County Level Population Data, 1995-2013

[Jeff Wooldridge]

Does anyone have, in Stata format (up through Stata 14 is fine), annual county population data for the United States for about the past 20 years? I know that years between censuses are estimates but that's fine with me. Thanks.

[Dimitriy V. Masterov]

I recently pulled annual residential non-seasonally-adjusted pop data for all the counties together from FRED for 1970-2014. This data contains every county on this list other than "Petersburg Census Area, AK" and "Hoonah-Angoon Census Area, AK", so N=3,141 and T=45. You can get the data here.

[Greg Falls]

Jeff
Below is a link which might help.
http://www.census.gov/popest/data/historical/index.html
Greg

[Dimitriy V. Masterov]

In case anyone is curious or wants to update, here's the code:

[Jeff Wooldridge]

Many thanks for this, Dimitriy and Greg.

[Steve Samuels]

What is "FRED", Dimitriy?

[Dimitriy V. Masterov]

FRED stands for Federal Reserve Economic Data. It's nice data repository of US and some international data, plus the internet interface has some plotting capabilities. David Drukker wrote a handy command called freduse that allows you to pull in this data directly into Stata. The catch is that you need to know the name of each time series to do this. Unfortunately, each county has its own name, which is generally a two-letter state abbreviation plus first four letters of the county name plus some integer vaguely related to the county FIPS code (but sometimes missing) plus the suffix POP. freduse does not accept wildcards, so the trick was getting the names all the relevant series from somewhere.

Other commands/repositories, like quandl and wbopendata, also suffer from this problem.

[Steve Samuels]

Thanks, Dimitriy.

[William Lisowski]

The Population Estimates Program of the U.S. Bureau of the Census (http://www.census.gov/popest/) is another source of this sort of data. The current data starts from the 2010 Census, but historical datasets are available for previous decades. The data is in CSV format, which is "Stata format" in the sense that Stata can easily read it. On the upside, the full country is a single file, and the geographic identifiers are standard FIPS codes and text state and county names.

[Ariel Karlinsky]

Sorry to bump an old thread, but I think people might find this useful as well:
SEER from the National Cancer Institute has great data on a county-year level. these can be easily "converted" (I have a stata do file for that if you want it) to measures such as total county population, county population by gender, race, Hispanic origin, age groups, etc. See more here: http://seer.cancer.gov/popdata/