Home > Data Services > Catalog . Restricted Data . Census . ACS

Search Data Services

Meta Search
search across all the following databases:

Data Catalog
Data and documentation

Common questions and answers.

Entire collection of data resources.

Latest Data News

RSS Feed icon

A questionnaire change: politics or not?

Finding “war brides” in the ACS

Census Bureau is considering adding a new Middle East/North Africa ethnic category

Big Data Initiative at NIH-OBSSR

UM Now Has Access to IndiaStat

Data Services Knowledge Base


I am using a restricted data file that has zip codes for the geocode ID. I need to add some race-specific characteristics to the zip codes. However, zip codes are not iterated by race the way other geographies are (e.g., states, counties, census tracts). Is there a way around this?


You are correct that zip codes do not allow you to get characteristics by one or more race/ethnic groups.

There are two solutions. First, some of the zip code tables in SF3 are race-specific. But, if you are interested in specific Asian groups such as Asian Indian, Chinese, Korean, Japanese, etc., you need to use SF4.

Pull off the tables you are interested in*group using census tract as your geographic unit. Then using a cross-walk from the Mable/Geocorr correspondence engine, you can assign the characteristics of the census tracts to the appropriate zip code.

Many census tracts will be combined to form your zip code characteristics. Your statistical package will allow you to create a weighted average so that you end up with a file of zip code characteristics.

There are two flaws with this procedure. First, in the case of a census tract split across one or more zip codes, one splits the census tract as though there is no geographical diversity. Thus, if the mean income for the census tract is $60,000, you are making the assumption that both parts of the census tract have the same mean income. In fact, it is quite possible, that one part of the census tract is wealthier than the other part.

The second problem is that your race groups will be fairly small in population. SF4 iterates across lots of race groups, but if the population group is small, for reasons of confidentiality (and reliability) there will be no characteristics reported for that group. It is possible that the total for the zip code would be large enough to be reported, but not for each individual census tract that comprises the zip code. I am assuming that due to the sampling frame, most of the zip codes in the geocode file are in areas with large Asian concentrations.

Annotated Resources:

Direct Links:


Related Question Groups: