R

The 2000 One Percent PUMS also allows the development of data to compare the racial and ethnic identity of spouses. Percentages of married persons in each of the 11 non-white groups who have a white spouse are obtained. Because the 2000 census data measure the incidence of marriage regardless of when the marriages were contracted, the analysis is limited to persons less than 35 years of age in order to capture more recent marriages as opposed to those that occurred in the more distant past. In addition, intermarriage percentages are computed separately for native- and foreign-born individuals. Moreover, because of the small size of native-born individuals in the six specific Asian groups, these are aggregated into a single native-born Asian group. Finally, the intermarriage percentages are broken down by gender.

Table 6.3 reports the intermarriage rates for the 11 groups broken down by nativity status and gender. Among native-born married individuals, American Indians (males, 47.2%; females, 45.1%), Cubans (males, 41.2%; females, 36.7%), and Asian females (39.5%) are the most likely to be married to whites. About one-fifth to one-fourth of Mexicans, Puerto Ricans, and Asian males have a white spouse. African Americans, particularly females, are the least likely to be married to whites, reflecting the rigid color line in American society. By and large, foreign-born individuals tend to be less likely to be intermarried with whites. Nonetheless, three groups of foreign-born Asian women— Japanese (33.0%), Filipinas (30.4%), and Koreans (19.2%)—are the most likely to have exchanged wedding vows with a white male. It is likely that some of these marriages consist of "war bride'' unions involving U.S. servicemen and Asian women (see Saenz, Hwang, and Aguirre 1994). In addition, among foreign-born males, Filipinos (12.7%), Cubans (9.8%), and Japanese (8.6%) are the most likely to be married to white women. In contrast, seven foreign-born groups have intermarriage rates below 5%: Vietnamese men (2.5%), Mexican women (3.2%), Asian Indian women (3.6%), Chinese men (3.9%), Asian Indian men (4.3%), Mexican men (4.6%), and African American females (4.7%).

MEASURES OF INEQUALITY. A few measures of inequality that are commonly used to assess differences between racial and ethnic groups are now discussed. Readers interested in a wider array of measures of inequality should consult Iceland, Weinberg, and Steinmetz (2002), Lieberson (1975), Massey and Denton (1988), and Massey, White, and Phua (1996). Attention is focused here on two dimensions of inequality, namely, spatial inequality and earnings inequality.

MEASURES OF SPATIAL INEQUALITY. Two popular measures for assessing spatial inequality are the dissimilarity index and the isolation index. The dissimilarity index is the most common measure used for assessing distributional differences between two groups (Iceland et al. 2002). Although this index is commonly used to measure differences in residential patterns between groups, it is quite flexible and may be used to assess distributional differences in a wide variety of phenomena (e.g., income categories, educational categories, occupational categories, industrial categories, etc.). The dissimilarity index may be defined as:

where pai refers to the proportion of members of a given minority group in unit i, and pbi refers to the proportion of members of the majority group in unit i. The absolute differences are summed across the units and divided by two. The dissimilarity index ranges from 0 to 1. It represents the proportional amount of one or the other group needing to change to certain other units in order to achieve the same distributions.

The isolation index is defined as:

where x, refers to the minority population size in a unit, X refers to the total minority population size across all units, and tt is the total population in a given unit. The isolation index ranges from 0 to 1 and represents the probability that a given minority group member is likely to come into contact with a member of one's own group.

We use data from the analysis of Iceland and associates (2002) that is based on the construction of five indices, including the dissimilarity index and the isolation index, for Metropolitan Areas (MAs) in the United States, using census tract data from the 1980, 1990, and 2000 censuses. For an example of research using these two indexes in another country, China, see Poston and Micklin (1993). Iceland and his colleagues obtained weighted average indices based on all MAs, as well as others based on selected MAs that meet given population size criteria. For the sake of simplicity, weighted average indices based on all MAs are used. Figure 6.9 shows the weighted average dissimilarity and isolation indices for 2000. The results indicate that African Americans are the most segregated from whites. The dissimilarity index of 0.640 for African Americans indicates that, on average, in any given MA in the United States, nearly two-thirds (64%) of either African Americans or whites would need to move to certain other census tracts in order to bring about equal spatial patterns for the two groups. Latinos also tend to have a high level of spatial segregation, with approximately half of Latinos or whites having to shift to certain other census tracts to bring about the same geographical distributions for the two groups. The isolation indices for African Americans and Latinos indicate

0 0

Post a comment