Earnings of Mexican male workers in the U.S. labor market: role of English language proficiency

ABSTRACT

Using the data drawn from the 1990 U.S. Census, this paper reaffirms the importance of English language proficiency (ELP) in determining the earnings of Mexican male workers in the U.S. labor market. The returns to work experience and educational attainment are higher for workers with higher ELP than workers with lower ELP. Although the exact nature of reciprocal relationship between ELP and educational attainment is not clearly identified in this paper, it appears that the acquisition of ELP and educational attainment gives access to better occupations, allowing workers to earn more in the labor market. Focus on the Mexican youth for "drop in" and "stay in" the U.S. school system, and flexible schedules of English language classes to accommodate adult workers will be important considerations for future policy tasks.

1. INTRODUCTION

Mainly due to the continuing influx of immigrants and the high fertility rates, the Hispanic population in the United States has recently grown substantially. According to a Census Bureau projection, by 2050, Hispanic population will reach 97 million and constitute 24.5 percent of the U.S. population, thereby replacing blacks to be the first minority group. The low educational attainment and economic status of the Mexican Americans, the largest group in Hispanic population, have been a focus of concern. While Mexican Americans are less educated than non-Mexican Americans in general, the pattern differs across generations. According to Gonzalez (2002), first-generation Mexican Americans have an average educational level of less than 9 years, whereas the second generation averages about 11.5 years, and the third generation, a high-school diploma or 12 years. However, important in this pattern is that even third-generation Mexican Americans lag behind the average of 13 years of schooling for the rest of the population. In particular, the low earnings of Mexican Americans are quite disturbing and provide a focus point of this paper. This paper investigates factors that influence the determination of earnings of Mexican American male workers with a particular emphasis on one specific human capital: English language proficiency (ELP). ELP is important not only for immigrants' economic assimilation, but also for their social integration. The data analysis that will be explained in more detail in the next section shows that 47 percent of Mexican male immigrant workers are not proficient in the English language. Even among native-born Mexican American population, 3 percent is reported not to be proficient in the English language. The limited ELP of many Mexican Americans is likely to serve as a significant deterrent to success in the U.S. school system and in the labor market.

Although not being the first study to analyze the economic status of Mexican American workers in the U.S. labor market, this paper differs from other research by shedding a different light in the roles of nativity and ELP to explain the earnings patterns of Mexican American workers. Knowledge in English language is an important component of human capital necessary in the United States. In general, language capital is very specific to the host economy, since it is usually not transferable from the immigrants' home economy (Dustmann and Fabbri, 2000). In order for the educational attainment accumulated in foreign countries to be translated into earnings power in the U.S. labor market, immigrants should acquire proficiency in the English language. Some native born populations who reside in ethnic enclaves under heavy ethnic influences may encounter difficulties for English acquisition.

Without proper ELP, immigrants who arrive in the United States at school age will have difficulty in catching up in the U.S. school system, and adult immigrants who intend to increase their educational attainment in the United States will have similar problems. ELP is likely to be a decisive factor in determining employment probabilities, since ELP may help workers acquire information about optimal job search strategies. Immigrant workers who are not proficient in the English language will have difficulties convincing prospective employers of their qualifications. For instance, many jobs in the service sector require good communication skills. Likewise, literacy in the dominant language is a crucial prerequisite for many unskilled occupations (Dustmann and Fabbri, 2000). For Mexican Americans, ELP is particularly pertinent since the lower educational attainment of Mexican Americans is correlated with the lower ELP.

As the earnings gap between skilled and unskilled workers have widened as a result of deindustrialization and technology development with the trend being expected to persist in the future, the improvements of ELP and educational attainment of Mexican American workers take on more importance than ever. The improvement also addresses a question of economic assimilation debate on Mexican American population. There is an understanding that economic assimilation of immigrants is socially desirable, and policy makers should support this process, either by programs aimed at the immigrant population, or by selection of incoming immigrants (Dustmann and Fabbri, 2000). As Chavez (1991) argues, "analyzing all Mexican workers under one umbrella obscures as much information as it illuminates." Although limited data availability often leads researchers to treat Hispanics as if they were a homogeneous group, the U.S. Hispanic population is diverse. While Mexican Americans, Puerto Ricans, and Cubans constitute 3 major subgroups, smaller groups are increasing from immigrants from Central America and South America. These subgroups are substantially different in their residential concentration, economic status, and the circumstances of arrival in the United States (Sabel and Stennett, 1998). This paper shows that native born and immigrant Mexican American workers differ quite substantially in their economic status, substantiating this paper's attempt to distinguish the two groups in order to identify the impact of ELP and educational attainment on earnings.

