1940 census

The images with details from the 1940 U.S. census went public online today (I wrote about this over the weekend) and immediately swamped the servers at the National Archives and Records Administration. About 1.9 million users — 22.5 million page hits — deluged the site within the first three hours after it opened at 8 a.m. central.

Tweets from the National Archives staff this afternoon said their vendor was adding servers to make the site work better.

I’ve been trying to pull up the image for a particular enumeration district in Minnesota all day and have not yet succeeded (although I quit trying around 1 p.m.-ish when it became clear that my efforts were useless). I’m starting to wonder if I need to get up around 2 a.m. or 3 a.m. and try to access the site while everyone else is sleeping!

Other interesting news today is that the Minnesota Population Center at the University of Minnesota is teaming up with Ancestry.com to create a complete database from the 1940 census images.

This is huge news for researchers and data geeks like me.

It’s going to take four years until they have all of the records keyed.  In the interim, they will release a 1 percent file later this year, and another file sometime in 2013 that will have all the records, but a limited number of variables. By 2016, they expect to have everything keyed.

They will also be doing extensive coding for things like occupation, industry and geography. Steven Ruggles, the director of the Minnesota Population Center, told me the file will have latitude and longitude (some, though, might be the best guesses on the location), as well as the 1940 Census tract and enumeration district codes. It will also be possible, using mapping software, to determine which 2010 Census tracts each record fall in, which makes for lots of interesting analysis options.

The Minnesota Population Center has been making “microdata” available for 20 years, but this dataset is the biggest one they’ve ever had.