Read an Excerpt

Mexican Immigration to the United States
The University of Chicago Press Copyright © 2007 the National Bureau of Economic Research
All right reserved.
ISBN: 978-0-226-06632-5



Chapter One The Evolution of the Mexican-Born Workforce in the United States

George J. Borjas and Lawrence F. Katz

1.1 Introduction

The population of Mexican-born persons residing in the United States has increased at an unprecedented rate in recent decades. This increase can be attributed to both legal and illegal immigration. During the entire decade of the 1950s, only about three hundred thousand legal Mexican immigrants entered the United States, making up 12 percent of the immigrant flow. In the 1990s, 2.2 million Mexicans entered the United States legally, making up almost 25 percent of the legal flow (U.S. Immigration and Naturalization Service 2002). In addition, it is estimated that (as of January 2000) there were 7 million illegal aliens residing in the United States, with 4.8 million (68 percent of this stock) being of Mexican origin (U.S. Department of Commerce 2004). As a result of the increase in the number of legal and illegal Mexican immigrants, nearly 9.2 million Mexican-born persons resided in the United States in 2000, comprising about 29.5 percent of the foreign-born population (U.S. Bureau of the Census 2003).

It is instructive to place the Mexican immigrant influx of the late twentieth century in the context of earlier immigrant flows. In 1920, toward the end of the first great migration, the largest two national origin populations enumerated by the 1920 Census were Germans and Italians, and together these two populations comprised about 23.7 percent of the foreign-born population at the time (U.S. Bureau of the Census 1975). From this perspective, it is clear that the Mexican-born population of the late twentieth century is historically unprecedented, being both numerically and proportionately larger than any other immigrant influx in the past century.

This paper analyzes the evolution of Mexican immigration as a component of the U.S. workforce during the twentieth century. As a result of the rapidly increasing Mexican immigrant influx described earlier, the fraction of the workforce composed of Mexican-born workers increased rapidly after 1970. As the top panel of figure 1.1 shows, only 0.4 percent of the workforce aged eighteen-sixty-four in 1970 was composed of Mexican-born workers. By 2000, the Mexican immigrant share had increased to 4.0 percent. The increase is even larger in the male workforce, where 0.5 percent of working men were Mexican-born in 1970 and 5.1 percent in 2000.

It is of interest to contrast the explosion of Mexican workers in the U.S. workforce in the late twentieth century with the demographic trends at the beginning of the century. Although Mexican immigration was relatively small in the early 1900s, the relative number of Mexican immigrants in the U.S. workforce increased to 0.6 percent in 1920 (and continued rising until the late 1920s). The halting of European migration to the United States with the outbreak of World War I followed by Congressional action to restrict immigration combined with strong labor demand in the booms of the late 1910s and the 1920s engendered substantial efforts by U.S. employers to recruit Mexican laborers through private labor contractors (Massey, Durand, and Malone 2002). Remarkably, the Mexican immigrant share went into a long steady decline after the 1920s that lasted for several decades. It was not until the 1970s that the Mexican immigrant share of the workforce was at least as large as it was in the 1920s!

The reasons for the declining Mexican share in the workforce are not entirely clear. Until 1965, there was not a numerical limitation on immigration from countries in the Western Hemisphere. In theory, at least, legal migration from Mexico was guided by a "first-come, first-served" approach. Potential immigrants applied for entry and local consular officials had a great deal of discretion in determining which applicants would be provided entry visas.

To ease the labor force shortage caused by World War II in the agricultural industry during the early 1940s, the Bracero Program was launched on August 4, 1942. This guest-worker program brought almost 5 million Mexican-born farm workers to the United States between 1942 and 1964, when it was abruptly terminated by the United States. The main reason given for the discontinuation of the program at the time was the assertion that the Bracero Program depressed the wages of native-born Americans in the agricultural industry (Massey and Liang 1989; Marcell 1994).

The latest wave of illegal immigration from Mexico began in the late 1960s, after the discontinuation of the Bracero Program. There is, in fact, a clear link between the end of the Bracero Program and the beginning of the illegal alien flow, at least as measured by the number of Mexican nationals aliens apprehended as they attempt to enter the United States illegally. The number of Mexican illegal aliens apprehended by the Border Patrol began to increase soon after the Bracero Program ended. In 1964, for example, the Border Patrol apprehended only 41.6 thousand Mexican illegal aliens. By 1970, apprehensions were up to 348.2 thousand annually. In 1986, about 1.7 million Mexican illegal aliens were apprehended (U.S. Immigration and Naturalization Service, various issues).

