« More on pre-accepted academic articles | Main | New at CJR: Obama "evolves," Romney "flip-flops" »

April 30, 2012


At the risk of repeating myself, let me note that when social science research that is based on techniques whose validity is at least open to doubt (e.g., questionnaires that may suffer from framing or ambiguous language or responses that reflect underlying attitudes toward the subject rather than real answers to the question), based on samples that are uncharacteristic of the general population, and subject to acknowledged caveats and limitations, begins to be cited as establishing fundamental truths about the general population, that is just another form of truthiness.

Rob, like all my research, my findings are of course necessarily provisional. With that said, I hope you'll look at the body of evidence presented in the New America report that I co-authored - I'm not the only scholar to reach these conclusions: http://mediapolicy.newamerica.net/sites/newamerica.net/files/policydocs/Misinformation_and_Fact-checking.pdf

Brendan, you've generally been pretty conscientious about representing your research fairly (for example, in the New America Foundation piece, you state that your 2005 and 2006 research was limited to undergraduates), and my comments weren't intended to be directed specifically to you.

Berinsky's research, cited in the New America Foundation article, may be a useful example here. First, his findings are based on an Internet survey. That's not necessarily bad, but it is controversial. The validity of the results to the general population is subject to at least some doubt. Second, the extent to which people's responses are less about their true beliefs in the (usually) very carefully framed questions than about their general attitudes toward illegal immigrants or John Kerry or the Bush Administration or Obamacare is, to my mind, open to considerable question. It takes a respondent capable of parsing these questions carefully and highly motivated to answer only the specific questions asked to get good responses; otherwise a lot of what you're likely to get is attitudinal.

That's not to say that these research efforts aren't useful. Given their limitations, however, perhaps the most accurate way of characterizing them is that they are suggestive of certain propositions, and even that conclusion is, to use your term, provisional. (Regrettably, in the absence of replicating studies, many such research findings are forever provisional.)

Yet it's rare for research results to be characterized this way. Instead, the limitations and caveats are forgotten, and the provisional results become articles of faith. In other words, the "truthy" version of the research is halfway around the world while the caveats are pulling their boots on.

As I pointed out a few years ago, Brendan's claim that federal income taxes collected went down after the Bush Tax cuts is a half-truth, at best. There were two sets of Bush tax cuts -- 2001 and 2003. I assume that they were effective in 2002 and 2004, respectively.

As you can see, after the 2001 tax cuts, federal income tax receipts did decline in 2002 and 2003. However, FIT then rose in 2003 and 2004 and rose sharply in 2005 and 2006. For each of the years 2004 - 2007, FIT collected was far above any prior year.

So, the truth of Brendan's claim depends on the meaning of "after." IMHO the appropriate definition is "several years." Here's why. This bears on the question of whether or not the Bush tax rate cuts caused an increase in taxes collected. Those who claim that lower tax rates caused higher tax dollars collected didn't expect an immediate impact. They predicted that the lower tax rates would cause the economy to grow so much that FIT collected would eventually be higher than before. Actual results were consistent with this prediction (although IMHO that's not enough to prove causality.)

Regarding the Bush 2003 tax cuts, Brendan's claim is even farther from the truth. FIT didn't even go down in 2004, the first year the new rates were effective. And, as I mentioned, FIT kept going up through 2007.

It was ironic to hear Brendan and his host discuss the mistreatment of facts while they themselves were helping to perpetuate a myth.

The comments to this entry are closed.