This paper uses 1990 Census Public Use Microdata Sample (PUMS) to analyze different patterns of labor market returns to educational attainments and work experiences of Mexican American workers with different ELP. This paper is organized in the following way. Section 2 surveys the assimilation of Mexican Americans in the U.S. labor market. Section 3 includes empirical works and estimation results along with interpretation of the empirical results. Finally, section 4 concludes and summarizes the results.

2. ASSIMILATION OF MEXICAN AMERICANS IN THE U.S. LABOR MARKET

According to Borjas (1995), substantial initial earnings disadvantages of immigrant workers are often a result of lower levels of human capital when arriving in the United States. The process of economic assimilation of immigrants depends then on the intensity with which they acquire U.S.-specific skills such as ELP. It may be in the interest of the United States to support the process of economic assimilation of immigrants. As immigrants attain parity with native workers, they will contribute more by paying taxes and will be less burdensome in terms of transfer payments. To achieve this, it is important to understand the factors that determine the economic performance of immigrant workers (Dustmann and Fabbri, 2000). Despite the general consensus that the improvements of the educational attainment and ELP are important factors that will enhance the economic status of Mexican workers (Trejo, 1997), researchers disagree about the prospects of Mexican Americans' attaining economic parity with the non-Hispanic white workers in the U.S. labor market.

Regarding Mexican Americans' economic progress, Chapa (1990) argues that if the notion of steady progress is applied to Mexican Americans, then three factors should be considered: similar educational attainment, occupational distribution and earnings with non-Hispanic white workers. Chapa (1990) presents evidence that strongly suggests Mexican Americans as a group fail to satisfy any of these criteria, and further argues that the Mexican Americans who are relatively uneducated and economically disadvantaged may fall into the trap of vicious cycle of lower class. Borjas (1990) estimates that Mexican immigrants would narrow the earnings gap with comparable native-born workers at a rate of about 0.6 percent a year. However, their earnings may never completely reach those of non-Hispanic whites, in an opposition to this pessimistic view, however, Chavez (1991) argues that lumping native-born Mexican workers with immigrant workers would result a misleading belief that Mexican workers in general fail to achieve economic viability in the United States. Chavez opines that when the Mexican immigrants who are mostly poorly educated with minimal skills are added to the pool to be measured, the achievement levels of the whole group will obviously decline. She further argues that Mexican American workers born in the United States have experienced greater socioeconomic mobilities in the last few decades. Their educational attainment has steadily improved, their earnings no longer reflect wide disparities with those of non-Hispanics, and their occupational distribution begins to take closer resemblance to that of the general population.

Trejo (1997), using CPS data from November 1979 and 1989, finds that Mexican Americans earn low wages primarily because they possess less human capital than other workers, not because they receive smaller labor market returns for their skills. In 1989, third and higher-generation Mexican American male workers earned on the average 21 percent less than non-Hispanic whites, roughly similar to the wage deficit for blacks. More than 75 percent of the wage gap of Mexican American workers is due to their younger age, the lack of ELP, and their pervasive educational deficit. These same variables, by contrast, explain less than a third of the black-white wage gap. These findings, as Trejo indicates, render support to a pessimistic view by Chapa (1990) as well as an optimistic view by Chavez (1991). Trejo finds a middle ground to these opposing arguments by suggesting that the large improvements in human capital and earnings of first and second generation Mexican Americans and similar labor market returns to their human capital when compared to white workers support the optimistic view of Chavez (1991), while the finding that intergenerational progress stalls after the second generation, leaving third and higher-generation Mexicans behind the educational attainments and earnings of the average Americans support the pessimistic view of Chapa (1990).

Improvements of educational attainment and ELP have broad economic ramifications that are associated with economic assimilation. Recent evidence points to a much higher level of intergenerational transmission of economic status than was previously thought to be the case (Bowles and Gintis, 2002). To the extent that parental income and wealth are strong predictors of the likely economic status of the next generation, the lower economic position of Mexican American workers will be a source of various problems in the future. Policymakers need to identify what factors and how they influence earnings process of Mexican American workers before designing intervention programs that would loosen the connection between the economic success from generation to generation.

