What Does the Public "Really Want"?
Can we, as social scientists, know what the public "really wants" in terms of public policy? Can public opinion surveys enlighten this line of research? I answer these questions respectively as "uhhh, maybe" and "yes, to a limited extent." I will elaborate briefly. A quick review of the blog literature to catch everyone up. In April 2007, I presented Eva Mueller's 1963 article as a challenge to Bryan Caplan's challenge. In February-March 2008, Dr. Caplan responded with kind praise (making me blush, seriously) and an on target critique.
I understand Caplan's response as consisting of two parts. Part one: even when framed in terms of paying more taxes, the American public still supports bigger government (or at least the status quo). Part two: question variations involve information effects, which can have significant but ambiguous (from a policy interpretation standpoint) results. I see the two parts as intimately linked, and I still maintain that support for libertarian policy changes exists.
Dr. Caplan directs us to the GSS. Yes, the question Caplan refers us to (variable TAXSPEND, for those that want to check) indicates that the public prefers bigger government, but I don't think the implications are what he infers. First, this is not "overall government spending" as Caplan indicates, but social spending (which is perhaps less than half of spending, depending on who you ask). More importantly, the question is very unconstrained. It asks about three categories of spending, in an open-ended way, as well as three (plus) categories of taxes. (I use "open-ended" to refer to interpretation of the question, rather than "fill in the blank" as many surveys do.)
One could just as easily point to the GSS result (variable CUTGOVT) that the median position is "in favor" of cutting government spending as something "the government might do for the economy" (the result is very strong, almost 9:1, if we group favorable vs. non-favorable responses). There appears to be a social intransitivity, with the public favoring both more spending and fewer taxes. But I would direct your attention to the unconstrained nature of Caplan's favored GSS question, especially on the taxation side. As the Mueller article indicates, merely asking individuals to reevaluate their response "if the government had to raise taxes to finance the additional expenditures" achieves a noticeable decline in expressed preference for policies.
To me, this is a huge issue. Can the respondent simply "make up" the method of finance or is she constrained, to some degree, by the question? Dr. Caplan doubts that "the public is that stupid" to not realize that more spending requires more taxes. I agree. But taxes can take many, many forms. I would suspect decreasing support (this is empirically testable) among survey respondents over the following three taxation options for increased government spending: 1. unspecificied (possibly "free" or taxation of the others); 2. taxes (in general) increase; and, 3. your taxes increase. Mueller tests option 1 vs. 2, but alludes to the importance of option 3 (p. 216 fn. 6). Furthermore, I would suspect decreasing support on option 3 for increasing dollar amounts of taxation (probably the same for option 2).
Which option is the accurate one, in terms of what the public (or more simply, an individual) "really" prefers? Is it number 1, 2, 3, or one of the various subsets (2a, 2b, 3a, 3b, ...)? I honestly do not know, but in my very junior estimation, no one knows the correct answer (my hunch is that the most constrained and accurate wording, in terms of actual tax policy, may be best). And how the question is framed can significantly affect the median voter result. Many, many others factors must be considered (preference falsification comes immediately to mind, in a major way), but the question is far from being resolved. I eagerly anticipate the future debate, as to my young mind this (the post title) is the central question of public choice and political science.
(Caplan's suggestion of a dissertation on this topic is absolutely awesome, but unfortunately I will not pursue it. I have now locked-in to my dissertation topic: an economic analysis of the origins and consequences of voting rights expansions, with an historical application to the early 19th century U.S. [much, much more to come, I promise!]. However, this is a great topic which I intend to further research and eventually write about. Someone should definitely claim as their own for a dissertation).
