What does it mean to maximize utility?

In two debates this week, one online and one in per­son, the ques­tion of utility-​​maximization has become a focal point. I addressed it briefly in a recap of one of them, but I feel that it’s use­ful to go back, rethink, and expand on these com­ments now that I’ve stewed over it for a bit — and since the impor­tance of the point has become increas­ingly clear to me.

A great deal of the nor­ma­tive side of eco­nom­ics depends on the ques­tion of our ulti­mate goal. Usu­ally that means decid­ing to max­i­mize some­thing — and one com­mon con­cep­tion in mod­ern eco­nom­ics is the idea of max­i­miz­ing util­ity - that is, sat­is­fy­ing the most num­ber of wants in society.

Cen­tral to almost any lassez-​​faire argu­ment, be it neo­clas­si­cal or Aus­trian, is the idea of mar­ket effi­ciency. Given per­fectly com­pet­i­tive mar­kets (and ignor­ing the like­li­hood of this assump­tion), the First Fun­da­men­tal The­o­rem of Wel­fare Eco­nom­ics demon­strates that out­put will be allo­cated so as to ful­fill the most num­ber of wants (as mea­sured essen­tially by a con­cept of sur­pluses — or prof­its). This is both intu­itive and con­ve­nient, which explains the fer­vor with which it is repeated.

But as I dis­cussed in that pre­vi­ous post, the sat­is­fac­tion of indi­vid­ual wants in the mar­ket­place does not tell the whole story. Say we care about the dis­tri­b­u­tion of income, but don’t think that we as indi­vid­u­als can do any­thing about it — which hap­pens to be true. One char­i­ta­ble con­tri­bu­tion does not move the nee­dle appre­cia­bly, and so nobody is inclined to give of their own income to solve what every­body believes is a problem.

This means that we have two tiers of pref­er­ences. We have what neo­clas­si­cals term ‘revealed pref­er­ences’, which is the actions that indi­vid­u­als take in the free mar­ket­place, and basi­cally reflects the align­ment of all their eco­nomic incen­tives. But it is also clear that there exists a second-​​order pref­er­ence, in which indi­vid­u­als col­lec­tively believe that some­thing ought to be dif­fer­ent about their incen­tives so that they could effect a dif­fer­ent aggre­gate outcome.

The way I see it, max­i­miz­ing util­ity is about sat­is­fy­ing these col­lec­tive desires, even if it means a lower level of ‘revealed pref­er­ences’ get­ting sat­is­fied. As I dis­cussed the other day in the Pris­on­ers’ Dilemma prob­lem, max­i­miz­ing social hap­pi­ness in fact means that nobody’s revealed pref­er­ence is allowed — that is, every indi­vid­ual is frus­trated that he is not allowed to defect in the game (because doing so would improve his pay­out), but on the whole, every­one is bet­ter off since nobody is allowed to.

In a world where Arrow’s Impos­si­bil­ity The­o­rem means that we have no reli­able way of express­ing col­lec­tive pref­er­ences, the task set us is much more dif­fi­cult than sim­ple com­par­isons between var­i­ous Nash and Pareto equi­lib­ria. Our social aggre­ga­tors (often mean­ing rep­re­sen­ta­tive gov­ern­ment) are sim­ply unable to express a con­sis­tent pre­ferred goal and then guide us to it. What we have to do is, to take the Chi­nese expres­sion, ‘cross the river by grop­ing at stones’.

In other words, eco­nom­ics and gov­er­nance is not just hard, it’s impos­si­ble. We don’t even have a way to define end goals, much less reach them. We have to con­stantly reeval­u­ate our loca­tion and our progress, know­ing that what we do next depends on what we did last, and not know­ing how it will affect our options even in the very near future.

I can’t think of any way to sum this up but to say that eco­nom­ics will never get any eas­ier. It’s less like a nat­ural sci­ence in which each dis­cov­ery lays ground­work for more advanced work, and a lot like lit­er­a­ture where we have to solve very sim­i­lar prob­lems over and over, while still show­ing marked devel­op­ment across time.

For­get equi­lib­rium. This is the age of dynamism.