3. EMPIRICAL WORK AND ESTIMATION RESULTS

The data used in this paper are drawn from the 1990 U.S. Census of Population 5 percent PUMS. Census data include both legally admitted immigrants and illegal immigrants. In line with Betts and Lofstrom (2000), this paper uses the census data in the belief that they provide the most detailed information of all immigrants, regardless of legal status. To obtain accurate estimates of ethnic minorities, it is necessary to use a large representative sample containing detailed individual background information on variables such as earnings, ELP, education, occupation, geographic location, age, and the time of arrival in the United States. The only currently available data set that satisfies these criteria is the U.S. Census PUMS (Lofstrom, 2000).

To focus on the workers with strong attachment to the labor force, the empirical analysis includes civilian male workers between the ages 25 and 65 who are not residing in group quarters, not in school, and report positive earnings in the year prior to the census. Since most men work full-time, labor supply decisions do not pose as significant an issue as they would for an analysis of women. For workers who are older than 65, there is a risk that nonrandom mortality would bias the sample of older workers in favor of the more healthy (Lubotsky, 2001). Many people under age 25 have not yet completed their formal education. Since the 1990 Census is not a random sample of the population, sampling weights are used. To investigate how labor market returns to education vary across workers with different ELP, this paper uses a slightly different method in comparison with other research where the variables measuring ELP are directly included as explanatory variables. In a similar vein with Park (1999), this paper measures the impact of various categories of educational attainments and labor market experience on earnings by running the following earnings regressions for immigrant Mexican American workers and the native-born Mexican American workers by their ELP: proficient workers and non-proficient workers. The earnings regression can be written as

ln[W.sub.ij] = [[alpha].sub.j] + [[beta].sub.j][X.sub.ij] + [[epsilon].sub.ij]

where the subscript j denotes the proficient or non-proficient group; [W.sub.ij] is the hourly wage of the individual i in the language group j; [[alpha].sub.j] and [[beta].sub.j] are the intercept and vector of coefficients to be estimated; [X.sub.ij] is a vector of human capital and socioeconomic characteristics; and [[epsilon].sub.j] is the error term. The dependent variable in the earnings regression is logarithm of hourly wage that is obtained by dividing the wages or salary income in 1989 by the product of the usual hours worked per week in 1989 and the weeks worked in 1989. To eliminate outlier, we drop top and bottom 1 percent of hourly wage from the sample. Sample statistics of the variables in [X.sub.ij] are listed in Table 1.

Census PUMS includes the information on how well a respondent speaks English; the responses consist of "n/a, only English is spoken at home," "very well," "well," "not well," and "not at all." This paper combines the "n/a, only English is spoken at home," and "very well" categories to indicate the proficiency in English language, and "not well" and "not at all" for non-proficiency. According to Table 1, there are 39,132 immigrants and 35,603 natives. 53.1 (97.1) percent of the immigrant (native) sample reported that they are proficient in English. Immigrant workers are at an enormous disadvantage in education. Since the 1990 census codes educational attainment using a mixture of point and interval, this paper recodes years of educational attainment in the same fashion as Borjas (1995). No school completed, nursery school, and kindergarten are recoded as 0 years of schooling; first through fourth grade are recoded as 2.5 years; fifth through eighth grade as 6.5 years; ninth grade as 9 years; tenth grade as 10 years; eleventh grade or twelfth grade without a high school diploma as 11 years; highs school graduate as 12 years: some college, no degree as 13 years; associate degree as 14 years; bachelor's degrees 16 years; master's degree as 17 years; and professional or doctorate degree as 20 years. 89.6 percent of non-proficient Mexican immigrant workers have not completed high school. The situation is slightly better for proficient immigrant workers. But the percentage of not completing high school is still quite large at 65.5 percent.

