Technical Information

Identification of black and white subgroups
The Census offers two ways to identify a person's geographic origin: ancestry and country of birth. We identify geographic origin by ancestry. Where ancestry is not available, we use country of birth.

Census respondents were able to list a first and second ancestry. To avoid double-counting, we use only the first ancestry here. A large majority of each ethnic group, identified by either first or second ancestry, is captured this way. For users interested in one specific group, it may be preferable to use both, which would yield a larger group size.

American ancestry includes all people identifying their ancestry as American or United States.

European ethnic groups are identified solely by first ancestry.

Most Muslim-origin groups are also identified by first ancestry. Among the ancestries included in North African are Ethiopian and Sudanese. Although Ethiopia and Sudan are actually sub-Saharian African countries, they are included in the Muslim-origin category because they have a Muslim majority. But some important origins are not identifiable in Census summary files. For these groups, we rely on country of birth, which has the disadvantage of capturing only the first generation. These include people from Afghanistan, Indonesia, Malaysia and Pakistan, who are identified as South Asian in our data tables.

Among blacks, the best method of identifying Afro-Caribbeans is by first ancestry. The count of Africans, however, would be inflated by using ancestry, because many blacks who are descended from several generations of forebears in the U.S. list their ancestry as African. Therefore, we again rely on country of birth for this group, with the accompanying disadvantage of counting only the first generation.

In some other census sources, such as the Current Population Survey (CPS), parents' countries of birth are also reported. National estimates of groups' population size that include both the first and second generations can be compared to the figures in these web pages. However the sample size of the CPS is insufficient to be a basis of estimates of group populations in most metro areas.

Comparing black ethnic total to non-Hispanic black total
The black-origin total is computed by summing the total number of Afro-Caribbeans, African, and African-Americans in a given metro area. Although each of the black-origin categories is comprised of non-Hispanic individuals, often the black-origin total will not match the non-Hispanic black total in a given area. In most cases the difference in values will be negligible, but we have identified two potential sources for the observed discrepancies:

1. While the Census provides data for ancestry and place of birth (used to construct the Afro-Caribbean and sub-Saharan African categories, respectively), these variables are not available by race. As a consequence, a small number of whites are represented in our Afro-Caribbean and African totals.

2.The ancestry and nativity classifications are not mutually exclusive. It is possible for an individual to be born, for example, in sub-Saharan Africa and to be of Afro-Caribbean ancestry. In such cases individuals will be represented twice in our black-origin total (once as Afro-Caribbean and once as African). Such cases are admittedly rare, but it is impossible to derive a precise estimate using currently available 2000 census data. However, the 1990 IPUMS data show that such cases comprised only .32% of the nationwide sample.

Metropolitan Statistical Areas
The data on our pages are reported for metropolitan regions. The Census Bureau uses a standard set of definitions of the area included in each "metropolitan statistical area" (MSA) or "primary metropolitan statistical area" (PMSA). In most cases both a central city (or sometimes two or more central cities) and the ring of surrounding suburbs are included. At the top of the page for each metro area, we have listed the counties and the city (or cities) that the Census has designated as the central city (or cities) for that area. After viewing data for the entire metropolitan area, you may elect to see information for only the central city or suburban portion. If there is more than one central city, please remember that what we report is for all these cities combined.

Isolation Index
The isolation index is the percentage of same-group population in the census tract where the average member of a racial/ethnic group lives. It has a lower bound of zero (for a very small group that is quite dispersed) to 100 (meaning that group members are entirely isolated from other groups). It should be kept in mind that this Index is affected by the size of the group -- it is almost inevitably smaller for smaller groups, and it is likely to rise over time if the group becomes larger.

Exposure Index
The exposure indices (P*) describes a group's exposure to all racial groups. We use it to determine the racial composition of the neighborhood of the average person of a given race. When we compare exposure indices for different parts of a metro area (for example the suburbs) with the metro area as a whole, we can see whether a racial group is evenly distributed across the neighborhoods in the metropolitan area, or whether the group is concentrated in neighborhoods.