B I N A T I O N A L S T U D Y

E S T U D I O B I N A C I O N A L

The demographic, educational, and income characteristics of Mexican-born

migrants form a picture of their likely achievement and impact in both countries. These characteristics also mirror the forces driving migration. Unfortunately, no single data set from either side of the border provides information on all the major characteristics or on all of the varied Mexican-born subgroups. Thus, the characteristics of Mexican-born migrants have to be pieced together from a number of data sources in the United States and Mexico. These are examined here to describe how migrant groups have changed over time and how they differ from one another today.

We utilize well-known data sources and only those that reliably reflect the characteristics of the Mexican-born. All provide data on different migrant subpopulations. However, they either do not distinguish legal status or imperfectly capture the varied "types" of migrants (by time spent in the United States). For example, the U.S. Census has excellent information on characteristics, but does not capture legal status perfectly.¹ Thus, Census data do not permit contrasting the characteristics of, for example, legal permanent resident aliens and legal temporary aliens, to say nothing of unauthorized persons.

Analysis of the incomplete yet complex wealth of information available leads to three broad conclusions:

· Mexican-born migrants differ systematically along two fundamental

dimensions: legal status in the United States (from unauthorized persons to naturalized citizens), and a basic migration pattern or "type" (from short-term visitors to occasional sojourners to settlers).

· Mexican-born migrants tend to have low skill levels, relative both to the U.S. population at large and to other migrant groups. These low skill levels reflect the demand for labor in sectors where Mexican-born

migrants are employed and, in turn, are reflected in the low incomes and high poverty rates of Mexican-born settlers in the U.S. This situation is exacerbated by the unauthorized status of many of these migrants.

· The characteristics of migrants show increased diversity over time, consistent with the increasingly diverse demand, supply, and network factors that are shaping migration flows.

III.

CHARAC-TERISTICS

OF

MEXICAN-

BORN

MIGRANTS

MEXICOESTADOS UNIDOS SOBRE MIGRACION

MIGRATION BETWEEN MEXICO & THE UNITED STATES

Character-

istics of

Migrant

Types

Though data constraints are considerable, we can contrast, with representative and primarily national-level data, the following types of migrants to the United States:²

· Sojourner migrant (legal or unauthorized whose principal residence is in Mexico);

· Settled resident (legal or unauthorized who habitually reside in the U.S.); and

· Naturalized U.S. citizens (who have met five-year legal residence and other requirements).

Sojourners may be of any legal status and encompass those who come for short periods of stay for reasons varying from brief family visits and tourism, to work in unauthorized status or with legal visasalthough most migrate to work. We draw on several sources to describe these mobile individuals. The primarily ones from Mexico are the 1992 national survey (ENADID), surveys at the northern border (EMIF), the Zapata Canyon project, a special survey in the state of Michoacán, and the Mexican Migration Project.

Table III-1.

Demographics & Education for the

United States & Mexican Populations & Mexican-Born

Migrants

MEXICO

RESIDENT

POPULATION (1992)¹

U.S. RESIDENT POPULATION (1990)³

MEXICAN-BORN

TOTAL

RESIDENTS

SETTLERS

MEXICAN-BORN SOJOURNER

(1992-1994)²

NATURALIZED

33 YRS

49%

56%

57%

3%

28%

72%

6%

¹Source: Encuesta Nacional de Indicadores Demográficos [ENADID].

²Multiple sources. See Bustamante et al. 1997.

³Source: 1990 United States Census.

⁴Population 25 years and older.

⁵ means data not available.

