December 6, 2005

Survey of Mexican Migrants, Part Three

V. Migrants’ Experiences in the U.S.: A Narrow Range of Familiar Industries and Low Wages

The industries into which migrants are hired in the U.S. are mostly familiar ones to them, but local demand plays a role as well in determining where they find work. Two-thirds of employed respondents can be found in only four industries—agriculture, construction, manufacturing and hospitality. This degree of concentration does not vary much by the respondents’ background and origin in Mexico or their year of arrival. Thus, there appears to be a steady, and strong, level of demand for these workers in certain segments of the U.S. economy. However, there are some notable regional variations. For example, demand for construction workers is high in Atlanta, Dallas and Raleigh. Meanwhile, hospitality jobs appear plentiful in New York City, and manufacturing stands out in Chicago. All of these cities, except Chicago, are relatively new destinations for migrants from Mexico and the fact that these are new migrations shows in the employment patterns of the latest arrivals. Thus, Mexican migrants appear responsive to specific needs in different regional labor markets.

Although survey respondents had great success in finding jobs in the U.S. that did not translate into job stability or high wages for them. Relatively long spells of unemployment—one month or more— are not uncommon for these workers, and earnings for many appear to fall below federal poverty guidelines. Women are especially vulnerable to both unemployment and low earnings. Recently arrived migrants, who tend to be young, lacking in English ability and without a U.S. government-issued ID, also lag the rest in earnings.

Industries of Employment Display Familiar Patterns

The industry distribution of survey respondents in the U.S. resembles the one they left in Mexico. Most of them work in agriculture, construction, manufacturing and hospitality. This concentration has endured over time, suggesting that there is a persistent and high level of demand for the services of these workers in those sectors. Nonetheless, there are some variations over time in the way migrants are distributed among these industries as the latest arrivals are more likely to work in hospitality and construction. At least to some degree that is a consequence of the dispersal of Mexican migrants to new destinations and regional variations in demand. Migrants are moving into new areas of settlement, such as Atlanta, Raleigh and New York, where the demand for construction and hospitality workers is higher than in the older destinations such as Los Angeles.

Table 13 shows the distribution of employed respondents across industries in the U.S. The agriculture, hospitality, construction and manufacturing industries account for two-thirds of the employment for survey respondents. Using data from the Current Population Survey (CPS), the Pew Hispanic Center also finds the same degree of concentration in these four industries among all foreign-born Mexican workers who have been in the U.S. for five years or less. Looking at all foreign-born Mexican workers in the U.S., the Center estimates that 57% of them were employed in these four industries in 2004. Thus, the industry distribution of migrant workers in the survey closely corresponds to the distribution of their counterparts in the U.S. labor force.11 The construction industry is especially important to male migrant workers as 28% of them work in that industry alone. In contrast, migrant women are more likely to work in cleaning and maintenance, domestic service and commerce (or sales).

The nature of demand for the services of migrants has been relatively steady over time. It is clear from the data in Table 14 that both recently arrived and long-term migrants from Mexico have followed the same general pathway into the U.S. labor market. Approximately 70% of respondents who entered the U.S. at various times within the past 10 years are employed in agriculture, hospitality, construction and hospitality. About 40% of these workers, whether they came two years or eight years ago, work in hospitality and construction alone. The main variation over time is that agriculture was more important to respondents who arrived more than 15 years ago. Many of these earlier arrivals reside in the agricultural area around Fresno. However, the overall nature of demand for migrant services appears largely unchanged over time.

Even if the demand for migrant workers appears to emanate from a small number of industries, the assimilation of migrants into the U.S. economy is shaped considerably by variations in local demand. Table 15 shows the industry distribution of respondents by survey city. The data are presented for all respondents and for those who arrived in the past two years. It is evident that construction dominates in Atlanta, Dallas and Raleigh. About 30% to 40% of respondents, regardless of when they arrived, work in construction alone in these cities. Similarly, hospitality is the key employer in New York. These four cities represent relatively new areas of settlement for migrants from Mexico. As more and more migrants from Mexico have moved into these cities, the importance of hospitality and construction work has increased over time.

In contrast, older settlement areas around Chicago, Los Angeles and Fresno continue to offer different opportunities to migrant workers. Agriculture dominates in Fresno. Even among migrants who entered the U.S. in the past two years, 60% of those who went to Fresno are found to work in agriculture alone. Similarly, manufacturing is important in Chicago to both recent and other arrivals. Los Angeles offers the most diverse set of opportunities. Thus, local economic conditions have an important influence on the employment opportunities available to migrant workers.

