Issues in the degree of economic dependence between Texas and Mexican economies
Three sets of variables illustrate the high degree of economic interdependence: One is the disproportionately large share of Texas' exports in US-Mexico merchandise trade. Two is the high degree of dependence of the Texas economy on trade with Mexico. And three is the growth of global value chains and the coordination of intermediate production stages internationally in the United States and Mexico. Each of these three sets of variables are explained in slightly more detail.
First, it is well known that the state of Texas is the origin of a disproportionately large share of US merchandise goods exports to Mexico. In 2018, 41.3 percent of US merchandise goods exports to Mexico originated in the state. The pattern has been robust over time. In 1999, Texas accounted for a similar 43.5 percent of US merchandise trade with Mexico (Department of Commerce, 2019).
Several factors cause the state of Texas to play this role. It is located on the border with Mexico, it has multiple infrastructure linkages in the form of highways, pipelines, and rail, it is close to Mexico's industrial center in Monterrey, it has direct infrastructure connections to Mexico City and other production centers in central Mexico, and the Texas port of Huston is a convenient transshipment node for shipping goods on water to and from Mexico's east coast. In combination, these factors cause the Texas economy to be extremely dependent on international trade in general, and trade with Mexico specifically. Texas' ratio of merchandise exports-to-GDP is 17.8 percent.[1] For the US as a whole, the ratio is 8.1 percent. More specifically, merchandise exports from Texas to Mexico, as a share of Texas GDP, is 6.2 percent and by far the largest percentage for any state (International Trade Administration, 2019; Bureau of Economic Analysis, 2019).
Another key variable in the growth of both US-Mexico trade and Texas-Mexico trade is the development of global value chains. The ability of manufacturers to locate different stages of production in different places and even different countries has grown enormously with the communication and transportation revolutions that were building over the last decades of the 20th century and especially in the 1990s. Communications revolutions enabled new movements of large quantities of precise information and made it possible to coordinate production in multiple plants simultaneously. A key result was the cutting up of manufacturing and agricultural value chains and the location of different stages of production in different places where each stage and location reflects the specific comparative advantages of its production site.
Economists and policy makers are only beginning to explore the consequences of the development of global value chains. The OECD and WTO, for example, have collaborated to measure international trade using value added concepts rather than gross output concepts (World Trade Organization, 2019a). From this work, we know that 46.8 percent of the value of Mexico's manufactured exports is created abroad, and mostly in the United States. In the highly integrated automobile industry, the 52 percent of the value of Mexican car and car parts exports is created outside the country (World Trade Organization, 2019b).
The creation of a robust and profound Texas-Mexico relationship is implied but not demonstrated conclusively by the fact that the top three exporting federal entities (Mexican states) are on the Texas border: Chihuahua, Coahuila, and Nuevo Leon (INEGI, 2019). Furthermore, the single largest category of merchandise exports from Coahuila and Nuevo Leon is from the subsector 336, fabricación de equipo de transporte, which is to say cars and car parts. This is also the second largest export category from Chihuahua, after subsector 334, Fabricación de equipo de computación, comunicación, medición y de otros equipos, componentes y accesorios electrónicos (computer, communication, measurement and other electronic equipment, components and accessories), which, like cars and car parts, also has a high share of foreign value added. While these patterns do not prove a high degree of integration between the state of Texas on the one hand and Mexico and a number of Mexican border states on the other, they are highly suggestive and worthy of further exploration.
Deep and important ties between the Mexican economy and Texas did not just happen accidentally. Public policies at the state level, along with national policies in the United States and Mexico, played important roles. For example, Gerber (2007) shows how at the end of the 1980s, Texas policy makers began to re-orient state policies and businesses towards trade with Mexico as a partial solution to the economic problems the state experienced in the 1980s. This took place in the context of growing US concerns about international competitiveness and legislation to promote exports, along with the historical opening of Mexico's economy.
[1] Note that this is for merchandise exports to all markets combined.
First, it is well known that the state of Texas is the origin of a disproportionately large share of US merchandise goods exports to Mexico. In 2018, 41.3 percent of US merchandise goods exports to Mexico originated in the state. The pattern has been robust over time. In 1999, Texas accounted for a similar 43.5 percent of US merchandise trade with Mexico (Department of Commerce, 2019).
Several factors cause the state of Texas to play this role. It is located on the border with Mexico, it has multiple infrastructure linkages in the form of highways, pipelines, and rail, it is close to Mexico's industrial center in Monterrey, it has direct infrastructure connections to Mexico City and other production centers in central Mexico, and the Texas port of Huston is a convenient transshipment node for shipping goods on water to and from Mexico's east coast. In combination, these factors cause the Texas economy to be extremely dependent on international trade in general, and trade with Mexico specifically. Texas' ratio of merchandise exports-to-GDP is 17.8 percent.[1] For the US as a whole, the ratio is 8.1 percent. More specifically, merchandise exports from Texas to Mexico, as a share of Texas GDP, is 6.2 percent and by far the largest percentage for any state (International Trade Administration, 2019; Bureau of Economic Analysis, 2019).
