Perspectives on trade and structural transformation

Abstract

ABSTRACTThis paper surveys macroeconomic and microeconomic perspectives on the role of international trade in structural transformation. We start by describing canonical frameworks that have been used to quantify how trade influences sectoral shares of employment and value added. We then pivot to survey micro-empirical evidence on the impact of changes in trade on the allocation of labor across sectors and productivity at the firm level. In this, we put special emphasis on the role of participation in global value chains and inward foreign direct investment in mediating these effects. Next, we evaluate evidence on the barriers to trade faced by low-income countries, with special attention to recent work that measures these costs taking firm dynamics into account. We conclude by discussing how these micro-perspectives can be integrated into macro models to advance our understanding of structural change.KEYWORDS: Global value chainsForeign direct investmentFirm-level costs of tradeMicro-evidence and macro modelsJEL CLASSIFICATION: F1F23F43F62 Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1. Several papers have developed models to explain them. One of these papers includes international trade; see Sposi et al. (Citation2021).2. That said, measurement of GVCs has typically employed a more narrow definition: imported inputs are used in production and some of the resulting output is exported, i.e. some stages of production cross multiple national borders.3. We acknowledge that our survey here cannot cover the full range of perspectives; for example, we do not directly address industrial policy, so refer the reader to Chang (Citation2003) and Oqubay et al. (Citation2020) for complementary discussion.4. To clarify the scope of our discussion, we focus on the allocation of factors across sectors. Thus, we do not engage research on the allocation of factors across firms within a sector, which plays an important role in models of trade with heterogeneous firms. To the extent that this within sector reallocation alters sectoral productivity, it may be an additional mechanism through which trade influences structural change.5. A separate set of questions are related to the welfare effects of worker displacements due to trade. One might be concerned that displacements are particularly costly when workers do not have access to social insurance programs, as in most low-income countries. We are not aware of work that addresses these concerns explicitly, though they would be important for understanding the political economy of trade reforms.6. Dix-Carneiro et al. (Citation2021) shows that incorporating the informal sector into models is important quantitatively, and also qualitatively for wage inequality, productivity, and welfare.7. Our emphasis here is tailored to highlight particular issues pertaining to participation in global value chains and trade in low-income countries. See De Loecker and Goldberg (Citation2014) for a more comprehensive review of the literature on the impact of trade on firm performance.8. In addition to expanding import variety, imports (particularly from high income countries) are likely to be higher quality than domestic alternatives. Halpern et al. (Citation2015) links import variety and quality to firm productivity in Hungary.9. Looking forward to discussion below regarding inventories and trade costs, we also raise a point about measuring productivity gains here. Khan and Khederlarian (Citation2020) argues that omission of inventory holding costs in computing revenue based productivity leads to overestimation of the productivity gains of import liberalization. This is related to a broader issue in this literature: productivity changes hinge on accurate measurement of firm-level input price indexes (equivalently, input quantity). Input price and quantity measurement is challenging in most micro-data sets, and it is particularly difficult when the mix of inputs used in production changes rapidly due to liberalization itself. As such, many studies include changes in input prices in the firm’s measured productivity residual.10. Dai et al. (Citation2016) show that Chinese firms in the processing trade sector are less productive than ordinary exporters. Whether this pattern holds elsewhere is unknown; if it does, it suggests that countries are effectively subsidizing unproductive firms via their processing trade regimes. If these subsidies simply reallocate activity within the manufacturing sector from more productive to less productive firms, then they would seem unwise. If instead they serve to reallocate labor from lower productivity outside sectors, then they may raise economy-wide productivity overall, though not as much as subsidies that would have directed workers to the more productive non-processing firms. Work that investigates these alternatives, both in China and non-Chinese contexts, would be valuable.11. Newman et al. (Citation2015) uses a unique firm survey from Vietnam to unpack the role of direct sourcing relationships and technology transfers between firms in explaining the role of aggregated sector-to-sector linkages (as studied by Javorcik, Citation2004) in mediating spillovers.12. In the specific Bangladeshi context, there is potential for rich data sources could be used in research. For example, Grossi et al. (Citation2023) exploits a rich data set on transactions between domestic garments firms and foreign buyers, together with information about input sourcing by domestic firms.13. Low-income economies have relatively high domestic trade costs and import barriers. See high-quality surveys such as the World Bank’s Doing Business Survey or Logistic Performance Index, for example.14. For instance, Baier and Bergstrand (Citation2007) and Baier et al. (Citation2014) show that preferential free trade agreements only gradually increase bilateral trade flows. Jung (Citation2012) estimates that less than half of the trade expansions from a free trade agreement occurs within the first 10 years. Likewise, the sharp movements in trade over the cycle do not reflect large changes in policy or trade barriers, but rather dynamic aspects of trade that are outside the static gravity model.15. Mix (forthcoming) develops a multi-country model with transition dynamics to study how the dynamics effects of trade reforms are shaped by financial openness, geography, country size, and development.16. See the recent survey by Alessandria et al. (Citation2021).17. These types of models are also needed to make sense of the aggregate trade data we discussed earlier.18. Though we do not review it here, the empirical macro-literature on trade and economic growth is almost entirely disconnected from this micro-evidence. See Irwin (Citation2019) for a recent survey.19. See Martin and Warr (Citation1993) and Friedberg (Citation2001) for evidence on Rybczynski effects. Sposi et al. (Citation2021b) provides a model with such feedback effects involving capital accumulation, but it does not have structural change.20. As discussed above, McCaig and Pavcnik (Citation2018) is one of the few papers that has addressed the role of trade in reallocation across formal and informal firms.21. Pahl and Timmer (Citation2020) presents related evidence on the positive correlation between GVC participation and manufacturing productivity growth for a wide set of countries.Additional informationFundingThe work was supported by the Department for International Development (DFID) [STEG LOA PP10 Alessandria Johnson Yi].