Today I had the pleasure of speaking over the phone with someone from the Census Bureau. This was my first time speaking with someone who works there, and it was fun to move from thinking of the Census Bureau as “a government agency” to “a place where someone I know works.”
The conversation revolved around how R users can get better at working with Census data. In my experience the R community has a lot of interest in Census data. But the details can be hard. Here are some of the questions I had:
- I’m familiar with the American Community Survey (ACS). But what other data is available? The Census director recently said that in 2015 alone the Bureau conducted over 130 surveys. That’s enough surveys for me to not be sure where to start.
- Why is only some data available via the API? Will all the data eventually be available via the API?
- How can I be sure that I am using the data correctly? For example, sometimes I’m unsure how to deal with margins of error. And there are so many ways to measure income (e.g. household vs. per capita, median vs. mean) that I’m not sure which one I should use.
- PUMS sound important, but I don’t know how to use them!
- Why are there so many tables?!
Not surprisingly, a single phone call did not answer all my questions. But I did leave the conversation with a list of recommendations. The following should help people who want to “take the next step” with Census data:
- Watch a Free Webinar. The Census Bureau regularly hosts webinars on a host of niche topics. Who better to teach you about Census data than the Bureau itself?
- Join the Census API Mailing List. The Census rolls out new datasets via its API all the time. This is something that I just learned. If you want to stay on top of these changes, the Census API’s mailing list is a good place to start.
- Learn to use American Fact Finder. This is the main tool that the Census Bureau provides for accessing its data.
If you’re new to working with Census data in R, a great place to start is my free course Learn to Map Census Data in R.