Introduction

Trade between the United States and Mexico encompasses a great variety of goods and services. One of the main characteristics of that trade is the large share of intra-industry trade, concentrated mainly in the automobile industry, electronics, and telecommunications. Also, the size of the economies, together with the development of value chains emerging from vertical foreign direct investment (FDI) and transportation costs in the manufacturing sector have been important sources of regional economic integration between the two countries.

The establishment of the North American Free Trade Agreement (NAFTA) was a factor in promoting trade and investment among the member countries. Specifically, NAFTA gradually reduced the tariff structure and established rules of origin for input content in the North American region, as well as rules for the protection of foreign direct investment. The effects of the establishment of NAFTA have been extensively studied.

Regarding the effects of NAFTA on trade among the three countries, the rapid increase in trade and the growing importance of trade from Mexico and Canada in total trade with the United States stand out. Likewise, it is noteworthy that, with the establishment of NAFTA, important changes occurred in trade and investment patterns, generating a greater synchronization of trade with the economic cycles of the United States and Mexico (^{Villarreal and Fergusson, 2014}). The implementation of NAFTA accelerated the development of intra-industrial trade between the United States and Mexico. Particularly, vertical intra-industrial trade has been very significant, reflecting the trade pattern of the region based on the differentiation of traded goods in terms of value, quality, and their role in the production process (^{Ekayanake, Veeramacheneni, and Moslares, 2009}). ^{Hillberry and McDaniel (2002)} analyzed the characteristics of trade growth among the U.S., Mexico, and Canada. They concluded that trade in the North American region has grown intensely, indicating that two characteristics of trade between the U.S. and Mexico are that U.S. industries have faced competition from Mexican imports, and that U.S. consumers and manufacturers have had access to imports from Mexico at a lower cost. In their study, ^{Burfisher, Robinson, and Thierfelder (2001)} indicated that for the United States, the establishment of NAFTA had a positive but limited impact on its trade, pointing out small but positive effects for the U.S. economy. ^{Waldkirch (2010)} analyzed the impact of NAFTA considering foreign direct investment; the results suggest that there are positive effects on productivity and wages in Mexico. ^{De la Cruz and Riker (2014)} estimated the impact of NAFTA on U.S. labor markets and found small but positive effects on the real wages of skilled workers in the United States.

The recent changes in U.S. trade policy leading up to the U.S.-Mexico-Canada Agreement (USMCA) introduced a scenario of increased value-added content in the North American region for the automobile, electronics, aluminum, and steel industries. The establishment of the USMCA would modernize aspects of trade in services and digital commerce and would change the rules of origin, which in this agreement would demand a larger percentage of inputs from the North American region. The modifications of the new agreement could have positive or negative effects, depending on the manufacturing sector’s ability to adapt, both in the U.S. and Mexico.

This article is aimed at evaluating the determinants of trade between the U.S. and Mexico at the regional level. Specifically, the research seeks to estimate the effects of tariffs and transportation costs on trade between Mexico and U.S. states. The research methodology consists of an empirical adaptation of a gravity model for estimating the determinants of trade between countries and regions. This econometric model, based on the gravitational equation, has been extensively used in empirical studies on trade relations between countries. From this perspective, the assumptions are that bilateral trade between the U.S. and Mexico at the regional level depends on income levels, population, distance, and the tariff structure, as well as additional control variables.

The article is structured as follows: first, I introduce the objectives. In the second section, I discuss the characteristics of NAFTA and its effects on trade between the U.S. and Mexico. In section three, I present the characteristics of the manufacturing sector in North America and the new provisions of the USMCA. Section four discusses the major determinants of the U.S.-Mexico trade and the methodology of the empirical model. In section five, I explain the econometric results, and section six presents the conclusions.

NAFTA And U.S.-Mexico Trade

NAFTA and U.S.-Mexico Trade

After the establishment of NAFTA, trade among the three members of the agreement grew exponentially, increasing from US$343.7 billion in 1994 to US$1.16 trillion in 2019 (Figure 1). The covid-19 pandemic in 2020 resulted in a break in the expansionary trend of total trade in the region, with negative growth observed in that year. The share of the U.S.-Mexico trade within the NAFTA region expanded from 29.1 percent in 1994 to 51 percent in 2019.² The empirical evidence has created a consensus among economists and policymakers that NAFTA has created an important volume of trade and economic integration among the three countries of the agreement.

Source: Developed by the author using the ^{United Nations Comtrade data base (n. d.)}.