CHARACTERISTICS

DEMOGRAPHICS

Age (average)

Male Proportion

Married⁴

Men

Women

SCHOOLING⁴

Years (average)

Fewer than 5

Fewer than 12

12 or more

ENGLISH

Not speaking well

or very well

25 YRS

49%

83%

72%

5 YRS

46

90%

10%

28 - 32 YRS

73 - 94%

56 - 85%

43 - 66%

6 YRS

39%

91 - 99%

1 - 9%

93%

30 YRS

55%

59%

61%

8 YRS

28%

76%

24%

71%

42 YRS

54%

76%

80%

⁵

24%

67%

33%

57%

B I N A T I O N A L S T U D Y

E S T U D I O B I N A C I O N A L

Settlers establish a usual or permanent residence in the United States, although many return regularly to Mexico and as many as one-half reestablish residence in Mexico after a ten-year stay or longer in the U.S. Eventually, many legal settlers become naturalized citizens, a process that takes time and commitment; they tend to be older and more assimilated. To describe these populations we draw primarily upon the 1990 U.S. Census, the 1996 Current Population Survey [CPS], public and special analyses of INS administrative data, and several other sources.³

Selected characteristics are presented in Tables III-1 and III-2 as ranges, when drawn from more than one data source, and are compared with the total or average U.S. population (native and foreign-born). The text below summarizes and explains the tabular results and presents additional information.

Mexico to United States migrants have tended to be selected from the middle-to-lower segments of Mexico's socioeconomic hierarchy, a selection process that originated at the start of this century with the recruitment of low-skilled workers by U.S. employers for seasonal jobs, mostly in agriculture (Bustamante et al. 1992). This flow was facilitated by the long and historically porous border that put U.S. labor markets within reach of individuals with limited financial

Table

III-2.

Labor Force & Income/Earnings

for the

United States

& Mexican Populations & Mexican-Born Immigrants

MEXICO RESIDENT

POPULATION (1992)¹

MEXICAN-BORN SOJOURNER

(1992-1994)²

U.S. RESIDENT POPULATION (1990)³

MEXICAN-BORN

SETTLER

NATURALIZED

TOTAL

RESIDENTS

65%

75%

59%

6%

3%

25%

72%

$24,408⁷

$38,940

13%

¹Source: 1990 Mexican Census, note that these figures reflect the greater enumeration of formal sector activity.

²Multiple sources. See: Bustamante et al. 1997.

³Source: 1990 United States Census.

⁴United States population 16 years and older and Mexican population 12 years and older.

⁵means data not available.

⁶Source: Escobar Latapí 1996.

⁷Source: Special tabulations by Jeffrey S. Passel, Urban Institute, persons ages 25 and over.

CHARACTERISTICS

LABOR FORCE⁴

Total Participation

Male

Female

Unemployment Rate⁴

Employment Sector⁴

Agriculture

Construction/Manufacturing

Services

INCOME & POVERTY

Individual Earnings U.S. $ (year)

Household Income, Mean U.S. $

(year)

Poverty

43%

68%

20%

3%

23%

29%

48%

$8,880⁶

36%

83%

91%

58%

6 -11%

47 - 53%

25 - 26%

23 - 26%

$185 - 240

(week)

70%

85%

50%

11%

13%

37%

51%

$14,138⁷

$27,120

27%

69%

82%

53%

9%

10%

36%

54%

$16,553⁷

$28,210

25%

MEXICOESTADOS UNIDOS SOBRE MIGRACION

MIGRATION BETWEEN MEXICO & THE UNITED STATES

resources, by an expanding demand for low-skilled migrant workers in and out of agriculture, particularly in the southwestern United States (Alba 1992), and by extensive migration networks connecting families with low-skill U.S. jobs.

Sojourner migrants tend to be young males with little schooling who work in agriculture. Today, more than one-half of the highly mobile sojourners, particularly the seasonal workers captured in the data sources shown in Table III-2, still find employment in agriculture. Their short-term employment, however, is

associated with very low earnings of as little as $185 per week, and more than one-half of their families have incomes below the poverty line.

Settled or permanent residents tend to look more like the U.S. population as a whole, even though differences between Mexican-born migrants and the total U.S. population remain substantial. Many, if not most, settlers begin their stay in the United States as circular or sojourner migrants, often as legal entrants (but also in unauthorized status) to work or live with their families. As their time in the U.S. lengthens, they gain in experience. Over time they and their families and households develop greater resources and their ability and willingness to adapt to the economy increases.

