The effect of external and internal trade costs on regional development

International trade is an important driver of economic growth and a potential force that can change the geographic distribution of economic activity within countries. Recent trade liberalization in several countries has led to a reallocation of economic activity to coastal areas. Historically, coastal areas have also been favored for regional and international commerce - the global distribution of employment is skewed toward coastal areas where half of the world population lives within 100 kilometers of coastlines or navigable rivers, and 19 out of the 25 largest cities in the world are on the coast (Cosar and Fajgelbaum, 2016). With the growth of international trade, the question is, does globalization make regional inequality worse?

In a recent World Bank working paper, we examine if a change in external trade costs - cost of importing - on economic activity across regions within a country depend on internal trade costs, or how well these regions are connected to international ports. To answer this question, we examine the effect of two policy interventions in India on the geographic distribution of organized manufacturing activity from 1989 to 2009. These policies include trade liberalization during the early 1990s and the construction and upgrade of the large -scale Golden Quadrilateral (GQ) highways program. To analyze the internal geography of India, we grouped Indian districts into three regions, based on their distance from the nearest port: the primary region (districts within 200 kilometers of the nearest port), the secondary region (districts between 200 to 400 kilometers from the nearest port) and the interior region (districts located farther than 400 kilometers from the nearest port). Our research finds:

• A huge persistence in economic activity. Organized manufacturing activity was highly concentrated in a few districts, mainly in the primary region - a pattern visible over the entire sample period. 
• The impact of external trade on economic concentration in primary and secondary regions depends on the interactions with domestic trade costs. Coinciding with the timing of trade liberalization, there was a strong reallocation of manufacturing activity from the primary to the secondary region. This temporary dispersion in activity reversed in the decade following the Golden Quadrilateral (GQ) highway upgrades that may possibly have reduced the internal trade costs.
• Irrespective of the period, the interior region continued to perform poorly. Relative to the secondary region, the share of manufacturing activity in the interior region remained low, and continued to fall.

To explain these findings, we developed an economic geography model of trade and production where regions within a country differ in their distance from the port. We show that lowering import tariffs changes the geographic distribution of economic activity - with regions that are further away from the ports improving their outcomes relative to those close to ports. The distance from the ports is what insulates the former regions from import competition. By comparison, lowering domestic trade costs causes a shift in manufacturing activity towards the ports that are relatively more attractive for production due to their lower cost of engaging in international trade.

Exploiting the tariff liberalization of the 1990s, together with the upgrading of the national highway system in the 2000s in a difference-in-difference setting, we show that while the average growth rate of districts in the three regions were not significantly different during the pre-liberalization period, 1989-1994, the secondary region did grow faster relative to the primary region from 1994 to 2000. Figure 1 shows this pattern for output, but the finding is similar for employment and establishment count.

_{Note: The figure presents rope-ladder plots of regressions of change in output on the interaction of the two phases of trade liberalization periods, using primary districts during the period 1989–94 as the reference category. All regressions include state-period fixed effects and standard errors are clustered at the district level. CL = confidence level.}

Moreover, the relatively higher growth rates in the secondary region were most evident for industries that experienced large tariff cuts on their output, but not for industries experiencing smaller tariff cuts. This is consistent with our model which shows that the secondary region is protected from import competition relative to the primary region. By comparison, trade liberalization had no effect on interior hinterland districts located farther than 400 kilometers from the ports. Unlike most of the existing research, we did not find an increase in spatial concentration coinciding with globalization.

So how does external trade cost interact with internal trade cost? Following the massive infrastructure investments in the GQ highways post-2000, activity in secondary locations along the GQ highway increased. Irrespective of their distance from the GQ network, remote locations were largely unaffected, suggesting that neither trade openness nor infrastructure connectivity by themselves changed their growth prospects or competitiveness (See Figure 2). Our findings are echoed in other papers that find that remote locations lose out when connected to more “attractive” locations (Faber, 2014; Baum-Snow et al., 2020; Lall and Lebrand, 2020). Complementary conditions, usually missing in remote areas, are needed to make a potentially viable place-based intervention work in practice (Grover, Lall and Maloney, 2022).

_{Note: The figure presents rope-ladder plots of regressions of change in outcome variables. The presented coefficients capture the effect on secondary and interior districts within 10 km of the GQ relative to districts farther away. 1994–2000 is the pre-treatment period.}

While globalization is appealing for expanding external market opportunities, economic activity in open economic systems tends to be concentrated near international gateways and large urban-centers. Further, internal trade costs limit the extent to which economically and physically distant regions can be integrated into trade and production networks. The interaction of scale economies in production and in transport make it harder - not easier - for economically distant places to enter production chains; trade tends to be vigorous between regions that are close. The challenges in spreading growth across many regions in a country often provides the motivation for place-based policies, such as investment in transport infrastructure. Our work suggests that such investments may not necessarily favor the dispersion of activity within countries, especially when the global trading environment is more open.