Natives earn more than immigrants, and there are large differences in hourly wages between the proficient and non-proficient groups for both immigrant and native workers. Recent cohorts of immigrants in general earn much lower throughout most of their working lives than do their native counterparts. According to Padilla and Glick (2000), this pattern appears to be particularly true among Mexican immigrants. The gap is 31.3 percent for immigrant workers ($9.961 vs. $7.586) and 26.8 percent for native workers ($12.066 vs. $9.514). Although about a quarter of Mexican immigrants reside in the south, Mexican natives are more concentrated in the south. Natives are better educated than immigrants. Among immigrants, those in the proficient group are slightly younger and arrived in the United States about 5 years earlier than those in the non-proficient group. Those in the proficient group have approximately 3 more years of schooling than those in the non-proficient group. Lower educational attainment of Mexican American workers, particularly immigrant workers, is likely to be associated with lower labor market skills and lower paying occupations. To the extent that Mexican American workers are disproportionately located in the lower end of the skill distribution and in the lower paying occupations, then increasing disparity in the returns to skill along with lower paying occupational status will result in wider earnings gap between Mexican American workers and non-Hispanic white population. According to Couch and Daly (2002), Mexican American workers will be penalized by increases in the prices of such measured skills as education and returns to particular sectors of the economy. In addition, if labor market discrimination or actual differences in unmeasured skills of Mexican American workers are present, then increasing returns to unmeasured skills will put additional downward pressure on the relative wages of Mexican American workers. Consistent with age and the differences in educational attainment, non-proficient immigrant workers have approximately 4 more years of labor market experience. Proficient workers, both immigrants and natives, are more likely to have professional and managerial occupations than non-proficient workers.

The finding that proficient immigrant workers are less experienced in the labor market, and have resided in the Unite States longer than non-proficient immigrant workers implies that proficient workers have immigrated at younger ages than non-proficient workers. These findings confirm the general notion that children who are exposed to foreign language early in life tend to acquire the language more successfully than adults. The exact nature of the reciprocal relationship between educational attainment and ELP is unclear since formal schooling reinforces the acquisition of English. It appears that the knowledge of English language helps students in schools. Although non-proficient workers, for both immigrants and natives, have longer labor market experience, proficient workers' higher educational attainment, occupational status (managerial or professional), and ELP are more important factors in determining the earnings in the U.S. labor market.

A large proportion of Mexican American male workers are employed in labor-intensive sectors such as farming, craft and operation: the tendency being more prevalent for immigrants than for natives. For both immigrant and native workers, the proportion is higher for non-proficient group (79.9 percent for immigrants and 69.8 percent for natives) than for proficient group (69.1 percent for immigrants and 54.5 percent for natives). As expected, natives are more heavily represented in managerial and professional occupations, and for both immigrants and natives, proficient workers are more strongly represented in managerial and professional occupations. Proficient immigrant workers are more than twice likely to be U.S. citizens. This result is not surprising since the acquisition of citizenship not only can be used as a proxy for the willingness to assimilate in the U.S. society, but also requires certain level of knowledge in the English language. Table 2 lists the estimated coefficients of the earnings regression.

Most of the estimated coefficients conform to the theoretical expectations. The relatively low adjusted [R.sup.2] of non-proficient immigrant workers, however, implies that other factors not discussed in this paper may play important role to explain their earnings. Estimated coefficients of variables related to educational attainments reveal important differences as well as some similarities between groups. While educational attainment is a significant predictor of earnings, its impact on earnings is greater for proficient workers than for non-proficient workers. This result applies to both immigrant and native workers. For every additional year of education, the hourly wage of proficient immigrant workers increases by 1.5 percent while the increase for the non-proficient immigrant workers is 0.8 percent. For native workers, the hourly wage of proficient workers increases by 2.8 percent for every additional year of education, but the increase is only 1.5 percent for non-proficient workers. In sum, rewards to education are greater for natives than for immigrants, and within each group, the returns to education are approximately twice greater for proficient workers than for non-proficient workers. This finding confirms the importance of communication skills and suggests that educational attainments per se are sufficient enough to be translated into earnings power when they are not associated with ELP. This result confirms the hypothesis that immigrants' education is discounted in the U.S. labor market (Betts and Lofstrom, 2000). Although there are many explanations for this practice, ranging from the poor quality of education received outside the United States to the blatant labor market discrimination, the exact reason is not known.

To account for nonlinearity in the returns to education, two categorical variables "High school" and "College" are included in the earnings regression. The baseline group is the workers who did not complete high school. The estimated coefficients of these variables show that the U.S. labor market rewards college education substantially more than it does high school education. However, this interpretation does not apply to non-proficient immigrant workers for whom the premiums of high school education and college education are almost identical, although the estimated coefficient of "College" variable is statistically insignificant. This puzzling result may be due to the insufficient number of college graduates in non-proficient immigrant group. According to Table 1, only 1.2 percent of non-proficient immigrant workers have college or higher degrees. In addition to this, their lower ELP implies that they most likely would have received their higher education outside the United States, subject to a discount in the U.S. labor market. Results in Table 3 show that regardless of nativity, proficient workers earn more than non-proficient workers in all categories of educational attainment.

