Tuesday, August 23, 2011

On "Efficiency"

(Neo-classical) economists love to talk about efficiency. Markets, we're told, are nothing if not efficient. Businesses that succeed, supposedly do so because they're efficient. Talk to an economist, and you'll hear a lot about "efficient allocations" of resources, and "maximizing efficiency", and so on. What does all this talk mean? Why is it billed as such an important criterion for evaluating social institutions and public policy?

It's worth noting that economists themselves haven't always taken "efficiency" to be the most important goal of policy. As Andrew Levine notes: "For economists who came of age during the Great Depression and their students, unemployment was the major theoretical and practical concern. Nowadays, efficiency has taken its place." Levine is correct. It hardly needs to be said that full employment has long since ceased to be even a rhetorical goal of public policy. But if the old demand for full employment was concrete and easy to endorse, the new lingo of "efficiency" is abstract and unconnected with the needs and interests of working people. Again, as Levine puts it, "it is hard, at least for strangers to the economics profession, to understand why economists today focus so doggedly on a structural property of economic systems that represents real world concerns in only the most indirect and attenuated fashion."

So what is that economists mean by "efficiency"? For all their talk about being a rigorous science, conventional economists are actually quite unclear about what their favored criterion means. As Levine points out, "sometimes it is used loosely as a synonym for productivity...sometimes it invokes a cluster of non-technical ideas- delivering the goods, avoiding waste (of time, effort and resources), running smoothly, and so on." But in more technical contexts, it means "Pareto optimum". In the reigning jargon, an allocation of "resources" is "efficient" if it is "Pareto optimum", that is, "if any change would make someone worse off (holding technologies, preferences and the (initial) distribution of resources constant)." Or, put another way, a "Pareto optimum" or "Pareto efficient" allocation is one in which no individual can be made better off by redistributing resources without making another worse off.

Now, I want to draw your attention to two things at this point. First off, take note of the fact that the ordinary, non-technical uses of "efficiency" do not necessarily coincide with the technical definition embodied in the idea of "Pareto efficiency". That is, there is a substantial gap between our ordinary understanding of the word "efficiency" and the way that the term is used in technical contexts in neo-classical economics.

Second, note that "Pareto efficiency" isn't necessarily always a good thing. As Levine points out, "the most vile situations can be Pareto-optimal if a change would make someone worse off". That is, Pareto-efficiency is compatible with any degree of equality or inequality. Thus, an extremely unequal distribution could be Pareto-efficient just in case it is not possible to make one person better off without making another person worse off. Pareto efficiency also fulfills a subtle ideological role in justifying capitalism, often being invoked as a theoretical tool for justifying the dogma of "trickle down". It can be used to argue against re-distributive taxation on grounds that huge increases in the wealth of the rich (and thus huge increases in overall economic inequality) either increase or have no effect on the welfare of the worst off.

It's also worth pointing out that the way mainstream economists interpret "better off/worse off" is problematic. They're supposed to be measuring human well-being, but they use problematic surrogates such as possession of money or the satisfaction of "revealed preferences". And individual's "preferences", according to economists, are "revealed" in market transactions. It is assumed that these preferences are fixed and to be taken as given. But in reality, individual "preferences" are shaped by and formed within social conditions. Moreover, they are open to rational criticism and democratic debate; surely no critical approach to understanding society can take preferences as mere givens. So, the economist's metric for measuring how "well off" a person is is deeply flawed. Contrast it with the Marxist/Aristotelian view of human well-being, where a person is well-off to the extent that she is able to develop and exercise her talents and natural powers (e.g. creative faculties, physical capabilities, capacity to form meaningful relationships, intellectual powers, etc.). The juxtaposition makes clear that mainstream economists work within a framework for understanding human welfare that is thoroughly pro-capitalist, in that it contorts the definition of human well-being to fit the structural needs of the capitalist economy. It has the ideological effect of leaving us unable to articulate grievances or unmet needs, because it defines our needs and welfare in terms friendly to market society.

So much for Pareto efficiency. But what about the "ordinary" conception of efficiency? Why shouldn't that be the fundamental measuring stick by which we evaluate economic and political institutions?

