Economics Strategy Nov 1, 2007

Nobel Designs

Under­stand­ing mech­a­nism design

The 2007 Nobel in Eco­nom­ics raised smiles in the Man­age­r­i­al Eco­nom­ics and Deci­sion Sci­ences (MEDS) depart­ment at Kel­logg for two rea­sons. First, the prize hon­ors a body of work, mech­a­nism design, which has been and con­tin­ues to be a cot­tage indus­try with­in the depart­ment. Sec­ond, it hon­ors three indi­vid­u­als, Leo Hur­wicz, Eric Maskin, and Roger Myer­son, with ties to MEDS. Hur­wicz is a par­tic­u­lar friend of Stan­ley Reit­er, a founder of the depart­ment. They have co-authored sev­er­al papers as well as a book on mech­a­nism design. Maskin is a co-author and for­mer advi­sor of two cur­rent fac­ul­ty mem­bers in MEDS (Sandeep Bali­ga and Péter Eső). And, of course, Roger Myer­son was a fac­ul­ty mem­ber for twen­ty-five years in the MEDS department.

The award raised hack­les among the athe­o­ret­i­cal unwashed because it ele­vates an abstract sub­ject with roots in game the­o­ry. The New York Times felt com­pelled to report this in its Octo­ber 20, 2007, issue. The arti­cle could be sum­ma­rized thus: X says the award was unde­served, and Y said oth­er­wise. Fair, bal­anced, and vacuous.

The lau­re­ates them­selves have game­ly tried in sound bites to describe their con­tri­bu­tion and its impor­tance. How­ev­er, mech­a­nism design does not lend itself to such a task. Its impor­tance comes not from a col­lec­tion of take-aways, to-do lists, or ten-point plans. Rather, it is an ana­lyt­i­cal frame­work for think­ing clear­ly and care­ful­ly about the most fun­da­men­tal of social prob­lems: what exact­ly can a giv­en insti­tu­tion achieve when the infor­ma­tion nec­es­sary to make deci­sions is dis­persed and pri­vate­ly held? The range of ques­tions to which the approach can be applied is strik­ing. To achieve a giv­en reduc­tion in car­bon emis­sions, should one rely on tax­es or a cap and trade sys­tem? Is it bet­ter to sell an IPO via auc­tion or the tra­di­tion­al book­build­ing approach? Would juries pro­duce more informed deci­sions under a una­nim­i­ty rule or sim­ple major­i­ty? Mech­a­nism design helps us under­stand how the answers to these ques­tions depend on the details of the under­ly­ing envi­ron­ment. In turn, this knowl­edge helps us under­stand which details mat­ter and which do not. To para­phrase an old proverb, Hur­wicz, Maskin, and Myer­son gave us not fish for a day but fish­ing rods.

To get a sense of what mech­a­nism design is, we begin with a sto­ry recount­ed by an ear­li­er Nobelist, Ronald Coase, more than forty years ago. It involves a rail­road with a coal-burn­ing loco­mo­tive that runs near a stretch of land owned by a farmer.

The loco­mo­tive emits sparks that set fire to the farmer’s crops. Sup­pose run­ning the train yields $1,000 worth of prof­it for the rail­road but caus­es $2,000 worth of crop dam­age. Should the rail­road be made to pay for the dam­age it causes?

The sparks by them­selves do no dam­age. One might say the farmer caused the dam­age by plac­ing her crops next to the rail­way line. It is the jux­ta­po­si­tion of sparks and crops that lead to the $2,000 worth of dam­age. Per­haps, then, the farmer is liable.

If you think this strange, sup­pose it costs the farmer $100 to ensure the safe­ty of her crop. If we make the rail­road liable for dam­age to the crop, what hap­pens? The rail­road stops run­ning (assum­ing the absence of tech­nol­o­gy that would elim­i­nate the sparks). Why spend $2,000 to get a return of $1,000? The farmer takes no pre­cau­tions to secure her crop. As a soci­ety we are out $1,000, the prof­its the rail­road would have made had it run. Now sup­pose we make the farmer liable. The train runs. The farmer pays the $100 to safe­guard her crop rather than the $2,000 in crop dam­age. On bal­ance, soci­ety is out only $100.