The settler population is more equally balanced between the sexes and appears to be better educated than sojourners (Donato 1993). Although, as Table III-1 shows, they generally complete only six years of education and three-fourths of those 25 years and older have not completed high school, 31 percent of those aged 18-24 are high school graduates. Moreover, 6.5 percent of all Mexican-born persons 18 and older were enrolled in college at the time of the 1990 Census. Within the settler population, legal immigrants are better educated. Among 1996 legal immigrants aged 25 and over, 35 percent are high school graduates and 15 percent are college graduates (Jasso et al. 1997). Fewer settlers than sojourners work in agriculture and, with longer-term residence and more resources, their households bring in more income. Nevertheless, as Table III-2 shows, settler household income ranges from two-thirds to about three-quarters that of the average U.S. household and twice as many live below the poverty line.

The legalization programs of IRCA formalized and speeded the transition from sojourner, to settler, to legal resident, and, finally, to naturalized citizen. There is limited information on the Mexican SAW workers who, upon legal admission, typified the seasonal agricultural or sojourner type migrant. Primarily males, most were employed in agriculture while many were likely employed in casual urban-based jobs as well. SAWs have little education and income, but with time

B I N A T I O N A L S T U D Y

E S T U D I O B I N A C I O N A L

they are moving out of agriculture to potentially more stable households and more secure jobsthey comprised 33 percent of seasonal U.S. farm workers in 1989, but only 19 percent by 1995.

The pre-1982 legalization population typifies a settler population with a

relatively balanced proportion of males and with only 7 percent employed in agriculture. Yet even in 1992, five years after legalization, they remained a population with little education; only 45 percent have completed primary education. Research finds that legalization has afforded some modest improvements, partly due to increased use of English and job training: the rate of earnings growth of legalized men 1989-1992 was greater than other U.S. workers. Two-thirds reported that legalization afforded them "easier" advancement at work. Earnings increased from U.S. $7.14 to U.S. $9.43 per hour in 1992 (constant dollars). Median family earnings were U.S. $19,112 (U.S. DOL 1996).

The number of Mexican-born becoming naturalized U.S. citizens is rapidly

increasing as shown in Chapter II. Historically, the Mexican-born, along with Canadians, had among the lowest rates of naturalization of any foreign-born group. Proximity to their country of origin is a key factor. Mexican-born

migrants who live in the Southwestern United States are even less likely to naturalize than their counterparts who settle in the Midwest or in New York or New Jersey, although there has been a recent surge in naturalizations among the Mexican-born. One of the more important of the many other reasons for recent trends⁴ is the legacy of IRCA, those granted legal permanent resident status first became eligible to naturalize starting in 1994. Forty-three percent of legalized Mexican-born adults reported an intention to naturalize as of 1992. The surge in naturalizations, however, is not limited to the IRCA cohorts.

Longitudinal INS data on naturalization for Mexican-born persons admitted to permanent residence in 1977 and 1982 indicate that those who naturalize are drawn from among the more highly-skilled. They also experience occupational upgrading during the years between admission and naturalization. Moreover, for women, those who report work upon admission are more likely to naturalize, and labor force participation also increases between admission and naturalization. According to the 1990 Census, more than 42 percent of naturalized citizens speak English "very well" compared with 25 percent of those who do not report being naturalized. Similarly, the naturalized citizens are better schooled: 33 percent of those 25 years and older are high school graduates, compared with 24 percent of those not naturalized.

MEXICOESTADOS UNIDOS SOBRE MIGRACION

MIGRATION BETWEEN MEXICO & THE UNITED STATES

Changes in Charac-

teristics

over Time

There are some indications that the characteristics of migrantsin terms of origins and destinations, gender, age, education, and employmentare becoming more diverse over time. Migration, thus, is a dynamic process: the forces selecting people into and out of migration change over time. Migrant characteristics partly reflect the characteristics of the population-at-large from which migrants are drawn and partly are determined by selective migration forces. For example, if migrants are drawn randomly from a population that becomes increasingly educated over time, the average education of migrants will increase. The spread of migration networks may make migration "less selective" of individual characteristics over time. Increased diversity may also partly reflect changes in the kinds of worker in demand by U.S. employers. Studies document cases in which differences in characteristics between migrants and nonmigrants from specific locales in Mexico, at one time pronounced, have faded or disappeared. They also show increasing diversity of migrant origins, destinations, and demographic

characteristics.