Because "efficiency" in the ordinary sense is of merely instrumental value. It has no intrinsic value. Doing some task efficiently is good, but only insofar as the task is itself a good one. So, if the goal of some institution is to treat human illness, then doing so efficiently would appear to be a good thing. Efficiency, in this case, would be instrumentally valuable for achieving the goal of treating human illness. But the metric of efficiency doesn't tell us whether or not treating human illness is a good or bad thing. This is made clear by way of another example. Suppose the goal of some institution, imagine a colonial military venture, was to oppress some group of people. We can, of course, imagine the goal of oppression being carried out more or less efficiently. And we wouldn't want to say that efficiency is a good thing in this case, because the goal that is being efficiently achieved is abhorrent. So what we come to see is that the metric of efficiency tells us nothing of the worthiness of the goals of some process or institution. But the overall goal or end that some process aims at is crucial, indeed far more important than whether it is efficient. Because if the goal is unworthy, it doesn't matter whether its efficient.

So the fetishism of efficiency has the effect of obscuring the goals of certain social processes or institutions. Within the idiom of efficiency, in fact, we can't even have a discussion about what the goals are; they are taken as given. Thus, if you think that the function or goal of the health care system should be to treat sickness and ill health, you'll quickly find that there is no space for you to express such a position within a narrow debate about whether the present for-profit health care system is "efficient" or not. The same could be said of public transportation. It is sometimes said that public transit is not "efficient" because it doesn't run the way that a for-profit firm runs (e.g. fare prices are set lower than their market value, services are allocated more according to need than effective demand, etc.). But that judgment typically smuggles in and ascribes irrational goals/functions to public transit systems. The basic function or goal of public transit should be to create a maximally accessible system that helps everyone get where they need to go in a safe, comprehensive way. But if that's so, it hardly makes sense to complain that it doesn't produce profits. That's not the point of it! So, to ask whether it is efficient or not would really be to ask whether or not it fulfilled its function well. But more often than not, talk about "efficiency" smuggles in tacit goals and preferred functions. Thus, public transit can be attacked as "inefficient", where the thing it doesn't do efficiently is accumulate surpluses.

So where does this leave us? Efficiency, in the ordinary sense, clearly matters. For whatever it is that we're doing, it would be irrational for us to want it to be done inefficiently. Even if we're aiming to lay about on the couch and veg out, there are more or less efficient ways of doing it (e.g. it wouldn't make sense to do sixteen jumping jacks every 10 minutes in order to achieve the goal of laying on couch). But efficiency is not, as we've seen, intrinsically valuable. It is only valuable insofar as what we're doing is valuable. So what really matters is the choice of goals or ends. And that holds both for individual goals (e.g. how to live one's life, what sorts of relationships to cultivate, what career path to choose, etc.) as well as collective/social goals (e.g. what should the basic goal of social institutions be? what kind of life do we want to live together? what forms of association do we want to enter into with one another?, etc.). All of this drops out of discussions about mere efficiency. In order to know what goals are worthy of our assent, we need robust democratic debate between free and equal citizens. But, as economists will quickly point out, such a radical democratic vision is not compatible with structuring social life around market forces. Economists take this to be a problem with democracy; but socialists see things precisely the other way around. That is, socialists see that as a reason to reject the fetishism of the market as the central organizing structure of society.

1 comment:

Anonymous said...

I'm not entirely sure you understand how economists treat preferences. Economists don't treat preferences as given in the sense that "everybody is born with x preferences and they will never change or be influenced by anything!" They treat preferences as given in a specific situation that they are looking at. Generally most economists assume what influences preferences is outside of their realm of thought. This actually doesn't have all that much to do with Pareto efficiency either.

Furthermore most "neoclassical" economists are not market fundamentalists. New Classical maybe (neo means new in linguistic technicality, but the two groups have specific names because they are different groups...) but neoclassical is a broad brush, and includes plenty of people who know markets can be flawed from time to time.

As for Pareto efficiency, I have problems with it too, but they have little to do with individual preferences or the WARC.

I believe you may unfortunately fall into the trap of calling anyone who rejects Marx as "free market fundamentalist."

I reject Marx generally, but I also know markets can fail, and advocate full employment.