Want Useful Data on Obamacare? The Census Bureau Won't Help You.



One of the big unknowns about Obamacare is what effect it has had on reducing the number of uninsured in America. Over the last few months, we've seen a number of attempts to measure this, but they've all been relatively narrow surveys that for a variety of reasons don't provide a complete picture. As a result, many analysts on both sides of the debate generally agreed that the best way to answer this question would be to use information gathered by the Census Bureau, which provides some of the most consistent, comprehensive data on the nation's uninsured.

Well, now it looks like we won't be able to use the Census to determine the health law's effects on insurance coverage over time. That's because the Bureau is in the process of radically altering the way it asks questions about insurance status. The update will render useless comparisons between data reported last year and data reported this year, according to The New York Times

The Census Bureau, the authoritative source of health insurance data for more than three decades, is changing its annual survey so thoroughly that it will be difficult to measure the effects of President Obama's health care law in the next report, due this fall, census officials said.

The changes are intended to improve the accuracy of the survey, being conducted this month in interviews with tens of thousands of households around the country. But the new questions are so different that the findings will not be comparable, the officials said.

An internal Census Bureau document said that the new questionnaire included a "total revision to health insurance questions" and, in a test last year, produced lower estimates of the uninsured. Thus, officials said, it will be difficult to say how much of any change is attributable to the Affordable Care Act and how much to the use of a new survey instrument.

The Times reports that the update was not politically motivated, and that it was made in the interests of gathering more accurate data. The old survey methodology is said to have been prone to measurement errors that overestimated the number of uninsured; the new data supposedly corrects that problem.

Now, to be clear, the change will give us one year of pre-expansion data: The new measurement methodology will start with 2013, which means we'll be able to compare last year's insurance pre-coverage-expansion stats with this year's. That will give us some sense of the changes under the Affordable Care Act. But there are still big problems with the timing of the transition. As Aaron Carroll notes, it still destroys the long-term trendline. The trend info would have had a lot of value generally, and also would have set the 2013 results in a helpful context. Now it's going to be difficult to discern the impact of the upheavals that we saw in the health insurance market in 2013, which featured, among other things, large unexplained spikes in coverage, and the cancellation of millions of individual insurance policies. 

This is infuriating and unfortunate. Ignore, for a moment, how politically convenient it is. It's just poor social measurement.

If the interest is in getting the best and most useful data possible, then we should want a time series without breaks or alterations, one that can be used to reliably compare the years before and after the single largest and most controversial policy change intended to affect the provision of health insurance in the last four decades. One might think that creating a consistent measure for this effect would be a priority for people whose job is to create socially useful guides to the makeup of the nation's population. Apparently not. The data may be more accurate going forward, but it will be far less useful for measuring this crucial, once-in-a-generation transition.

As to whether this is politically motivated, there's no evidence that it is. And I would be surprised if this move were intentionally cooked up simply to obscure the figures. But back in 2009, the White House moved to exert more direct control over the Census Bureau and its operations. What that means is that, at the very least, the White House should have been involved. If the Obama administration was truly concerned about transparency in measuring the impact of the health law, then its officials should have known about the change, should have noted it publicly before the end of the open enrollment period, and should have pushed to stop or delay it.

Because as it stands now, we simply won't have the robust information set we need to confidently assess the impact of the Affordable Care Act on the uninsured over time. Nor will we have a good picture of exactly how Obamacare is affecting various types of coverage—the particular ways in which people are, or aren't, accessing health insurance under the law.

These aren't the sort of questions you can answer based on first principles, or what feels right. You need robust measurements. Solid Census Bureau data, taken consistently over time, would have helped answer these questions. But it appears that's not something we're going to have. Instead, we'll be left with ongoing arguments and unanswered questions, especially if the administration continues to keep mum on other critical unanswered questions about the law. A decade from now, we'll be able to definitively determine that Obamacare was a thing, that happened, and did something, and cost some money. But that may be as much as we can know.