Similar patterns are found for work experiences. As summarized in Table 4, years of work experience are better rewarded for native workers than for immigrant workers. Every year of work experience increases the hourly wage of proficient native workers by 1.42 percent but the increase for non-proficient workers is smaller at 0.69 percent. For immigrant workers, each additional year of work experience rewards proficient workers by 0.68 percent increase in hourly wage while the increase for non-proficient worker is slightly smaller at 0.55 percent. This finding reaffirms the importance of ELP. Workers with ELP are more likely to be better educated and engage in managerial and professional occupations. Given that the premium earned by workers with more education has grown dramatically during the past two decades, the higher returns to work experiences of proficient workers are not surprising. As Juhn et al. (1993) point out, while returns to education have increased for almost all workers, recent entrants to the labor market have experienced the most dramatic increase. Therefore, as Mexican American youths attain educational attainment and ELP as they stay longer in the United States, the labor market reward to their work experience is expected to improve.

Estimated coefficients of other explanatory variables also generally conform to theoretical predictions. Workers who have professional or managerial occupations earn significantly more than the worker in the baseline group (operators). But this pattern does not hold for non-proficient "professional" workers. The estimated coefficients of the variable "professional" for non-proficient workers are statistically insignificant for both immigrant and native workers. One possible reason may be a vertical segregation: even if a worker has a professional occupation, the lack of ELP may place the worker in a lower position within the broadly defined category of professional occupations.

Estimated coefficients for the variables "Years since migration," and its square show the concavity of the length of stay in the United States on the hourly wage of immigrant workers. The variable "Married" was included to test the hypothesis that married workers are more mature, stable, and less likely to quit their jobs than unmarried workers, and therefore likely to earn more. For all categories, the estimated coefficients of the variable "Married" are statistically significant at 1 percent level to prove the validity of the marriage premium hypothesis. Residence in the south depresses the wage of native workers more. This result implies that native Mexican American workers who reside in the south may have less labor market skills than those living in other areas, since they would have migrated to other higher paying areas within the United States if they had marketable skills that are valued elsewhere.

Empirical findings and their implications are summarized as follows. First, regardless of nativity, proficient workers earn more than non-proficient workers. The interpretation follows mainly from viewing language as human capital since the ability to speak English have an effect on productivity. Second, the returns to education and to work experience are lower for non-proficient workers. This implies that some of the educational attainments of non-proficient workers may be accumulated prior to immigrating to the United States, and this portion of schooling may not be as productive as schooling received in the United States. Although the empirical findings are instructive and conform to theoretical expectations, following caveats should be considered. First, the measure of ELP used in this paper is based on self-assessment, which is often problematic in accuracy. To the extent that there exists uncertainty in the process in which census respondents assess their own English proficiency, the empirical results presented in this paper should be interpreted with care. Second, as Park (1999) indicated, the lower returns to educational attainment for non-proficient workers may be related to their unobservable ability. Non-proficient workers earn less probably because of their lower productivity and lower educational attainment that are the result of their lower abilities.

4. SUMMARY AND CONCLUSIONS

Using the data drawn from the 1990 U.S. Census PUMS, this paper reaffirms the importance of ELP in determining the earnings of Mexican American workers in the U.S. labor market. While the exact relationship between ELP and educational attainment is unknown, the reciprocal relationship of ELP and education complements each other, helping workers to be qualified for better occupations and enhance their earnings in the labor market. This mechanism is also reinforced by the higher compensation for work experiences for proficient workers. Therefore, in order for Mexican American population to make a progress and move into the U.S. social and economic mainstream, particular attention should be focused on improving educational attainment and ELP. The empirical findings in this paper raise several questions for policy makers: What are the factors behind the lower educational attainment of Mexican Americans workers? What programs are needed to improve their educational attainment? Although the prospects of Mexican American workers economic success lies in the improvement of educational attainment and ELP, there are no "magic bullets" that will eliminate all disparities between Mexican Americans and other workers (Chavez, 1991). Improving educational attainment is a long and complex process that requires a change in values, and a willingness to defer gratification.

