Friday, April 30, 2010

I Believe, I Trust, I Promise

Over at the Monkey Cage, John Sides has another of a series of excellent posts about trust in government.  He shows, once again, that pretty much everything that people say about trust in government is wrong.  If you're tempted to think that survey questions about trust in government yield important answers, he has the evidence that should convince you. Or, you can just take my word for it:

"Trust in Government" survey questions don't work.  They don't measure what we want them to measure...some sort of general sense of whether people think the government works, or something like that.  Instead, the questions tap into people's feelings about the economy and the president.  As far as I know (and I haven't read the literature on this for some time, but everything Sides says is consistent with what I knew about it from back then), the answers to "trust in government" questions don't actually tell us anything at all, not once we know about how the economy is doing and presidential approval ratings (which are partially, but not entirely, caused by the economy).  It may be that people don't actually distinguish between the president and the government in general; it may be something else; but really, who cares?  All non-specialists need to know is that the questions don't tell us anything.  My suggestion is that any time anyone relies on survey results about trust in government, that you just run the other way; you're not going to learn anything from whatever it is they are trying to say.


  1. I think that's a bit much. Sure, responses to the question "do you trust the government" might track very closely with the electorate's views about the economy and the president, and there is almost certainly a causal relationship but I don't think it follows that they "don't measure what we want them to measure."

    Is there any reason to think that, even though this inchoate sense of "whether the government works" is a dependent variable, it's not a genuine attitude that can be measured by telephone polls? We probably aren't able to infer many useful political maxims from the responses to this poll question, but I have no hesitation in saying that people who claim to distrust the government genuinely don't, from the IRS to the Social Security Administration. It's useful to have a number to measure and digest this dyspeptic, antigovernment sentiment.

    By analogy, if we already know the temperature and the relative humidity, we don't need to listen when the weatherman tells us the dewpoint -- but it sure is handy when you want to know how hot it feels.

  2. And actually, now that I come to think of it, it's not at all clear that a lot of the causation doesn't run in the opposite direction. The common story is that the falling approval ratings of Johnson and Nixon in the 60's and 70's were to a great degree spurred by the increasing public awareness that the government was untrustworthy. There were undoubtedly other factors (especially in the case of Nixon, who presided over oil shocks), but I haven't seen much to rebut the theory that "trust in government" is (at least in this case) what was driving presidential approval ratings -- not the other way around.


Note: Only a member of this blog may post a comment.

Who links to my website?