Places of Origin in Mexico. The traditional sending states continue to dominate, but Mexican-born migrants increasingly have come from other states in Mexico. In 1926, more than one-half of all monetary remittances from the United States were directed to the three core states of the west-central region Guanajuato, Michoacán, and Jalisco. The role of these three has diminished somewhat and today they are joined by Durango, Zacatecas, the state of México, México City, Chihuahua, Tamaulipas, and Guerrero: these ten entities represent 49 percent of Mexico's total population, but more than 70 percent of the migrant total.⁵

A broader picture of migrant sources can be constructed by dividing Mexico into six regions based on the geographic distribution of migrant birthplaces and using 1992 data:

(1) West-central core states38 percent of all migrants: Guanajuato, Michoacán, Jalisco, and Colima;

(2) Northern-border states21 percent of all migrants: Baja California, Sonora, Chihuahua, Coahuila, Nuevo León, and Tamaulipas (including Baja California Sur);

B I N A T I O N A L S T U D Y

E S T U D I O B I N A C I O N A L

(3) States between regions One and Two listed above22 percent of all migrants: Sinaloa, Durango, Nayarit, Zacatecas, San Luis Potosí, and Aguascalientes;

(4) Interior states9 percent of all migrants: in and surrounding the Valley of México, the Federal District, the state of México, Querétaro, Hidalgo, and Tlaxcala;

(5) Four southern states8 percent of all migrants: Oaxaca, Guerrero, Puebla, and Morelos;

(6) Six southwestern states2 percent of all migrants: Veracruz, Tabasco, Chiapas, Campeche, Yucatán, and Quintana Roo.

In recent years, the share of migrants from rural areas appears to have fallen and a progressively larger share comes from urban areas. Traditionally migrants predominantly originated from rural areas (places with population less than 20,000). As of 1992, 59 percent of the Mexican-born who had lived in the U.S. reported coming from rural areas.⁶ Nevertheless, 25 to 31 percent of migrants born in rural areas had changed their homes to urban areas.

Places of Destination in the United States. For historical, geographic, and labor-market reasons, Texas was the leading destination for

Mexican-born migrants prior to the 1920s, joined much later by California and Illinois. While the southwest remained the core sending area, during the Bracero program Mexican-born migrants fanned out across a broader geographic area from Texas and California to Arizona, Indiana, Delaware, Michigan, Arkansas, Montana, and Washington. The connections to California and Texas are widespread across Mexico, but vary depending on sending traditions and networks: case studies in Michoacán indicate that communities "channel" their migrant streams to particular U.S. destinations.

Mexican-born migrants in the United States remain concentrated in a few states and localities, but increasingly they are found in new destinations. Border surveys find that the traditional predominance of Texas, California (70 percent of experienced migrants), and Illinois has continued for sojourner and settler alike. The 1990 U.S. Census indicates that the Mexican-born are highly concentrated in California, Texas, and Illinois: about 85 percent of all Mexican-born

immigrants resided in these three states compared to 45 percent of all immigrants

MEXICOESTADOS UNIDOS SOBRE MIGRACION

MIGRATION BETWEEN MEXICO & THE UNITED STATES

to the U.S. California is the single major destination with 50 percent of all Mexican migrants.

Los Angeles is the single most important urban destination of all Mexican-born migrants. It is followed by San Antonio, the south Texas Rio Grande Valley, Houston, Dallas-Fort Worth, El Paso, Fresno, and Phoenix.

