The Problem with Welfare Economics

Uwe Reinhardt repeats a familiar refrain from critics of welfare economics, that compensation tests do not erase the value judgments implicit in ignoring the distributional effects of policy. This goes way, way back, and Reinhardt might have been more forthcoming about the attempts by economists to bring distribution back in via refinements to benefit-cost analysis and other techniques. His point remains pertinent, but it misses the real problem with welfare economics.

Welfare econ rests on utility (or “preference”) theory, the idea that a person’s well-being has a stable and predictable relationship to the consumption choices they make. Reinhardt worries that economists are simply adding up utility gains and losses without taking into account who’s winning or losing, but the more fundamental issue is whether utility (or “preference satisfaction”) has any validity to begin with.

There are two deep problems with welfare economics. The first is that it actually assumes that well-being is identical to consumption choices. The absurdity of this proposition was demonstrated decades ago in several pungent articles by Amartya Sen, and no one, to my knowledge, has successfully rebutted him. The empirical failure of this assumption has more recently been exposed by “happiness studies”. There are raging disputes between happyologists (on the Easterlin paradox, for instance), but there is no doubt any more that, on an individual level, a chasm has opened up between “preference satisfaction” via consumption and empirical measures of well-being.

The second problem is that welfare economics depends on the assumption that choices are rational, that utility is actually maximized by each choice made by each individual in each situation. If you’ve paid any attention to behavioral economics, you’ll know that one went out the window some time ago.

So the problem is not just that economics fudges the distribution of utility, but that utility itself has become a sort of phlogiston, a make-believe substance that once served to prop up a theory, but got put to death by the evidence and is now no more than a curiosity studied by antiquarians and professional Scrabble players.