The transitional gains trap

SAVE OFFLINE

The Transitional Gains Trap Author(s): Gordon Tullock Source: The Bell Journal of Economics, Vol. 6, No. 2 (Autumn, 1975), pp. 671-678 Published by: RAND Corporation Stable URL: http://www.jstor.org/stable/3003249 Accessed: 30-10-2015 19:18 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

RAND Corporation and Wiley are collaborating with JSTOR to digitize, preserve and extend access to The Bell Journal of Economics.

http://www.jstor.org

This content downloaded from 134.184.26.108 on Fri, 30 Oct 2015 19:18:51 UTC All use subject to JSTOR Terms and Conditions

The transitional gainstrap Gordon Tullock UniiversityProfessorVirginiaPolytechnicInstitute

Many governmenlt programiswhvhich appear to be designed to lielp some particullarindulstry or grolupdo not seem to be sUcceeding. The explanationofferedhere is thattheprogram, vhen inaugulrated,generated transitionalgains for the individlualsor companies in the inidustry,blut that these have been-fiully capitalized, vith the result that the people in the indlustry now are doing no betterthan normal. On the otherhand, the termination of theparticularscheemewvould, in general, lead to large losses for the entrenchedinterests. a One of the major activities of modern governmentsis the grantingof special privileges to various groups of politically influentialpeople. Air transportation,for example, has been cartelized by the CAB, surface transportationby the ICC, and we are preventedfrom receiving "too-high" interestrates on our bank accounts by the Federal Reserve Board and various other governmentagencies. On the whole, however, the profit record of these protected industries does not seem to differ systematicallyfromthe nonprotected.This raises questions of why these special privilegesdo not seem to do much good.'

1. Introduction

* The purpose of thisarticleis to discuss thisapparentlong-run unprofitability of governmentaid programs.To give a previewof the general plot, it will be my thesis that there are only transitionalgains to be made when the governmentestablishes special privilegesfor a group of people. The successors to the original beneficiarieswill not normally make exceptional profits;but, unfortunately, they usually will be injuredby cancellationof the originalgift.2Indeed, we shall be on what, paradoxically,seems to be an inefficientportionof the Paretian frontier. Althoughsome governmentrestrictionsare clearly designed to benefitspecificgroups of people, thisis by no means the only way in which such special privileges can arise. The pharmaceuticals industry,for example, foughthard against the 1962 legislation.Since it has come into effect,they have become, on the whole, satisfied with it, because it turns out that it has reduced the vigor of competition, with the result that their profitsare about as big as they were before, withoutthe risk

2. General argument

I All of thishas led David Friedmanto propoundwhat he calls - Friedman's Second Law," and what others call "Friedman II's Law," which is "the governmentcan't even give anythingaway." 2 See Buchanan and Tullock [3] for a similaranalysis of privatemonopoly. For an earlier analysis in somewhat the same spirit,see Stigler [2].

This content downloaded from 134.184.26.108 on Fri, 30 Oct 2015 19:18:51 UTC All use subject to JSTOR Terms and Conditions

THE BELL JOURNAL OF ECONOMICS

/ 671

involved in the introductionof new and improveddrugs.3Even in those cases where the governmentregulation was fairly clearly motivatedby a desire to help some particulargroup, it usually turns out that at least some features are not to the advantage of the benefited group. Apparently government moves in mysteriousways its wonders to perform,and is not completely controlled even by the best organized pressure group. 3. Taxi medallions