If we cared about avoid­ing dam­age in the most cost-effec­tive way pos­si­ble, we should make the farmer liable. Sup­pose, instead, the rail­road has access to tech­nol­o­gy that would elim­i­nate the sparks for a price of $50. In this case, since it is cheap­er for the rail­road to avoid the dam­age, it should be made liable. If cost effec­tive­ness is our pur­pose, it puts us in a pick­le because the assign­ment of lia­bil­i­ty depends on the details of the par­tic­u­lar sit­u­a­tion. Coase’s insight is that it mat­ters not how lia­bil­i­ty is assigned as long as the par­ties are per­mit­ted to trade the lia­bil­i­ty amongst themselves.

Sup­pose the rail­road is made liable. What mat­ters is if the rail­road can pay the farmer to shoul­der the lia­bil­i­ty. Assume as before that the rail­road can­not reduce the sparks emit­ted with­out shut­ting down the loco­mo­tive and that the farmer can avoid the crop dam­age at a cost of $100. Observe that the rail­road is bet­ter off pay­ing the farmer at least $100 (and no more than $1000) to move the crops. The farmer will also be bet­ter off. In effect, the rail­road pays the farmer to assume the lia­bil­i­ty; a win-win” sit­u­a­tion. Thus, as long as we allow the par­ties con­cerned to trade their lia­bil­i­ties, the par­ty with the least cost of avoid­ing the dam­age will shoul­der the lia­bil­i­ty. In terms of eco­nom­ic effi­cien­cy, it mat­ters not who is liable for what. It mat­ters only that the lia­bil­i­ties be clear­ly defined, eas­i­ly trad­able, and enforced. It is true that the farmer and rail­road care a great deal about who is held liable for what. If it is the rail­road, then it must pay the farmer. If it is the farmer, then the rail­road pays noth­ing. One may pre­fer, for rea­sons quite sep­a­rate from eco­nom­ic effi­cien­cy, to hold one par­ty liable rather than the oth­er. But the out­come in terms of who does what remains the same.

Coase rec­og­nized that there are has­sle costs asso­ci­at­ed with bar­gain­ing over the trans­fer of lia­bil­i­ties. Fur­ther, these costs might over­whelm the gains to be had from bar­gain­ing. There­fore, it is of fun­da­men­tal impor­tance that such costs be min­i­mized. Nev­er­the­less, mutu­al­ly ben­e­fi­cial bar­gains fail to be struck even when has­sle costs are nonex­is­tent. Per­son­al­i­ty, ego, and his­to­ry can con­spire to pre­vent agree­ment. These are unsat­is­fy­ing expla­na­tions for why mutu­al­ly ben­e­fi­cial agree­ments are unmade because they are idio­syn­crat­ic and sit­u­a­tion-spe­cif­ic. Mech­a­nism design sug­gests anoth­er rea­son: the actu­al cost incurred by each par­ty to avoid the dam­age is pri­vate infor­ma­tion known only to themselves.

To see why, sup­pose the rail­road incurs a cost $R of avoid­ing the dam­age while the farmer incurs a cost of $F to do the same. I have pur­pose­ly cho­sen to use let­ters rather than con­crete num­bers to empha­size that only the rail­road knows the actu­al val­ue of $R and only the farmer knows the actu­al val­ue of $F. If $F > $R, we (mean­ing soci­ety) would like the rail­road to incur the cost of avoid­ing the dam­age. If $F < $R, we would like the farmer to incur the cost of avoid­ing the dam­age. In the event that $F=$R, we are indif­fer­ent between which one incurs the cost.

