H ISPANIC American men have lower average wage rates than white non-Hispanics. In 1975 the average white non-Hispanic male wage-earner in the United States earned $5.97 an hour. Mexican men earned $4.31, 72% as much as white non-Hispanics; Puerto Rican men earned $4.52, 76% as much; and Cuban men earned $5.33, 89% as much as white non-Hispanics. By way of comparison, black men's average wages in 1975 were $4.65, 78% of the white male wage.' Several possible reasons for the Hispanics' lower wages come to mind. Among them are age and education, geographic location, immigration, language difficulties, and discrimination. For example, as shown in table 1, Mexicans and Puerto Ricans are younger, on average, than the white non-Hispanic population, and earnings tend to rise with age. Hispanics have lower average levels of education than white non-Hispanics, and wages are positively associated with education. Many Mexican Americans live in the Southwest, where prices are relatively low. Moreover, Hispanics are more likely to be recent immigrants and to lack fluency in English than white non-Hispanics, and so to be at a disadvantage in the labor market. In addition, there is a widespread belief that Hispanics suffer from employment discrimination, and cannot obtain the wages that their human capital would command if they were non-Hispanic whites. How much of the wage differentials described above are due to each of these factors, and to other inter-group differences in wage-related characteristics? In particular, how much impact does labor market discrimination have on the average Hispanic man's wage and how does this compare with discrimination against blacks? This paper provides answers to these questions. A few other efforts have been made to analyze the relative earnings of Hispanic and white nonHispanic men, using 1960 and 1970 Census data (Fogel, 1966; Poston and Alvirez, 1973; Poston, Alvirez, and Tienda, 1976; Long, 1977; and Gwartney and Long, 1978). We use more recent data from the 1976 Survey of Income and Education. This data set enables us to measure wage rates more accurately and to specify the wage function more completely than does the Census. Unlike previous analysts, in estimating the wage function we take account of possible selectivity bias due to the distinction between average wage offers and average observed wages. Thus, we hope to obtain a more accurate and up-to-date measure of labor market discrimination against Hispanic men. Section II explains the method used to separate the minority-white non-Hispanic wage differential into the portions due to differences in average characteristics and the portion due to differences in unobserved factors and discrimination, taking into account the possibility of selectivity bias in the observed wage sample. Section III describes the data used in the study and the specification of the wage equation. The breakdown of the observed wage differentials into components due to differences in participation in the wage and salary sector, local price levels, average characteristics, and discrimination are presented in section IV. Section V summarizes our findings and discusses their implications for efforts to improve the economic situation of Hispanics in the United States.

Both the distribution of income and the role of ethnicity in economic behavior can be illuminated by an analysis of ethnic differences in married women's labor supply. Differences in wives' labor supply are the main source, beside differences in rates of female headship and wages, of the disparities in family income among racial and ethnic groups in the United States (see my 1984 article). Moreover, differences among ethnic subcultures may affect the labor supply of wives more than they influence many other types of economic behavior. Ethnic groups are distinguished by, among other things, views about male and female roles in the family and about wives and mothers working outside the home, as well as by the value placed on children, family size, household composition, and the education of women. These differences may give rise to systematic differences in utility functions that lead to systematic differences in behavior by women in different ethnic or nativity groups who face the same constraints or opportunity set. Such cultural differences in utility functions no doubt are historically shaped by economic as well as other circumstances, and they evolve, but more slowly than the economic conditions. Ethnic differences in attitudes are, therefore, presumably more pronounced in the first generation of immigrants than in their American-born descendants. These cultural attitudes may have both direct and indirect effects on wives' labor supply. They directly affect the allocation of time between home and market work by women with the same education, number of children, etc. They also affect decisions about education, fertility, and other choices which in turn influence the market work opportunities and value of home time, and so indirectly affect labor force participation. Virtually all of the numerous studies of black-white differences in female labor supply have found that black wives have higher labor force participation rates (LFPR) than whites, even after adjusting for differences in measured variables such as age, children, education, location, other family income, and wages. (For a summary of the results, see Mark Killingsworth, 1983, pp. 122, 195, 202, 404; and Phyllis Wallace, 1982, ch. 2.) Several explanations have been suggested-such as blacks' greater marital instability, their extended-family households, black husbands' lower wages and less stable employment-but none has proved satisfactory. It seems that black wives' higher labor force participation is in large part a cultural difference, rooted in the historical experience of blacks in America, and not explainable by current conditions alone. No one has yet attempted to measure and account for the differences in wives' labor supply among the other ethnic and nativity groups, as this paper will do. These differences are large, with the variation in annual LFPRs among ethnic and nativity groups being greater than the variation in annual hours worked for those in the labor force, as shown in Table 1. The ranking of groups in terms of annual hours worked by those in the labor force differs from the ranking in terms of LFPRs. This suggests that different parameters govern the participation and hours worked decisions, perhaps because the groups face different fixed costs of working. These two aspects of labor supply therefore need to be analyzed separately. In this paper I focus on the differences in labor force participation rates. *Department of Economics, Hunter College, and Graduate School of the City University of New York, 695 Park Avenue, New York, NY 10021. I thank Cecilia Conrad for helpful discussions. A PSC-CUNY Research Award helped support this research.