I will try my best to engage in close reasoning in response to Anthony de Jasay’s reply to my points. Let me then be very careful, even if this means being a somewhat pedantic. From his previous comments I thought that de Jasay did not wish to engage in a methodological dispute. Obviously I was wrong.
Anthony de Jasay says:
Professor Gaus’s contributions of February 21 and 22 to the debate do not seek to answer my charge that his explanation of political outcomes via a holdall “utility function” (whose principal variables are moral sentiments) is both unverifiable and unfalsifiable.
As we all know a utility function is simply a formal representation of consistent choice over outcomes, with consistency understood in terms of completeness, transitivity, etc., and, if the function is cardinal, let us say also the standard von Neumann-Morgenstern axioms (I warned that this might get pedantic). The axioms that must be met by a well-formed utility function provide formal, not substantive, constraints, on choice. Some have argued that it will always be unverifiable whether a person demonstrates consistent choice because, say, it could never be shown that a person violates the transitivity axiom because preference change might always explain choices that appear intransitive. If that is what de Jasay has in mind, his is a subtle point. More generally, a similar deep question confronts any rational actor analysis: are we stipulating rationality or showing it? This is a long-standing debate in the philosophy of social science. I must say that I do not think that the hypothesis of consistent choice is really immune to evidence and so is unverifiable: there is evidence that can be brought to bear on the question of whether a person has changed preferences or violated transitivity. (Do the preferences tend to “change” back again? If so, the hypothesis of preference change rather than intransitive choice looks less plausible.)
My impression, though, is that de Jasay’s concern was not with this point. What he didn’t seem to like was my claim that, in order to accurately model people’s choices, we often have to suppose that their preferences are affected not only by their interests but by certain notions of fairness. I believe that it is what he thinks is unverifiable and unfalsifiable. Well, there is now a large body of experimental literature, the upshot of which is that in many contexts consistent choice over options is best explained and predicted by the hypothesis that people’s preferences are influenced by norms of fairness and other moral sentiments. I cited some of this research in my reaction essay, since I agree with de Jasay that evidence, not an a priori claim, is needed here.
As far as I can see, de Jasay is dismissive of this body of research, but that he is dismissive of certain empirical evidence does not, in my view, show that any of these claims about people making choices on the basis of moral norms and sentiments are “unverifiable and unfalsifiable.” Perhaps de Jasay is calling attention to the leap from experimental findings to general claims about politics. Fair enough: we need to do more research. But why should these claims be unverifiable in politics when they have been so well tested in experimental settings? And why are they less verifiable than the claim that people act only on their interests? (Which, indeed, I believe has been shown to be simply false.) Experimentalists have been able to formulate relevant hypotheses. That is the crucial point. A claim that can be, and has been, tested, simply cannot be called “unverifiable and unfalsifiable.” Again, the only thing de Jasay’s assertion could plausibly refer to is the general claim, considered in the previous paragraph, that the hypothesis that people are consistent or rational choosers cannot be tested. But, as I said, this is a general problem of all rational actor explanations, including those based on interests.
Next, de Jasay says:
For Gaus,”…the welfare state is not based on a power grab based on the pursuit of interests but on the majority’s sense of fairness.” Fairness is desired for others no less than for oneself. It is the chief variable in the “utility function”. [I actually never made such a strong claim: I held that “Our sense of fairness is as much part of our utility function as is our aim for wealth or other goods” — a rather more modest claim.]
de Jasay comments:
If this were true of everyone, better-off and worse-off alike, any collective choice rule would produce the “fair” result. Minority rule would generate much the same welfare state as majority rule. (The reason is that conflict over distribution would not exist). Why, then, does Gaus need majority rule? The sole necessary reason I can see is that while a majority can be found “for” fairness, a minority will be against it. There will be distributional conflict. Government will buy “fairness” for some or all in the majority with the money of some or all in the minority.Since rich-to-poor redistribution can always trump (has a greater potential “yield”) than poor-to-rich redistribution, a bid offering “fairness” to the worse-off will beat a rival bid offering “unfairness” to the better-off. Hence the worse-off and the middle will form the majority and the better-off the minority.This stylized result is obtained by using the fairness motive as represented by Gaus. Precisely the same result has been obtained long ago by using the interest motive. It is puzzling to find Gaus arguing passionately that one is right and the other wrong.