The lack of ELP impedes immigrants to succeed in the U.S. school system and labor market. Bilingual education can at least make transition period less turbulent for Mexican youth in the U.S. school system in the short run. The long run efficacy of bilingual education, however, is controversial and is beyond the scope of this paper. Since investment in English language depends on its potential future benefits, on the cost of acquisition, and on the individuals' efficiency in producing it, policies should be implemented in different manner to different group of people. Attempts to limit educational service to the children of illegal Mexican immigrants may not be a good idea. Denying such children the benefits of a free public education would cost society more in the long run, by making it likely that such children would grow up to be dependent on welfare.

While the proficiency in colloquial English is important per se, reading and writing skills are more important to enhance economic status of immigrant workers and such skills take more systematic efforts to acquire than speaking proficiency. Although, for example, a colloquial proficiency in English is sufficient for workers who have labor-intensive occupations, it takes more than just colloquial ability to work for professional and managerial occupations. Since reading and writing skills in English language take more systematic efforts to acquire, workers should be afforded the well-structured educational assistance to enhance ELP in reading and writing. However, given that a majority of Mexican immigrant workers are not proficient as a result of socioeconomic circumstances, it is impractical at this time for workers who are currently engaging in lower status occupations to stop working to attend school to enhance ELP. Therefore, flexible schedules of English language classes to accommodate adult workers should be an important consideration in future policy tasks.

A particular focus should be directed at Mexican youths at school age. Since "drop in" the U.S. school system is known to be more problematic than "stay in" or "drop out," policy attention should be focused on ensuring that Mexican American youths are introduced into the U.S. education system. Such early intervention as organized under the auspices of preprimary programs in various forms or Head Start programs will be beneficial. As the recent fiscal profligacy concerns many, a large number of states consider cutting back education budgets. Affirmative action that had helped to increase the college enrollment of Mexican American youth, or Hispanic youth in general, is under a heavy public scrutiny as the social climates have grown in favor of a diversity based on merit. For these reasons, it is unpredictable whether more financial resources can be allocated to enhance educational attainment and ELP of Mexican Americans or even ethnic minority population in general.

TABLE 1. SAMPLE STATISTICS

                                     Immigrants

                             Proficient       Non-proficient

Variables                    Mean (SD)          Mean (SD)

Wage (dollar)               9.961 (5.906)      7.586 (4.871)
Age                        36.277 (8.826)     37.062 (9.664)
Experience                 21.556 (10.378)    25.354 (11.398)
Education                   8.721 (4.452)      5.708 (4.073)
Less than high school        .655 (.475)        .896 (.306)
High school degree           .163 (.369)        .062 (.240)
Some college                 .132 (.338)        .031 (.174)
College degree               .030 (.171)        .006 (.077)
More than college            .021 (.141)        .006 (.076)
Years Since Migration      15.483 (8.792)     10.759 (7.424)
Married                      .752 (1.479)       .848 (1.578)
South                        .236 (.425)        .249 (.432)
Citizen                      .312 (.463)        .158 (.365)
Speak English Very well      .485 (.499)            --
Speak English Well           .515 (.499)            --
Speak English Not Well           --             .694 (.461)
Speak No English                 --             .306 (.461)
Managerial                   .046 (.209)        .013 (.115)
Professional                 .027 (.163)        .005 (.073)
Technical                    .014 (.119)        .003 (.054)
Sales                        .044 (.205)        .018 (.132)
Clerical                     .040 (.196)        .018 (.134)
Service                      .138 (.344)        .144 (.351)
Farming                      .112 (.316)        .223 (.416)
Craft                        .246 (.431)        .193 (.395)
Operators                    .333 (.471)        .383 (.486)
N                              20,784             18,348

                                         Natives

                             Proficient       Non-proficient

Variables                    Mean (SD)          Mean (SD)

Wage (dollar)              12.066 (6.585)      9.514 (6.148)
Age                        38.609 (9.984)     43.004 (12.038)
Experience                 20.703 (10.955)    29.232 (15.173)
Education                  11.906 (2.902)      7.772 (4.823)
Less than high school         .283 (.451)      .660 (.474)
High school degree            .321 (.467)      .178 (.383)
Some college                  .287 (.453)      .126 (.332)
College degree                .075 (.263)      .025 (.157)
More than college             .034 (.181)      .011 (.103)
Years Since Migration            --                 --
Married                      .815 (1.465)      .874 (1.514)
South                        .409 (.492)       .558 (.497)
Citizen                          --                 --
Speak English Very well      .865 (.341)            --
Speak English Well           .135 (.341)            --
Speak English Not Well           --             .893 (.309)
Speak No English                 --             .107 (.309)
Managerial                   .094 (.291)        .036 (.186)
Professional                 .065 (.246)        .024 (.154)
Technical                    .035 (.184)        .018 (.135)
Sales                        .073 (.261)        .042 (.201)
Clerical                     .087 (.282)        .051 (.219)
Service                      .101 (.302)        .130 (.337)
Farming                      .036 (.187)        .139 (.346)
Craft                        .236 (.425)        .227 (.419)
Operators                    .273 (.445)        .332 (.471)
N                              34,574              1,029