Figure 1 VALUE OF TRADE IN THE NORTH AMERICAN REGION (1994-2020) (U.S. dollars) 

The trade growth between Mexico and the U.S. was characterized by a more rapid growth in Mexican exports than U.S. exports. Consequently, Mexico’s trade deficit vis-à-vis the U.S. before 1994 became a surplus starting in 1995, when its trade surplus reached US$11.525 billion (Figure 2). The rapid growth of Mexican exports resulted in a surplus of US$149.682 billion in 2018, more than ten times the 1995 surplus. However, the success of Mexican exports to the U.S. was, in large part, the result of increasing manufacturing exports from U.S. companies operating in Mexico. It is worth mentioning that by 2020, the contraction of economic activity in the U.S., and particularly in Mexico, markedly reduced Mexican imports, thus increasing its surplus vis-à-vis the U.S. to US$162.57 billion.

Source: Developed by the author with information from the ^{United Nations Comtrade data base (n. d.)}.

Figure 2 MEXICO-U.S. TRADE (1994-2020) 

Table 1 shows that, between 1994 and 2019, the industries with the highest growth rate for Mexican exports to the U.S. were railway parts (16.2 percent), vehicles (11.6 percent), beverages (11.84 percent), optical and medical instruments and parts (11.0 percent), and machinery and mechanical appliances (10.02 percent). However, the bulk of Mexican exports were concentrated in vehicles and auto parts. In 2019, the value of those exports was US$100.6 billion. Another important group of Mexican exports is made up of electrical machinery, equipment, and parts, with a 2019 export value of US$68.9 billion. Those exports are closely related to the North American supply chain encouraged by U.S. multinational firms in Mexico. Other outstanding exports of Mexico are machinery and mechanical appliances, medical instruments, and fruits and vegetables. It should be noted that the average annual growth of Mexican exports in the period was 7.8 percent between 1994 and 2019. However, this significant expansion experienced an unforeseen break due to the covid-19 pandemic, showing a significant annual decrease in 2020.

U.S. exports to Mexico grew at an average annual rate of 5.3 percent between 1994 and 2019 (Table 2). Specifically, the following products grew rapidly: mineral fuels (13.8 percent), vehicles (8.5 percent), chemical products (7.9 percent), and iron and steel (6.8 percent). In 2019, the major imports from the U.S. in terms of value were mineral fuels (US$36.0 billion); nuclear reactors, machinery, and mechanical appliances (US$28.49 billion); followed by electrical machinery and parts (US$21.16 billion) and vehicles and auto parts (US$21.16 billion), respectively (Table 2). An analysis of Mexican imports from the U.S. indicates that they are based on comparative advantages and the rapid development of global supply chains. Those imports corroborate that U.S.-Mexican trade is predominantly based on intra-industry trade in the electrical, automobile, and oil industries. However, the covid-19 pandemic had a negative effect on the trend of U.S. exports to Mexico in 2020.

The success of NAFTA for Mexico and the U.S. was related to three important factors: the development of production supply chains in the manufacturing sector, the two countries’ different natural endowments, and the differential in the levels of education, labor skills, and wages between the workers of the U.S. and Mexico.

The wage differential is an incentive for trade and investment and compensates for the growing difference in labor productivity between the U.S. and Mexico.³ The new technology developments in communications, computers, and the Internet have resulted in a segmentation of the production process; firms and businesses have taken advantage of the wage differentials to establish production processes using intensive unskilled labor in Mexico (^{Robertson, 2018}).

As a result, FDI and trade substantially increased after the establishment of NAFTA. The most important segment of that trade has been related to manufacturing. Exports from Mexico have complemented the U.S. manufacturing industry by trading manufacturing parts and components that have been used in the U.S. to gain competitiveness. This process has encouraged the development of integrated supply chains among NAFTA members.

Mexico and U.S. Trade at the Regional Level

Many studies in the literature delve into the positive effects of NAFTA on trade between the U.S. and Mexico and on both countries’ economic performance. ^{Hillberry and McDaniel (2002)} analyzed the characteristics of trade growth between the U.S., Canada, and Mexico with results that revealed a broadening of international trade activity in North America. Although U.S. industries have faced competition from Mexican imports, U.S. consumers and manufacturers have had access to imports from Mexico at a lower cost. ^{Burfisher, Robinson, and Thierfelder (2001)} pointed out small but positive effects for the U.S. economy. ^{Waldkirch (2010)} considered that NAFTA and FDI have positively impacted productivity and wages in Mexico. ^{De la Cruz and Riker (2014)} studied the impact of NAFTA on U.S. labor markets, using a CGE model with data on NAFTA preference margins. They found small but positive effects on the real wages of skilled workers in the United States.

However, to understand the full impact of the establishment of NAFTA, it is important to study its regional impacts on trade. In 2019, the total trade of the most important U.S. state trading partners with Mexico reached US$438.1 billion. The total value of exports to Mexico was US$180.5 billion, while imports were US$257.6 billion. Around two-thirds of the trade between the U.S. and Mexico is concentrated in four states: Texas, California, Michigan, and Illinois, which accounted for 61.7 percent of the total. In particular, Texas and California stand out as Mexico’s major trading partners (Table 3). The factors explaining the intense trade of those states with Mexico are related to the size of their economies, their relative proximity to Mexico, and the characteristics of the commodities traded with that country.

