality. Imperfect labor mobility is prevalent due to frictions that arise from various sources, including housing and land use policies. Traditionally, trade economists have focused on the compensation principle—the idea that the gains from trade are larger than the losses—to argue in favor of some sort of economic transfer in favor of those who lose. But practice has typically lagged behind. The US is one of relatively few economies that has a dedicated Trade Adjustment Assistance policy, which provides special allowances to those negatively affected by trade. But as the literature on the China trade shock suggests, it has proved inadequate to recent challenges. Indeed, a formal evaluation of the program found that while it was effective in some dimensions, participants still generally suffered some negative effects years after the change in trade policy (SPR and Mathematica, 2012). More broadly, the fact that some workers lose jobs or wages due to changes in trade costs reinforces the need for a social safety net (see e.g., World Bank, 2009a, for an overview focusing on low and middle-income countries). The form that such a set of policies takes naturally varies across countries. But the general idea is to ensure that public funds supply some level of resources to those unable to find work, including when the reason is related to displacement due to increased trade integration. Direct payments are common in high income countries, while India, for example, relies more heavily on guaranteed work schemes in rural areas. Whatever the mechanism, the objective in relation to trade is the same: using the fiscal system to move economic resources from those who gain from trade to those who lose. Of course, in low and middle-income countries where fiscal capacity is greatly constrained, it may be very challenging to develop such systems, with the result that people rely more on informal networks or family relationships as a safety net. Social institutions can play an important role in determining which social groups bear the brunt of the losses associated with increased trade integration. From an inclusion standpoint, it is important to ensure that laws and regulations favor the employment of historically marginalized groups, including women and ethnic minorities (e.g., OECD, 2021). Beyond legal mechanisms, of course, social practices loom large in decisions to exit the labor market, or to seek employment in the informal sector. Changing attitudes requires a long-term approach, and economic measures are only part of the equation. But in the short term, active enforcement of rules prohibiting discrimination in employment can help ensure more equitable sharing of the burdens associated with the losses from trade. A final aspect of this side of the problem is that trade can put downward pressure on wages and working conditions more broadly in sectors and locations that experience intensified import competition. While labor market rules and institutions differ substantially across countries, it is important to ensure that pressure is kept within socially acceptable boundaries, through the enforcement of social standards such as minimum wages and regulations governing working hours, safety, and other employment conditions. For countries with well-established systems, the emphasis is more on enforcement of existing rules. But for countries with only basic systems, or those where conflicting practices are not always detected and sanctioned, the emphasis needs to be on developing core labor protections as embodied in the ILO’s Fundamental Conventions (see Brown, 2000, for an overview of the issues). Education and retraining are economically and socially good in their own right. Education shapes the supply of skills and therefore comparative advantage. Countries that are relatively more endowed with human capital have a comparative advantage in knowledge-intensive sectors, and thus show higher backward and forward GVC participation (Fernandes et al. 2022). But education spending may also act as a deterrent for inward migration. Evidence for Mexico suggests that export expansion over the period 2004-14 led to lower inward migration in regions with higher spending on education. It may 12 be possible that the availability of high wage labor in a region acts as a deterrent for inward migration, given the evidence that trade has predominantly attracted low wage migrants (Vazquez and Winkler 2023). Policy interventions to promote education and skills development look different depending on country features, including the general level of education. Investing in basic education is often a priority of lower income environments, while in middle income countries, the emphasis is more often on higher education and vocational training. Enabling workers to access these kinds of programs mid- career is a crucial part of ensuring that they can be economically mobile. Retraining or reskilling initiatives, on the other hand, help mitigate disaggregated job or wage losses after a trade shock and thus accentuate the aggregate gains from trade. For instance, preliminary estimates suggest that retraining programs in Germany increased the gains from trade by 7% in the aggregate (Conwell and Williams 2022). From a labor market perspective, access to opportunities to retrain and reskill are one kind of active labor market policy (Levy Yeyati et al., 2019). This set of policies together is an important part of the toolbox that can support a resilient labor market, including in the face of shocks related to changes in trade policy. While the mix of active policies naturally varies across countries according both to fiscal space and regulatory capacity, measures like public interventions to assist matching between employers and employees, provision of incentives to people to seek new employment, or time-limited subsidies for hiring unemployed workers, can help boost the employment prospects of those displaced by trade. From a social perspective, policies to protect people from the negative consequences of trade cost changes, and to allow them to take full advantage of the benefits, have much to recommend them. Even from a strictly economic point of view, the level of distortion involved in these policies is typically minimal, though the use of some measures, like employment subsidies, needs to be carefully monitored. Consequently, this set of policies should be the starting point for dealing with labor market disruptions linked to trade integration, both because they are closest to the locus of the problem—the labor market—and because they can achieve major social goods at relatively low economic cost. The need to mobilize real economic resources to support some of these policies is discussed further below, in terms of the relationship between trade policy reform and fiscal space. 3.2 Facilitating Movement of Workers It is not enough to ensure that systems are in place to mitigate the negative effects of trade. Trade can also have positive effects on labor markets, either locally or in the aggregate. But to do so, it again depends on the operation of a reallocation mechanism. First, competitive firms have to be able to grow: expanding exports implies greater demand for labor. Second, people have to be able to move relatively easily from one place and sector to another. At a finer level of detail, people also migrate from one activity or occupation to another as economies undergo structural change (Kruse et al., 2023). Facilitating these movements is an important part of the set of complementary policies required to ensure that trade integration produces beneficial labor market outcomes. This section discusses two policy options to support this goal, (i) labor market flexibility, and (ii) mobility support. Restrictive labor market regulations can inhibit the reallocation of workers. For instance, Hasan et al. (2007) show that labor demand elasticities are larger for Indian states where labor regulations are more flexible. Specifically, the extent of labor market flexibility determines how quickly firms and workers react to trade shocks in terms of adjustments in the quantity of workers. Evidence for Mexico over the period 2004-14 finds that the quantity of labor reacted more strongly to export expansion in municipalities with high labor market flexibility, as suggested by the larger positive effect on the number of immigrants. Contrasting these findings, the price of labor, as measured by average labor incomes, 13 reacts more strongly in municipalities with low labor market flexibility (Vazquez and Winkler, 2023). Combining labor market flexibility with the types of labor support discussed in the previous section has shown to matter for trade-and-labor linkages. In a sample of 15 OECD countries, a higher share of manufacturing imports from low- and middle-income countries between 1991 and 2008 was associated with more positive labor market outcomes in countries following the “Flexicurity” and “Rhineland” models characterized by higher labor market flexibility and high labor support, compared to countries with low labor support (Anglo-American, Mediterranean, East Asian models) (Milberg and Winkler, 2011). Promoting mobility of people within the economy is a multi-faceted problem. While it implies some degree of labor market flexibility, the problem is by no means solved just by loosening regulations around hiring and firing. It is also important to look at issues like portability of social benefits such as health insurance and retirement savings across jobs, and across administrative boundaries within countries (cf. Holzmann, 2018, on portability in the context of international migration). At a deeper level, workers need to be equipped with skills and training that enable them to adapt to a new work environment. In addition, a range of policies affect the willingness and ability of workers to change geographical location, not least of which are housing and land use regulations, as well as factors such as occupational licensing. The changes involved can be profound, in particular in cases where movement is not just from one occupation to another or one sector to another, but across economic aggregates, such as from industry to services. 