“Are Republicans healthier than Democrats?”

6 responses to ““Are Republicans healthier than Democrats?””

1. 
   Brad

   November 4, 2008 at 5:25 am | Permalink

   I must be missing something.

   Time and time again unhealthy states are listed are the deep south states (Louisiana, Arkansas, etc) which vote overwhelmingly Republican. The studies usually show the northern states as being the most healthy, which are usually Democratic voting.

   I don’t understand how the usual studies relate to this one.

2. 
   Steve Kass » Are Republicans healthier than Democrats?

   November 5, 2008 at 8:10 am | Permalink

   [...] (which contained several other wrong numbers caused by the same mistake) was reported here, here, and here, except that it was reported as surprising research, not wrong research. These reports [...]

3. 
   Steve Kass

   November 5, 2008 at 8:30 am | Permalink

   The explanation is that the study results are just plain wrong. The researchers got their crosstabs backwards and reported the wrong percentages. I’ve explained it, with screen shots of the GSS results they used, here (http://stevekass.com/2008/11/04/are-republicans-healthier-than-democrats/).

4. 
   Lauredhel

   November 6, 2008 at 4:39 pm | Permalink

   Thanks, Steve. Though that changes the numbers, there still seems (at a glance, without statistical analysis) to be at least some correlation between health and political leanings, though the numbers are somewhat fuzzy in the middle.

5. 
   Steve Kass

   November 8, 2008 at 3:01 pm | Permalink

   True, there’s still something going on. However, my guess (also at a glance, without statistical analysis) is that most of the relationship can be explained by a known correlate of party identification, say economic status.

   If it turns out that party identification does help explain respondents’ self-reported health beyond what other factors explain, I hope the authors will let us know.

   While strong Democrats reported poor health more frequently than strong Republicans in the General Social Survey, for respondents at the same economic status, things could even be the other way around. Strong Democrats might have reported poor health less often than strong Republicans (of the same economic status).

6. 
   Lauredhel

   November 10, 2008 at 1:13 pm | Permalink

   Another failure of correlation/causation assessment is found in today’s WA Today. In “Low wage earners happy to have jobs”:

   The results, based on an analysis of more than 7000 people surveyed in the Household, Income and Labour Dynamics in Australia series, show that unemployment is extremely bad for well-being. Not only are the unemployed less happy, their opinion of their health is much lower than those with jobs.

   Later in the article they quote a chap saying that the causation may run the other way, but I guess the author didn’t believe him.