At the same time, Mexican-born migrants have become attracted to new geographic destinations. Midwestern and eastern states with few foreign-born

workers have become destinations for thousands of Mexican-born persons

employed in agriculture, food processing, construction, and manufacturing operations. Mexican data show sojourners are increasingly headed to Florida, Arizona, New Mexico, and North Carolina. Although the majority of legal immigrants continue to list California, Texas, and Illinois as their selected places of residence, increasingly they go now to destinations such as Florida and Georgia.

Demographic Profile of Migrants. Mexico to U.S. migration flows traditionally were dominated by young, solo males. For example, a classic study published in 1975 analyzing apprehended migrants found that more than

90 percent were 40 years of age or younger, 92 percent were males, and 62 percent were single. The Zapata Canyon border survey, dominated by unauthorized circular migrants and sojourners, finds high concentrations of males: 97 percent in 1996.

The importance of males persists at least among sojourner migrants, although there is a trend toward more female migrants, and women dominate among new legal immigrants. The 1992 ENADID survey found that 21 percent of the Mexicans had lived or worked in the United States were women.⁷ The share of males in Michoacán surveysa traditional sending areawas 63 percent in 1983 and 56 percent 1993, indicating that more women are joining the migration stream. In contrast, the U.S. Census and other standard surveys, undoubtedly better at capturing settlers, show Mexican-born migration to be much more

gender-balanced [see Table III-1]. INS admissions data on legal Mexican immigrants to the United States show that women outnumbered men for eight consecutive years between 1964 and 1971, and they outnumbered men again in 1993 and 1994.

Mexican migrants appear to be a young group, younger than migrants to the U.S. from other countries and than the U.S. population at large [see Table III-1]. This

B I N A T I O N A L S T U D Y

E S T U D I O B I N A C I O N A L

relative youthfulness may partly reflect the presence of young pioneer migrants in early migration streams and children accompanying legal immigrants. ⁸ The INS legal admissions data also show Mexican immigrants to be a young group with a lower mean and median age than other immigrants in the U.S.

Mexican community surveys and INS data indicate, however, that migrant ages may be increasing. The average age of Michoacán migrants rose from a mean of 29 in 1983 to 32 years in 1993.⁹ Border crossing surveys for 1988 to 1996 reveal sharply rising shares of unauthorized migrants between the ages of 25-29 and decreasing shares of younger migrants. The age of women in U.S. legal admissions increased from 21 to 26 years from 1971 to 1994. Among these legal immigrants, increases in age may reflect longer waits for visa issuance and/or fewer children.

One last demographic characteristic: most migrants are married [see Table III-1]. Among sojourners, men are more likely to be married, while among settlers, women are more likely to be married. Additionally, a substantial proportion of persons who become legal U.S. immigrants do so as the spouses of U.S. citizens. Although the Mexican-born spouse flow has declined, from more than 30,000 per year in 1986-1988 to less than one-half of that in 1995, Mexico remains the leading source country for spouses of U.S. citizens. While most spouse immigrants from other countries are women, a majority of those from Mexico

are men.¹⁰

Educational Profile. Mexican migrants have less schooling relative to the U.S. population and other immigrant groups. Nevertheless, it appears that the average schooling of northbound migrants is increasing over time. Sojourners average around six years of schooling. In the early 1980s, relatively well-educated villagers were likely to migrate, but their destinations were urban areas of Mexico. Over the decade, several Mexican data sources indicate increases in the schooling of U.S. migrants.¹¹

Among migrants in the Michoacán surveys, average schooling increased from 4.1 to 5.8 years between 1983 and 1993. At the same time, average schooling among nonmigrants who remained in Mexico decreased from 4.5 to 4.3 years. These changes reflect shifts in migrant selectivity. During Mexico's crisis years in the mid-1980s, better-educated migrants appear to have shifted destinations to the United States.