The major states of Mexico receiving exports from the U.S. in 2018 were the northern border states of Chihuahua, with 14.1 percent; the state of Nuevo Leon, with 8.4 percent; and Tamaulipas, with 5.7 percent. Other important trading regions were the state of Mexico, with 9.4 percent and Mexico City, with 7.5 percent (^{BTS, n.d.}).⁴

Regarding the major exports of U.S. states to Mexico in 2019, exports of oil and derivatives from the state of Texas have the highest value, with US$24.6 billion (Table 4) and represented 9.6 percent of the U.S. exports to Mexico (^{USCB, n.d. b}). The increasing energy trade and investment reflects the integration of that industry in the North American region. NAFTA Chapter 6 eliminated tariffs and quotas in the energy sector but left special provisions for Mexico, allowing it to prohibit foreign direct investment in exploration and distribution of crude, natural gas, and electricity (^{Hufbauer and Jung, 2017}). In 2013, Mexico passed constitutional reforms allowing FDI in exploration, refining, transport, and storage of crude and natural gas. The establishment of the USMCA could deepen the legal certainty for investments, thus encouraging further growth in that sector.

A second group of U.S. exports that reflects the increasing integration of global supply chains involves the automobile industry. In 2019, Texas was the major exporter of vehicles, engines, and auto parts followed by Michigan, which mainly exported piston engines, with a value of exports of US$7.1 billion and US$3.0 billion, respectively (Table 4). The automobiles and auto parts exported by U.S. states showed that they were also involved in the supply chain for auto production in the North American region. Mexico also exports automobile components and vehicles to the United States. In 2018, Mexico’s automobile industry exported US$29.8 billion to Michigan, US$12.2 to Texas, and US$11.5 billion to California (Table 5).

Auto industry trade between the U.S. and Mexico demonstrates the significance of intra-industry commerce as a part of U.S.-Mexico economic integration. The increasing interconnection of trade among countries arising from global supply chains is derived from the fragmentation of production. Since the mid-1990s, global economic integration has accelerated, through increasingly complex trade relations and global supply chains that create value throughout the production and distribution processes in different countries of the world, particularly in the auto, electronics, and computer industries. According to data estimated by the Organisation for Economic Co-operation and Development (OECD), significant trade exists in value added, which represented 15.6 percent of total U.S. trade and of 46.9 percent of Mexican trade.

Another group of U.S. exports to Mexico consists of electronics, computers, and electronic circuits. Parts and accessories for computers and electric machinery and parts were Texas’s second and third major exports and California’s most important exports to Mexico (Table 4). This underlines the relevance of the states of Texas and California in the production of electronics and computer-related goods in the United States and the Mexican economy’s increasing demand for these manufactures.

In 2019, U.S. states’ main imports from Mexico included automobiles and auto parts for Michigan, Texas, and California, valued at US$42.6 billion, US$15.9 billion, and US$12.5 billion, respectively. In addition, Mexico also exported electronic products to the U.S. In particular, computers and parts and electric machinery, equipment, and parts were exported, with a value of US$20.5 billion and US$10.1 billion, respectively (Table 5). Like exports, in 2020, U.S. imports from Mexico dropped sharply as a result of the coVid-19 pandemic. However, toward the end of that year both exports and imports tended to increase.

Major North American Manufacturing Sectors Under the USMCA

Econometric Results

The econometric model estimated the underlying determinants of trade between Mexico and the United States, in particular structural aspects such as the size of the economies of the U.S. states, the distance between the states and the border with Mexico, and other control variables, such as fdi in Mexico and the average tariff rates applied in the period under analysis. Given the characteristics of both countries’ economies, it is assumed that factors such as transportation costs and the size of economic activity are determinants for the growth of trade, in a context of trade integration and low tariff rates.

The methodology of estimation was based on a panel data model with fixed and random effects. To determine which econometric model was more appropriate for the panel data set, a Hausman test for the correlated random effects was estimated. The null hypothesis assuming that the effects are not correlated with other regressors was rejected. Therefore, the fixed effect model was preferred to the random effect model. In addition, a Breusch-Pagan test to check for heteroscedasticity was estimated. The null hypothesis of equal variance and distribution of the residuals was rejected, and therefore the test corroborated the pertinence of the fixed model (Table 6).

The results indicated that trade between the states of the U.S. and Mexico was positively impacted by the size of the GDP of the U.S. states. The coefficients were statistically significant at 5 percent and 1 percent for the fixed and random effects models (Table 7). This result highlights the importance of the U.S. states in trade with Mexico. Particularly, those that have a significant commercial flow based on proximity and global value chains.

