How to get counts of variables for 1880 Census full-count


I’m trying to download a data set that gives me counts of certain characteristics by census enumeration district. For instance, a data set that includes a column for the number of black people in an enumeration district; or, the number farms in an enumeration district; etc.

Using the data extract system, I’ve selected the 1880 full count, the State ICPSR code variable, the County ICPSR code variable, the Enumeration District Variable, and the Race variable. When I download this extract, however, the data set does not give me counts — i.e. it does not tell me how many black people lived within a census enumeration district within a county. Instead, it seems to be giving me individual responses within the enumeration district. It’s telling me whether a person reported themselves as white (RACE = 1) or black (RACE = 2) within the enumeration district.

How can I create an extract that gives me a data set with counts instead of individual responses within an enumeration district?

Thanks in advance for your help.

Hi Casey,

The extract system can only provide you with microdata where either each observation is an individual respondent or with responses on the household level. Once you have the microdata, the only way to obtain aggregate counts is for you to aggregate these responses to the enumeration district level yourself using a statistical software such as Stata or R. The egen command will help you sum observations by group in Stata, while aggregate will do the same in R. This is my recommendation since the alternatives I outline below may not be ideal. Note that ENUMDIST needs to be combined with STATEICP to uniquely identify districts across states.

One of these alternatives is the online analysis tool. It requires you to analyze the 1880 10% sample rather than the full count file. Since there are many enumeration districts across states, you will likely need to generate separate tables for each state. To do so, enter ENUMDIST into the Row field, RACE into Column, and STATEICP(1) into the Selection Filter. Make sure that perwt is selected and click to run the table. This will generate weighted counts by race by enumeration district for the given STATEICP filter. You will need to re-run the table for each state of interest. Another alternative is to get aggregated data from IPUMS NHGIS, however the lowest level for which data is available for 1880 is on the county level.