Impoverished conditions on the U.S. side of the Mexican border have recently been identified and analyzed by a host of social scientists and policy makers. Ensuing policy prescriptions generally hinge on promoting business and population expansion in the border region. The presence of local economic growth, however, may not necessarily improve the socioeconomic conditions along the U.S.-Mexico border. For historical perspective, some scholars have noted that the border region experienced economic prosperity in the early 1900s. Indeed, around the turn of the century, an abundance of agricultural and mining opportunities along the U.S.-Mexico border, aided by the expansion of the railroad, created an economic boom that lasted until the late 1920s (Lorey 1999; Fernandez 1989; Foley et al. 1988; Hansen 1981). What remains unclear is whether this expansion improved the labor market income of workers on the U.S. side of the border.
This study utilizes data from the 1900, 1910, and 1920 U.S. decennial censuses in the Integrated Public Use Microdata Series (provided by Ruggles and Sobek 2003) to compare the occupational earnings of male workers along the U.S.-Mexico border to those in the U.S. interior. The empirical results indicate that while workers had similar occupational income between U.S. counties along the Mexican border and the rest of the United States in 1900, a border-earnings penalty developed over the next two decades. Further evidence suggests that this penalty was not solely driven by the relative shift toward agriculture or the increased military presence along the border during that time. These findings provide historical evidence that local economic growth may not be a sufficient means to alleviate poverty along the U.S.-Mexico border.
Background: Border Growth in the Early 1900s
Socioeconomic conditions along the U.S.-Mexico border currently lag behind the rest of the United States. According to the U.S. Census Bureau (2002), for example, twenty-three of the twenty-four counties directly adjacent to Mexico had smaller median family income in 1999 than the U.S. median of $41,994 (the exception being San Diego); median income in seventeen of these twenty-four counties fell below $30,000. Similarly, twenty-three of these twenty-four border counties had larger poverty rates than the U.S. national average of 12.4 percent, and eighteen border counties had poverty rates exceeding 20 percent that year (ten of which were over 30 percent). Finally, the U.S. Bureau of Economic Analysis (2004) located the three metropolitan statistical areas with the lowest per capita personal income in the nation in 2002 on the Texas-Mexico border (McAllen, Brownsville, and Laredo). Explanations for these relatively impoverished conditions include low education levels, cost-of-living differences, hedonic wages reflecting regional preferences, a highly elastic supply of low-skilled Mexican labor, and relative labor immobility of the border's workforce (e.g., Brown and Mora 2004; Flota and Mora 2001; Fullerton 2001; Davila and Mora 2000; Sharp 1998; Peach 1997; Davila and Mattila 1985; Davila 1982; Hansen 1982, 1981; Smith and Newman 1977).
Some border scholars have nevertheless produced qualitative evidence that the lagging socioeconomic conditions might not have existed around the beginning of the twentieth century. For example, the arrival of the railroad in the late 1800s and early 1900s brought tremendous opportunities for growth at that time because it allowed the manufacturing and mining industries (including copper, silver, and zinc) to flourish from El Paso westward through New Mexico and Arizona (Hansen 1981; Fernandez 1989; Lorey 1999; Sumner 2001; Garza 2002). The railroad also increased the accessibility of relatively inexpensive land and labor, which lured many Americans seeking farming and ranching occupations to the lower Rio Grande Valley in Texas (Garza 2002; Garza and Long 2002; Lorey 1999; Foley et al. 1988; Fernandez 1989; Hansen 1981). In fact, the period between 1910 and 1930 has been referred to as the "rainbow era" of that region (Pope 1971).
The growth of the U.S. military stationed along the border starting with the Mexican Revolution in 1910 and continuing with the "border raids" in the mid 1910s (particularly with Pancho Villa's raid on Columbus, New Mexico, in 1916) is another cited factor that contributed to the region's economic growth; indeed, there were over 40,000 troops stationed in the El Paso area alone during that time (Fox 1999). The presence of the U.S. military along the border was furthered by the growing suspicions and distrust of Mexico's relations with Germany (especially after the decoding of the infamous January 1917 Zimmerman telegram) during World War I (Fox 1999; Foley et al. 1988), while the naval base in San Diego, California, provided that city with additional growth at the time (Fernandez 1989).
