Labor Force Issues
This chapter summarizes the workshop discussions on the labor force issues of wage trends and newly emerging industries.
Economic Aspects of Immigration
Economists have two interests in immigration: the labor market impacts of immigration and the progress of immigrants in the U.S. economy. It should be stressed that labor markets are not static; they are always moving toward a new equilibrium. Because there are both short-run effects and second-order effects, depending on the speed of adjustment in the market, cross-sectional data can fail to detect changes. The study of labor market substitution and complementary changes requires longitudinal data, which may tell stories different from those of the cross-sectional data.
There are two important areas for research on immigrants and the labor force. First, more work is required on the bounds and nature of labor markets. Labor markets have changed considerably in recent decades, presenting a changing context for recent immigrants. Metropolitan labor markets are now a prominent feature for immigrants, yet we lack a clear idea of the edges of urban labor markets and how metropolitan labor markets work. Conceptual work is also needed on the role of employers in metropolitan labor markets, how they recruit workers from within and outside urban areas, and how they affect labor force participation.
Second, there are issues involving ethnic groups in the study of labor markets. Current research often uses pan-ethnic labels, such as Asian or Hispanic,
without a sound rationale for including many different nationality groups within a common category. Also, studies frequently use the group white non-Hispanic as a reference for empirical comparisons. Important questions for research are: What groups should be used and what is the proper reference group for empirical research on immigrants? If labor markets are structurally divided, as is often argued, what ethnic or country-of-origin aggregation is meaningful for comparisons?
Although discussions of the economic assimilation of immigrants encompass several alternative perspectives, most agree on the beginning fact: in cross-sectional data, immigrant earnings are typically lowest for the most recent arrivals and highest for those who came long ago, compared with native-born persons. The first explanation for this pattern is that it reflects labor market assimilation (this is the view associated with Chiswick, 1978). According to this view, earnings of the typical immigrant rise quickly after arrival and eventually equal or even overtake the wages of natives. A critical concept for this first perspective is the idea of location-specific capital. Immediately after immigration, immigrants suffer the disadvantage of knowing far less than natives about the cultural, institutional, and economic characteristics of the U.S. labor market. In addition, their skills may not be readily adaptable to the new market. Consequently, their initial wages will be below those of otherwise equally qualified native-born residents. As the immigrants become better acquainted with the U.S. labor market, their wages catch up with natives, usually within 10 to 15 years after arrival. Empirical research with cross-sectional data suggests that the wage gap between immigrants and natives is most pronounced for immigrants whose culture and language are most dissimilar to the United States. The wage gap between immigrants and natives is smaller for immigrants from Canada and the United Kingdom, for example, for whom the initial differences are smaller.
A second perspective is that economic assimilation is largely a mirage. Instead, it is argued, the declining cohort quality of immigrants is the primary reason for the observed cross-sectional decline in wages for the more recent arrivals (this view is associated with Borjas, 1990). In this view, the reason that more recent immigrants do poorly is not because they have not yet assimilated into the U.S. labor market. Rather, their wages are low because more recent waves of immigrants are of lower "labor market quality" (i.e., have lower education qualifications and poorer job skills) than earlier immigrant cohorts. If the quality hypothesis predominates, the future outlook for recently arrived immi-
grants is dim. The wage deficiency that they now face will continue with them throughout their U.S. labor market careers. This contrasts with the assimilation hypothesis, which promises a far more optimistic future for recent immigrants as their wages rise as they adjust to the U.S. labor market.
An examination of empirical evidence on the wage trends of male Hispanic workers from 1940 to 1987 reveals that they earned about two-thirds of the earnings of all white men during the period. There was also an increase since 1970 in the proportion of Hispanic men living in poverty, following a long-term decline since 1940. At present, there is a higher proportion of Hispanic men living in poverty than black men. In 1940 black men earned about two-thirds of what Hispanic men earned in 1940; they now earn more. The first explanation for this pattern is that it indeed reflects labor market assimilation, with the earnings of the typical Hispanic male immigrant rising quickly after his arrival and eventually equaling the wages of natives. The alternative view is that their wages are low because more recent immigrant cohorts are of lower labor market quality than earlier ones.