4 POLICIES FOR SECTORS: TRADE AND COMPLEMENTARY POLICIES The next policy dimension in which specific measures may be necessary to deal with the labor market effects of trade integration is sectoral, that is to say, policies that are targeted at specific sectors. Of the policies discussed thus far, these pose the greatest risk from a purely economic standpoint: sector- specificity encourages lobbying by affected industries, which makes it difficult for the government to focus on overall economic and social welfare. Similarly, once put in place, it can be difficult to undo sector-specific policies because of these kinds of political economy effects. Nonetheless, there is evidence that at least in environments with strong governance—where budget constraints are real, and where commitments to future withdrawal of measures are credible—sector-specific policies can be effective in some dimensions (e.g., Choi and Levchenko, 2021). The following section looks at the role of trade policy in reducing trade costs, while section 4.2 focuses on complementary policies which aim at reducing uncertainty, thereby also lowering trade costs. 4.1 Trade Policy to Reduce Trade Costs The first part of the causal chain running from trade costs to labor market disruption is, of course, a trade cost change itself. From a policy perspective, the issue is not about how to move a country to a “textbook perfect” trade policy, but about how to unlock the largest possible boost in aggregate economic welfare for a given investment of political capital. This is necessary because a reduction of trade costs unleashes potentially strong reallocation effects and adjustment costs, which induce firms, workers, and consumers to organize and lobby in favor of or against the change. Against that background, what are some general policy approaches that could help low and middle-income countries obtain significant aggregate economic benefits from trade policy reforms, at the same time as managing the disaggregated losses due to labor market disruptions? The following discussion first introduces different types of trade policy (section 4.1.1), followed by a discussion of how trade policy can be designed to manage labor market outcomes (section 4.1.2). The final sub-section highlights the tension between trade policy and fiscal space (section 4.1.3). 14 4.1.1 Different Types of Trade Policy Liberal trade policies, including reductions in tariffs and NTMs, and engagement in preferential trade agreements (PTAs), help reduce trade costs, and thus enlarge effective market size (World Bank 2020a). While the positive trade effects of reduced trade barriers on imports and exports have been well- documented, the gains can be even larger in a GVC world where value-added crosses multiple borders, as they reduce the price of both exported goods and input costs (e.g., Yi 2003, 2010; Caliendo and Parro 2015, Antràs and DeGortari 2020, Fernandes et al. 2022). Reduced tariffs on intermediates have also been shown to raise export performance at the firm level (Bas and Strauss-Kahn 2015; Pierola et al. 2018). Similarly, reciprocal tariff reductions by partner countries reduce the costs of exporting. Liberalizing trade is about more than just tariffs, however. Recent evidence for Indonesia suggests that higher exposure to selected protectionist NTMs significantly reduces the survival probability of firms in export markets and negatively affects exporters’ extensive and intensive margins. Importantly, the negative effects of NTMs are larger than for import tariffs (Calì and Montfaucon 2021). However, other research on Indonesia emphasizes that some NTMs on imported inputs like anti pest treatments have beneficial welfare effects, while protectionist NTMs such as pre-shipment inspections have negative ones (Calì et al. 2021). Reforming non-tariff measures is often more important in creating welfare gains, as evidenced by recent estimates for the African Continental Free Trade Area (AfCFTA) using a Computable General Equilibrium (CGE) approach. In fact, tariff reductions are estimated to raise the welfare of AfCFTA member countries by 0.2 percent, whereas lowering NTMs in goods and services by half would increase it by 1.6 percent (World Bank 2020b). The estimated gains are even larger accounting for improvements in trade facilitation (see section 5.2). In addition, deep PTAs have been shown to enhance GVC participation, as they also include policy areas such as movement of capital, investment, visas, competition and intellectual property rights – beyond mere market access (see, e.g., Orefice and Rocha 2014, Kowalski et al. 2015, Johnson and Noguera 2017, and Laget et al. 2018). Facilitating imports of intermediates is not only about goods: it is also important to have access to producer services, such as transport, finance, and research and development. Where the local economy does not have comparative advantage in these areas, as is common in low and middle-income countries, it is advantageous to import these services with a view to upgrading productive activities in other parts of the economy. There is evidence that liberalizing services policies can produce significant benefits for manufacturing exports within subnational regions (Hoekman and Shepherd, 2017) and for the productivity of manufacturing firms downstream (Arnold et al. 2011, Arnold et al. 2015), which is important in terms of helping the economy generate employment gains following a trade policy reform. Rodrik (2022) goes further, to argue in favor of an industrial policy for services, focusing in particular on support for good quality jobs in that sector. The latter point illustrates that trade policy has a sector-specific component with implications for labor market outcomes. Lowering tariffs and NTMs in sectors that play a strong role as input suppliers for other parts of the economy can boost the gains from trade liberalization. The rationale for this approach is that, as shown by Ngoma (2022), improved access to inputs can drive productivity growth in sectors using these imported inputs downstream, which in turn increases demand for labor. Since imported intermediates are often not perfect substitutes for locally produced varieties, the displacement effects of trade liberalization in intermediates are not as strong as they are for final goods, where substitution effects may be stronger. Recent evidence suggests that the impact of imported inputs on labor market outcomes is more favorable than that of final imports, both at the cross- 15 country sectoral level (Winkler et al., 2023) and at the municipality level for Mexico (Vazquez and Winkler, 2023). However, countries benefit differently from trade liberalization. Firm-level evidence for China and Bangladesh, for instance, suggests that liberal trade policies have been an important factor for upgrading along GVCs (Kee 2015, Kee and Tang 2016). The expected trade gains are even larger for countries with high trade protection, especially in South Asia, but also in Sub-Saharan Africa (SSA) and the Middle East. These also tend to be the regions having a lower reliance of imported inputs for export production (World Bank 2020a). Evidence for the food sector confirms that while a reduction in import tariffs increases GVC participation in SSA countries as well as globally, the gains are stronger for SSA countries (Balie et al. 2019). Such differences also have implications for labor market outcomes. An event study suggests that wages increased most strongly in countries after they entered limited manufacturing GVCs compared to other forms of upgrading (World Bank, 2020). 