Writing, publishing, geekdom, and errata.

Healthcare Reform and Crappy Research Design

No comments
soc_econ.pngIn the ongoing frothiness about healthcare, I've got something important to point out. It's pretty easy to find reports showing what Americans think about their current insurance and healthcare. But I don't think those reports are accurate - and it's not because of the surveys themselves.

I ran across this article (from ebri.org's Notes in September 2009 1) (PDF). While the actual report is relatively straightforward, the introduction seems to be slanted against any kind of reform (slightly edited for space and formatting):
"...most Americans are satisfied with their health coverage, and have been historically. But the question of whether Americans will be able to keep what they have has been raised by a number of analysts, and while Americans are satisfied with the coverage they have now, they are concerned about the future.

And this gem:
While these sharp differences in attitudes are not surprising, shoring up the system for the uninsured and individuals with chronic conditions means changing the health care system for everyone—notably for those with insurance coverage and in good health. Ultimately, will the needs of the few outweigh the satisfaction of the many?

They go on to note that not only has the satisfaction with current plans stayed level for a decade, but actually started to go up in 2009. (Click charts to embiggen.)

And they're right. Among their sample in The EBRI/MGA Health Confidence Survey, that's absolutely correct. But why did it go up in 2009 - and if it's going up (they imply), why should we get reform?

I think there's a spurious variable in there. The Health Confidence Survey was conducted between 8 May and 2 June 2009. Presumably, the other administrations of the survey in prior years happened at the same time. Between 2 June 2008 and 2 June 2009, President Obama was elected after making the reform of healthcare a major issue. It's a few months after he made it a key point of a joint address - and two months before wingnut mobs started harassing elected officials about HCR.

I think what we see in that chart above (statistical correlation here, if you're interested) is the impact of the right-wing propaganda machine between November 2008 and May 2009.

I notice the same thing with winter weather - since I stopped watching TV news, my anxiety over "snowpocalpyses" has declined dramatically. In my region, these supposed "-pacalypses" 2 would've been "winter storms" five years ago, now they are causing runs on grocery stores.

As researchers and citizens, it's important to cut through the spurious data and try to get to the truth of things.

Or maybe it's just that the truth of things is that many people are easily manipulated.

1 I had trouble finding consistent trend data quickly; I didn't spend much time on it. If anyone else has better data, that'd be awesome.
2 I use the term still, but ironically.

No comments :