Source: 1990 Census PUMS

TABLE 2. ESTIMATED COEFFICIENTS

                                              Immigrants

Variables                          Proficient       Non-proficient

Constant                         1.45 (.027) **     1.49 (.029) **
Exp                               .018 (.002) **     .009 (.002) **
[Exp.sup.2] / 100                -.026 (.003) **    -.014 (.003) **
Years since migration             .026 (.003) **     .029 (.002) **
Years since [migration.sup.2]    -.004 (.001) **    -.001 (.001) **
Education                         .015 (.003) **     .008 (.002) **
High school                       .068 (012) **      .048 (.017) **
College                           .197 (.025) **     .049 (.042)
Married                           .111 (.009) **     .077 (.009) **
South                            -.189 (.009) **    -.136 (.009) **
Managerial                        .126 (.019) **     .148 (.033) **
Professional                      .219 (.027) **     .055 (.059)
Technical                         .102 (.028) **    -.021 (.058)
Sales                            -.009 (.019)        .008 (.033)
Clerical                          .015 (.016)       -.011 (.027)
Service                          -.188 (.012) **    -.152 (.012) **
Farming                          -.274 (.014) **    -.171 (.011) **
Craft                             .104 (.011) **     .083 (.011) **
Adjusted [R.sup.2]                   0.2185             0.1143
  N                                  20,784             18,348

                                               Natives

Variables                          Proficient       Non-proficient

Constant                         1.49 (.024) **     1.77 (.113) **
Exp                               .030 (.001) **     .008 (.006)
[Exp.sup.2] / 100                -.039 (.002)**     -.002 (.011)
Years since migration
Years since [migration.sup.2]
Education                         .028 (.002) **     .015 (.008) *
High school                       .113 (.011) **     .197 (.063) **
College                           .302 (.019) **     .528 (.121) **
Married                           .137 (.007) **     .123 (.043) **
South                            -.234 (.006) **    -.289 (.037) **
Managerial                        .139 (.011) **     .104 (.098)
Professional                      .129 (.015) **     .175 (.099)
Technical                         .127 (.015) **     .306 (.099) **
Sales                            -.003 (.012)       -.006 (.095)
Clerical                          .015 (.011)
Service                          -.171 (.011) **    -.174 (.057) **
Farming                          -.292 (.018) **    -.213 (.066) **
Craft                             .096 (008) **      .144 (.051) **
Adjusted [R.sup.2]                   0.2338             0.2220
  N                                  34,574              1,029

* p<.05. ** p<.01.

Source: 1990 Census PUMS

TABLE 3. MEAN HOURLY WAGE BY EDUCATIONAL ATTAINMENT

                                       Immigrants

                             Proficient        Non-proficient

Less than highs school     $9.15 (65.5%) *    $7.54 (89.6%)
High school degree        $10.45 (16.3%)      $7.79 (6.2%)
Some college              $11.68 (13.2%)      $8.18 (3.1%)
College degree            $14.01 (3.0%)       $7.46 (0.6%) **
More than college         $14.95 (2.1%)       $9.43 (0.6%)

                                      Natives

                            Proficient      Non-proficient

Less than highs school    $10.08 (23.8%)     $8.36 (66.0%)
High school degree        $11.31 (32.1%)    $11.02 (17.8%)
Some college              $13.09 (28.7%)    $11.37 (12.6%)
College degree            $16.32 (7.5%)     $15.74 (2.5%)
More than college         $19.03 (3.4%)     $17.77 (1.1%)

* 65.5 percent of proficient immigrant workers have less than high
school education.