Now, let us quite arbi­trar­i­ly make the rail­road liable for the dam­age and trust that bar­gain­ing between rail­road and farmer will result in the per­son with the low­er cost of avoid­ing the dam­age under­tak­ing the bur­den to avoid the dam­age. If $R > $F, the rail­road should pay the farmer to take on the lia­bil­i­ty. Fur­ther­more, it would want to pay as lit­tle as pos­si­ble, ide­al­ly no more than $F. How­ev­er, the rail­road does not know the val­ue of F. So how much should it offer? The low­er the offer, the less like­ly it will be accept­ed. On the oth­er hand, it is more prof­itable to the rail­road if it is accept­ed. The farmer, how­ev­er, has every incen­tive to bluff the rail­road into think­ing that $F is larg­er than it actu­al­ly is so as to make a tidy prof­it. If the farmer is too aggres­sive in this regard, the rail­road may walk away think­ing that $R < $F. One can con­ceive of a vari­ety of bar­gain­ing pro­ce­dures that might mit­i­gate these dif­fi­cul­ties. Both could simul­ta­ne­ous­ly reveal their actu­al costs and split the dif­fer­ence, or they could rely on some trust­ed third par­ty as a medi­a­tor. Is there a bar­gain­ing pro­to­col that will always lead to the par­ty with the low­er cost of avoid­ing the dam­age assum­ing the liability?

Mech­a­nism design approach­es this ques­tion using the tools of game the­o­ry. Any such pro­to­col can be thought of as a game that encour­ages each par­ty to reveal truth­ful­ly its cost of avoid­ing the dam­age so that the cor­rect assign­ment of lia­bil­i­ty can be made. The encour­age­ment to truth­ful­ly reveal this pri­vate infor­ma­tion is bought with mon­ey. (Econ­o­mists, unlike lawyers, believe you have to pay peo­ple to tell the truth.) The mon­e­tary rewards must be gen­er­at­ed inter­nal­ly, i.e., there is no rich uncle wait­ing on the side­lines to come to the aid of either farmer or rail­road. Thus, the ques­tion becomes a pure­ly math­e­mat­i­cal one: is there a game with these prop­er­ties? Roger Myer­son and his Kel­logg col­league, Pro­fes­sor Mark Sat­terth­waite, proved that the answer to this ques­tion was a resound­ing, de Gaulle-like NON.” There is no bar­gain­ing pro­to­col or trust­ed medi­a­tor that is guar­an­teed in all cir­cum­stances to ensure that the par­ty with the low­er cost of avoid­ing the dam­age assumes the lia­bil­i­ty. None — no mat­ter how imag­i­na­tive, elab­o­rate, or con­vo­lut­ed. Hence, there is always the pos­si­bil­i­ty that no bar­gain will be struck even when it is in the mutu­al inter­est of both par­ties to come to terms.

Thus, Coase’s orig­i­nal obser­va­tion that the assign­ment of lia­bil­i­ty is irrel­e­vant since an incor­rect assign­ment would be cor­rect­ed by bar­gain­ing in the mar­ket­place (pro­vid­ed has­sle costs are small) is not true in the shad­ow of pri­vate infor­ma­tion. Mech­a­nism design also sug­gests how lia­bil­i­ty should be assigned. Specif­i­cal­ly, to ensure that the lia­bil­i­ty is assigned to the par­ty with the low­est cost for avoid­ing the dam­age, the right to avoid the lia­bil­i­ty should be auc­tioned off to the high­est bid­der. How is this possible?

