Lack of work does not appear to be the main reason why migrants leave Mexico for the United States. Most migrants in the survey reported that they were gainfully employed before they left and, in fact, their employment experience reflects that of the overall labor force in Mexico. The data show that unemployment in Mexico played an even lesser role in the migration decision of more recently arrived respondents. To the extent that work experiences in Mexico are a motive for migration it is more likely that underemployment, not unemployment, is the main reason.

The work history of migrants in Mexico also reveals a strong concentration in a handful of industries. The primary background of these migrants is in agriculture. That is especially true of men, the elderly, those who did not attend or graduate from high school, and those who arrived in the U.S. more than 15 years ago. Newly arrived respondents are more likely to have worked in construction, manufacturing and commerce in Mexico. Despite the changing mix of industries, the overall background of respondents is in farming, production and other blue-collar work.

The employment experience in Mexico

The employment experience of respondents before migration was determined through questions on their principal industry and occupation of employment in Mexico. Those who answered that they “did not work” constituted the pool of migrants who were potentially unemployed before they left for the U.S. Of those migrants who provided valid answers to the question on industry of employment, 11% of men and 29% of women indicated they did not work in Mexico. For men and women combined, 18% reported not working in Mexico. The proportion who apparently did not work in Mexico appears high at first glance but it is necessary to rule out respondents who do not meet the normal criteria for unemployment as explained below. When that is done, no more than 11% of respondents, perhaps fewer, are estimated to have been unemployed in Mexico before migrating to the U.S.

To be counted as an unemployed worker a person must be actively looking for work. Otherwise, the worker is deemed to be outside the labor force. Examples of adults who might not be members of the labor force are students, housewives and retirees. Other criteria, such as, hours worked in the survey week, are also normally applied before a worker is counted among the unemployed in government-run surveys used to calculate unemployment rates. However, this is more difficult to do when exploring work experience in the past, sometimes years past, as was the case in this survey when Mexicans in the United States were asked about their pre-migration experiences. Among those who did not work in Mexico, it is possible to identify the respondents who are students, housewives and retirees. Excluding them from the labor force yields a more accurate estimate of potential unemployment among respondents to the survey.

The occupational status of workers who responded that they did not work to the question on industry in Mexico is shown in Table 1. As shown, 738 respondents indicated not working in Mexico. Nearly one-half of these workers—45%—said that they were housewives, students or retirees. Those workers can be excluded from the labor force as well as from the ranks of the unemployed. When that is done, the estimated unemployment rate among respondents, prior to the move to the U.S., drops from 18% to 11%.

It is also worth noting that 30% of workers failed to provide a valid response to the occupation question. The assumption made in the calculation above is that these workers were all unemployed in Mexico. But it is likely that at least some of them were housewives, students or retirees. If a share of these workers is assumed to have been out of the labor force, the estimate of unemployment among migrant workers would drop even further below 11%.

The estimated rate of unemployment among respondents before they left for the U.S. puts them in the middle of the range of official estimates of unemployment in Mexico. The Mexican government publishes not one but eleven estimates of rates of unemployment and underemployment. The low end of these estimates is based on narrow definitions of unemployment that exclude many workers in unstable, marginal jobs, such as street vending. While these workers are not “openly” unemployed they are severely underemployed.4 In 2004, official estimates of the unemployment and underemployment rates in Mexico ranged from a low of about 4% to over 20%.5 Thus, unemployment among survey respondents while they were in Mexico is not unusually high for the labor market in that country.

But is the unemployment experience of newly arrived migrants while still in Mexico any different from that of long-term migrants? That question is worth asking because the flow of migrants from Mexico has been increasing. U.S. Census bureau data shows that the number of Mexicans residing in the U.S. has nearly doubled in the past decade, and the Pew Hispanic Center has estimated that in recent years the unauthorized flow from Mexico was significantly greater than the number coming through legal channels (Passel, 2005a, b). Many have been in the U.S. for less than five years and they differ from their predecessors with respect to gender, education, state of origin and other demographic characteristics. Thus, the unemployment experience in Mexico and the motivation to migrate to the U.S. for recent arrivals might be different from that of long-term migrants.

The evidence indicates that unemployment in Mexico played an even lesser role in the migration decision of the most recently arrived respondents than it did for those who have been in the United States longer. Table 2 shows the proportions of respondents who reported not working in Mexico arranged by the number of years they have been in the U.S. The table shows two unemployment rates, one before and one after adjustment for the respondents who were students, housewives or retirees in Mexico. In either event, it is the case that the role of unemployment as a potential motive for migration diminishes according to the amount of time a migrant has been here.

Unemployment in Mexico for respondents who have been in the U.S. for two years or less was quite low. Only slightly more than 5% of these migrants, almost all of whom lack a U.S. government-issued ID, were unemployed in Mexico (Table 2). In sharp contrast, more than 15% of respondents who migrated more than 10 years ago had been unemployed in Mexico. Those respondents would last have been in Mexico prior to 1995. Many would have been in Mexico during the 1980s, a decade that witnessed two economic recessions in that country. The turnaround in the Mexican economy since then, except for a sharp but brief downturn in 1995, appears to be reflected in the reduced incidence of unemployment among more recently arrived migrants.