4.1.2 Trade Policy Design to Manage Labor Market Disruptions Besides a country’s type of trade protection and trade specialization there are aspects of the design and implementation of trade policy that can have labor market implications. Managing trade policy reform appropriately is an important part of an overall recipe for reconciling desirable economic ends—increased efficiency due to lower trade barriers—with the need for such changes to be politically sustainable, in the sense of not causing so much displacement and individual loss that attitudes become entrenched against reform, or even in favor of undoing it. This section discusses four areas of trade policy design that policy makers can influence, namely (i) prioritizing reform sectors, (ii) actively facilitating growth of the export sector, (iii) introducing reform gradually, and (iv) allowing for corrective action. A fundamental question for any policy maker to ask is what kind of trade policy could both unlock significant aggregate economic gains and at the same time limit localized economic losses? The research cited above highlighted the importance of access to imported intermediates for trade and labor market outcomes, especially for when firms participate in GVCs. Focusing liberalization efforts on input markets can not only limit labor market adjustment costs—still an issue to the extent that there are substitutable rather than complementary inputs produced domestically—but even promote strongly positive labor market outcomes, as evidenced in Ethiopia (Ngoma 2022). A first key lesson is therefore that there may be important advantages in concentrating trade policy reform efforts on intermediate goods (and services) markets, but also on imported technology and machines required for domestic and export production more broadly. Indeed, some countries have already taken this lesson to heart through the use of tariff band systems, where intermediates and capital goods are either duty exempt or subject to lower rates. The East African Community uses such a system, for example, and while it is not free from issues due to dual-use goods and the potential for misclassification, it nonetheless has the potential to provide both a trade policy win and a boost to local labor market conditions. Secondly, and linked to the first point, it is important to ensure that as import competition intensifies pressure on local producers, potentially with negative local labor market implications, the economy is capable of shifting resources to sectors in which it has a comparative advantage, so that exporters can grow and take on more workers. From a trade policy perspective, this approach means giving some scope to reciprocal reforms in partner countries, so that reduced import barriers are exchanged for reduced export barriers abroad. But it also means that it is important to encourage investment in export sectors. While such sector-specific complementary policies are addressed in more detail below, it is important to cite examples like investment promotion (Harding and Javorcik, 2011) and export 16 promotion (Cadot et al., 2015): there is high-quality evidence that both types of policies can be beneficial in low and middle-income countries, albeit with some limitations in the case of export promotion. A third way in which the design of trade policy reforms can make movements of workers across sectors and places less disruptive is to proceed gradually, with a clearly announced timetable. In other words, rather than cutting tariffs rapidly, the government could proceed to the desired end point of reform by making a smaller proportional reduction each year during a phase-in period. In the past, external forces, such as structural adjustment, imposed rapid tariff reductions that led to significant labor displacement, in the way that the relatively rapid changes associated with China’s WTO accession led to substantial labor market issues in at least some high-income countries. Countries typically take this kind of approach with trade agreements, which phase in over a number of years. But there is a case for gradually introducing unilateral and multilateral reforms as well. That case has become more apparent in recent years as the nature and extent of labor market disruptions linked to trade policy changes has been better quantified. Finally, recognizing that trade policy reform is, in part, a learning process rather than a one-off event can help governments adjust course and take corrective action to prevent undue labor market disruption during a liberalization episode if necessary as new information comes to light. Before undertaking a reform, its impacts are subject to uncertainty, even in an environment with access to top quality data and economic modeling. It is always possible that increases in import competition are much stronger than expected, which would put more stress on local labor markets than the government intended. As part of an experience of learning about market responses, and labor market dynamics, it is potentially important for even reform-minded governments to be open to temporary pauses or even changes in course if that is necessary to preserve the political economy conditions necessary for ongoing reform. Corrective actions are easier to implement when reforms have been introduced gradually rather than abruptly, highlighting that both trade policy design elements are intertwined. At the WTO level, contingent protection measures like the Agreement on Safeguards are designed to provide a transparent, temporary, and non-discriminatory way of limiting the political economy pressures that arise when the extent of import response is misjudged in this way. However, the procedural requirements of the WTO Agreement on Safeguards are relatively strict, and jurisprudence has tended to interpret the agreement narrowly (e.g., Sykes, 2003), all of which means that low- and middle-income countries can experience difficulties in accessing this mechanism in a way consistent with WTO law. But it is also legal for low- and middle-income countries to increase tariffs above their currently applied level to some level below their bound tariff level which is a de facto safeguard. Such a step is possible because the current reality of trade policy in most countries is that there is a substantial gap between applied and bound tariffs, commonly referred to as tariff “water”. Of course, any change in applied tariffs cannot discriminate among trading partners, consistent with WTO rules. While such measures should be used judiciously, as they risk undoing productive liberalization episodes, they can also provide an important safety valve in cases where import response is unexpectedly strong and the government needs “breathing room” in a political economy sense before fully implementing desired reforms (Crowley, 2010). Of course, the principle that such measures are transparent, temporary, and non-discriminatory should also apply to these de facto measures. The U.S.-China trade war illustrates that non-alignment of trade policy reform with such principle involves high costs. Comparing the aggregate economic benefit for the US through increased trade with China from Caliendo et al. (2019) with the losses due to the trade war initiated by President Trump 17 (Fajgelbaum et al., 2020; 2021 update cited in Fajgelbaum and Khandelwal, 2022) shows that the latter actions essentially fully undid the former. 4.1.3 The Tension between Trade Policy and Fiscal Policy Space Classical trade policy reforms typically do not involve large real resource costs: lowering tariffs or liberalizing services policies does not require financial investments of comparable scale to infrastructure upgrades or fiscal transfers. But many of the policies examined in this paper, or subsumed under the heading of reducing trade costs, do not fall into that framework. Implementing active labor market policies or improving infrastructure and connectivity, as well as many other people-, sector-, and place-oriented policies, involve significant real resource costs. As such, they require the government to have sufficient fiscal space to contemplate responsibly financing the chosen policy measures. This issue is all the more salient at the present time, when necessary fiscal support measures, as well as activity-based reductions in tax takes, due to the COVID-19 pandemic, have reduced fiscal space for countries at all income levels. In high income countries, and some middle-income countries, the question of financing public spending has little if any connection to trade policy, because trade taxes are not a significant source of government revenue. 6 In terms of averages, the proportion is 6.4% in lower-middle income countries, 3% in upper middle-income countries, and 0% for high income countries (unavailable as an average for low-income countries). There are many countries, however, where the connection is much tighter. Although data availability is only partial, the World Development Indicators record eight low and middle-income countries where the proportion of trade taxes in government revenue is 20% or higher, and 35 where it is 10% or higher. The largest proportion is 37%, in Somalia. So for a significant number of countries, there can be a tension between liberalizing trade by lowering effective tax rates— which tends in turn to lower tax revenues—and the ability to put in place policies that manage the reallocation effects of that liberalization—which requires additional tax revenue. Can this tension between trade policy reform and the need for revenue be resolved, and if so, how? One approach is a broader shift in policy away from a reliance on trade taxes, towards alternative tax bases (typically income, corporate, and consumption). In theory, governments can implement a trade policy reform and still maintain their revenue if it is accompanied by a more comprehensive domestic tax reform over the long term. In reality, however, many countries have found such moves difficult to implement effectively over a long period. Evidence for a sample of 60 countries across all income levels for the period 1980-1999 found that globalization, including trade openness, led to a major decline in revenue from ‘easy to collect’ taxes including tariffs, whereas revenue from ‘hard to collect’ taxes (value-added and income) increased, although at a smaller rate. There is, however, substantial heterogeneity across countries, with countries characterized by a low institutional quality being more likely to see their net revenue decline (Aizenman and Jinjarak 2009). In addition, firms are more difficult to tax in a globalized world, as they are internationally mobile and can shift their profits to lower-tax jurisdictions. A similar logic applies to top income earners. Consequently, corporate tax rates and personal income tax rates for the top percentile declined in 65 countries over the period 1980-2007, whereas the median income tax rate rose, both of which have been attributed to rising globalization (Egger et al. 2019). This is 6 “Trade taxes” is defined by the World Development Indicators as follows: “Taxes on international trade include import duties, export duties, profits of export or import monopolies, exchange profits, and exchange taxes.” As such, it does not include value added taxes levied on imported goods. 