FIGURE 1 TAXI MEDALLIONMONOPOLY

LC-

0-

PI

I

_ _

I

M

I 1 C QUANTITY

* Let us consider a very simple example of government monopoly creation. This example will be the taxi medallion system,althoughmy simplifieddescriptionwill not correspond exactly to that in New York. In Figure 1 we show the demand for taxicab services with the usual downward-slantingline, labeled D, and theircost as a horizontalline, labeled P. In a competitiveenvironment,the price charged would also be at cost and C units of taxi service would be purchased at a price, P. Withsupply of taxis restricted,the price rises to P' and, of course, thereis now a significant monopolyprofit,shown by the usual rectangle.We shall, forthe moment,ignorethequestion of what happens to the taxis which are idled by this change and simplyconcern ourselves withthe taxi owners and driverswho have the rightto drive under the new dispensation, i.e., who have the taxi medallions. It is clear thattheyhave gained a great deal. Suppose, however, we wait fora numberof years. By now, the capital value of the monopolyprofithas been fullytakeninto account in the industry.New entrantsenteronly by purchasing the medallion,withthe resultthattheyget only normalprofits.4 Further,the survivingoriginal owners have opportunitycosts equivalent to the value of the medallions upon which they receive normalreturns.The customers,of course, are worse off. Can we suggesta compensationscheme which would get us out of the mess? Normally,the answer to thisquestion is "no," because the impliedtransactioncosts are excessive. If we could somehow identifythose people who are now not usinga cab but who would at a slightlylower price and tax them so that the of the burdenfollowedthe dottedline in Figure 1, it distribution would be possible to graduallybuy back the monopolyfromits currentowners, and hence benefiteveryone. Unfortunately, we have no way of doingit, and hence myreferenceto the situation as being an "inefficientpoint on the Paretian frontier." There is a possible way in which we mightget out of the trap. It may be that the monopolisticorganizationof the industry is not efficientfor one reason or another. The number of taxicab medallions in New York has remained unchanged for verylong periods of time. It seems unlikelythatthe demand and cost of taxicabs duringthat period of time have continuously been such that the unchanged number of cabs maximized the monopoly gains. A brief examination of the other regulated industriesor specially privilegedareas will usually lead to the I

672 / G. TULLOCK

4

See Peltzman [4], especially pp. 1086-1089. See Auster [1].

This content downloaded from 134.184.26.108 on Fri, 30 Oct 2015 19:18:51 UTC All use subject to JSTOR Terms and Conditions

are notefficiently sameconclusion.The ownersofthemonopoly the consumer,not because theyare stupid,but beexploiting governcause the politicalproblemsinvolvedin manipulating mentare great.Any changein the numberof cabs otherthan themwouldrequirethatthe medallionsbe doublingor tripling givenfractionalvalue, unless it were intendedto changethe medallions.The problems ownershipof different proportionate ownerofa medallionsuddenlyfindthat ofhavingtheindividual he had the rightto operate 1.07 cabs (or 0.93 cabs) would of politicalsuccessforsucha probablyprecludeanypossibility sucha proposal costforgetting thetransactions move.Certainly approvedby all the cab ownerswouldbe immense. or thegainsobtainedby theorganizers Note thatfrequently in the even kind arrangement, of cartel a of this beneficiaries period,are quitesmall.On Figure1 whenthecartel transitional in thetotalnumber therewas a sharpreduction was introduced, ofunitsproduced.Clearly,thecost ofreducingcapacityin this and mustbe borneby someone. waycouldbe quiteconsiderable York medallions,theywere In the particular of the New case actuallyprovidedforall existingcabs and, in theshortrun,had verylittleeffect.Later,whenthedemandforcabs wentup,they the growthof the cab business.In orderto achieve restricted it wouldhave been necessaryto dispose of immediateprofits, someproducing assets in orderto getthemonopolygains.This of this type of is, of course, not a special characteristic monopoly.