Whether or not migrants are in possession of a U.S. government-issued ID also has some influence on their employment pattern. Respondents with a U.S. ID are more likely to be employed in agriculture (Table 16). This is not just because those with a U.S. ID have, on average, been in the U.S. for a longer period of time. Even respondents who entered the U.S. less than two years ago are more likely to be employed in agriculture if they have a U.S. ID. This is most likely a reflection of the temporary labor certification program (H-2A visa) for agricultural workers. Not coincidentally, the Mexican industry of origin for respondents with a U.S. ID, regardless of year of entry, is also more likely to be agriculture. Respondents who do not have a U.S. ID, and, therefore, are almost certainly in the U.S. without work authorization, are more likely to be found in the hospitality industry.12 Another difference between workers with or without a U.S. ID is that the former, while still concentrated in agriculture, hospitality, construction and manufacturing, are somewhat more likely to be dispersed across other industries. Thus, there are some differences in the employment opportunities available to respondents who are in the U.S. without apparent authorization.

Finally, education also shapes the industrial destiny for Mexican migrants in the U.S. The least educated workers are the most likely to be employed in agriculture (Table 17). More than one-third of working respondents who have no schooling whatsoever are employed in this field. In contrast, this is true for only 3% of college-educated respondents. A college education opens the door to white-collar work. About 25% of migrants who attended college are located in professional services and commerce (or sales). Nonetheless, a substantial share—47%—of college-educated migrants work in hospitality, construction and manufacturing. And since there are very few college educated respondents, the overwhelming destination for these migrants remains blue-collar industries.

Spells of Unemployment Among Migrants are not Uncommon

It is not uncommon for respondents to experience spells of unemployment lasting one month or more. That is despite the fact that they have had great success in locating jobs in the U.S. There is evidence, however, that Mexican migrants may be more likely than average to be in temporary or contingent work arrangements. That is beneficial for the economy to the extent that contingent work arrangements provide flexibility and lower costs for employers. But for migrant workers it means that the stability or permanence of their jobs is in some doubt.

A very high percentage of respondents—38%—reported they had a spell of unemployment in the previous year that lasted more than one month (Table 18).13 The instability in employment is very high among women (48%). Lacking a high school education increases the likelihood of going jobless for long spells as over 40% of migrants with only a primary school education or less reported having the experience. It is notable that even a college degree did not insulate migrants from the unemployment rolls since 35% of them went without work for a month or more. Similarly, the likelihood of unemployment spells is high regardless of year of arrival, possession of a U.S. government-issued ID, survey city or other attributes of respondents. Reflecting the seasonal nature of agricultural employment, 54% of Fresno residents reported they had experienced lengthy spells of unemployment.

The high probability of experiencing some unemployment most likely results from the nature of jobs held by the respondents. While most respondents work full time (76%) and at only one job (89%) they do so at rates below the norm for the labor force. Looking at foreign-born Hispanic workers in the U.S., the Pew Hispanic Center estimates that 88% of them usually worked full time and 97% held a single job in 2004.14 Another reason might be that migrants are more likely than average to be in contingent work arrangements. Contingent workers may be full-time or part-time, but the key characteristic is that either their job is temporary or they do not expect it to last. The contingent workforce is currently estimated to be approximately 4% of the U.S. labor force (Bureau of Labor Statistics, 2005). Hispanics, who are 13% of the overall labor force, make up over 20% of the contingent workforce.15 The elevated probability of contingent employment would contribute to a higher likelihood of experiencing spells of unemployment.

The Earnings of Mexican Migrants are Among the Lowest in the U.S.

The weekly earnings of Mexican migrants are among the lowest in the U.S. Earnings for women migrants are especially low and respondents who are recently arrived also earn less than average. As shown in Table 19, the midpoint or median of the earnings distribution for migrants is $300 per week—one-half of migrants earned less than $300 per week while the other half reported higher earnings. That is well below the median wage of $360 per week for all foreign-born Mexican workers and $384 per week for all foreignborn Hispanic workers in the U.S. in 2004 based on the government-collected Current Population Survey. The Pew Hispanic Center also estimates a median wage of $320 per week for foreign-born Hispanic workers who entered the U.S. in 2000 or later years based on government-collected data (Kochhar, 2005). Thus, the survey respondents’ median earnings closely reflect what the CPS shows to be the earnings for recently-arrived foreign-born Latino workers.

The wages earned by survey respondents potentially place many of them at or below the federal poverty level. If a migrant worker is assumed to work 50 weeks in a year, the median weekly wage translates to $15,000 per year. The 2005 federal poverty guideline for a family of three is $16,090.16 The poverty guideline for a family of four is $19,350. Given the low median wage reported by the survey respondents, it is possible that a fair proportion of migrant families, especially those with two or more children, are living at, or below, the poverty level. However, the Survey of Mexican Migrants also shows that many respondents live in households with multiple earners, not all of whom are members of a nuclear family. Living arrangements such as these may help keep many migrant households above the poverty level.