18 especially problematic because workers face not only adjustment costs from reduced trade costs, but under this scheme also a higher tax burden. A more viable option may therefore be to prioritize trade policy reform of non-tariff measures. Although discussions of trade policy reforms often focus on tariffs, there is a wide range of non-tariff measures that also increase trade costs. For instance, rationalizing import and export licensing procedures or simplifying rules of origin in trade agreements boosts trade integration by lowering trade costs. An attractive feature of these kinds of reforms is that they boost imports without lowering tariff rates, so their net effect is to increase, not decrease, tariff revenue. If accompanied by a tariff reform, net revenue could be maintained. The lesson here is that smart reforms of non-tariff measures can help support fiscal space in countries where trade taxes account for a significant proportion of government revenue. Given that the tariff equivalents of non-tariff measures tend to be much higher than tariff rates of protection (Arvis et al., 2016), there is also an efficiency argument for focusing on non-tariff measures when prioritizing trade reforms. A relatively underexplored avenue of reform relates to tariff changes that are efficiency-enhancing but which minimize the potential for negative revenue effects. While the structure of such a reform may be complex, it would consist of some mix of maintaining rates in some areas, while reducing them in others, typically focusing on the highest rates. Detailed modeling would be required to ascertain the balance between enhanced import response (which tends to boost revenue) and decreased tariff rates (which tend to reduce revenue), but there is the possibility of designing a reform that promotes efficient reallocation across sectors, while not impinging unduly on the government’s ability to design and implement complementary policies that involve significant real resource costs. 4.2 Complementary Policies to Facilitate Sector Reallocation Trade policy reform is a necessary but not sufficient condition for successful trade integration, let alone for managing its labor market adjustment costs. Complementary policies are needed, which at the sector level could include export promotion, and investment, competition, and innovation policy, to name a few (Bartók and Miroudot 2008). While such complementary policies can lower trade costs and uncertainty, one key contribution is that they help facilitate sectoral reallocation of resources to the growing export sectors. But sectoral policy measures can also help mitigate the labor market disruptions accompanying trade policy reforms, by helping export sectors expand– either directly through export promotion (Cadot et al., 2015) or indirectly by promoting certain types of foreign investment (Harding and Javorcik, 2011). Different complementary policies at the sector level are discussed in more detail below. FDI has been identified as a significant driver of GVC participation across countries (Kowalski et al. 2015, Buelens and Tirpak 2017, Fernandes et al. 2022). FDI is especially relevant when countries are scarcely endowed with capital, technology, and knowledge (World Bank 2020a). Another study specifically measures the role of FDI restrictiveness and finds that it is associated with reduced GVC participation in low-tech manufacturing (Cheng et al. (2015). So taking steps to encourage foreign direct investment in export industries, including with a focus on particular sectors, is a way of helping expand output in those industries, which in turn increases demand for labor. That increased demand can take up part or all of the slack in the labor market created by contraction of import-competing sectors. Investment promotion can help attract certain types of foreign investment and indirectly stimulate export growth. Harding and Javorcik (2012) find that sectors that have been prioritized by investment promotion agencies in low- and middle-income countries export relatively higher-quality products. 19 Special Economic Zones (SEZs) have often been put in place as a means of both investment and export promotion. SEZs have been shown to attract export-oriented foreign investors, stimulate the growth of imported inputs used for export processing, and to create jobs. However, SEZs are typically locally confined and have resisted linkages to domestic suppliers and thus indirect job creation, because imports are duty-free (Farole and Winkler, 2014). The evidence on export promotion policies is more mixed in a dynamic sense (Cadot et al., 2015). Chile’s export success in the agribusiness GVC (e.g. salmon, wine, horticulture) since the early 1990s can be attributed to the strong role of its export promotion agency, which is the Chilean Trade Commission or ProChile. Another example is Japan’s External Trade Organization which also successfully promoted exports partly because of its emphasis on providing market research information to Japanese firms (Taglioni and Winkler, 2016). Competition policy aims at overcoming anticompetitive practices in a country that can distort relative prices. They therefore also bias the sectoral reallocation process (or allocative efficiency) because of market exit of the least productive firms, low market entry of more productive firms and reduced movement of productive resources across firms and sectors. Such practices consequently also affect the distribution of the welfare gains from trade. Anticompetitive practices include monopolistic, duopolistic and oligopolistic market structures, but also government subsidies or other regulations that distort prices and markets. For instance, subsidies in a sector can inhibit factors of production (capital and workers) from reallocating to export sectors of higher comparative advantage. Strong competition in a sector or location can thus magnify the effectiveness of other policies that facilitate sector relocation such as the movement of workers or investment promotion. On the other hand, Autor et al. (2020) suggest that if the sector reallocation process favors the most productive firms within sectors, product market concentration increases as sectors become dominated by superstar firms. As these superstar firms are more capital-intensive, this affects the labor share in value added. Finally, this section concludes with policies that stimulate innovation. There are many ways to support innovation, for instance through subsidies to research and development, which is crucial to unlocking the type of growth that can move countries from middle to high-income status. These kinds of subsidies are not strictly sector-specific, as research and development can take place anywhere in the economy. But in practice, they tend to be most used in advanced and innovative manufacturing and services sectors where technology and design are important components of competitiveness in international markets. Innovation support also extends to facilitating access to foreign research and development (R&D) through GVCs. GVC-mediated access to foreign R&D is associated with more domestic innovation and higher local productivity spillovers. Positive effects, however, are strongly linked to capabilities of workers, firms, and institutions (Asian Development Bank et al., 2021). While this section discussed a range of complementary policies that facilitate sectoral reallocation, countries need to prioritize different policies. While reform of tariffs is relatively straightforward, the same is not true of non-tariff measures or services policies, as well as the broad range of other policies that affect trade costs: they require a higher level of technical capacity, which is sometimes scarce in countries at very low-income levels. Countries that aim to diversify their exports into limited manufacturing can therefore start with measures focusing on traditional trade policies with a view to facilitating imports of intermediates. Countries aiming to move into advanced manufacturing and services exports can focus on the broader range of complementary policies discussed above. For countries that want to become innovative, in particular the upper-middle income group, there are additional policies available that aim to foster innovation in specific sectors. 