Another key variable in the growth of both US-Mexico trade and Texas-Mexico trade is the development of global value chains. The ability of manufacturers to locate different stages of production in different places and even different countries has grown enormously with the communication and transportation revolutions that were building over the last decades of the 20th century and especially in the 1990s. Communications revolutions enabled new movements of large quantities of precise information and made it possible to coordinate production in multiple plants simultaneously. A key result was the cutting up of manufacturing and agricultural value chains and the location of different stages of production in different places where each stage and location reflects the specific comparative advantages of its production site.
Economists and policy makers are only beginning to explore the consequences of the development of global value chains. The OECD and WTO, for example, have collaborated to measure international trade using value added concepts rather than gross output concepts (World Trade Organization, 2019a). From this work, we know that 46.8 percent of the value of Mexico's manufactured exports is created abroad, and mostly in the United States. In the highly integrated automobile industry, the 52 percent of the value of Mexican car and car parts exports is created outside the country (World Trade Organization, 2019b).
The creation of a robust and profound Texas-Mexico relationship is implied but not demonstrated conclusively by the fact that the top three exporting federal entities (Mexican states) are on the Texas border: Chihuahua, Coahuila, and Nuevo Leon (INEGI, 2019). Furthermore, the single largest category of merchandise exports from Coahuila and Nuevo Leon is from the subsector 336, fabricación de equipo de transporte, which is to say cars and car parts. This is also the second largest export category from Chihuahua, after subsector 334, Fabricación de equipo de computación, comunicación, medición y de otros equipos, componentes y accesorios electrónicos (computer, communication, measurement and other electronic equipment, components and accessories), which, like cars and car parts, also has a high share of foreign value added. While these patterns do not prove a high degree of integration between the state of Texas on the one hand and Mexico and a number of Mexican border states on the other, they are highly suggestive and worthy of further exploration.
Deep and important ties between the Mexican economy and Texas did not just happen accidentally. Public policies at the state level, along with national policies in the United States and Mexico, played important roles. For example, Gerber (2007) shows how at the end of the 1980s, Texas policy makers began to re-orient state policies and businesses towards trade with Mexico as a partial solution to the economic problems the state experienced in the 1980s. This took place in the context of growing US concerns about international competitiveness and legislation to promote exports, along with the historical opening of Mexico's economy.
[1] Note that this is for merchandise exports to all markets combined.
Methodology and preliminary result
Given the last issues about the economic dependence, in this research we study the spillover effects between the GDP growth rates of Texas against the economic activity of the states in Mexico (this can also be extended to particular industries and sectors of these regional economies as mentioned before). However, the spillovers if any, can depend on the phases of the economic cycle fluctuations (recession or expansion). Moreover, the magnitude effects might be different among the states of the Mexican economies versus Texas, e.g., both can be in recession, both in expansion or both at different phases. This evidence can be useful to evaluate the magnitude of shocks that can affect the persistence of the economic dependence between these regional economies, e.g., breaking down the free trade agreement or strong depreciation/appreciation of the coins.
To this goal, the suggested econometric model considers bivariate autoregressive process (VAR) between the growth rates of Texas GDP (i) against the production activity of states in Mexico (j). To capture the changes on the mean and volatility by the different phases of the time series, Markov switching parameters are allowed in the autoregressive dynamics growth rates, and in the conditional heteroskedasticity (ARCH) of the processes as follows:
To this goal, the suggested econometric model considers bivariate autoregressive process (VAR) between the growth rates of Texas GDP (i) against the production activity of states in Mexico (j). To capture the changes on the mean and volatility by the different phases of the time series, Markov switching parameters are allowed in the autoregressive dynamics growth rates, and in the conditional heteroskedasticity (ARCH) of the processes as follows:
The ARCH effects and cross volatility spillovers are given by:
The stochastic hidden indicators of the states (Sit for i=1, 2) capture the different phases of the time series by Markov process of order one. The transmission or spillover effects from Texas (i) to Mexican states (j), are given by the GDP growth rates (Øij) , and their cross volatilities (γij) for i≠j. However, both are state dependent on the economic phases. To estimate the model, we suggest Bayesian simulation methods to overcome issues in estimation by maximum likelihood (sample size, hidden indicators, and model selection among Markov switching models).
Preliminary results
The growth rates of Texas GDP (blue lines) are plotted against the highest correlated (p) growth rates of economic activity from the Mexican states (red lines). The sample covers Q1-2005 to Q1-2019.
In the second plot, the posterior densities of the spillover effects ( and ) are plotted. The effect from Texas to Mexican states (blue lines) and Mexican states to the Texas (red lines), provide the next preliminary results: in most of the pairwise results, there might not be spillover effects in the growth rates, since might no be different from zero (see the first ten plots). However, important volatility spillover effects from Texas to Mexican economies are present (blue lines). Moreover, Jalisco, Baja California and Mexico City, can also affect the volatility of the Texas GDP (red lines). The mixture distribution of these densities supports different magnitude in the spillover effects by the dependence on the phases of the economic cycle fluctuations.