Sup­pose our auc­tion works as fol­lows. We have a price clock ini­tial­ly set at zero. We then raise the price. At each price we ask the bid­ders (rail­road and farmer) whether they wish to buy the right to avoid lia­bil­i­ty at the cur­rent price. If both say yes,” con­tin­ue rais­ing the price. Stop the instant one of them drops out and sell the right to the remain­ing active bid­der at the ter­mi­nal price. Observe that the farmer will stay active as long as the cur­rent price is below $F. The rail­road will stay active as long as the cur­rent price is below $R. If the farmer drops out first, it must be because $F < $R, in which case the farmer assumes lia­bil­i­ty and the rail­road pays the auc­tion­eer $F. In short, the farmer, who had the low­er cost of avoid­ing the dam­age, is sad­dled with the lia­bil­i­ty. If $R < $F, the reverse hap­pens. The exam­ple of the rail­road and the farmer involved the allo­ca­tion of a lia­bil­i­ty. It could just as well as have involved the allo­ca­tion of a prop­er­ty right. One is the obverse of the other.

What is the punch line of this brief excur­sion into mech­a­nism design? When gov­ern­ments cre­ate new prop­er­ty rights or asset class­es, to ensure they are allo­cat­ed in an eco­nom­i­cal­ly effi­cient man­ner, they should be auc­tioned off. It is exact­ly that rea­son­ing that sup­ports the allo­ca­tion of spec­trum rights by auc­tion. It is exact­ly that rea­son­ing that sup­ports the allo­ca­tion of per­mits to pol­lute by auc­tion. It is exact­ly this rea­son­ing that will even­tu­al­ly pro­pel the FAA to use auc­tions to allo­cate arrival and depar­ture slots at air­ports. John May­nard Keynes said it best: I am sure that the pow­er of vest­ed inter­ests is vast­ly exag­ger­at­ed com­pared with the grad­ual encroach­ment of ideas.”

Mech­a­nism design is a young scion of an unbro­ken line sprung from the belief that math­e­mat­ics is a pow­er­ful engine of inquiry into the work­ings of com­merce and gov­ern­ment. One can trace the line back to the hand of Con­dorcet and the Enlight­en­ment philoso­phers and prob­a­bly beyond. I think they would have been proud of what has been wrought. It is a mon­u­ment, to quote Horace,

.….more last­ing than bronze and far high­er
than that roy­al pile of Pyra­mids,
which the gnaw­ing rain and furi­ous
north wind can­not destroy, nor the chain
of count­less years and the flight of time.”

Cen­ter for Math­e­mat­i­cal Stud­ies in Eco­nom­ics and Man­age­ment Sci­ence at Kellogg 

For close to half a cen­tu­ry the Kel­logg School has been a strong believ­er in the pow­er of math­e­mat­i­cal analy­sis to yield insights into the work­ings of the pri­vate and pub­lic sec­tor. That belief found sub­stance in the estab­lish­ment of the Cen­ter for Math­e­mat­i­cal Stud­ies in Eco­nom­ics and Man­age­ment Sci­ence (CMS-EMS) about forty years ago. The found­ing direc­tor was Stan­ley Reit­er, Mor­ri­son Pro­fes­sor of Eco­nom­ics and Math­e­mat­ics. Reit­er him­self has made impor­tant con­tri­bu­tions to mech­a­nism design for which he was hon­ored with, among oth­er things, a Guggen­heim award as well as mem­ber­ship in the Amer­i­can Acad­e­my of Arts and Sciences.

Since its incep­tion the cen­ter has played an impor­tant role in turn­ing North­west­ern into one of the world’s largest and most impor­tant cen­ters for the math­e­mat­i­cal study of social phe­nom­e­na. It has been a forum for the exchange and dis­sem­i­na­tion of ideas. Indeed, a num­ber of the sem­i­nal papers on game the­o­ry, mech­a­nism design, and auc­tion the­o­ry were first pre­sent­ed at Cen­ter sem­i­nars and cir­cu­lat­ed as CMS-EMS tech­ni­cal reports. Myerson’s prize is the first to hon­or some of the work that arose from those activ­i­ties but not the last. There are at least two more wait­ing in the wings.

