Regional adjustment to trade liberalization

https://doi.org/10.1016/S0166-0462(98)00006-4Get rights and content

Abstract

In this paper, I examine the effect of trade reform on regional employment in Mexico. Three factors condition regional labor demand: (1) transport costs, which encourage firms to locate in regions with good access to foreign markets; (2) backward–forward linkages, which encourage firms to locate near buyers and suppliers, and (3) agglomeration economies, which reinforce the pre-trade pattern of industry location. The results suggest transport costs and backward–forward linkages influence regional employment. Post-trade employment growth is relatively high in regional industries that are close to the United States and near upstream and downstream industries. Trade reform appears to have contributed to the breakup of the Mexico City manufacturing belt and the formation of new industry centers in northern Mexico.

Introduction

How do regions adjust to trade liberalization? In this paper, I study the effect of trade reform on regional industry employment in Mexico. I disentangle the effects of three opposing forces on regional labor demand: transport-cost considerations, which encourage firms to relocate their activities to regions with relatively low-cost access to foreign markets; backward–forward linkages, which give firms an incentive to locate near their buyers and suppliers; and agglomeration economies, which tend to reinforce the pre-trade pattern of industry location.

That international trade causes a sectoral reallocation of resources is a basic insight of trade theory. Whether the motivation for trade is relative factor abundance, increasing returns to scale, or imperfect competition, the transition from an open economy to a closed economy alters a nation's pattern of specialization. The effect of trade on the spatial allocation of resources has received less attention. Economists recently have begun to examine regional specialization within countries, but there has been little empirical work on how the transition to an open economy affects the location of economic activity.

One reason that industry location matters for trade is transport costs. Krugman (1991) shows that the interaction between plant-level scale economies and transport costs can explain the formation of cities. Krugman and Livas (1992) extend this framework to show that the size and location of cities is conditioned by the openness of an economy. Similarly, in Rauch (1993a) transport costs determine the volume of trade within and between countries.

A second reason that location matters for trade is the existence of backward and forward linkages between industries. Recent work by Venables (1996) and Krugman and Venables (1995) formalizes Hirschman's (Hirschman, 1958) concept of how vertical relationships between industries create a pattern of interdependent industry location. In this body of work, expansion in one industry contributes to the expansion of upstream and downstream industries.

A third reason that location matters for trade is the existence of externalities, which link the productivity of agents to the local agglomeration of resources. Dynamic externalities, due to knowledge spillovers or learning by doing, figure prominently in recent theories of economic growth (Romer, 1986, Romer, 1990, Lucas, 1988, Young, 1991). To the extent such externalities are localized, economic activity becomes geographically concentrated over time. Several recent studies have found evidence that is consistent with dynamic localized externalities (Glaeser et al., 1992Jaffe et al., 1993Henderson et al., 1995).

The importance of trade liberalization is that it changes the reference market for firms in a country. Given transport costs, we expect trade to shift resources to locations with relatively low-cost access to foreign markets, such as border regions and port cities. The existence of backward–forward linkages and agglomeration economies complicate the picture. They imply that the size or mix of industries in a region may also affect how it adjusts to trade. By this logic, as trade shifts employment across industries, we expect locations with large concentrations of firms, all else equal, to expand. To the extent closed-economy industry centers have relatively poor access to foreign markets, transport costs, backward–forward linkages, and agglomeration economies may have opposing effects on how regions adjust to the opening of the economy.

The regional effects of trade liberalization have important policy implications. The current process of economic integration is likely to reorganize the location of economic activity in developed and developing regions alike. The formation of the European Union and the fall of communism in Eastern Europe imply a substantial increase in regional resource mobility on the European continent. The spread of trade reform in the developing world has reoriented producers in these countries towards a new set of markets. Wei (1993) finds that in China the fastest growing cities are those with large export sectors. For obvious reasons, many of these cities are located in coastal areas near Hong Kong. Regional movements in response to trade strain existing infrastructure and change relative regional economic fortunes.