* To take an exampleof the way thiskindof thingworks,let forthereasonthatBlue Laws are me offermyownexplanation They are supportedby the mercantilecommunity. frequently not,of course,themainforcebehindtheBlue Laws. Supposea whichare community has a certainnumberof supermarkets open sevendaysa week. It is proposedthattheybe requiredby law to close all day Sunday. Merelyhavingthe store open imposescertaincosts thatcannotbe totallyeliminated.There haveto be a certainnumberofpeopleon duty,thelightshaveto have be turnedup to a certainlevel,heatingand airconditioning thanwhenthe storeis closed, to be handleda littledifferently etc. Thus, a reductionin thenumberof hoursthestoreis open does reducethetotalcost. In essence,each sale is accomplished are less in thewayofcostbecausethecustomers withsomewhat compelled to come more closely grouped, and various thatthe economiescan be derivedfromthat.In a way,requiring storesclose downon Sundayis somewhatsimilarin itseffectto of the stores. one-seventh destroying If a privatebusinessacquiredall of the storesin orderto extractmonopolyrents,it wouldfindit necessaryto reduceits it woulddo thisin partby closingdown capacity.Presumably, storesand,to someextent,by reducingthetimetheyare open. combination The privatebusinesscouldseek themostprofitable ofthesetwomethodsofreducingcapacity.Butthisis notreally of supporters availableto the mercantile a politicalopportunity This content downloaded from 134.184.26.108 on Fri, 30 Oct 2015 19:18:51 UTC All use subject to JSTOR Terms and Conditions

4. Blue laws

THE BELL JOURNAL OF ECONOMICS

/ 673

Blue Laws; hence, theirmonopoly will be less efficientthan a private monopoly. The result, then, of forcingthe stores to close on Sunday should be, temporarily,a rise in profits.If they are in a highly competitiveenvironment,this profitwill be completely competed away; but many retail stores are in monopolistic competitiveenvironmentsand this will partiallyprotectthem from that contingency.Still, a good deal of the cost-savingwill inevitablypass on to the consumers,who presumablywould prefer to pay a little more and have the greater convenience of having the stores open all week. Once the Blue Laws have been enacted, however, the stores are operatingat somewhatabove normalreturnon theirassets, and the customersfindthemselvesa littlemore crowded in their shoppingthan theyreally want to be. Under the circumstances, the constructionof new stores is highlylikely. Eventually,then, we will have the situationin which there are more stores than there were before and the returnon all stores is, once again, normal. The customersare not being served quite so well, so we have a social cost; but it is not practicallypossible to put a tax on the customersand use thattax to buy the stores back to their originalposition. On the otherhand, althoughthe store owners are now making normal profits,the repeal of the Blue Laws would be quite inconvenientfor them. With the Blue Laws repeaied, some of them would surely startopening on Sunday and the costs saved by going to a six-day week would be reimposed upon them. They would suffer a considerable transitionalloss.

5. Related phenomena

674

(l llOOCKIC (G. TJI

* For Blue Laws the transitionalgain was extremelysmall. In other cases, there may be quite large initialgains. The organization of a new labor union, which raises wages by 15 percent forits currentemployees,5 does bringa verylarge improvement in the well-beingof the currentlaborers who remainemployed. Essentially,however, this is simplya capital giftto the existing benefitthe new laborers who laborers; it does not significantly will, in the long run,replace themin the industry.Withwages in this industryaveraging 15 percent above wages elsewhere, the numberof people who want to become employees immediately becomes much larger than can be employed. Under the circumstances, some methodof rationingjobs is necessary. If this method is a straighthereditarymethod,i.e., in order to get in, you have to be childrenof the previous membersof the union, then the gain is retained permanentlywithinthese families. More normally,however, the methodof rationingthese high wage jobs is throughthe seniorityclause. Senior employees ;Normally, of course, with some reduction in the total number of employees, so some people get hurteven at that time. 6 In some of the lower paid occupations, the opposite phenomenon is currentlyoccurring.Unemploymentinsurancerates are now highenoughso that there is a desire on the part of the senior employees to take paid vacations periodically,withlthe result of what is called "'reverse seniority."