Prohibition may have been another source of economic prosperity in cities along the U.S.-Mexico border in the late 1910s and 1920s. Many U.S. residents flocked to Texas-Mexico border towns for drinking and gaming entertainment starting in 1918 when the state of Texas went dry and to other border cities with the passage of the Volstead Act (national prohibition) the following year (e.g., Martinez 1996). A few years earlier, Tijuana became increasingly popular when the state of California banned horse racing (Taylor 2002; Fernandez 1989). Despite being located on the Mexican side of the border, many of the saloons and casinos were owned by U.S. citizens (Taylor 2002; Martinez 1996; Hansen 1981), which likely enhanced income on both sides of the border. Moreover, cities such as El Paso were able to capitalize on their location near these "vice" activities in terms of other related events, such as frequently being selected for conventions (Lorey 1999).
As noted above, however, economic growth per se may not automatically translate into improved socioeconomic outcomes. Indeed, an abundance of studies illustrates that economic growth might result in greater income inequality. Following Kuznets' 1955 classic work, for example, the higher savings rates (hence assets) and human capital levels among upper income groups imply that technological advances and urbanization initially lead to greater income inequality. Whether such inequality spawns or hinders additional economic growth (and the redistributive policy implications thereof has been debated in the literature (e.g., Welch 1999; Aghion, Caroli, and Garcia-Penalosa 1999; Alesina and Rodrik 1994; Perotti 1993; Barro 1991), but this issue goes beyond the scope of this paper.
The fact that an established research strand links economic growth with inequality raises questions as to whether the growth that occurred along the U.S.-Mexico border in the early 1900s enhanced labor market earnings vis-a-vis the U.S. interior. There are at least three reasons to suspect it did not. First, much of the growth in the region occurred in the agricultural sector-traditionally a low-wage sector. As a result, even with the creation of additional jobs, average earnings in the border region conceivably declined vis-a-vis the rest of the United States. Second, the major influx of workers into the region pursuing employment or avoiding the Mexican Revolution might have stifled earnings growth. In fact, if the supply of labor increased by more than the demand for labor along the border, wages may have fallen in the region relative to the rest of the United States. Finally, a variety of instances have been noted in which undesirable social conditions, such as segregation and discrimination against Mexican Americans, became widespread along the border after 1900, as increased numbers of settlers from the Midwest and the eastern sections of the United States moved to the region (Garza 2002; Garza and Long 2002; Lorey 1999). Observed outcomes of the segregation include the growth in schooling inequality along ethnic lines, where many Mexican American children were assigned to overcrowded and poor quality schools and often were not expected to continue beyond elementary school (Meier and Stewart 1991, 60-64). The deterioration of both the quality and quantity of human capital levels among the relatively large Mexican American population could have induced earnings inequality between the border and the rest of the United States, particularly if the available jobs along the border demanded few specific skills.
In sum, despite qualitative evidence that American cities along the Mexican border experienced an economic expansion in the early 1900s, the overall effect of such a boom on labor market income remains ambiguous. On one hand, it is often assumed that local economic growth enhances earnings because of rising labor demand. On the other hand, if such growth occurs (1) mainly in low-wage sectors, (2) in conjunction with an even larger increase in the supply of labor, or (3) at the expense of reducing average human capital, relative earnings growth may stagnate or decline.
Insight into this ambiguity for the early 1900s can be provided empirically. One reason for the dearth of quantitative research on border earnings is presumably the absence of reliable historical data. Indeed, large-scale earnings data were not readily available until the 1940 U.S. census; research based on 1950 census data indicates that an earnings penalty existed between states along the border and those in the U.S. interior (Barrett 1966). A recent breakthrough in census data, however, now permits an analysis of a proxy for labor market earnings at the turn of the century in counties on the U.S. side of the Mexican border. The Integrated Public Microdata Series (IPUMS), provided by Ruggles and Sobek 2003, contains information on occupational earnings and socioeconomic status in 1900, 1910, and 1920. Public-use microdata samples based on the 1930 census were not available during...