20 5 POLICIES FOR PLACES: INSTITUTIONAL QUALITY AND CONNECTIVITY Economic activity is not distributed evenly in a geographical sense: it has a natural tendency to agglomerate in particular places, and that process can be beneficial in terms of generating local spillovers (World Bank, 2009b). It is likely that trade liberalization reinforces that process, in particular for geographical areas that are relatively close to international markets (Crozet and Koenig, 2004). The analysis above, in which trade policy changes give rise to a migration decision, makes clear that there is the potential for people to move from one sector and location to another, which has direct implications for the dispersion of economic activity within countries. This also has implications for other labor market outcomes. Evidence for Mexico confirms a strong geographical concentration of trade activity in Northern Mexico, driven by GVCs, close to the U.S. border. The export expansion between 2004 and 2014 affected labor markets in the North through higher total labor incomes and increased inward migration, while it affected municipalities in Mexico’s South through higher employment and reduced informality rates. The lack of effects through the employment channel in the North is partially explained by the increase in labor supply due to larger net migration compared to the South. The absence of the employment channel in the South may be explained by fewer job opportunities and lower labor demand. The informality rate could reflect worker preference for the benefits of being formal (such as contributions to pensions or retirements) over the prospect of higher incomes (Vazquez and Winkler, 2023). On the one hand, an important part of the structural change process inherent in development is fostering increasing returns to scale and local spillovers. Agglomeration is therefore an important part of that effort, which explains why cities in low and middle-income countries have seen rapid growth over recent decades. This growth poses challenges but also opportunities, in particular in terms of job creation (World Bank, 2015). In this context, this section discusses the role of two key policies for places, namely (i) institutional quality to increase competitiveness and (ii) connectivity to facilitate movement of traded goods and services. While both facilitate the reallocation of resources – capital, and labor – towards export sectors of comparative advantage, thus creating more and higher-paid job opportunities, institutional quality has also been shown to shape comparative advantage directly and thus increase competitiveness. At the same time, a development path that is too unbalanced between cities and the hinterland risks not only generating migration pressures that are difficult to manage, but also may be less than ideally inclusive from a social point of view. So in the context of reforming trade policies, it is important to think about how those changes will play out in terms of changes in economic activity, and employment rates and conditions, across different parts of the national territory. 5.1 Institutional Quality to Increase Competitiveness Institutional quality, which includes competition policy, contract enforcement mechanisms, and the protection of property rights with the aim of increasing market efficiency and investor confidence, shapes how countries engage in trade and GVCs in the first place. Several studies confirm the role of institutional quality for comparative advantage and thus export patterns (Acemoglu et al. 2007, Levchenko 2007, Nunn 2007, Costinot 2009, and Chor 2010). While a large part is determined by institutional quality at the country level, countries with stronger institutional quality have a comparative advantage not only in exports but also in GVC participation in contract-intensive sectors. 21 The hypothetical example of Mozambique suggests that if the country increased its rule of law index to the cross-country median, its backward GVC participation would grow by 40 percent and its exports would grow by 42 percent, at the sample mean of sectoral contractual intensity (Fernandes et al. 2022). Institutional quality therefore also facilitates sector reallocation towards sectors that are more contract intensive. In addition, the role of institutional quality matters more strongly for the import content of exports (backward GVC participation) than for total exports across countries (Fernandes et al., 2022). One key factor is that GVC trade is characterized by intense firm-to-firm interactions including contracting, specialized products and investment (such as for the customization of products) and the exchange of large flows of intangibles (such as technology, intellectual property, and credit) (Antràs 2016; 2020). The protection of intangibles, such as patents, trademarks, copyrights, industrial processes, and designs, is particularly relevant for factory-less and innovative goods producers, such as Nike and Apple (Asian Development Bank et al., 2021). There is also a role for regional trade agreements, as discussed as part of countries’ trade policy options, in improving institutional quality, especially those with deep provisions in areas such as regulating competition law or services and protecting intellectual property rights. This is because such trade agreements import both reform and technical and financial assistance and have shown to strengthen GVC participation. The number of provisions in deep trade agreements increases for countries participating in more sophisticated GVCs and is highest for those specialized in innovative activities, and so does political stability (World Bank, 2020). These findings reinforce the role of institutional quality in shaping comparative advantage. 5.2 Connectivity to Facilitate Movement of Traded Goods and Services Physical and digital connectivity and infrastructure and other measures facilitating trade flows at the border or behind help lower trade costs, thus facilitating the movement of traded goods and services and lowering effective distance to key markets. High transport costs remain key obstacles to GVC participation, according to surveys of developing country suppliers (OECD and WTO, 2013). Other findings suggest that unpredictable land transport hinders most Sub-Saharan African countries from participating in the electronics GVC (Christ and Ferrantino, 2011). Better infrastructure – including communication, electricity, roads, and power – and GVC participation in manufacturing are positively associated (Cheng et al., 2015). Estimates of the ad valorem equivalent of transit delays highlight the huge potential for connectivity improvements to reduce trade costs. One day of delay in transit due to a different transport mode choice has an ad valorem equivalent of 0.6 to 2.1 percent, with trade in parts and components being the most sensitive (Hummels and Schaur, 2013). Similar magnitudes for the cost of a one-day delay in inland transit are suggested by Djankov, Freund, and Pham (2010). Similarly, research finds that logistics performance matters more strongly for trade in parts and components than for trade in final goods (Ansón et al., 2017). Improving trade facilitation by streamlining border procedures boosts trade integration by lowering trade costs. Estimates suggest that one third of the observed increase in GVC trade—which focuses on the role of intermediates—over the 2015 to 2019 period can be attributed to improvements in trade facilitation (Shepherd 2022). These improvements focused on increased implementation of the WTO Agreement on Trade Facilitation, emphasizing the potential to move forward on trade facilitation measures in order to promote access to imported intermediates. 22 Connectivity extends beyond the physical supply chain of goods to effective communication between the participants in GVCs, including technology and linguistic proximity. Information and communication technologies link markets in distant locations by providing real time information for participants. They can also facilitate transactions through digital marketplaces (e.g., ITC, 2020). Internet access at the firm level has been shown to determine export performance (Fernandes et al., 2019). Provided the necessary infrastructure is in place—laws to facilitate online transactions, internet access, and access to formal financial services—they can enable producers to ship goods to distant consumers, either within or outside the national territory. As such, digital marketplaces have potential to help moderate the necessary agglomeration effects referred to above, by contributing in the reverse sense, to some degree of dispersion in a geographical sense. But linguistic proximity, such as use of the English language, also matters for bilateral GVC links (Buelens and Tirpak 2017, Ignatenko et al. 2019), especially for business process outsourcing services such as information and communication, marketing, after-sales or other services. This part of the place agenda overlaps with the people agenda. Extending access to technology is only worthwhile if people have the education and skills necessary to make use of it. Building capacity in the area of technology use—from basic computer literacy to website design, digital payments, and the ability to use digital platforms—is therefore an important emerging area of education and skills training policy in countries at all income levels. While the barriers to moving forward in this area are substantial in low-income environments, higher income developing countries, particularly larger ones