From now on, I’m going to assign Mike O’Hare’s essay every time I teach policy analysis. If you haven’t read it, I suggest that you stop and do so now; what follows is just a pair of footnotes to it.
Briefly, then:
1. It’s not necessary to invoke “compassion” to justify redistributive measures. Diminishing marginal utility means that great inequality is almost always inefficient. Insurance markets have their own characteristic failures, and it will often be the case that the market can’t sustain some risk-spreading arrangement that would be mutually beneficial. That’s what the idea of “social insurance” is about. (See Richard Zeckauser, “Risk Spreading and Distribution,” in Redistribution Through Public Choice, Harold M. Hochman and George E. Peterson (eds.), New York: Columbia University Press, 1974, 206-28.)
2. The principle of benefit-cost analysis is pretty bullet-proof: it merely insists that public choices be efficient, substituting for the market as a mechanism for ensuring that resources aren’t wasted. The practice of benefit-cost analysis is more problematic; since hard-to-measure and hard-to-value benefits and costs are often omitted, thus effectively valuing them at zero, it’s likely to be somewhat biased against intervention.
But the supposed moral outrage against valuing lives (properly, valuing the prevention of statistical deaths) reflects mere confusion. As Schelling pointed out a long time ago (see “The Life You Save May Be Your Own” in his Choice and Consequence), in making private decisions we don’t put an infinitely negative value on every risk of death, so why should we do so publicly? In practical terms, using the right value - $7M-$10M under current U.S. conditions - will identify lots of opportunities to save lives.
Now that the Obama Administration has decided to reverse the Bush Administration trend of de-valuing statistical deaths, the business community has suddenly decided that benefit-cost analysis - which it loved when the result was anti-regulatory - is far too “subjective” for daily use, and wants the decisions returned to its tame politicians.
If I said that the main difference between liberals and conservatives (as those terms are used in contemporary political discourse in the U.S.) is that liberals value human life more than conservatives do - especially, but not only, if the human lives in question belong to poor people - that would be rude. So I won’t say it.
Footnote Naturally, the “Right-to-Life” folks, including the Catholic Bishops, have nothing to say on the topic. It seems that “life” is infinitely valuable only when ending it might relieve suffering, but not if saving it would cost money. Or, as Barney Frank said, they believe in a right to life that begins at conception and ends at birth.
“The principle of benefit-cost analysis is pretty bullet-proof . . .”
I’ve done a few of these, and I have never quite known what to do, when the kernal of the policy — as is often the case — is a strategic gambit, to change costs and induce behavioral change. Or, to induce behavioral change, which will change costs.
Cost-benefit analysis, as the half-brother of the economic “trade-off”, sometimes suffers from the same genetic defects inherited from the faulty economic theory of production and allocative efficiency, that is, an inability to correctly value, or account for, technological improvements that result in efficiency improvements from gains in total factor productivity and the reduction of risk.
With regard to safety, for example, what we want is a long-run improvement in technical efficiency, not a refined allocative efficiency. We want the “production process” of transportation to throw off fewer errors, so we want better control. In regard to air traffic safety, for example, over the very long-run, a program for reducing errors in the system, has achieved a remarkably high level of safety. Cost-benefit analysis, I’m sure, has played a part along the way, but I’m not sure that one could make sense of trying to compare the existing system with a counter-factual system, which, by dint of a different allocation of resources, would be “cheaper” but somewhat more wasteful of lives.
There’s been a real strategic value in letting an absolute goal of safety drive evaluative analysis of accidents and the long-run enginneering development of the system.
“Government” and “regulation”, if we trace the root of these words, literally, means control.
The economic analysis of allocative efficiency, which gives rise to the idea of a trade-off, suggests to our intuition, that there is a trade-off available in stasis, a choice between more control and less control. I think that intuition is faulty and misleading.
In the framework of economic trade-offs, control is a technical matter. Improvements in technical control are not movements along the frontier of production possibilities (the definition of “trade-off”), but, rather, movements along a vector orthogonal to the frontier.
The economic theory of production assumes away the technical problems of management, engineering and control — assumes that output is “maximized” in a technical sense — in order to focus on the problem of allocative efficiency. That serves a purpose in abstraction, but misleads us back here in messy reality, where the technical problems of organizing are not anymore “solved” or at a maximum, than are the problems of resolving the allocation of resources.
Libertarian political philosophies, as rationalizations for conservative political desiderata, exploit a number of these basic misunderstandings, beginning with the false idea that, at any particular stage of economic development, there is a choice between “more rules” and “fewer rules”, that there is some feasible, counterfactual political regime, where there would simply be less authority and fewer rules, but this wouldn’t entail regression to a more primitive economic state. Allied with this false notion, that disconnects control from technical progress, is the equally false idea that power is a zero-sum game, in which reducing power at the center is necessary and sufficient to increasing power at the periphery.
It may be that conservatives, generally, as the “selfish” defenders of vested interest and exploiters of externalities, are careless about the interests and lives of others, particularly mere abstractions, standing in for putative others. But, I’d be careful about pushing too far on the ol’ self-righteous liberal idealism. Concrete self-interest and pragmatism have a legitimate place against the abstract and ideal, in political competition and contention.
The libertarian confusion, on the other hand, is a genuine social pathology, that threatens our ability to arrive at practical progressive compromises through democratic means. It is not an accident that libertarian arguments seem to derail political arguments, with “constitutional” questions; there’s a basic misunderstanding of power and rules at the core of the libertarian conceit.
Bruce’s comment elegantly ties two very different things together. First, his (generally true) insight: in the medium-to-long run, technical efficiency beats allocative efficiency. The wealth of nations is in their engineers, not their economists. I’ve always believed this to be so, but I had never tied this into regulatory goals before. Hence the value of his second insight: a regulatory system that focuses on improving technique is more likely to be more long-term effective than one that improves allocation of existing resources. This argues for a realistic regulatory absolutism, rather than tradeoffs. NEAT!
I’m just repeating what Bruce said (I think), but imitation, after all, is the sincerest form of flatery. Nice comment, Bruce.
The wealth of nations is in their engineers
Not to derail, but this is the clear subtext of “Sky Captain and the World of Tomorrow”, in which our ostensible hero puts himself and others into grave danger in order to save the engineer, a national treasure.
Diminishing marginal utility is on pretty sound grounds, but it doesn’t actually drive any policy decisions without further premises which are contested. I have a nice turkey dinner, my neighbor is having instant ramen for thanksgiving, I have no doubt whatsoever that he’d enjoy a plate from my dinner more than I’d miss the second wing.
But absolutely nothing forces me to refrain from saying, “So what?”. And substantial competing values would drive me to say, “Hell, no!” if you proposed forcing me to send a plate over.
In short, while it’s not necessary to invoke “compassion” to justify redistributive measures, neither is it sufficient to invoke marginal utility. I think this is something liberals tend to miss because the shared values, and relative absence of competing values, which makes you liberals in the first place, is the water in which you swim. It doesn’t tend to occur to you that the relatively high value you place on utilitarian considerations, and low value you place on such considerations as property rights, are far from a universal combination, and in no way logically demanded.