Discussion: (11 comments)
Comments are closed.
The public policy blog of the American Enterprise Institute
View related content: Economics
One of the most important goals of Obamacare is to increase the number of people with health insurance coverage. Has it worked?
As we all know, the White House reported that over 7 million Americans signed up for private health plans on Obamacare’s newly created and IT-troubled exchanges, and Medicaid enrollment is up by 3 million since October. So can the White House declare victory and exit the battlefield?
No. Why? Because we don’t know how many of those 7 million Americans are newly insured. If, say, 5 million were on private health plans before Obamacare took effect, had their plans canceled, and subsequently signed up for health insurance through the Obamacare exchanges, then the effect of the law will be to increase health insurance coverage by only 2 million. That ain’t a lot.
So how will the government know whether Obamacare has worked? It’s simple: The plan was to ask people through statistical surveys whether they had coverage in 2013, 2014, 2015, etc., and compare that to the number in 2012, 2011, 2010, etc. Using multiple years allows you to study the trend over time in health insurance coverage, which really is more important than any one year’s estimate.
This plan was complicated this week because the Census Bureau will change the questions they ask about health insurance in one of their flagship surveys, the Current Population Survey (CPS). The changes to the CPS will make it very difficult to compare Obamacare-affected years with the long series of previous years.
Some on the right have suggested that this change is evidence of a conspiracy — of the White House trying to rig the statistics in such a way that makes Obamacare look more effective than it actually is.
I would be shocked if any such thing was taking place. I see no evidence of a conspiracy here.
1. For several reasons, the current insurance questions in the CPS are badly designed and need to be improved. Efforts to improve the questions have been underway for years — years before anyone in Washington ever heard the word Obamacare.
2. The new survey questions will be used to gather data about 2013. The first year of coverage through the exchanges and Medicaid expansion will occur this year, in 2014. So we can still compare 2013 and 2014. Ideally, we could compare a longer time series. But if this were a conspiracy then you would think the conspirators would have obscured the ability of researchers to make any before-and-after change.
3. As any potential conspirators would surely know, there are several other sources of data that one could use to determine Obamacare’s effect on health insurance coverage. The economist Justin Wolfers suggests Gallup is particularly good — and is free of government influence.
4. Finally, changing a question on the CPS requires a lot of bureaucratic maneuvering. It seems unlikely that the White House could have pulled this off even if it wanted to.
None of this is to say that the Census Bureau’s decision to change to a new methodology in the midst of Obamacare’s rollout isn’t a dumb decision. It is. And Census should have known better. Either the new methodology should be delayed, or the old and the new methodologies should be employed concurrently to ensure comparability across years.
But it’s a long walk from bad decision to conspiracy.
The process of crafting sound public policy requires excellent data. Economists, analysts, and government officials need the best information they can get, and that means the highest quality government statistics. It is crucial that the public trusts the federal statistical agencies, if for no other reason than to ensure the quality of the data being collected.
Census knows this as well as any organization in the country. They should have employed that knowledge, known that Caesar’s wife must be above suspicion, and not changed the health insurance questions on the CPS in what is the most politically volatile year to do so in decades. It may happen that even a year from now suspicion will still cloud the numbers. If so, then everyone loses.
So this decision is evidence of very poor judgment. But until there’s any evidence of a conspiracy, we should not make accusations of conspiracy.
— Michael R. Strain is a Resident Scholar in Economic Policy Studies at the American Enterprise Institute. Previously, he worked for the Census Bureau. You can write to him on Twitter at @MichaelRStrain.
Comments are closed.
1150 17th Street, N.W. Washington, D.C. 20036
© 2015 American Enterprise Institute for Public Policy Research