like China and India, have shown that it is possible to make rapid progress in this area. ITC (2020) also shows that the digital marketplace environment is very active in Africa, so this policy domain is clearly one that is of potential interest to a wide range of low and middle-income countries. In terms of labor market effects, one important aspect of the digital marketplace revolution is that it substantially reduces trade costs facing micro, small, and medium enterprises (Lendle et al., 2016). As such, it makes it easy for these small companies—sometimes consisting of just one person—to move goods and services within the national economy, and internationally. In terms of the migration options for workers (section 3.2), it is necessary to add a fourth category: self-employment or entrepreneurship. Integrating markets by reducing trade costs, including through the use of technology, can make it easier for those facing negative labor market pressures from trade policy changes to move into solo- entrepreneurship, or small-scale business. While additional training may be necessary, there is clear scope for complementarities between policies for people in that area, and place-related policies that emphasize tighter geographical integration of markets through the use of digital technologies. In summary, improving infrastructure and connectivity—which has significant real resource costs, and which therefore requires fiscal space—is an important way of both improving the access of well-placed regions to world markets, and enhancing the positive effects of trade, but also improving linkages between people and markets in those places and the national hinterland. In low and lower-middle- income environments that aim to strengthen their manufacturing exports and GVC participation, the emphasis is typically on developing physical infrastructure, along with the regulatory “software” required to make it work efficiently. But in upper middle-income countries targeting exports of more advanced manufacturing goods and services, there is more of an emphasis on development of advanced logistics capacities, such as cold chain management and express delivery, with an emphasis on service quality and reliability (World Bank 2018, 2020). 23 6 CONCLUSION This paper has shown that trade and labor markets are intimately connected. While labor market functioning is an important factor in a country’s ability to respond to trade opportunities or shocks, there can also be potentially important labor market disruptions when trade integration changes. The problem for governments of countries at all income levels is to reconcile the desire for aggregate economic gains from increased trade integration with the need to limit disaggregated labor market losses and disruptions. There is no single solution to this problem. The answer depends on country circumstances, and should be the outcome of a deliberative process in accordance with national institutions. The purpose of this paper is not to provide a “silver bullet” answer, but rather to set out a framework for thinking through the issues, and to identify aspects of particular policies that may be useful to governments in various settings in resolving this inherent tension. The first part of a broader agenda for managing the intersection between trade and jobs is to focus on the human dimension: that is, policies for people. As the effects of trade costs are primarily felt, in an economic sense, at the level of sectors and locations, there is a role for discussing policies that are specific to sectors and places. Policies discussed in this framework focus on movement – of traded goods and services (connectivity), of workers (labor market flexibility), and of capital (foreign direct investment) – but also supporting sector reallocation by lowering trade costs, putting in place complementary policies and ensuring a high quality of institutions. From a worker perspective, however, the most immediate type of policy is the labor support extended to workers to mitigate the losses they are facing. In terms of policies for people, it is important to recognize that the human costs of local labor market disruptions can be both real and large. As such, social safety nets are important, as are active labor market policies such as retraining, education, and reskilling. Trade policy creates an environment in which substantial numbers of people can have an incentive to move across occupations or tasks, sectors, and places, so it is important that such movements not be unduly restricted. While a flexible labor market may be helpful, this recommendation is distinct from the standard flexibility formula. It emphasizes measures like portability of social benefits, and development of human capital, rather than just facilitating hiring and firing. Similarly, policies to ensure that changes in employment do not have particularly negative effects for historically marginalized groups, or that competitive pressures do not turn into declines in working conditions and safety, are important from a social point of view. Sectoral policies have the highest potential to introduce economic distortions, or to be difficult to unwind in a dynamic sense, so they should be used judiciously. From a trade policy perspective, the paper’s conclusions can be summarized from three directions. 1. First, in designing and implementing trade policy reforms, it is important to focus in the first instance on input markets. Doing so limits labor market disruption, due to lesser substitution effects. But it also gives a competitive boost to downstream producers and exporters, which can take up part or all of the labor market slack generated by increased competition from imported intermediates. 2. The second aspect of trade policy design that is important is that it should proceed gradually, potentially even with standstills or temporary rollbacks. The rationale for this approach is that it is better to maintain the political economy consensus for reform over the long term than to adopt a “big bang” approach to reform that is subsequently undermined. Low- and middle- income countries typically have substantial tariff “water” that they can use to design de facto 24 safeguard measures to deal with import surges, provided that they do so in a non- discriminatory way. 3. Third, trade taxes are part of revenue policy for some low- and middle-income countries. So lowering them can decrease the government’s fiscal space at a time when it may need access to a range of complementary policies that can involve real resource costs. Renewed focus should be given to reform of non-tariff measures, which are typically revenue positive, as well as design of revenue neutral but economically beneficial tariff reforms. As a general rule, however, trade policy is not the most appropriate way of managing major labor market disruptions. Following the theory of domestic divergences (Corden, 1997), the most effective and efficient solution is likely to be as close to the locus of the problem as possible. That is why this paper recommends a suite of complementary policies focusing on labor markets themselves, to a large extent. The mix of individual policy measures and the extent to which they are specific or general will of course vary from one country to another depending on national circumstances. Having said that, there is evidence that investment and export promotion, competition policy, and innovation policy can be effective ways of supporting the sectoral reallocation process. While support for research and development is not sector specific in a legal sense, it often has that aspect in a de facto sense. Since economic activity is not spread evenly across the national territory, and political economy has a strong territorial link, it is important to think about the role of place in designing policy as well. On the one hand, agglomeration economies—which drive the growth of cities—need to be fostered with a view to taking full advantage of the opportunities trade offers. This critically hinges on a location’s institutional quality, including its ability to attract domestic and foreign firms that export as part of the sector reallocation process. But increasing connectivity, both between key cities and overseas markets, and between those cities and the hinterland, is also an important part of integrating markets and allowing distant producers and consumers to trade goods and services. Promoting technology— rollout but also literacy—can help with this issue, as digital marketplaces reduce trade costs in particular for small traders. Clearly, countries at different income levels and with different economic structures will need to prioritize different policies from among those listed. The key concern is to ensure that both trade policy and complementary policies are well aligned with the economic mechanisms set out in the paper’s framework. This alignment can help design policies that are both effective in achieving their economic and social objectives, and efficient in the sense of doing so at minimum economic cost. Having said this, it is possible to give some indications as to the prioritization of measures in different contexts. Given the real resource costs involved in many of the policy interventions discussed above, the key variable for delineating groups of countries is fiscal space. 