At the beginning of this year, I wrote a blog post talking about the changes that the Trump Administration was making to publicly available data. Going back and reading this blog, it’s interesting to see which of my fears were warranted and in what cases I was being reactionary to what was a very uncertain time where data was being removed seemingly at random.
The good news is that we still have access to a lot of data. The 2024 Behavioral Risk Factor Surveillance System is available for download thanks to a court order, the Census Bureau has released 2024 American Community Survey Data and there is a scheduled release date for the 5-year estimates. Additionally, the team over at IPUMS is working to get the individual microdata ready for release!
Not everything has been smooth however. We asked our Ohio Economic Experts Panel about the firing of Erika McEntarfer, the former Bureau of Labor Statistics commissioner and they agreed that reduced trust in that organization would hurt Ohio’s economy. However, because of the government shutdown, the BLS has said they won’t be releasing some employment data from October. While these events might not be related, this does mean that we don’t have all of the federal data we are used to having.
Near the end of my blog post about public data, I mentioned that in 2024 the Census Bureau was going to be changing the way they collect race data by adding categories for Middle Eastern/North African, as well as adding Hispanic as a race instead of asking a separate ethnicity question.
Recently, NPR reported that the current Census Bureau is reviewing those changes, opening the door for them to be reversed. While this doesn't necessarily mean that these standards will be undone, it seems likely based on the administration's past track record with data pertaining to race, gender, etc. that it will.
I want to note here that how we collect race data is not an easy decision. Any time we try to categorize people into neat boxes we invariably have to make assumptions and decisions that harm the overall quality of the data. However, some options are better than others.
The old way the Census collected race data was worse than the 2024 standard. The only argument that I can think of against it is that changing that question makes comparisons to past years of data more difficult, but I think the improved understanding of the present and future far outweighs those costs.
I don't even think it would be too difficult to compare to previous years if you take the time to look at the microdata. WIth a couple of assumptions about how people would be classified under the old system, you could seemingly reconstruct those estimates and get those backward comparisons. It wouldn’t be perfect, but I bet that if dedicated researchers took their time they could produce fairly accurate estimates.
While I don’t expect this to happen, it has crossed my mind that a review of the race standard might mean some other change as opposed to a revision back to the previous rule. The worst-case scenario would be if the Census Bureau decided to not ask a question about race and ethnicity at all. As I said before, there is no “right” way to collect race data. However, we would all be worse off if a change diminished our understanding of race in this country.
The 2030 census is still a long way away. However, projects that big and ambitious require immense amounts of planning and preparation. Decisions made now about the survey will have major ramifications that we will feel for a very long time. I hope that the Census Bureau’s review doesn’t undo this particular change. It might not be perfect, but it’s better than what we had before.