On Jun 8, 2010, at 8:32 AM, Joseph Catron wrote:
> "Several academic studies have found that there is a link between education
> levels and civic behavior. But a new study from the Federal Reserve Bank of
> New York <http://www.newyorkfed.org/research/staff_reports/sr450.pdf> has
> concluded that how much economics people study can influence their political
> activity and how they spend their spare time."
> http://economix.blogs.nytimes.com/2010/06/07/does-studying-economics-make-you-more-republican
It's nice to see several of the comments call them out on it, but the shamelessness of the methodological gloss in this study is actually breathtaking. The title is:
"Is Economics Coursework, or Majoring in Economics, *Associated with* Different Civic Behaviors?" (emphasis added; the point is association)
but the first line of the abstract is:
"Studies regularly link levels of educational attainment to civic behavior and attitudes, but only a few investigate *the role played* by specific coursework." (emphasis added; causation tacitly invoked)
As the very first commenter points out, the authors drop a CYA in footnote 11: "To the extent that our approach for controlling for ability bias may not completely address the potential endogeneity of our measures of economics training, caution should be used in interpreting the estimated relationships as being necessarily causal." But the whole gist of the paper runs the other direction - as is clear even simply from their use of the word "necessarily" in the note.
PS: Didn't Jon Stewart do a piece not so long ago on how Fox News was using the question form of headline ("Does X make you Y?") to mask blatantly sloppy / biased imputations?