1. For countries with limited fiscal capacity and/or limited space, the starting point is reform of non-tariff measures in particular in input markets (goods and services) and restrictions to imported services. Complementary sectoral policies to implement could focus on competition policy to overcome barriers to sectoral reallocation. This includes opening protected markets to stronger competition, such as removing government subsidies and distortions that regulate the number of firms and access to inputs, among others. From a labor market perspective, those countries can take steps to implement basic social protections, focusing on measures like enforcement of basic labor standards and minimum wages, which do not have direct fiscal costs. Similarly, they can remove barriers to mobility across places and sectors. From a human capital standpoint, they will typically need to invest in basic education. 25 2. For countries with moderate fiscal capacity and/or moderate space, the above measures remain important. But they can add to them in a number of ways. From a trade policy standpoint, they can seek to engage in deep trade agreements and potentially introduce complementary export and investment promotion and competition policy in a limited way. While the latter policy measures help attract capital from abroad, these often take the form of fiscal incentives with implications for government revenue. Investment attraction also requires basic political stability and institutional quality. Complementary competition policy can emphasize reducing the role of state-owned enterprises which distort prices and markets. Labor market priorities would include more comprehensive active labor market policies, such as commitments to higher labor standards, offering worker training, supporting the self- employed and start-ups, and helping workers with job search. Building human capital by ensuring universal secondary education, and some development of the tertiary sector, would also be priorities. Policies for places would look to develop core infrastructure and connectivity, and to increase the institutional quality to include contract enforcement and protection of tangible assets. 3. For countries with significant fiscal capacity and/or significant space, all of the measures discussed above are potentially available. They can adopt a wide range of active labor market policies to support movements of workers across tasks, occupations, sectors, and places, as well as to protect them from the consequences of sectoral contraction. Besides the active labor market programs discussed above, these could also include passive labor market programs such as unemployment insurance and early retirement. Sectoral interventions would focus on building capacity in more sophisticated product offerings by supporting research and development and protecting intangibles, in addition to targeting export and investment promotion, ensuring competition. Human capital development would focus on life-long learning, including tertiary and vocational education. Policies for places would additionally focus on fostering advanced logistics performance and putting in place advanced information and communication technologies. Recognizing the symbiotic relationship between trade and labor markets is not new. But the recent academic literature has highlighted that the costs of disaggregated labor market disruptions have historically been underestimated by economists. This paper attempts to correct this imbalance by taking those costs seriously, and identifying policy options for mitigating them while not sacrificing the gains from trade, which have demonstrated potential to support development and structural change. 26 REFERENCES Acemoglu, D., P. Antràs, and E. Helpman. 2007. “Contracts and Technology Adoption.” American Economic Review 97 (3): 916–43. Acemoglu, D., D. Autor, D. Dorn, G. Hanson, and B. Price. 2016. “Import Competition and the Great US Employment Sag of the 2000s.” Journal of Labor Economics, 34(S1): S141-S198. Aizenman, J. and Y. Jinjarak. 2009. “Globalisation and Developing Countries – a Shrinking Tax Base?”, The Journal of Development Studies, 45:5, 653-671 Anderson, J. and E. Van Wincoop. 2004. “Trade Costs.” Journal of Economic Literature, 42(3): 691-751. Ansón, J., J. Arvis,M. Boffa,M.Helble, and B. Shepherd. 2017. “Time, Uncertainty, and Trade Flows.” Working Paper 673, Asian Development Bank, Manila, Philippines. Antràs, P. 2016. Global Production: Firms, Contracts and Trade Structure, Princeton University Press. Antràs, P. 2020. “Conceptual Aspects of Global Value Chains.” World Bank Economic Review 34 (3): 551–74 Antràs, P., and A. De Gortari. 2020. “On the Geography of Global Value Chains.” Econometrica 88 (4): 1553–98. Arnold, J., B. Javorcik and A. Mattoo. 2011. “Does services liberalization benefit manufacturing firms?: Evidence from the Czech Republic,” Journal of International Economics, 85(1): 136-146. Arnold, J., B. Javorcik, M. Lipscomb and A. Mattoo. 2016. “Services Reform and Manufacturing Performance: Evidence from India,” Economic Journal, 126(590): 1-39. Artuc, E., Lopez-Acevedo, G., Robertson, R. and D. Samaan. 2019. Exports to Jobs : Boosting the Gains from Trade in South Asia, South Asia Development Forum, Washington, DC: World Bank, World Bank and International Labour Organization. Autor, D., D. Dorn, L. F. Katz, C. Patterson, J. Van Reenen. 2020. “The Fall of the Labor Share and the Rise of Superstar Firms.” The Quarterly Journal of Economics, 135(2): 645-709. Arvis, J.-F., Y. Duval, B. Shepherd, C. Utoktham, and A. Raj. 2016. “Trade Costs in the Developing World: 1996-2010.” World Trade Review, 15(3): 451-474. Asian Development Bank, Research Institute for Global Value Chains at the University of International Business and Economics, World Trade Organization, Institute of Developing Economies – Japan External Trade Organization, and China Development Research Foundation. 2021. Global Value Chain Development Report 2021 – Beyond Production. https://www.wto.org/english/res_e/booksp_e/00_gvc_dev_report_2021_e.pdf Autor, D., D. Dorn, G. Hanson, and K. Majlesi. 2020. “Importing Political Polarization.” American Economic Review, 110(10): 3139-3183. Balié, J., D. Del Prete, E. Magrini, P. Montalbano, and S. Nenci. 2019. “Does Trade Policy Impact Food and Agriculture Global Value Chain Participation of Sub-Saharan African Countries?”. American Journal of Agricultural Economics, 101: 773-789. 27 Bartók, C. and S. Miroudot. 2008. “The Interaction amongst Trade, Investment and Competition Policies.” OECD Trade Policy Papers No. 60. Paris: OECD. Bas, M., and V. Strauss-Kahn. 2015. “Input-Trade Liberalization, Export Prices, and Quality Upgrading.” Journal of International Economics 95 (2): 250–62. Ben Yamed, S., and P. Bombarda. 2020. “Gender, Informal Employment, and Trade Liberalization in Mexico.” World Bank Economic Review, 34(2): 259-283. Brown, D. 2000. “International Trade and Core Labor Standards: A Survey of the Recent Literature.” OECD Paper DEELSA/ELSA/WD(2000)4, https://www.oecd- ilibrary.org/docserver/677200103808.pdf?expires=1669821860&id=id&accname=guest&checksum =0DB5A1738B1EBCB197F0B6F91869FC10. Buelens, C., and M. Tirpak. 2017. “Reading the Footprints: How Foreign Investors Shape Countries’ Participation in Global Value Chains.” Comparative Economic Studies 59 (4): 561–84. Cadot, O., A. Fernandes, J. Gourdon, and A. Mattoo. 2015. “Are the Benefits of Export Support Durable? Evidence from Tunisia.” Journal of International Economics, 97(2): 310-324. Calabrese, L. and N. Balchin. 2021. “Foreign Investment and Upgrading in the Garment Sector in Africa and Asia.” Global Policy, 13(Suppl. 1): 34–44. Calì, M. and A. Montfaucon. 2021. “Non-Tariff Measures, Import Competition, and Exports.” Policy Research Working Paper No. 9801. Washington, DC: World Bank. Calì, M.; M. Le Moglie; G. Presidente. 2021. “Gain without Pain Non-Tariff Measures, Plants’ Productivity and Markups.” Policy Research Working Paper No. 9654 Washington, D.C.: World Bank Group. Caliendo, L., M. Dvorkin, and F. Parro. 2019. “Trade and Labor Market Dynamics: General Equilibrium Analysis of the China Trade Shock.” Econometrica, 87(3): 741-835. Caliendo, L., and F. Parro. 2015. “Estimates of the Trade and Welfare Impacts of NAFTA.” Review of Economic Studies, 82(1): 1-44. Cheng, K., S. Rehman, D. Seneviratne, and S. Zhang. 2015. “Reaping the Benefits from Global Value Chains.” IMF Working Paper WP/15/204.Washington, DC: International Monetary Fund. Cherkashin, I., S. Demidova, H. L. Kee, and K. Krishna. 2015. “Firm Heterogeneity and Costly Trade: A New Estimation Strategy and Policy Experiments.” Journal of International Economics 96 (1): 18–36. Choi, J., and A. Levchenko. 2021. “The Long Term Effects of Industrial Policy.” NBER Working Paper No. 29263, http://www.nber.org/papers/w29263.pdf. Chor, D. 2010. “Unpacking Sources of Comparative Advantage: A Quantitative Approach.” Journal of International Economics 82 (2): 152–67. Christ, N., and M. Ferrantino. 2011. “Land Transport for Exports: The Effects of Cost, Time, and Uncertainty in Sub-Saharan Africa.”World Development 39 (10): 1749–59. Citino, L., and A. Linarello. 2022. “The Impact of Chinese Import Competition on Italian Manufacturing.” Review of International Economics, 30(3): 702-731. 28 Conwell, L., and T. Williams. 