Hi, I’m trying to get the following data on voting-age citizens for each congressional district, census block, or census tract going back to 1910 (blocks/districts where it’s possible, and tracts I can interpolate where it’s not):
-age*race (age breakdown by race, counting Hispanics as a separate race)
-education*race (counting Hispanics as a separate race)
I don’t need to link this data across years. It’s also fine if not all of this data is available; the most important data for me is probably just education, race, age, and income data. What is important is being able to limit the universe to voting-age citizens.
However, I’m having trouble finding similar tables providing this information across years. Is there a way to request all this data at once for all available years?
You should be able to find most of this information from IPUMS NHGIS. NHGIS has geographically consistent time series tables covering 1990-2020 census data for CDs, but only for the 110th-112th Congress (districts used from 2007 to 2013). This only includes information for the census short-form, which will give you sex, age*race, homeownership rate, marriage rate, but not income or education*race. You might not be able to get every combination of AGE*RACE*HISPAN for each year.
For the income and education tables, it will be a bit trickier if you’re looking for geographically consistent series. The problem is that congressional districts change regularly, and different datasets present data using different CD vintages. So for example you can find median household income for CD’s from the 1990 census in Table NP80A. But this is only available for the 1987-1993, 100th-102nd Congress CD boundaries. In contrast, Table B19013 gives median household income in 2010, but only for the Congressional districts that existed in 2010 (the 110th-112th Congress). These are not directly comparable since the boundaries have changed. Alternatively, you might be able to create a consistent series by finding your statistics of interest at the block or block-group level, then using a crosswalk (available at IPUMS NHGIS) to a base year (such as 2010), and then aggregating these to the CDs in place in that year, using the crosswalks available at GEOCORR.