This content downloaded from 134.184.26.108 on Fri, 30 Oct 2015 19:18:51 UTC All use subject to JSTOR Terms and Conditions

have a prescriptiverightto theirjobs.6 Granted the work force is not entirely stable, this means that the entire burden of unemploymentfalls on the new workers. Since in the early employment years, in any event, the employerwill be restricting and investingin labor-saving devices (because of the higher wages he is paying), this phenomenonmay mean that the new entrantto the union spends a good deal of time as a very low employment. seniorityworker with only intermittent Presumably,the situationis broughtintoequilibriumwiththe discountedlifetimeincome streamforbecominga union member employmentat first (which means a long period of intermittent and then high wage employmentlater) equal to the returnon taking a nonunionized job. When this equilibrium has been reached (and it surelyhas been forall of the older unions), once again we are in a situationwhere no one is betteroffas a result of the existence of the union than he would be had the union never been organized. Unfortunately,there is a deadweight social loss and thereare a numberof people, the membersof the union, who will lose very considerablyfromthe ending of the union. The transitionalgain has been completely "sunk," but the possibilityof large transitionallosses remains. A somewhat similar situation exists with respect to agriculture.The very large gains in value of farmproducts,as a result of the AgriculturalAdjustment Administration'sinitial work, were quickly capitalized into the value of land. Whoever owned farmland at the time the programwas institutedmade immense gains, but these gains were fullycapitalized. We are social now stuck withthe situationin which thereare significant losses, but a powerfulpolitical group would suffertransitional losses by the terminationof the program.No one is gainingnow in the sense that his income is higherthan it would be had the institutionnever been established, except, of course, those people who have received largerinheritancesfromtheirparents than they otherwise would. The higher-than-competitive wages that are now paid in much of the governmentsector are anotherexample. They depend essentially on the fact that families of government employees now make up somethingon the order of one-thirdof all voters.7Governmentemployees use a somewhatodd method of capitalizingthe gains. Suppose some categoryof government employee is paid more than is necessary to attractan adequate numberof suitablyqualifiedpersonnel. There will be an excess of people who want the job. They must be rationed in some way, and the method used by the United States government (and, indeed, by many local governments)is the civil service examination. If the prices being paid are at the marketlevel, then the civil service examination would be more or less unnecessary. The numberof candidates who would turnup forthe 7This estimate is a little shaky, but governmentemployees do make up about 19 percentof all employed persons. They appear to vote about 50 percent employees, and hence the computation more frequentlythan do nongovernment of one-third.For investigationof the evidence on thispoint,see Borcherding[2]. Currently,Richard L. Moss is makinga much more carefulinvestigationof the matteras part of his doctoral dissertationat Virginia Tech.

This content downloaded from 134.184.26.108 on Fri, 30 Oct 2015 19:18:51 UTC All use subject to JSTOR Terms and Conditions

THE BELL JOURNAL OF ECONOMICS

/ 675

jobs would be about thatwhichis needed, and the methodsused in private industryto select suitable employees could be used. On the other hand, if the pay is above that necessary to attract adequate labor, then the civil service exam takes the formof a rationingdevice. The exam is made harder than is necessary or it covers irrelevantmatterssuch as knowledge of American history. As a result, the employees are actually overqualifiedforthejob itself,althoughnot forthe examination. In equilibrium,individuals choosing to take a civil service job would expect about the same discounted lifetimeearningsas if they had taken a job in private industrywhich had an inherent requirementforhigherqualitylabor thanin the civil servicejob. There is a significant social cost in the sense thatthejobs are held by people whose native capacity is such that they could hold betterjobs equally well. For example, the private post officesthat are now so successfully competing with the governmentPostal Service, in those areas where they are legal, normallyhire much lower quality labor and pay a much lower wage than does the Postal Service. It should be emphasized that,althoughthis is true, they also get much more work from theiremployees. A custom of workingonly part of the day has developed in the Postal Service, and many letter carriers are done with their day's work in much less than an eight-hour stint.8

