Explaining differences in economic performance among racial and ethic groups in the USA: the data examined

American Journal of Economics and Sociology, The,  Oct, 1996  by William Darity, Jr.,  David K. Guilkey,  William Winfrey

• E-mail
• Print
• Link

I

Introduction

Ethnically and racially plural societies, like the United States of America, customarily are characterized by a high degree of intergroup inequality. Whether indexed by disparities in income, wealth, or occupational status, wide differences continue to be observed across ascriptively differentiated groups. Most research has addressed gaps in economic outcomes between blacks and whites, but the decennial censuses for 1980 and 1990 facilitate a more intricate look at America's ethnic heterogeneity. Respondents to both censuses were asked to provide answers to questions about both their race and ancestry which enable researchers to sketch in fine detail the extent of intergroup disparity in America in the late 20th century. Farley (1989, 1990) made use of the 1980 census to investigate economic performance based upon racial and ancestral identification. We extend Farley's study using both the 1980 and 1990 censuses.(1)

We proceed as follows: First, we partitioned the US population between non-Hispanics and Hispanics, the latter including individuals reporting Puerto Rican, Mexican, Cuban, or other Spanish speaking ancestry. Next, we divided the non-Hispanic population between whites and nonwhites. Non-Hispanic nonwhites include persons who are black and persons of Native American, Asian Indian, Chinese, Japanese, Korean, Filipino and Vietnamese ancestry. Then non-Hispanic whites were subdivided by reported ancestry ranging alphabetically from Armenian to Yugoslavian. We also examined differences in economic outcomes between black and white Hispanics and differences in economic outcomes among blacks who reported West Indian and non-West Indian backgrounds.

Our multivariate regression analysis for earnings involves a more expanded specification than that in Farley's original model. The nature of the expanded specification, which includes variables capturing English fluency, foreign or domestic site of birth, and an indicator of the extent to which a person is "assimilated," permits us, unlike Farley, to include persons born overseas in the sample. This has a dramatic effect on the number of observations of persons in racial/ethnic groups who have experienced substantial recent immigration. For example, the total number of observations for Cubans for the 1980 Farley regression was only 172 due to the exclusion of the foreign born. For our expanded specification the total number of observations for 1980 is 2728. There are only 6 observations in 1980 and 10 observations in 1990 for the Vietnamese when the foreign born are excluded; with the inclusion of the foreign born the numbers rise to 553 in 1980 and 1823 in 1990.

We also have run additional earnings regressions with the expanded specification for groups that still have comparatively small numbers of observations even after the inclusion of the foreign born. The additional earnings regressions make use of the 5 percent Public Use Micro Sample (PUMS) just for those groups. This allows us to address the robustness of our results.

Before turning to the multivariate results, the next section presents descriptive statistics for the subgroups that we have defined. We also are able to look at shifts that occurred between the 1980 and 1990 censuses. The data and the results of our statistical examination are presented in a comprehensive set of tables, available upon request from the first listed author. The appendix to this paper provides instructions about how to obtain the tables and a detailed description of the contents of each of the tables.

II

Rich and Poor Among America's Ethnic Groups

Table 1A presents general demographic information for the various ethnic groups while Tables 1B through 1E present economic indicators (see the appendix). Consider four of the indicators in these tables: mean per capita income, percent of population below the poverty line, the proportion of males holding professional/managerial jobs, and the proportion of females holding professional/managerial jobs. For the US population as a whole mean per capita income was $7167 in 1980 and $14,196 in 1990. Twelve percent of the population lived on incomes below the poverty line in 1980, and thirteen percent of the population lived on incomes below the poverty line in 1990. Twenty-six percent of men 25-54 years of age held professional/managerial jobs in 1980, and the same proportion held professional/managerial jobs in 1990. Twenty-three percent of all females between 25-54 years of age held professional/managerial jobs in 1980, while 29 percent held such jobs in 1990. There was a significant general rise over the decade in the percentage of women in professional/managerial occupations that also is evident across specific racial/ethnic groups.

Taking these national averages as the benchmark for comparison makes it possible to identify groups having a profile of economic achievement above or below the mean for the total population. For example, the group self-identified as racially white and ancestrally Russian had the highest estimated income of all groups in the study. In both years, Russian per capita income was nearly twice as high as the national mean. In 1980, we estimate that 7 percent of this population was below the poverty line, while, in 1990, only 6 percent of this population fell below the poverty line. In 1980, 57 percent of Russian males held professional/managerial occupations, and in 1990, 54 percent of Russian males held such occupations.