Economics, prioritisation, and pro-rich bias

tl;dr: Welfare eco­nomics is highly rele­vant to effec­tive al­tru­ism, but tends to rely on a flawed con­cep­tion of so­cial welfare, which holds that the more some­one is will­ing to pay for a good, the more util­ity or welfare they would get from con­sum­ing that good. (I use ‘welfare’ and ‘util­ity’ in­ter­change­ably here). This ne­glects the fact that differ­ences in will­ing­ness to pay are of­ten merely due to differ­ences in ini­tial re­source en­dow­ments. As a con­se­quence, welfare eco­nomics is bi­ased to­wards poli­cies that favour the rich. Effec­tive al­tru­ists should be aware of these prob­lems, and economists should adopt a re­vised con­cep­tion of so­cial welfare.

cross-posted from my blog.


Effec­tive al­tru­ism is the use of rea­son and ev­i­dence to pro­mote the welfare of all as effec­tively as pos­si­ble. Welfare eco­nomics is highly rele­vant to effec­tive al­tru­ism be­cause it aims to show which poli­cies or ac­tions would best max­imise so­cial welfare. The mod­ern dis­ci­pline of eco­nomics was heav­ily in­fluenced by early util­i­tar­ian thought, and eco­nomics has in­fluenced effec­tive al­tru­ism in nu­mer­ous ways with tools such as cost-effec­tive­ness-anal­y­sis and Dis­abil­ity Ad­justed Life Years. Welfare eco­nomics is, in my view, the most use­ful and prac­ti­cally ap­pli­ca­ble pri­ori­ti­sa­tion tool cur­rently available to gov­ern­ments. How­ever, as I will now ar­gue, main­stream welfare eco­nomics re­lies on a flawed the­ory of so­cial welfare, which leads to pro-rich bias in policy eval­u­a­tion.

I hope this post will im­prove un­der­stand­ing of welfare eco­nomics among effec­tive al­tru­ists. It would also be use­ful for economists to recog­nise these prob­lems and take a re­vised ap­proach.

Tout­ing and so­cial welfare

I will bring out this is­sue by dis­cussing the ques­tion of ticket ‘tout­ing’ or ‘scalping’. Economists are some­what un­usual in be­liev­ing that tout­ing is ac­tu­ally a good thing be­cause it cor­rects for un­der­priced tick­ets. Here is The Economist on the is­sue:

“Flint-hearted economists might note that a sec­ondary mar­ket sug­gests that the seats were un­der­priced. Cheaper tick­ets meant to boost equal ac­cess lure in touts, for whom low prices mean big­ger pre­miums. And more scalpers means more dis­ap­pointed fans in the queue.

Rather than al­low­ing touts to profit, the play’s pro­duc­ers could take a cue from “Hamil­ton”, a wildly suc­cess­ful Broad­way mu­si­cal, and raise prices for the pre­mium seats un­til de­mand falls in line with sup­ply (even at up to $849 per ticket, some ar­gue that “Hamil­ton” is too cheap). But the Pot­ter pro­duc­ers seem to be more wor­ried about im­pe­cu­nious wiz­ard­ing fans los­ing out than about the prospect of touts swiping sur­plus.

Stamp­ing out the sec­ondary mar­ket en­tirely means pre­vent­ing peo­ple sel­l­ing their tick­ets to those who value them more. This in­effi­ciency is wince-in­duc­ing for economists...” [em­pha­sis added]

Ac­cord­ing to some economists, ticket tout­ing im­proves al­loca­tive effi­ciency.

Alloca­tive effi­ciency oc­curs when there is an op­ti­mal dis­tri­bu­tion of goods ac­cord­ing to con­sumer prefer­ences, or, in other words, when so­cial welfare is max­imised.

The ar­gu­ment goes as fol­lows. By sel­l­ing tick­ets at a sin­gle price on a first come first served ba­sis, some peo­ple who re­ally want to go to the show will be un­able to go. When the ticket is un­der­priced, Pete, who is will­ing to pay no more than $50 for a Book of Mor­mon ticket, can get a ticket, but Rich, who is will­ing to pay up to $1000, doesn’t get a ticket.