Now, de Jasay seems to argue:
(1) Gaus “needs” majority rule because people disagree about fairness (they have “distributional conflict”);
(2) If people disagree about fairness this must mean that the majority’s notion of fairness rules;
(3) But (2) means rich-to-poor (and middle class) redistribution occurs.
(4) This is because “a bid offering ‘fairness’ to the worse-off will beat a rival bid offering ‘unfairness’ to the better-off.”
(5) But this same result has “long been obtained” by appeal to the interest motive, so why think that fairness is right and interest is wrong, and why should Gaus be so passionate (leaving aside, of course, my general passion for life)?
There is much that is perplexing here, at least for one who, like me, has such trouble with close reasoning. Leaving aside what is meant by the reference to my “needs” in (1), claim (2) looks plausible enough. What about (3)? The idea that fairness (for example, claims of desert) never benefits the better-off seems extreme. It is certainly interesting to ponder why, if politics is simply about the pursuit of interest, there is not a lot more redistribution. Think of John Stuart Mill’s worries about the poor plundering the rich. To be sure, there are rival explanations, but one is that the poor and middle class think that much of the income of the rich is deserved. Again, there is relevant evidence: empirical research indicates that the typical citizen does care a good deal about desert in the distribution of incomes. One of the things that drives some egalitarians to distraction is that so many low-income people think high-incomes earners deserve their high incomes. There is also strong experimental support for the claim that, while people do think that fairness requires an income floor, they reject an income ceiling. So disagreements about fairness cannot be simply translated into self-interested disagreements about distributive shares.
Let us accept, though, that there is indeed redistribution away from the rich, and let us accept that both a fairness and an interest account would predict this. So let us focus directly on (5), the conclusion. That the same consequence is explained by two different hypotheses does not show that there is nothing to choose between the two hypotheses, or that one should not be passionate that one is right and the other wrong. What could be more common? That, after all, is precisely what scientific dispute is about: debate about the correct explanation of the same event. This is not to say that the theories are extensionally equivalent, for in other contexts they may make different predictions. And, importantly for what the dispute between us was originally all about, they point to different diagnoses of why the welfare state has been able to so easily expand beyond the limits of classical liberal constitutional constraints. If the interest explanation is correct, one response is appropriate; if the fairness response is correct, that response may not succeed, since the aim is not to restrain the simple pursuit of interest but the expression of moral sentiments. I tried to make clear in my reaction essay that available evidence indicates that interests and moral sentiments respond differently to authoritative rule makers. Lastly, de Jasay’s claim (4) simply begs the question: (4) assumes that the state is taking “bids” and so the core of the matter is interest, but the argument was seeking to show that. (At best, it is a tendentious way of describing majority rule).
One of the things that has struck me during these exchanges was the hostility of some of the other contributors to the idea that a plausible explanation of the welfare state is that it appeals to the moral sense of the large majority, or the claim that this is morally relevant since democracy is a good thing. I do not believe that either of these claims are idiosyncratic or odd. The empirical claims I have made, I think, are reasonably well supported by the evidence; in another post I was so rash as to defend democracy, and the idea that it has been shown to be a reasonably effective way to prevent repression. What is quite astonishing to me is that my fellow classical liberals find all of this so outrageous. One can only wonder how an important part of the liberal tradition has gone so far astray, doubting that people are moved in political life by morality as well as interest, or that democratic rule has any legitimacy. I disagree with the moral judgments of the majority on most matters, but I do not have the disparaging attitude that sometimes seems to characterize today’s classical liberals.