The panel models exhibited a negative and statistically significant sign of the coefficients of the variable distance, suggesting that the distance between the U.S. states and Mexico plays an important role in trade with Mexico. Thus, the closer the U.S. states are to the Mexican border, the higher the level of trade. This is the case for Texas, California, and Arizona, which are relatively close to Mexico. This also follows the gravitational equation assumptions. The coefficients of the two models were negative and statistically significant and, therefore, support the assumption that if the distance is shorter there will be increased trade between regions, underlining the importance of transportation costs.

The coefficients of the average tariffs applied by both the United States and Mexico were positive. However, only the coefficient of the tariffs applied by the U.S. was statistically significant. These results suggest that tariff reduction has been an important factor in the growth of trade between the two countries, both during the NAFTA and the USMCA periods. In addition, per capita GDP of the United States and Mexico were included as explanatory variables to observe the effect of the size of both economies, considering the size of the population. The U.S. per capita GDP coefficient was positive and statistically significant, suggesting that the U.S. economy is a major driving force for growth for the two countries. For its part, Mexico’s per capita GDP coefficient for Mexico was also positive, but not statistically significant.

Finally, the amount of total foreign direct investment in Mexico was used as a variable that reflects the growing economic integration of the United States and Mexico and its impact on both countries’ trade. This variable’s coefficient was positive and statistically significant at 10 percent of the confidence level. For both the U.S. and Mexico, trade and FDI are closely related through intra-industry trade and global value chains.

Therefore, the econometric estimation suggests that the proximity of the U.S. states to the Mexican border, characterized by manufacturing and maquiladora activities, and the level of economic activity of the U.S. states have been important driving forces of U.S.-Mexico trade. In addition, the estimations support the conclusion that low tariffs will continue to promote trade and that the success of the USMCA will depend on the manufacturing sector’s capacity in the U.S. and Mexico to supply the inputs necessary to increase the North American value content of exports.

Concluding Remarks

The modification of NAFTA to establish the USMCA imposed new challenges and the need to restructure the North American supply chains to meet the new input content requirements proposed in the agreement’s rules of origin section. A major impact of NAFTA was the reciprocal dismantling of the tariff structure within the region composed of the three member countries. Mexico lowered its rates significantly, which, before NAFTA, were higher than those imposed by the United States and Canada. The tariffs were gradually eliminated in some cases, such as for chemicals, electrical parts, textiles, etc. The removal of tariffs was conditioned on the requirements of the rules of origin that established a minimum content of 62.5 percent of the value produced in the NAFTA region.

The research and statistical evidence indicate that NAFTA resulted in increased trade and investment in the North American region. Particularly, trade between the U.S. and Mexico expanded rapidly. Exports from the United States to Mexico focused on oil, electrical equipment, vehicles, and chemicals. U.S. imports from Mexico were mainly vehicles, electrical machinery, and mechanical appliances, as well as fruits and beverages. The result of this trade expansion was related to the development of value chains in manufacturing, differing endowments of factors of production and low transportation costs.

The states of Texas and California have traded predominantly with Mexico, based on their diverse economic activity and natural resources. The commercial relationship in this sector has been fundamental for the economic integration of those states with the Mexican economy. Also, the automobile trade between Mexico and the states of Michigan and Texas reflects the importance of intra-industry trade. Finally, the electronics and computer sector has seen an increase after the establishment of NAFTA, indicating its importance to trade between the two economies.

The new USMCA agreement modernizes intellectual property rules and government procurement and adds changes to the rules of origin. In particular, the automotive sector will gradually increase regional content from 62.5 percent to 75 percent. In addition, it introduces the concept of labor requirements, which requires that wages of US$16 an hour must be paid by at least 40 percent of Mexican automobile exports to avoid tariffs. This is still a factor that could potentially affect trade between the U.S. and Mexico, given the differences in labor productivity in both countries.

Tariff reduction under NAFTA has had a positive impact and has deepened supply chain integration between Mexico and the United States. The establishment of the USMCA could potentially increase investments in the automobile, steel, aluminum, and electronic industries in the U.S. and also could attract more FDI to Mexico, which would be necessary to comply with the agreement’s value content requirement. The USMCA tariffs could have a positive effect on trade, depending on the fulfillment of value content requirements.

The results of the panel model estimated suggest that the low tariffs have played an important role in North American region economic integration. The model confirmed that foreign direct investment has a positive effect on trade between the two countries through the link between investment and non-industrial trade. In addition, the model’s findings indicate that a determining factor for trade between the two countries is related to the size of economic activity in the United States at both the national and state level. Distance and economic activity have also contributed to the intensification of trade between the U.S. states and Mexico. Therefore, it can be concluded that polices to encourage trade between the U.S. and Mexico would require the development of communications and transportation infrastructure to take advantage of the relatively short distance between the U.S. border states and Mexico and to be able to reduce the transportation costs for trade.