That tra­di­tion of math­e­mat­ics as an engine of inquiry con­tin­ues today. The breadth of activ­i­ty is astound­ing, includ­ing the design of auc­tions for the sale of car­bon emis­sion per­mits, the choice of con­sti­tu­tions, the analy­sis of com­mu­ni­ca­tion, the study of open spec­trum, the per­for­mance of pre­dic­tion mar­kets, the allo­ca­tion of arrival slots at air­ports, the sale of online adver­tise­ments, the design of exchanges, and the study of rep­u­ta­tion systems.

Fur­ther reading:

Coase, Ronald H. (1960). The Prob­lem of Social Cost,” Jour­nal of Law and Eco­nom­ics, 3: 1 – 44.

Fed­der­sen, Tim­o­thy J. and Wolf­gang Pesendor­fer (1998). Con­vict­ing the Inno­cent: The Infe­ri­or­i­ty of Unan­i­mous Jury Ver­dicts,” Amer­i­can Polit­i­cal Sci­ence Review, 92(1): 23 – 35. Kel­logg Insight arti­cle: Judg­ing the Jury Vote.” April 2007

Jagan­nathan, Ravi and Ann E. Sher­man (2006). Why Do IPO Auc­tions Fail?” NBER Work­ing Paper No. 12151. Kel­logg Insight arti­cle: Why Do IPO Auc­tions Fail?” May 2007.

Myer­son, Roger B. and Mark A. Sat­terth­waite (1983): Effi­cient Mech­a­nisms for Bilat­er­al Trad­ing,” Jour­nal of Eco­nom­ic The­o­ry, 28(2): 265 – 281.

Featured Faculty

Rakesh Vohra

Faculty member in the Department of Managerial Economics & Decision Sciences until 2013

About the Writer

Rahesh Vohra is the John L. and Helen Kellogg Professor of Managerial Economics and Decision Sciences at the Kellogg School of Management

Editor’s Picks


Pod­cast: How to Be a Great Mentor

Plus, some valu­able career advice that applies to just about everyone.


A New Way to Per­suade Kids to Drink More Water and Less Soda

Get­ting chil­dren to make healthy choic­es is tricky — and the wrong mes­sage can backfire.


How Can Social Sci­ence Become More Solutions-Oriented?

A con­ver­sa­tion between researchers at Kel­logg and Microsoft explores how behav­ioral sci­ence can best be applied.


Buy­ing a Com­pa­ny for Its Tal­ent? Beware of Hid­den Legal Risks.

Acquir­ing anoth­er firm’s trade secrets — even unin­ten­tion­al­ly — could prove costly.


Take 5: Tips for Widen­ing — and Improv­ing — Your Can­di­date Pool

Com­mon bias­es can cause com­pa­nies to over­look a wealth of top talent.


Every­one Wants Phar­ma­ceu­ti­cal Break­throughs. What Dri­ves Drug Com­pa­nies to Pur­sue Them?

A new study sug­gests that firms are at their most inno­v­a­tive after a finan­cial windfall.


4 Key Steps to Prepar­ing for a Busi­ness Presentation

Don’t let a lack of prep work sab­o­tage your great ideas.


Video: How Open Lines of Com­mu­ni­ca­tion Can Improve Health­care Outcomes

Train­ing physi­cians to be bet­ter com­mu­ni­ca­tors builds trust with patients and their loved ones.


Here’s a Bet­ter Way to Sched­ule Surgeries

A new tool could dri­ve sav­ings of 20 per­cent while still keep­ing sur­geons happy.

Politics & Elections

Why Eco­nom­ic Crises Trig­ger Polit­i­cal Turnover in Some Coun­tries but Not Others

The fall­out can hinge on how much a country’s peo­ple trust each other.


Build­ing Strong Brands: The Inside Scoop on Brand­ing in the Real World

Tim Calkins’s blog draws lessons from brand mis­steps and triumphs.


How the Cof­fee Indus­try Is Build­ing a Sus­tain­able Sup­ply Chain in an Unsta­ble Region

Three experts dis­cuss the chal­lenges and rewards of sourc­ing cof­fee from the Demo­c­ra­t­ic Repub­lic of Congo.