Cru­cial Premise: Ne­c­es­sar­ily, the more some­one is will­ing to pay for a good, the more welfare they get from con­sum­ing that good.

[UPDATE: I ac­tu­ally mean this more pre­cise ver­sion of the premise. “Ne­c­es­sar­ily, if per­son A is will­ing to pay more for a good than per­son B, then per­son A gets more welfare from that good than per­son B. Thanks to ro­hin­mshah]

So, by meet­ing the mar­ket de­mand of those will­ing to pay more or, in other words, en­sur­ing that price is closer to marginal util­ity, touts en­sure that so­cial welfare is max­imised.

The vast ma­jor­ity (>68%) of economists be­lieve tout­ing in­creases so­cial welfare, as shown by this IGM poll (a good place to find the views of economists on lots of differ­ent top­ics). It’s some­what un­clear whether they do so on the ba­sis of the ar­gu­ment from al­loca­tive effi­ciency and the Cru­cial Premise, but I would bet that a sig­nifi­cant por­tion do en­dorse that ar­gu­ment.

What’s wrong with this ar­gu­ment?

I’m go­ing to ar­gue that the fore­go­ing ar­gu­ment fails be­cause the Cru­cial Premise is false. (Note that tout­ing might be jus­tified by other ar­gu­ments).

I’ll first clar­ify the as­sump­tions made in the ar­gu­ment.

Utili­tar­i­anism = Agents ought to perform the act which max­imises to­tal so­cial util­ity or welfare.

A large por­tion of economists ac­cept prefer­ence util­i­tar­i­anism, ac­cord­ing to which util­ity is con­ceived of as prefer­ence satis­fac­tion. When eval­u­at­ing policy, many economists like to say that they put moral­ity to one side, but this is sel­dom true. In ac­tual fact, they are ap­peal­ing to prefer­ence util­i­tar­i­anism. This is a moral the­ory.

Some economists be­lieve that al­loca­tively effi­cient out­comes might in­volve large in­equal­ities and there­fore be un­fair. Con­se­quently, they en­dorse an equity or fair­ness con­straint on prefer­ence util­i­tar­i­anism. In philo­soph­i­cal terms, this is equiv­a­lent to prefer­ence util­i­tar­i­anism with a welfare egal­i­tar­ian con­straint. Pro­po­nents of such a the­ory tend to recom­mend that gov­ern­ments cor­rect in­equal­ity through re­dis­tri­bu­tion.

The pro-tout­ing ar­gu­ment com­bines prefer­ence util­i­tar­i­anism and the Cru­cial Premise, con­clud­ing that tout­ing is jus­tified be­cause it max­imises so­cial welfare.

With this clar­ified, we can now ex­plore why the pro-tout­ing ar­gu­ment does not work. The Cru­cial Premise is false. It is not nec­es­sar­ily true that will­ing­ness to pay for a good is an in­di­ca­tor of how much util­ity one would get from a good. This is ob­vi­ous. For ex­am­ple, sup­pose that Pete is very poor and Rich is very rich. As a con­se­quence, Pete will­ing to pay up to $50 for a Book of Mor­mon ticket, but Rich is will­ing to pay up to $1,000. But this does not nec­es­sar­ily mean that Rich would get more util­ity from watch­ing the Book of Mor­mon than Pete. All it shows is that Pete doesn’t have as much money. It might be the case that Rich would mildly en­joy the show, but Pete would ab­solutely love it.

In­deed, imag­ine that Pete has no money at all. Ac­cord­ing to the view that, nec­es­sar­ily, the more one is will­ing to pay for a good the more util­ity one de­rives from it, Pete would not gain util­ity from the con­sump­tion of any good, even food or wa­ter. This is ab­surd.