MEXICOESTADOS UNIDOS SOBRE MIGRACION

MIGRATION BETWEEN MEXICO & THE UNITED STATES

The 1990 U.S. Census data present a similar picture of increasing educational levels: recent arrival cohorts have been better schooled. At the upper end of the educational spectrum, 4 percent of the 1980-1990 and 6 percent of the 1987-1990 migrants were college graduates. As already noted, 1996 legal immigrant

complete college at a rate approaching the native-born and a higher rate of postgraduate school (9 percent compared to 7 percent) (Jasso et al. 1997) Meanwhile, the absolute number of Mexican-born doctorate holders in the U.S. 1990 Census was 3,869 (of which 27 percent were women).¹²

Labor Force Characteristics. Work is the single most important attraction in the U.S.; however that does not mean Mexican migrants lack jobs in Mexico or combine jobs in the U.S. with usual residence there. Most migrants had some kind of work in Mexico prior to migrating, although the share who were unemployed before migrating may be rising. Border crossing data (which include large numbers of unauthorized migrants) find that while most had work prior to leaving, the majority migrated with the intention of working in the U.S. Data also show that among females, about one-third worked outside the home in Mexico prior to migrating.

Mexican data sources reveal that those who have lived in the United States have worked there. Of course, work experience may include periodic unemployment and U.S. data show Mexican migrants have nearly double U.S. unemployment rates. As of 1992, men are more likely than women to report a habitual residence in the U.S. and work experience (92 percent versus 54 percent). Still, the 1992 survey found more than one million Mexican-born migrants who declared having worked in the U.S. without having "lived" in or established a usual residence.

Mexican-born migrants in the United States have, over time, become less likely to be agricultural workers and are found in an increasing diversity of jobs. From the turn of the century and through the end of the Bracero program, Mexican-born migrants were employed primarily in agriculture. Today [see Table III-2], most settlers and Mexican-born citizens work outside of agriculture. Concurrently, even northward bound sojourners increasingly are found outside of

agriculture. Mexican data sources show evidence of an upward trend in U.S. urban-sector employmentparticularly in services and constructionfor unauthorized migrants [see Appendix B]. Still, Table III-2 shows that Mexican-born workers retain an industrial profile that differs from the U.S. average.

B I N A T I O N A L S T U D Y

E S T U D I O B I N A C I O N A L

Income Characteristics. Mexican-born households are much more likely than all U.S. households to be found at the bottom of the income distribution. The 1990 Census shows that the share of households in the lowest income groups was greater for the recently arrived 1980-1990 cohort and lower for the pre-1980 cohort. The share in the highest income groups was higher for the earlier cohort. This suggests the possible operation of two mechanisms: experience in the U.S. leads to higher earnings and/or the less successful

migrants in the earlier cohorts returned to Mexico, leaving the higher earners.

A 1996 U.S. survey finds that, while the average U.S. household saw income gains between 1990 and 1996, Mexican-born households lost income. On average, all U.S. households brought in $38,453 in 1990 and $44,938 by 1996. In contrast, the average U.S. Mexican-born household brought in $27,122 in 1990, but only $26,481 in 1996. Comparison of median household income shows the same relative loss in the incomes of Mexican-born households (all figures in current dollars). The 1996 data reveal a concentration of new (1990-1996)

migrant households at the bottom of the income spectrum: 11 percent of recently arrived migrant households have incomes less than $5,000, double the proportion of earlier arrivals. This may reflect differences in the legal status mix, for example, a larger proportion of undocumented among the recent arrivals. It also may reflect the lower earnings at entry among some of the IRCA-legalized aliens. Recent legal immigrants in 1996, for example, have higher earnings. A recent survey indicates earnings of $19,130 for adult male and $13,620 for adult women immigrants (Jasso et al. 1997).

¹ Information on the characteristics of citizens is biased because of substantial misreporting

of naturalization by noncitizen settlers (special tabulations by Jeffrey Passel, Urban Institute).

However, the systematic differences between settlers and citizens appear to be reasonable

approximations.