2022. “Schooled by Trade? Retraining and Import Competition.” Working Paper, https://trevorcwilliams.github.io/research/CW_Retraining_wp.pdf. Corden, W.M. 1997. Trade Policy and Economic Welfare Second Edition. Oxford: Oxford University Press. Costinot, A. 2009. “On the Origins of Comparative Advantage.” Journal of International Economics 77: 255–64. Crowley, M. 2010. “Why are Safeguards Needed in a Trade Agreement?” in Law and Economics of Contingent Protection, Cambridge: Cambridge University Press. Crozet, M., and P. Koenig. 2004. “EU Enlargement and the Internal Geography of Countries.” Journal of Comparative Economics, 32(2): 265-279. Dauth, W., S. Findeisen, and J. Suedekum. 2017. “Trade and Manufacturing Jobs in Germany.” American Economic Review, 107(5): 337-342. De Lyon, J., and J.P. Pessoa. 2021. “Worker and Firm Responses to Trade Shocks: The UK-China Case.” European Economic Review, 133(C): 103678. Djankov, S., C. Freund, and C. Pham. 2010. “Trading on Time.” Review of Economics and Statistics 92 (1): 166–73. Egger, P. H., S. Nigai, and N. M. Strecker. 2019. “The Taxing Deed of Globalization.” American Economic Review 109 (2): 353–90. Erten, Bilge, and Jessica Leight. 2019. “Exporting Out of Agriculture: The Impact of WTO Accession on Structural Transformation in China.” Review of Economics and Statistics. Published ahead of print, June 26. Fajgelbaum, P., P. Goldberg, P. Kennedy, and A. Khandelwal. 2020. “The Return to Protectionism.” Quarterly Journal of Economics, 135(1): 1-55. Fajgelbaum, P., and A. Khandelwal. 2022. “The Economic Impacts of the US-China Trade War.” Annual Review of Economics, 14: 205-228. Farole, T. and D. Winkler (eds.). 2014. Making Foreign Direct Investment Work for Sub-Saharan Africa: Local Spillovers and Competitiveness in Global Value Chains, Washington, DC: The World Bank Fernandes, A., A. Mattoo, H. Nguyen, and M. Schiffbauer. 2019. “The Internet and Chinese Exports in the Pre-Ali Baba Era.” Journal of Development Economics 138: 57–7 Fernandes, A., H.-L. Kee, and D. Winkler. 2022. “Determinants of Global Value Chain Participation: Cross-Country Evidence,” World Bank Economic Review, 36(2): 329-360. Fernandes, A. M., A. Forero, H. Maemir, and A. Mattoo. 2023. “Are Trade Preferences a Panacea? The Export Impact of the African Growth and Opportunity Act.” World Development, 162(106114), 1- 20. Grossman, G., and E. Rossi-Hansberg. 2008. Trading Tasks: A Simple Theory of Offshoring. American Economic Review. 98: 1978-97. 29 Harding, T., and B. Javorcik. 2011. “Roll Out the Red Carpet and They Will Come: Investment Promotion and FDI Flows.” Economic Journal, 121(557): 1445-1476. Hasan, R., D. Mitra, and K. Ramaswamy. 2007. “Trade Reforms, Labor Regulations, and Labor Demand Elasticities: Empirical Evidence from India.” Review of Economics and Statistics, 89(3): 466-481. Hayakawa, K., T. Ito, and S. Urata. 2021. “Impacts of Increased Chinese Imports on Japan’s Labor Market.” Japan and the World Economy, 59(C): 101087. Hoekman, B., and B. Shepherd. 2017. “Services Productivity, Trade Policy, and Manufacturing Exports.” World Economy, 40(3): 499-516. Holzmann, R. 2018. “The Portability of Social Benefits Across Borders.” IZA World of Labor, 452: 1- 12. Hummels, D., and G. Schaur. 2013. “Time as a Trade Barrier.” American Economic Review 103 (7): 2935–59. Ignatenko, A., F. Raei, and B. Mircheva. 2019. “Global Value Chains: What Are the Benefits and Why Do Countries Participate?” IMF Working Paper No. 19/18. ITC. 2020. Business and Policy Insights: Mapping E-Marketplaces in Africa. Geneva: International Trade Center. Johnson, R., and G. Noguera. 2017. “A Portrait of Trade in Value-Added over Four Decades.” Review of Economics and Statistics 99 (5): 896–911. Kee, H. L. 2015. “Local Intermediate Inputs, and the Shared Supplier Spillovers of Foreign Direct Investment.” Journal of Development Economics 112: 56–71. Kee, H. L., and H. Tang. 2016. “Domestic Value Added in Exports: Theory and Firm Evidence from China.” American Economic Review 106 (6): 1402–36. Kowalski, P., J. Gonzalez, A. Ragoussis, and C. Ugarte. 2015. “Participation of Developing Countries in Global Value Chains: Implications for Trade and Trade-Related Policies.” Trade Policy Papers 179, Organisation for Economic Co-operation and Development, Paris. Kruse, H., M. Timmer, G. de Vries, and X. Ye. 2023. “What you Do in Exports Matters for Growth: Evidence from Cross-Country Specialization in Occupations.” Working Paper. Laget, E., A. Osnago, N. Rocha, and M. Ruta. 2018. “Deep Trade Agreements and Global Value Chains.” Policy Research Working Paper 8491, World Bank, Washington, DC. Lendle, A., M. Olarreaga, S. Schropp, and P.-L. Vezina. 2016. “There Goes Gravity: eBay and the Death of Distance.” Economic Journal, 126(591): 406-441. Levchenko, A. 2007. “Institutional Quality and International Trade.” Review of Economic Studies 74 (3): 791–819. Levy Yeyati, E., M. Montane, and L. Sartoria. 2019. “What Works for Active Labor Market Policies?” Working Paper, Center for International Development, Harvard University, https://growthlab.cid.harvard.edu/publications/what-works-active-labor-market-policies. 30 Mansour, H., P. Medina, and A. Velasquez. Forthcoming. “Import Competition and Gender Differences in Reallocation.” Labor Economics. McCaig, Brian. 2011. “Exporting Out of Poverty: Provincial Poverty in Vietnam and U.S. Market Access.” Journal of International Economics 85 (1): 102–13. McCaig, Brian, and Nina Pavcnik. 2018. “Export Markets and Labor Allocation in a Low-Income Country.” American Economic Review 108 (7): 1899–1941. Milberg, W. and D. Winkler. 2011. “Offshoring and Economic Insecurity: Worker Perceptions and Labor Market Realities.” In Bacchetta, M. and M. Jansen (eds.), Making Globalization Socially Sustainable, Geneva: International Labour Office and World Trade Organization, 147-198. Ngoma, M. 2022. “Chinese Imports and Industrialization in Africa: Evidence from Ethiopia.” Working Paper, Tufts University, https://sites.tufts.edu/marinamavungu/files/2022/11/Ngoma_JMP2022_ChineseImports- industrialization-Ethiopia.pdf. Nunn, N. 2007. “Relationship-Specificity, Incomplete Contracts, and the Pattern of Trade.” Quarterly Journal of Economics 122 (2): 569-600. OECD. 2021. “Building Inclusive Labor Markets: Active Labor Market Policies for the Most Vulnerable Groups.” https://read.oecd-ilibrary.org/view/?ref=1112_1112680- u38fb4etmq&title=Building-inclusive-labour-markets-active-labour-market-policies-for-the-most- vulnerable-groups. OECD and WTO. 2013. Aid for Trade at a Glance 2013: Linking to Value Chains. Paris: Organisation for Economic Co-operation and Development. Orefice, G., and N. Rocha. 2014. “Deep Integration and Production Networks: An Empirical Analysis.”World Economy 37 (1): 106–36. Pierola, M., A. Fernandes, and T. Farole. 2018. “The Role of Imports for Exporter Performance in Peru.” World Economy 41 (2): 550–72. Robertson, Raymond, Mexico Alberto Vergara Bahena, Deeksha Kokas, and Gladys Lopez-Acevedo (2021). “International Trade and Labor Markets: Evidence from the Arab Republic of Egypt.” Policy Research Working Paper 9668, World Bank, Washington, DC. Rodrik, D. 2022. “An Industrial Policy for Good Jobs.” Policy Proposal, Hamilton Project, https://drodrik.scholar.harvard.edu/files/dani-rodrik/files/rodrik_- _an_industrial_policy_for_good_jobs.pdf. Shepherd, B. 2022. “Modeling Global Value Chains: From Trade Costs to Policy Impacts.” World Economy, 45(8): 2478-2509. SPR and Mathematica. 2012. “The Evaluation of the Trade Adjustment Assistance Program: A Synthesis of Major Findings.” Report Prepared for the US Department of Labor, https://wdr.doleta.gov/research/FullText_Documents/ETAOP_2013_08.pdf. Stolper, W., and P. Samuelson. 1941. “Protection and Real Wages.” Review of Economic Studies, 9(1): 58- 73. 31 Sykes, A. 2003. “The Safeguards Mess: A Critique of WTO Jurisprudence.” World Trade Review, 2(3): 261-295. Taglioni, D. and D. Winkler. 2016. Making Global Value Chains Work for Development. Trade and Development Series. Washington, DC: World Bank. Vazquez, E. and D. Winkler (2023), “Trade and local labor market outcomes in Mexico: Disentangling the channels and the role of geography, sectors, and trade types,” World Bank Policy Research Working Paper No. 103322, Washington, D.C.: World Bank. Winkler, D., L. Aguilar Luna, H. Kruse, and M. Maliszewska (2023), “Linking trade to jobs and activities: new stylized facts for low- and middle-income countries,” Washington, D.C.: World Bank. World Bank. 2009a. Conditional Cash Transfers: Reducing Present and Future Poverty. Washington, D.C.: World Bank. World Bank. 2009b. World Development Report 2009: Reshaping Economic Geography. Washington, D.C.: World Bank. World Bank. 2015. Competitive Cities for Jobs and Growth: What, Who, and How. Washington, D.C.: World Bank. World Bank. 2018. Connecting to Compete: The Logistics Performance Index and its Indicators. Washington, D.C.: World Bank. World Bank. 2020a. World Development Report 2020: Trading for Development in the Age of Global Value Chains. Washington, DC: World Bank. World Bank. 2020b. The African Continental Free Trade Area; The African Continental Free Trade Area: Economic and Distributional Effects: Economic and Distributional Effects. Washington, DC: World Bank. Yi, K.-M. 2003. “Can Vertical Specialization Explain the Growth of World Trade?” Journal of Political Economy 111(1): 52–102. Yi, K.M. 2010. “Can Multistage Production Explain the Home Bias in Trade?” American Economic Review 100 (1):364–93. 32