We can avoid this by cor­rect­ing for in­equal­ity in in­come or re­sources be­tween in­di­vi­d­u­als when as­sess­ing will­ing­ness to pay. We could, for ex­am­ple, ask what Pete would be will­ing to pay for a ticket if he had as much money as Rich. Thus, hy­po­thet­i­cal, rather than ac­tual, will­ing­ness to pay would de­ter­mine con­sumer prefer­ence. Con­sumer prefer­ence would not be re­vealed by ac­tual mar­ket de­mand. If so, then it is not nec­es­sar­ily true that tout­ing tick­ets at higher prices in­creases so­cial welfare by al­lo­cat­ing tick­ets to those who would get most util­ity from them.

Not only is it not nec­es­sar­ily true that ac­tual will­ing­ness to pay de­ter­mines con­sumer prefer­ence, it is not even usu­ally true. Differ­ences in will­ing­ness to pay are to a sig­nifi­cant ex­tent and in a huge range of cases driven by differ­ences in per­sonal wealth rather than by differ­ences in con­sumer prefer­ence. Rich peo­ple tend to holi­day in ex­otic and sunny places at much higher rates than poor peo­ple. This is en­tirely a product of the fact that rich peo­ple have more money, not that poor peo­ple pre­fer to holi­day in Black­pool. I think the same holds for the vast ma­jor­ity of differ­ences in mar­ket de­mand across differ­ent in­come groups.

In sum, the ar­gu­ment for tout­ing from prefer­ence util­i­tar­i­anism and the Cru­cial Premise fails.

Im­pli­ca­tions for welfare economics

This is one in­stance of a se­ri­ous gen­eral prob­lem for con­tem­po­rary welfare eco­nomics. Equat­ing mar­ket de­mand and util­ity with­out cor­rect­ing for in­equal­ity in in­come or re­sources leads economists to pro-rich bias. It is this same flaw that led the 1995 IPCC re­port to con­clude, on the ba­sis of a will­ing­ness to pay ap­proach, that In­dian lives were worth less than Amer­i­can lives.[1]

It is easy to see how this bias could come into play for pretty much all poli­cies as­sessed by welfare eco­nomics. Economists will ne­glect in­equal­ity and tend to recom­mend that goods be dis­tributed by mar­ket prices.

This is not a crit­i­cism of prefer­ence util­i­tar­i­anism from equity or fair­ness. I am not say­ing that only aiming to max­imise so­cial welfare is ine­gal­i­tar­ian, and I am not say­ing that equal­ity is in­trin­si­cally valuable. I am say­ing that prefer­ence util­i­tar­i­anism alone, prop­erly con­ceived and with­out an equity con­straint, favours more egal­i­tar­ian out­comes than economists ac­knowl­edge.

One ad­van­tage of hold­ing that ac­tual will­ing­ness to pay de­ter­mines prefer­ence is that it is eas­ier to mea­sure than hy­po­thet­i­cal will­ing­ness to pay. For this rea­son, in some cases it may be more prac­ti­ca­ble to ap­prox­i­mate prefer­ence util­i­tar­i­anism (prop­erly con­ceived) with the Cru­cial Premise + an in­de­pen­dent equity con­straint. This equity con­straint would be jus­tified on util­i­tar­ian grounds, rather than on the grounds that equal­ity is in­trin­si­cally im­por­tant.

The down­side of this is that economists would still be giv­ing an in­ac­cu­rate ac­count of what con­sti­tutes prefer­ence satis­fac­tion. The state­ment “tout­ing op­ti­mises the dis­tri­bu­tion of goods ac­cord­ing to con­sumer prefer­ence, but is in­equitable” is false be­cause the first con­junct is false.

Many thanks to Ste­fan Schu­bert for always helpful com­ments.

[1] The great John Broome dis­cusses this on p.15 here—http://​​​​~sfop0060/​​pdf/​​Valu­ing%20poli­cies%20in%20re­sponse%20to%20cli­mate%20change,%20some%20eth­i­cal%20is­sues.pdf