Recent changes in Mexico's trade policy make the country an ideal case study. In 1985, after four decades of import-substitution industrialization, Mexico began to open its economy to trade. This involved both a reduction of import barriers and a relaxation of controls on exports. The government enacted reform swiftly, eliminating most trade barriers in the following three years. Mexico's location in North America makes trade liberalization equivalent to economic integration with the United States. For Mexican firms, proximity to foreign markets means proximity to the U.S. market. Yet, Mexico's closed-economy production centers are located far from the United States. Since the 1950s, manufacturing capacity has been concentrated in the country's interior, around Mexico City. While foreign markets lure firms to the Mexico–U.S. border, the existing pattern of industry location may work against this shift.

I estimate the change in regional industry labor demand in Mexico before and after trade reform as a function of transport costs to the United States, the local geographic concentration of industry, and a series of control variables. If transport costs matter, employment growth will be higher in regions close to the United States; if backward–forward linkages matter, employment growth will be higher in regions with large concentrations of vertically-linked industries; and if agglomeration economies matter, employment growth will be higher in large production centers. Following Glaeser et al. (1992), I study the effects of two types of agglomeration: within-industry agglomeration, the concentration of firms in the same industry; and industrial diversity, the concentration of firms in a broad range of industries.

To preview the results, I find evidence of both transport-cost effects and backward–forward linkages on post-trade employment growth. After trade liberalization, employment growth is higher in regional industries that are relatively close to the United States or that are proximate to related industries. These findings suggest that the closed-economy manufacturing belt around Mexico City is breaking apart, as new industrial centers form closer to the Mexico–U.S. border. The Mexican economy is changing from one based on one large industry center to one based on a number of broadly specialized industry centers. The North American Free Trade Agreement is likely to reinforce this trend. Consistent with Glaeser et al. (1992), I find no evidence employment growth is higher where within-industry agglomeration is higher; contrary to their results and to those of Henderson et al. (1995), I find little evidence employment growth is higher where industrial diversity is higher.

Section snippets

International trade and industry location

In this section, I review briefly recent theoretical literature that addresses the relationship between international trade and industry location. Recent theories of trade have three common elements: increasing returns to scale, transport costs, and congestion costs. In Fujita (1988) and Krugman (1991) scale economies are internal to firms; in Henderson (1974) and Rauch (1989) scale economies are external to firms. The interaction of scale economies and transport costs creates a centripetal

The data

Data are from the Mexico Industrial Census, which is a comprehensive survey of manufacturing establishments by state and four-digit (ISIC) industry. Mexico has 54 industries and 32 states. I have data from the four most recent censuses, 1980, 1985, 1988, and 1993. Mexico initiated its liberalization of trade in 1985. I use two sets of observations on changes in regional-industry labor demand: 1980–1985, the period preceding trade liberalization, and 1985–1993, the period following the

Empirical results

If location-specific factors such as agglomeration and access to foreign markets condition adjustment to trade reform, then the most direct way to identify these effects is to study industries in different locations and verify in which locations industry employment grows faster. I study changes in state-industry employment during 1980–1985 and 1985–1993. There are 1728 (32 states×54 industries) potential observations per time period, but not all industries are present in all locations. I have

Concluding remarks

This paper studies the regional effects of trade liberalization. I focus on the roles of transport costs, which encourage firms to relocate to regions with relatively good access to foreign markets; backward–forward linkages, which encourage firms to locate near buyers and suppliers; and agglomeration economies, which encourage the growth of pre-existing industry centers. The particular case I consider is regional industry employment growth in Mexico before and after trade reform. Consistent

Acknowledgements

I thank Dan Hamermesh, Ann Harrison, Gerald Oettinger, Francisco Rivera-Batiz, and seminar participants at MIT, Georgetown University, Carleton University, University of Miami, and the Winter Econometric Society Meetings for helpful comments.

References (27)

  • J.V. Henderson et al.

    Industrial development and cities

    Journal of Political Economy

    (1995)
  • Hirschman, A.O., 1958. The Strategy of Economic Development. Yale University Press, New...
  • Jacobs, J. 1969. The Economy of Cities. Vintage, New...
  • Cited by (0)

    1

    National Bureau of Economic Research.

    View full text