When the New York Times reported yesterday that the U.S. Census Bureau is making significant changes to the part of its Current Population Survey that calculates how many Americans have health insurance — thus making it difficult to measure the effects of Obamacare — the initial media reaction on Twitter was outrage.
“Getting worked up into an increasingly heated health nerd rage about the Census changes,” wrote Vox’s Sarah Kliff. “I’m speechless. Completely inexcusable. The administration deserves all of the criticism it will get, and then some,” fumed Megan McArdle. Josh Barro called the Census Bureau’s change “insane” and asked, “Why can’t this wait?”
Why indeed? Such a change comes at seemingly the worst possible time vis-à-vis the health-care law. How are we to judge whether Obamacare has reduced the number of uninsured if we introduce a completely new way to calculate the uninsured rate? Census officials told the Times
“that it will be difficult to measure the effects of President Obama’s health care law in the next report, due this fall” using their new report. Changing the survey means “it is likely that the Census Bureau will decide that there is a break in series for the health-insurance estimates,” they say — meaning they won’t be able to compare data from previous years with the data collected under the new method.
It turns out, however, that the outrage on Twitter was a bit premature. As Patrick Brennan promptly noted, the White House later clarified to Sarah Kliff that the new survey questions — which ask for details about when people had coverage in the past year, and from what source, and are therefore more precise than the old survey method — would be used to collect data for 2013. That will make possible an apples-to-apples comparison of the uninsured from 2013 to 2014, when Obamacare’s exchanges and Medicaid expansions took effect. Kliff correctly notes, however, that the change “will make it difficult to compare the uninsured rate for 2012, the last year for the old questions, and 2013, the first year for the new questions.” It also means that the CPS will essentially have only one year of pre-Obamacare data, as opposed to decades’ worth, which is presumably why Census officials called the timing “unfortunate.”
So although the change is not as bad as it initially appeared to be, it still mangles one huge source of data on Obamacare and the uninsured, making it more difficult to assess the law’s effects on health-insurance rates. (Other large surveys, collected by the government and the private sector, do exist.) Because these surveys lag by a year, we’ll get 2013 numbers in the fall of 2014, and won’t be able to compare them with this year’s numbers until the fall of 2015, by which time whatever outrage is left will likely be much diminished.
But Census officials also said the unfortunate timing was “coincidental.” Arguably, measuring the net increase in health-insurance coverage is the only meaningful metric for Obamacare. Changing the way Census calculates this metric now, when we’re in the middle of implementing massive changes to health-insurance markets, seems like a good way to skew the numbers in the administration’s favor — especially if the new survey methodology is likely to yield a lower uninsured rate. The Times reports that the Obama administration was consulted on the new questionnaire and that it approved the changes — if the White House wanted the Census to produce data that could be used to assess Obamacare, presumably it could have asked that the Census Bureau hold off on the changes.