Efficiency: A Term to Use Sparingly

Microeconomic Theory II for PhD students starts again on Tuesday here at Temple University. I intend to talk about understanding economics as a way to (i) see clearly what the objectives are (whether they come from us as economists or not), (ii) see how the aggregation of many persons’ aims is fraught with terrible difficulties (Arrow’s impossibility theorem), and (iii) introduce the idea of incentives as something that needs alignment with the aims, which alignment may be very hard to achieve in a society (Gibbard-Satterthwaite theorem). I have made my lecture notes available here (warning: terse and notation-heavy stuff, be patient in reading).

Lest this sound like an impossible paragraph, let me hasten to follow it by one in plain English. To this end, I will point first to Princeton economics professor Uwe Reinhardt’s very recent posting in the Economix blog of the New York Times. Please do read it before continuing to read this post.

In this very good piece, Professor Reinhardt points out how often the term “efficiency” is trotted out to make audiences feel that the economist speaker/writer addressing them is making an objective, scientific claim. Not so fast! Efficiency, in the technical sense of economics (“Pareto efficiency”), means that a state of affairs prevails in the economy so that there is no feasible change that will improve the welfare of at least one person while harming no one.

Think about it for a second; it sounds appealing, at first. Why not aim to extinguish the “welfare waste” that inefficient states possess?

But it is not so easy. Nobel laureate Amartya K. Sen has classic criticisms (see his Nobel lecture, page 198 and onwards, or look up his 1977 classic paper, “Rational Fools”) of efficiency. (He is not the only one, but he is quite probably the most respected scholar, among economists, who has such criticisms.) The Rational Fools paper, if memory serves me right, is the one that pointed out that the state of affairs in which Roman Emperor Nero would have been prevented from having Rome set afire as a background to his lyre playing, would entail a loss of welfare for Nero (never mind that it would improve the welfare of thousands of Roman citizens) and therefore it would not be an improvement in efficiency over the historical state of affairs, in which Nero did enjoy performing his ode while Rome burned. Pretty much every person’s sense of fairness would find this repugnant, which points out one particular, often glaring, way that economic efficiency hides a strong value judgment, that says “don’t you dare hurt anyone, no matter privileged, to help any number of people, however non-privileged or deserving”.

Worse still, economists are blithely using “efficiency” to justify all sorts of policy suggestions, and as a benchmark in countless papers. I hope this post has helped at least one consumer of writings of economists become more critical of this practice.

For a bit more about this, please check out my writing on public economic on this page.