Thus, data from the Survey of Mexican Migrants underscore the diminishing importance of open unemployment as an economic motive for migrating to the U.S.6 This finding is relevant to discussions of policy proposals that would aim at reducing migration pressures by improving economic conditions in Mexico. Simply reducing overall unemployment might not have that effect. This survey suggests that in the realm of work additional factors such as the quality of jobs, wages, long-term prospects and perceptions of opportunity need to be considered in weighing the impact of economic development on migration.

Industries of employment for respondents in Mexico

The work history of respondents before they left for the U.S. reveals a strong concentration in farming and other blue-collar work. The principal industries of employment for respondents in Mexico were agriculture, construction, manufacturing and commerce (or sales). Male respondents, the elderly, those who did not attend or did not graduate from high school and respondents who arrived more than 15 years ago are especially likely to have a background in agriculture. The most recently arrived migrants show a more diverse background. Overall, the industrial experience of respondents is not unusual for Mexico as most of the labor force there is also employed in the same industries.

Among the migrants who worked in Mexico, 39% of men and 18% of women had been employed in agriculture (Table 3). Other major employers of employed male respondents were the construction (13%) and manufacturing (14%) industries. These two industries and agriculture were the source of employment in Mexico for two-thirds of the male migrants. Among employed women, two-thirds of migrants were in agriculture, manufacturing, commerce (or sales) and domestic service.

The relatively narrow concentration of migrants in a few industries reflects the distribution of the labor force in Mexico (Table 4). Just three industries—agriculture, manufacturing and retail trade—account for one-half of the employment in Mexico.7 Other industries of significance are construction for men and educational, health, recreation and other services for women. Over one-fifth (22%) of male workers in Mexico are employed in agriculture and nearly one in four (24%) women work in retail trade. Overall, blue-collar industries enjoy greater prominence within Mexico than in the U.S. which is more services oriented.

The most apparent difference between the industry distributions of survey respondents and the Mexican labor force is that the respondents are more likely to have a background in agriculture. That is due in part to the fact that the industry background of migrants in Table 3 reflects the state of the Mexican economy spanning a period of 15 years and more. On the other hand, the distribution of the labor force in Mexico as shown in Table 4 is a snapshot from 2004. If the data from the migrant survey are sorted by time spent in the U.S. it is evident that more recent arrivals are less likely to be drawn from agriculture than previous arrivals and are thus more reflective of the current Mexican labor force.

Table 5 presents the industry distribution of employed respondents in Mexico by years in the United States. Migrants who have been in the U.S. for two years or less are found to be only half as likely to have worked in agriculture in Mexico as migrants who arrived more than 15 years ago. In particular, only 20% of the latest arrivals were employed in agriculture in Mexico in contrast to 41% of the earliest arrivals. Not surprisingly, the industry distribution of the most recent migrants closely resembles the distribution of Mexico’s labor force in 2004. These migrants are only slightly more likely than the overall labor force in Mexico—20% versus 16%—to have been in agriculture and equally likely to have worked in construction and manufacturing in comparison with the Mexican labor force. Thus, current arrivals from Mexico appear to have been drawn not from the edges but from near the core of Mexico’s labor force.

The age and education of respondents are also related to their employment experiences in Mexico. The oldest Mexican migrants are most likely to have come from an agricultural background. Among those 55 and older, 40% had worked in agriculture prior to migration. The same is true of only 32% of those in the age group of 16 to 29 (Table 6). It is not a coincidence, therefore, that among the survey cities, Fresno, located in an agricultural area, had the highest proportions of migrants in the older age groups (Suro, 2005a). Moreover, nearly 60% of those surveyed in Fresno had left Mexico with a career in agriculture, compared to about one-third or less of the respondents in all other survey cities.

The younger migrants are more experienced than their older counterparts in construction and manufacturing work. For example, 24% of migrants age 29 or younger had jobs in construction and manufacturing before moving to the U.S. in comparison with only 17% of those 55 or older. The survey cities with especially high concentrations of young migrants are New York, Dallas, Raleigh and Atlanta. Of these, Raleigh and Atlanta were particularly attractive destinations for migrants seeking jobs in the construction and manufacturing industries (Kochhar, Suro and Tafoya, 2005).

The likelihood of having worked in agriculture diminishes sharply with a high school education. Only 16% of respondents with a high school degree and 9% of those with some college education were employed in agriculture before migration. In contrast, 59% of workers who did not attend or complete any school training had jobs in farming. High school graduates were most likely to have worked in commerce (or sales). College-educated workers were the most experienced in white-collar jobs; one-third of them had been employed in health and education services and professional services.

In sum, the respondents to the survey do not appear to have migrated to the U.S. because they were without work in Mexico. The vast majority of migrants were gainfully employed before they moved to the U.S. Their employment background reflects that of the labor force in Mexico—they worked principally in agriculture, construction, manufacturing and commerce (or sales). Male respondents, the elderly, those who did not attend or did not graduate from high school and respondents who arrived in the U.S. more than 15 years ago were more likely than average to have worked in farming in Mexico. The industry background of newly arrived respondents closely matches the current distribution of Mexico’s labor force. To the extent that unemployment has played a role in motivating workers to migrate, its role has steadily diminished over time. Underemployment rather than open unemployment is probably the more important factor behind the decision to migrate. Other economic incentives to migrate are likely to include earnings, job quality, long-terms prospects and perceptions of opportunity.