Wage analysis focused on the life cycle of Hispanic men who entered the United States between 1956 and 1960, classified by years of labor force experience, shows an overall pattern of little improvement with experience. Educational attainment improves over time, hut not as fast as it does for white men. The result is that the schooling gap relative to white men has widened. Data on English proficiency for Hispanic men demonstrates that there is a strong relationship between wages and English language skills.
Part of the explanation for the wage trends of Hispanic men may be employment discrimination. Hispanic men differ from other men in the labor force on a number of factors, including age, labor force experience, education levels, English language abilities, and place of residence. In work cited by Smith (1992), statistically adjusting for other factors explaining wage data suggests that a modest proportion (about 8 percent) of the variation in wage data may be possibly explained by discrimination. Apparently, discrimination is moderate, at best, for Hispanic men and accounts for little in the overall differences of wages between Hispanic men and others.
Percentile changes in wages for Hispanic men from 1971 to 1987, an additional useful analysis, lead to two conclusions. First, structural changes made it difficult to achieve wage increases during the period. Second, Hispanic immigrants experienced wage decreases during the period. Wage trends for immigrants can be affected by immigrants who have left the United States; in the analysis of male Hispanic workers, educational attainment was rising for the immigrant cohorts, so perhaps the poorly educated left. But if, for example, 30 percent exit, one would need to assume absurd changes (e.g., that only the poorly educated with very high wages departed) in order to produce the observed results (i.e., decreased wages, increased education) for selection alone—so emigration cannot explain the wage trends.
The main implication of this research is that excessive reliance on a single data set can be problematic for the study of wage trends. Because of the selection involved in immigration and emigration, it is important to study wage trends with data on individuals over time. Individual longitudinal data on immigrants would help by providing information on individual factors associated with wage rates, thereby enabling the direct study of selection. Improvements in the data collected would yield better interpretations, for example: (1) a census question about first and last entry to the United States; (2) maintenance of consistent data on Hispanic ethnic categories between censuses; (3) a census question about what work immigrants did before immigration; and (4) 4 or 5 retrospective census questions about prior work and occupational experience.
Immigrants and Emerging Industries2
The response of the U.S. economy to the internationalization of investments and the growing competitiveness of world markets raises the question for research of whether there is a connection between domestic industrial restructuring and the employment of immigrants. Electronics, a critical sector in these global trends, has been characterized as an industry shaped by advanced technology and giving rise to professional and highly skilled jobs that require a well-trained labor force. A close examination of the electronics industry in New York and southern California reported at the workshop suggests that the majority of jobs in this area require low levels of skill.
This effort to document the restructuring of the electronics industry and its impact on demand for and employment of Hispanic women reveals a major finding: there is increasing change in the U.S. work force as a result of international competition. In particular, the electronics industry has been changing in three interrelated ways: (1) shifting to subcontracting and decentralized manufacturing, (2) expanding the informal stratum with the proliferation of small shops for industrial work, and (3) tapping a labor force that is increasingly female, with many women from minority and immigrant populations.
The electronics industry in southern California now employs a significant number of foreign-born Hispanics, especially women, in the lower echelons of electronics production. Of the direct production workers in the southern California electronics industry, 45 percent are Hispanic and one-third of them are foreign-born; more than 60 percent of the direct production workers are female (these percentages are higher than for other California manufacturing industries). The electronics industry in New York also relies heavily on ethnic minorities and immigrants, particularly from East Asia and Latin America.
The growing presence of Hispanic women in the lower and middle echelons
of electronics production is partly due to changes in the composition of local labor markets in southern California and New York. The increasing number of Latin American and Asian immigrants makes it likely that many of them will seek employment in the various sectors of the economy, including industrial production. However, there is a high degree of selectivity of Hispanic women in the electronics industry.