** The number of observation is 110.

Source: 1990 Census PUMS

TABLE 4. MARGINAL RETURNS TO WORK EXPERIENCE EVALUATED AT THE MEAN

                                   Immigrants

                        Proficient        Non-proficient

Marginal returns to     0.68 percent *    0.55 percent
experience evaluated
at the mean

                                   Natives

                        Proficient      Non-proficient

Marginal returns to     1.42 percent    0.69 percent
experience evaluated
at the mean

* The marginal returns to experience evaluated at the mean are
calculated as the partial derivative of the natural logarithm of the
hourly wage with respect to work experience. For example, for
proficient immigrant workers, the marginal returns becomes 0.018 +
2(-.026 x 21.556/100) = 0.00679. A year of work experience increases
the hourly wage by 0.68 percent for proficient immigrant workers.

Source: 1990 Census PUMS

REFERENCES

Betts, Julian R. and Lofstrom, Magnus., The Educational Attainment of Immigrants: Trends and Implications. Pages 51-116 in George J. Borjas (ed.). Issues in the Economics of Immigrations. National Bureau of Economic Research Conference Report. The University of Chicago Press: Chicago, 2000.

Borjas, George J., "Assimilation Changes in Cohort Quality Revisited: What Happened to Immigrant Earnings in the 1980's?", Journal of Labor Economics, Vol. XIV, 1995, 201-245.

Borjas, George J., "The Economics of Immigration", Journal of Economic Literature, Vol. XXXII,_1994, 1667-1717.

Borjas, George J., Friends or Strangers: The Impact of Immigration on the U.S. Economy, New York: Basic Books, 1990.

Bowles, Samuel. and Gintis, Herbert., "The Inheritance of Inequality", Journal of Economic Perspectives, Vol. XVI, 2002, 3-30.

Chapa, Jorge., "The Myth of Hispanic Progress: Trends in the Educational and Economic Attainment of Mexican Americans", Journal of Hispanic Policy, Vol. IV, 1990, 3-18.

Chavez, Linda., Out of the Barrio: Toward a New Politics of Hispanic Assimilation, Basic Books, 1991.

Couch, Kenneth. and Daly, Mary C., "Black-White Wage Inequality in the 1990s: A Decade of Progress", Economic Inquiry, Vol. XL, 2002, 31-41.

Dustmann, Christian. and Fabbri, Francesca., "Language Proficiency and Labour Market Performance of Immigrants in the UK", IZA Discussion Paper No. 156, 2000.

Garcia, Eugene E., Hispanic Education in the United States: Raices Alas, Rowman & Littlefield Publishers, Inc., 2001.

Gonzalez, Arturo., Mexican Americans and the U.S. Economy: Quest for Buenos Dias, The University of Arizona Press, Tucson, 2002.

Juhn, Chinhui., Murphy, Kevin M. and Pierce, Brooks., Accounting for the Slowdown in Black-White Wage Convergence. Pages 107-143 in Marvin H. Koster (ed.): Workers and Their Wages: Changing Patterns in the United States, AEI Press. Washington, DC., 1991.

Lofstrom, Magnus., "Self-Employment and Earnings Among High-Skilled Immigrants in the United States", IZA Discussion Paper No. 175, 2000.

Lowell, Lindsay B. and Suro, Roberto., "The Improving Educational Profile of Latino Immigrants", Pew Hispanic Center Report, 2002.

Lubotsky, Darren., "The Effect of Changes in the U.S. Wage Structure on Recent Immigrants' Earnings", Working Paper No. 458. Industrial Relations Section, Princeton University, 2001.

Padilla, Yolanda C. and Glick, Jennifer E., "Variations in the Economic Integration of Immigrants and U.S.-Born Mexican", Hispanic Journal of Behavioral Science, Vol. XXII, 2000, 179-193.

Park, Jin Heum., "The Earnings of Immigrants in the United States: The Effect of English-Speaking Ability", American Journal of Economics and Sociology, Vol. LVIII, 1999, 43-56.

Sable, Jennifer., Stennett, Janis. and Pinkerton Computer Consultants, Inc., The Educational Progress of Hispanic Students. The Condition of Education 1998., http://nces.ed..qov/pubs98/conditions98.

Trejo, Stephen J., "Why Do Mexican Americans Earn Low Wages?", Journal of Political Economy, Vol. CV, 1997, 1235-1268.

Dr. Jongsung Kim earned his Ph.D. at the Johns Hopkins University in 1996. Currently he is an Assistant Professor of Economics at Bryant College, Smithfield, Rhode Island.

COPYRIGHT 2003 International Academy of Business and Economics
COPYRIGHT 2004 Gale Group