Although the discontinuation of the Bracero Program may help explain why illegal immigration accelerated in the 1960s and 1970s, there are several questions that remain unanswered. The wage gap between Mexico and the United States has been large for many decades, and it is far from clear that it is larger now than it was at the beginning or middle of the twentieth century. Why then didn't we observe large flows of Mexican immigrants prior to the 1970s? It is possible, of course, that the policy changes initiated by the 1965 Amendments and subsequent legislation, which made family reunification the central goal of immigration policy, could have eased the entry of Mexicans in the United States, but, at least in theory, Mexican immigration was not greatly restricted prior to the post-1965 policy shifts. Why did so few Mexicans take advantage of it? Or were the institutional barriers placed at the consular level in Mexico so forbidding that relatively few Mexicans even bothered to apply to enter the United States?

We do not know the answers to these questions. What we do know, however, is that the Mexican immigrant population today stands out from the rest of the immigrant population in two striking ways. It is well known, of course, that there has also been a sizable increase in the number of non-Mexican immigrants admitted to the United States. Nevertheless, Mexican immigrants comprise an ever-larger fraction of the foreign-born stock of the United States (see figure 1.2). Second, as we will document, Mexican immigrants tend to have demographic and socioeconomic characteristics that differ significantly not only from that of the native-born population, but from that of other immigrants as well. In general, the economic performance of Mexican immigrants lags significantly behind that of other immigrant groups, and this lagging performance is, to an important extent, transmitted to future generations of native-born workers of Mexican ancestry.

This paper differs from earlier contributions in the immigration literature by focusing specifically on the evolution of the Mexican-born workforce in the United States. We use data drawn from the Integrated Public Use Microdata Samples (IPUMS) of the U.S. Decennial Census throughout the entire twentieth century to describe the demographic and economic evolution of this population. The paper examines the evolution of the relative skills and economic performance of Mexican immigrants and contrasts this evolution to that experienced by other immigrants arriving in the United States during the period. The paper also examines the costs and benefits of this influx by examining how the Mexican influx has altered economic opportunities in the most affected labor markets and by discussing how the relative prices of goods and services produced by Mexican immigrants may have changed over time.

1.2 Data and Key Trends

The analysis uses data drawn from all of the available IPUMS of the U.S. Decennial Census between 1900 and 2000. This long-term look at the available data helps to provide a historical account of the evolution and economic performance of Mexican immigrants in the U.S. workforce.

Throughout the analysis, a person is classified as an "immigrant" if he or she was born in a foreign country; all other workers are classified as "natives." Persons who are immigrants and who were born in Mexico comprise the sample of Mexican immigrants. The pre-1970 censuses comprise (roughly) a 1 percent random sample of the population. Beginning with 1980, the data comprise a 5 percent random sample of the population. The entire available sample in each census is used in the empirical analysis.

In each census, the study is restricted to persons aged eighteen-sixty-four who work in the civilian sector, are not enrolled in school, and do not reside in group quarters. When appropriate, the sampling weights reported in the IPUMS data are used in the calculations.

1.2.1 The Geographic Sorting of Mexican Immigrants

Table 1.1 begins the empirical analysis by documenting how the geographic sorting of Mexican immigrants in the United States changed over the twentieth century. The top panel of the table reports the share of the stock of Mexican immigrants (both male and female) who reside in a particular state at a particular point in time, while the bottom panel reports the fraction of the state's workforce that is composed of Mexican immigrants.

The top panel of table 1.1 reports an important trend: a steady and substantial redistribution in Mexican immigration from Texas to California throughout much of the twentieth century. In 1900, for example, 62.5 percent of Mexican immigrants lived in Texas, and only 7.8 percent lived in California. By 1950, roughly equal numbers of Mexican immigrants lived in Texas (39.1 percent) and California (40.3 percent). By 1980, almost 60 percent of Mexican immigrants lived in California, and the fraction of those living in Texas had further declined to 21.2 percent. Between 1980 and 2000, however, California seemed to become a relatively less attractive destination for Mexican immigrants. By 2000, the fraction of Mexican immigrants living there had declined to 42.1 percent. Note, however, that this decline was not accompanied by an increase in the fraction choosing to reside in Texas; that share was relatively constant over the period.

Table 1.1 also shows that the recent decline in the relative share of Mexican immigrants who choose to live in California has been accompanied by a remarkable increase in Mexican immigration to states that had never been the recipients of large numbers of these immigrants. In 1980, for example, Mexican immigrants had, at best, a negligible presence in both North Carolina and Georgia. By 2000, however, almost 3 percent of the workforce in each of these states was composed of Mexican immigrants. Similarly, less than 1 percent of workers in Colorado were Mexican-born in 1980; by 2000, almost 5 percent of Colorado's workforce was Mexican-born. Although often noted in the popular press, this remarkable and sudden shift in the geographic sorting of Mexican immigrants in the United States has received little systematic analysis, and the reasons leading to the dramatic geographic redistribution are still not well understood.