Workshop discussion of this research suggested three implications. First, there is a need for further research into the process of immigrant selection within industries. The observed preference for particular labor pools—those formed by women, immigrants, and minorities—is an important factor in determining the locational choices of electronics companies. This is important because it suggests that not all workers benefit equally from the restructuring of domestic production and, similarly, not all will benefit equally from the future changes of U.S. industry. Moreover, there is considerable heterogeneity of immigrants, with divisions into various social classes.
Second, employment in smaller companies provides an alternative to the traditional forms of large industrial employers and warrants further attention. The restructuring of the domestic electronics industry has entailed subcontracting and the streamlining of production operations. Production has been translated into a proliferation of smaller, more flexible companies and an expansion of the informal stratum. The trend seems to be toward the growth of smaller companies, some of which illustrate the advantages of flexible, specialized niches, and others of which are sweatshops that depend on minimally paid workers. The relationship between these emerging new forms of production and immigration requires study.
Third, the North American Free Trade Agreement (NAFTA) is altering the employment context for immigration, which will need to be carefully considered. Some observers fear that the elimination of trade barriers among Canada, Mexico, and the United States will accentuate the trend toward Mexico-bound relocation of the manufacturing industry from Canada and the United States. The electronics industry reflects the growing polarization of the labor market in the United States, with the creation of some professional and highly skilled jobs at the top end and large numbers of unskilled jobs at the bottom end. Along with garment workers, electronics workers are paid the lowest wages of industrial workers in the United States. The availability of labor pools formed by immigrants from Asia and Latin America represents one factor affecting the decisions of companies to maintain or relocate their operations. Domestic operations, with a proximity to facilities where new technology is being developed and to the domestic market, may continue as a complement to operations abroad, rather than as a mutually exclusive choice. Workshop participants raised questions about the possible loss of specialized low-wage industrial jobs caused by NAFTA and the role of immigration in the creation of labor supply for low-wage employment.
Workshop discussion on the study of the labor force addressed the broad issues of the study of assimilation and the labor force impacts of immigrant entrants. For future research, the topic of labor assimilation involves the problems of defining the concept of quality and of interpreting aggregated and disaggregated data.
Discussion suggested three topical areas for further attention in immigration research on labor markets. First, period effects are important, albeit sometimes ignored, for work on the impact of immigrants on the labor force. Prior waves of immigration often paralleled economic cycles in the United States and Europe, with people moving in response to prospects for better employment. Recent immigration flows have not paralleled U.S. economic conditions; in recent years, the United States experienced high immigration (compared with the percentage of the labor force in prior years, although a substantial portion of immigration in the 1980s was a result of the legalization of previously illegal immigrants) during a period of weak job creation. Research is needed on employment experience for immigrants, for common periods of arrival, during different parts of the economic cycle.
Second, there has been a lack of study of the niche impacts of immigration. The industrial and regional complexity of the U.S. economy is great, so immigrants may have had a pronounced impact in specific local contexts. Immigrants differ in their labor force qualifications and, depending on the local area, may differ in their impact on the economy. Workshop participants raised questions about the impact of immigration in local areas, where there may be large or small numbers of immigrants. Some researchers suggested that comparative local studies would improve the understanding of specific labor market impacts of immigration.
Third, immigrants differ in a number of ways that are important for labor force studies, including their place of origin, skills upon arrival, and their immigrant status (illegal or legal, and visa status for legal aliens) upon arrival. In addition, there are important variations in education levels and English language abilities among immigrants. Asian and Eastern European immigrants are more likely to have higher levels of education. There has been a continuing flow of professionals with advanced degrees in recent decades from many developing countries to the United States. Although such selection by education complicates the analysis of occupational mobility of immigrants (among other topics), it must be taken into account. Too few empirical studies consider the diversity of immigrant types, categorized in several different ways, or consider different types of immigrants in the labor force.
Labor force studies of immigrants could be conducted using several types of major data sets: (a) field studies using multisite areas, (b) special studies, (c) administrative record systems, (d) national surveys, using the Current Population Survey as the key survey (with the idea of perhaps incorporating an additional sample of immigrants into these data), and (e) longitudinal surveys.