² To obtain a picture of the sojourner, we piece together data from numerous sources: the National Survey of Demographic Indicators [Encuesta Nacional de Indicadores Demográficos, or ENADID], the Survey of Migration at the Northern Border [Encuesta sobre Migración en la Frontera Norte, or EMIF], the Mexican Migrant Survey undertaken jointly by COLEF and a team of researchers from the University of Southern California [USC], the Mexican Migration Project, the Michoacán Project, INS Data on Nonimmigrants (Temporary Visitors), the UC-EDD Survey of California Farmworkers carried out by the University of California, Davis [UC], the State of California Employment Development Department [EDD], and the National Agricultural Workers Survey [NAWS] conducted by the U.S. Labor Department.

³ For pictures of the settlers and U.S. citizens, we rely on two main U.S.-side data sources: (1) the decennial censuses, including microdata as well as historical time series published in Historical Statistics of the United States; (2) official government information, including microdata compiled by the U.S. Immigration and Naturalization Service [INS], tabulations published by the INS in its

Endnotes

MEXICOESTADOS UNIDOS SOBRE MIGRACION

MIGRATION BETWEEN MEXICO & THE UNITED STATES

Annual Report (published 1943 to 1977) and the successor Statistical Yearbook (published since 1978), tabulations published by the U.S. Department of State in its annual Report of the Visa Office, and a special microdata sample assembled from INS data by the General Accounting Office [GAO]. Of these, only INS data provide information on legal status; and measurement of naturalization in the censuses is problematic.

⁴ These factors include: visa backlogs for spouses and minor children of permanent resident aliens, largely due to IRCA legalizations; the Green Card Replacement Program, initiated in 1992, which requires that permanent resident alien cards issued before 1978 be replaced and which leads some to naturalize rather than obtain a new green card; initiation of an expiration date, ten years after issuance, on green cards issued since 1988, which may similarly lead some to choose naturalization over green-card renewal; recent U.S. legislation which dramatically reduces the civil rights and social entitlements of nonnaturalized immigrants; and recent amendment of the Mexican Constitution that provides that Mexican nationals who become naturalized citizens of another country do not automatically forfeit Mexican nationality.

⁵ The share of migrants in a state's total population varies. It is highest in Zacatecas, where

9.7 percent of the population had lived or worked in the United States. It is 8.3 percent in Durango, 8.2 percent in Michoacán and 6.5 percent in Jalisco. In contrast, the share is only 0.9 percent in México City, and 1 percent in the state of México.

⁶ As of 1992, there were marked differences in rural versus urban origins across regions. While the Valley of México and border regions have small shares of rural migrants, 32 percent and 34 percent respectively, in other regions the rural share is more than 60 percent.

⁷ The share for those who had lived in the U.S. over the previous five years was 24 percent, suggesting that the female share is increasing slightly.

⁸ For example, in the nineteen communities studied by Massey and Durand, the average age of migrants in the United States at their most recent trip was 29 years. The average age of unauthorized migrants in the Michoacán surveys, who comprised just over 58 percent of all 1993 migrants,

was younger at 28.5 years than legal migrants at 37 years of age.

⁹ The 1983 migrants were somewhat younger than nonmigrants, whose average age was 32, but the 1993 migrants were considerably older than nonmigrants, whose average age was 20.

¹⁰ INS public-use data do not include sponsor's nativity, but a special study carried out by the General Accounting Office [GAO] on data from 1985 indicates that almost 78 percent of the U.S. sponsors of Mexican-born spouses were U.S. citizens by birth (worldwide, the rate was 80 percent). In an intergenerational twist on the international character of these couples, among the birth-citizen sponsors in the GAO sample, 4-5 percent were themselves born abroad to U.S. citizen parent(s).

¹¹ The Mexican Migrant Project show increased schooling and increased schooling is one of the more salient findings of the Zapata Canyon border survey. Mexican data for 1978-1979 found migrants had 4.9 years of schooling on average, slightly greater than the 4.7 years for the Mexican population over 14 years old.

¹² Note that Census data do not reveal whether these are persons undergoing further training or engaged in postdoctoral work with nonimmigrant visas as opposed to permanent residents of the United States.