As we close our 16th and final Job Market Paper Series blog post for 2024-2025, we want to thank you for your support this year. Wishing you a Merry Christmas and a prosperous New Year!
Anh Nguyen is a PhD candidate in the Dyson School of Applied Economics and Management at Cornell University. Her interests are in the intersection of International Trade, Development, and Labor Economics. You can find her JMP here.
For many Asian countries, the motto “supply from the East, consume in the West” embedded in the Factory Asia model of the late 20th century is a crucial recipe for the prosperity cookbook (Byung-il and Rhee, 2014). Within a country, such growth is facilitated by a large reallocation of production factors across sectors and regions, pulling workers toward the economic hotspots. One important determination for the migration decision is the existence of the pre-established migrant community at the destinations, which helps to solve the information friction and to provide a reliable referral (Munshi 2003). A stronger migration network to the fast-growing regions induces more outmigration and creates a negative labor supply shock at the origin, causing its labor market to readjust. On the other hand, migrants living at the manufacturing hotspots prior to the trade shock can earn more income and increase their remittances, thereby stimulating the local demand at the origin.
Both stories, while understudied in the trade literature, demonstrate how a less directly exposed region can still reap the benefits from trade expansion through migration network, which deserves to have a more rigorous empirical examination. This is the goal of my job market paper. I use Vietnam as an ideal study setting for two reasons. First, Vietnam underwent a monumental Bilateral Trade Agreement (BTA) with the U.S in 2001, which profoundly elevates its comparative advantage in low-skilled manufacturing production. Tariffs went down from 40% to 3% on average, and this drop is unlikely affected by any lobbying efforts from both parties (McCaig, 2011). By 2008, exports had reached $11,903 million which is equivalent to an 11-fold increase from that of 2001 (Fukase, 2009).
Second, high economic growth and relaxed household restriction has enabled a large reallocation of workers across regions, especially during the 2000s period. Nationwide, the number of internal migrants increased from 4.5 million in 1999 to 6.6 million people in 2009, which is approximately 8% of the total population and the primary reason for moving is due to economic reason (Coxhead et al., 2019).
Measuring exposure to trade:
The measurement of regional exposure to the BTA shock consists of two parts: direct and spillover effect. Region, or local labor market, is defined at a provincial level. To what extent a province is directly exposed to the shock will depend on (i) the magnitude of an industry’s tariff reduction and (ii) its regional employment composition prior to the BTA. I retrieve the tariff data from McCaig (2011) and the employment data from the 1999 Vietnam Population Census. From Figure 1a, we see that there is a significant spatial variation in the BTA exposure across the Vietnamese provinces. The most directly exposed provinces tend to concentrate on the manufacturing hubs in the Southeast region, which is also the main receiving regions for inter-provincial migration. This relationship is also found in McCaig (2022).
Turning to the spillover effect, as motivated, a less exposed province can still be affected if it has higher connections to the manufacturing hubs afar through the pre-established migrant network. Thus, a province’s exposure to this spillover effect is dependent on (i) the direct exposure faced by the migration destinations and (ii) how large the migrant stock between that province and other locations would be. The second item can be calculated from the 1999 Census as I can observe one’s current and previous location five years prior to the survey time. From Figure 1b, we see that the most (indirectly) exposed provinces are from the Mekong River Delta in the South and the Central Coastal region, which also experienced very high outmigration rate.
With the two exposures, we are ready for our empirical analysis. I extract individual employment and earnings information from the Vietnam Household Living Standards Survey, which is a nationally representative household survey conducted every two years. The regression results reveal three takeaways:
Both direct and spillover effects induce structural transformation away from agricultural and informal sector
Table 1 shows the key results for the BTA effects on the sectoral employment reallocation. We see that both types of exposure contribute to the industrialization process and the reallocation away from informality. On average, one standard deviation (sd) larger in provincial direct (spillover) exposure will lead to 2.74 (1.83) percentage points (pp) reduction in agricultural employment share. Compared to the mean, this is equivalent to a 5.3% and 3.5% decrease in agricultural share respectively. On the other hand, the estimates for the services and informal sector suggest that the remittance channel is not the dominating mechanism. Otherwise, we should see an increase in services employment share, and probably negative or null effects on manufacturing, due to the rising local demand from the influx of remittances.
The spillover effect persistently reduces the skill premium
In addition to facilitating structural change, the migration-induced spillover effect also helps to reduce earnings inequality between low and high-skilled workers. Regions more indirectly exposed to the BTA shock through the migration network would experience a significant and long-lasting decrease in the skill premium, while the direct effect only happens in the short run. Additionally, the spillover effect does not materialize immediately but rather taking off gradually, which is in line with the outmigration mechanism such that it takes some time lags for workers to reallocate spatially in response to the changes in local labor demand at the potential destinations.
Both types of exposure effectively reduce poverty incidence, although with different timing
The previous analysis shows that the BTA shock primarily improves earnings for low-skilled workers, who are more likely to reside in poorer households. Now, I will explore how the trade shock will further exert its impact on household expenditure and regional poverty incidence. Indeed, I show that regions more exposed to either exposure will enjoy an improvement in expenditure per capita at the household-level, as well as a faster reduction in poverty rate. One interesting pattern regarding the timing of each effect is that while the direct effect reduces poverty immediately after the BTA implementation, it takes some years for the spillover effect to grow in magnitude and become statistically significant.
Conclusion and moving forward
Over the last four decades since its reform in 1986, Vietnam’s trade policy has consistently proved to be a key instrument to deepen its global integration and transform the entire economy. My study leverages one of the most important trade agreements for Vietnam to uncover a novel channel that a developing country can benefit from trade expansion, ie. via the internal migration network to other manufacturing hubs. Through such linkages, the gains from trade can be shared beyond the geographical confinement of a local labor market. This methodology can be flexibly applied to other local shocks that also trigger a significant spatial labor reallocation, such as climate change or the recent COVID pandemic. As internal migration is a common phenomenon in a developing country context, the spillover effect augmented by migrant network deserves a more careful examination in the future. My paper is making a few first steps in that direction.
Feature image created by the blog author using ChatGPT Image Generator.
