Remember, People Lie to Pollsters

Many years ago when I worked at the Cato Institute (more on that later) I managed to so offend friend and foreign policy guy Justin Logan that he backed slowly out of the lunchroom. You may wonder, what bigmouthed thing had I said? That America was right to invade Iraq? That torture is A-OK? That I was thinking of getting a Woodrow Wilson tat?

None of the above. We had been discussing a poll finding that something like a majority of Republicans believed large caches of Weapons of Mass Destruction had been found in Iraq, contrary to the evidence. I argued many of those Republicans did not, in fact, believe that WMDs had been discovered in Iraq.

Rather, I explained, Republicans knew how those polls were going to used polemically and so answered a different set of questions. What they heard was more like “Do you support President Bush and/or the boots already on the ground in Iraq?” So of course they had answered yes.

To his credit, Logan asked a few questions before quitting the field. I explained my belief that people lie to pollsters all the time for a whole number of reasons. These reasons range from social disapproval (“Do I really want to tell the complete stranger on the other end of the line that I’ll vote for Jesse Helms?”) to confusion (the wording of the question can be all-important) to a cynical sophistication (“If I answer x, it will be used to argue y. So, let’s go with -x.”).

At the time, Logan really didn’t know what to make of such high-proof skepticism. I wonder if that’s changed.