Because there were relatively few Mexican immigrants living in the United States at the beginning of the twentieth century, it is worth noting that even though nearly two-thirds of Mexican immigrants lived in Texas in 1900, only 3.4 percent of the Texas workforce was Mexican-born. By 2000, however, nearly 14.8 percent of the California workforce and 10.9 percent of the Texas workforce were Mexican-born. The relative importance of Mexican immigration as a component of the workforce of the main immigrant-receiving states, therefore, now stands at a historic high. The growth has been most dramatic in California. In 1970, only 2.4 percent of California's workforce was Mexican-born. By 2000, this statistic had increased sixfold, to 14.8 percent.

1.2.2 Trends in Educational Attainment and Occupation

The skill composition of the Mexican immigrant workforce differs strikingly from that of the native workforce as well as from that of other immigrants. We begin the description of the skill composition of the various groups by comparing the trend in the educational attainment of native working men with that of Mexican immigrant men. The census provides data on educational attainment beginning in 1940, so that this phase of the study focuses on the trends in the post-1940 period. As table 1.2 shows, 67.3 percent of male native-born working men were high-school dropouts in 1940. This high native dropout rate was lower than that of Mexican immigrant men, where 94.6 percent had not completed high school. To provide a point of reference for these statistics, the table also reports that 84.4 percent of non-Mexican immigrant working men at that time were high school dropouts.

By 2000, the fraction of male native-born workers who are high school dropouts had fallen by almost 60 percentage points, to 8.7 percent. In contrast, the fraction of Mexican-born high school dropouts had fallen by only about 30 percentage points, to 63 percent. Again, as a reference point, note that the fraction of high school dropouts in the non-Mexican immigrant population had fallen by almost as much as in the native-born workforce, to 17 percent. As a result of these trends, the data indicate a remarkable fact: the population of male high school dropouts in the United States has become disproportionately Mexican-born. In 1940, 0.5 percent of all male high school dropouts were Mexican immigrants. Even as recently as 1980, only 4.1 percent of male high school dropouts were Mexican immigrants. By 2000, however, 26.2 percent of all male high school dropouts were Mexican-born.

The growing disadvantage of Mexican immigrants at the bottom of the educational attainment distribution is matched by an equally growing disadvantage at the top of the distribution, where a fast-growing number of native workers and non-Mexican immigrants are college graduates. In 1940, there was relatively little difference in college graduation rates among the three groups; by 2000, however, there is a wide gulf separating college graduation rates between Mexican immigrants and the other groups. In particular, 6.3 percent of native working men were college graduates in 1940, and this fraction had quadrupled to 27.4 percent by 2000. Similarly, 3.7 percent of non-Mexican immigrant men were college graduates in 1940, and this fraction had increased almost tenfold to 36.3 percent by 2000. In contrast, only 1.4 percent of Mexican immigrant men in 1940 were college graduates; by 2000, the college graduation remained a minuscule 3.4 percent in this group of workers.

The bottom panel of table 1.2 reports the trends in the education distribution for working women. The trends are similar to those reported for the various groups of working men, though not as dramatic. For example, the high school dropout rate of native women dropped by 44 percentage points between 1940 and 2000 (from 50.6 to 6.5 percent), as compared to the almost 60 percentage point drop experienced by native men. Similarly, the high school dropout rate for Mexican immigrant women dropped by 28 percentage points (from 84.5 to 57.0 percent), as compared to the 32 percentage point drop experienced by Mexican immigrant men. These data patterns presage a systematic finding in much of our analysis: the differences exhibited by the various groups of working women mirror those exhibited among the respective groups of working men but are less extreme. As a result of this similarity, much of the discussion that follows will focus on the trends observed in the sample of working men (even though many of the tables will report the respective statistics for working women). By focusing on the trends exhibited by working men, we can avoid the difficult conceptual and econometric issues introduced by the interpretation of skill and wage trends for working women during a period of rapidly rising female labor force participation rates.

Finally, table 1.3 illustrates the changing occupational distribution of Mexican immigrants by listing the "Top Ten" occupations employing these workers. The IPUMS data recode the very different occupation codes used by the various censuses into a single occupation categorization based on the 1950 Census definitions. We use this simplifying recoding to compare the occupation distribution of workers across censuses.

(Continues...)

Excerpted from Mexican Immigration to the United States Copyright © 2007 by the National Bureau of Economic Research. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.