Not surprisingly, low wages are characteristic of workers with the least stable jobs. Thus, the lowest wage earners were the most likely to report they had experienced lengthy spells of unemployment. Two-thirds of workers earning less than $100 per week had been unemployed for a month or more in the past year (Table 20). The same was true for only 19% of respondents earning over $500 per week. In the middle, 31% of workers making $300 to $399 per week had unemployment spells lasting over a month.

Wages for women are well below those earned by men. Whereas only 38% of men earned less than $300 per week the same was true for 74% of female respondents (Table 19). Contributing to this trend is the industry distribution of these workers. The construction industry is the primary source of jobs for male migrants and also the key source of decent earnings for them. More than 70% of migrant workers in the construction industry received a wage in excess of the overall median of $300 per week (Table 21). In contrast, much more so than men, women are employed in the commerce, cleaning and maintenance, and domestic service industries where the majority of workers earn less than the median wage.

Time spent in the U.S. is a key indicator of wages. More than 60% of respondents who arrived in the less than two years ago earn less than $300 per week (Table 22). At the other end of the spectrum, more than 60% of migrants who have been in the U.S. for at least 15 years earned a weekly wage higher than that amount. Thus, earning potential climbs steadily with time spent in the U.S. labor market.

Speaking English well or possessing a U.S. government-issued ID, attributes related to time spent in the U.S., adds significantly to the earning potential of migrant workers. Nearly two-thirds of those who speak a lot of English earn more than the median wage (Table 23). On the other hand, 68% of respondents who speak no English made less than $300 per week. Similarly, earnings of 57% of respondents without a U.S. government-issued ID are below $300 per week. The same is true for only 42% of migrants in the survey who have a U.S. government issued ID.

The survey also revealed a predictably strong relationship between earnings and education. More than two-thirds of respondents who did not attend school make less than the median wage while the opposite is true for two-thirds of workers who had at least some college education. However, there was little variation in earnings across age groups. It is likely that the greater concentration of elderly workers in agriculture negates their advantage in age and U.S. labor market experience. Conversely, as noted above, younger workers are more likely to be found in the relatively high paying construction industry and that helps negate their lack of experience in local labor markets.

The earnings distribution of Mexican migrants is fairly alike in the survey cities. The principal exception is Fresno where more than 50% of the survey respondents work in agriculture. A solid majority of respondents in Fresno—60%—live on less than the median wage of $300 per week. Migrants settled in Atlanta and Dallas are the best off, with 56% in each city receiving a weekly wage higher than the median.

  1. It should be noted that the presence of Fresno in the set of cities where the survey was conducted does elevate the role of agriculture in the industry distribution of respondents. In the aggregate, 11% of respondents have jobs in agriculture. This is twice as high as the proportion of all foreign-born Mexicans (5.4%) who are estimated to work in agriculture in the U.S. The over representation of migrants in agriculture is in part due to the fact that more than 50% of respondents in Fresno are employed in agriculture. Since Fresno accounts for 8% of the total sample, this means that 4% of the total sample consists of agricultural workers from Fresno alone. But even if the role of Fresno were discounted, the Mexican migrants in the survey would still be more concentrated in agriculture than their counterparts in the U.S. economy.
  2. Respondents from new states of origin in Mexico are also more likely to work in hospitality. That is a reflection of their greater concentration in areas such as New York City.
  3. A comparable statistic for the U.S. labor force is not available. What is known is that 19.4 million Hispanics (native and foreign born) worked or looked for work in 2003. Of this total, 2.5 million, or 12.9%, experienced some unemployment during the year. The median number of weeks of unemployment was 17.6 (Bureau of Labor Statistics, 2004).
  4. These are unpublished estimates made by the Pew Hispanic Center from CPS data. Similar conclusions emerge if the survey respondents are compared with foreign-born Mexican workers in the U.S. or with all Hispanic workers in the U.S. For example, 89% of all foreign-born Mexican workers in the U.S. usually worked full time and 98% held a single job in 2004.
  5. In part, that is due to the higher concentration of Hispanic workers in construction and business services (e.g. landscaping). Both industries make use of contingent arrangements at much higher rates than average.
  6. Poverty guidelines are published in the Federal Register each year by the United States Department of Health and Human Services. The 2005 poverty guidelines appeared in Federal Register, Vol. 70, No. 33, February 18, 2005, pp 8373-8375. They are also available at The poverty guidelines are similar to, but not the same as, the poverty thresholds published by the Census Bureau.