In the particular case of the civil service employees, the initialincrease in theirwages is essentiallya pure transfer.It is only withtime, as the higherwages attracthigherqualitylabor, that a social cost not connected with the transferitselfis generated. With respect to most of the other activities we have discussed, however, the benefitreceived by beneficiariescomes by way of a change in the economy whichis not, in and of itself, a pure transfer.They are given monopolyprivilegesof some sort in most cases. Any economist will, of course, say that pure transfersare betteras a way of transferring money than the establishmentof monopolyprivileges. Indeed, in the case of agriculturalsubsidy programs,it used to be quite common for economists to demonstrate that all the beneficiaries could be given the same benefitfor a small fractionof the cost to the rest of the population. This argumentis a little oversimple because it starts by assumingthe total cost to the transferis simplyits tax rate. Of course, thereis also an excess burdenattachedto each tax, and this excess burden,as in Figure 1, can be substantiallyidentical to the excess burdenof the monopoly. Nevertheless,it must be admittedthat ifyou look over the set of governmentinstitutions which are used to generatespecial benefitsfor special groups, it is usually true that a well designed set of taxes could generate the same revenueforthe beneficiariesat much lower cost to the victims. Why, then,is the less-efficient techniqueadopted? So faras I know, thereare only two explanationsforthis. The first,which 676 / G. TULLOCK

8 "Mail CarrierLikes His Job." WashingtonPost (June 14, 1974), p. A-7.

This content downloaded from 134.184.26.108 on Fri, 30 Oct 2015 19:18:51 UTC All use subject to JSTOR Terms and Conditions

is fairlyold, is based on the information cost.9 Not to repeat here the fulldetails of the argument,'0 it is simplythatthe pushingthrough ofsucha benefit fora specialclass requiresthat the cost of the benefitnot be obviousto the verymuchlarger collectionof voterswho will be injuredby it. This, in turn, requiresa certaindegreeofcomplexity in thesubsidyand direct raisedoutofdirecttaxeswouldnormally cash payment notmeet thatrequirement. In additionto thisargument, Rubinhas recentlysuggested another." He points out that budgetaryallocationsare reconsideredevery year in almost all countries.On the other hand,once an institution has been set up, it is notautomatically reexamined on a regularbasis, nordoes itscontinuedexistence requirea positiveaffirmative votein thelegislature. Underthe was set in handto giveme $1 circumstances, then,ifa program milliona yearoutofthe Treasury,I wouldhave to facea series of annualvotes on that$1 million.If, on the otherhand,the laws wererearranged in sucha waythatI gota monopolyworth $1 milliona year,thismonopolywouldremainin existence(and probablylargelyunnoticed) untilsuchtimeas positiveeffort was made to terminate it. In the firstcase, I wouldhave to have to thelobbyingeffort returned everyyear;whereasin the second case, once I get it, I keep it untilsomething untoward happens. Rubinpointsout thatunderthese circumstances pressure groupswouldpreferthetypeof preferential benefit whichdoes not requirecontinuousbudgetaryappropriations, simplybecause it is cheaper.Theyonlyhave to makeone lobbying effort and thenperhapsa defenseseven or eightyearslater,if the issue comes up forfurther discussion.This is quite different fromrunning a continuallobbyto raise theissue everyyearas partof the budgetprocess. Whetherone or the other(or both)of theseexplanationsis correct,or whetherthereis a thirdwhicheitherreplacesor supplementsthese, is not of vital importanceto our present concerns.It is certainly is very truethatthistypeof institution widelyfoundin oursocietyandthesocial costis great.It is also truethat,ingeneral,thebenefits are nowlongin thepast. They were transitional was first benefitsat the timethe institution founded.As ofnow,thereis no one whois positively benefiting fromthe organizationand thereis a large dead-weight loss. However,thereis a largenumberof people who wouldsuffer costs iftheinstitution These largetransitional wereterminated. transitionalcosts in many cases are large enough so that oftheloserswouldimposeuponsocietyan excess compensation burdenwhichwouldbe of the same orderof magnitude as the cost of the presentinstitution. * It is hardforan economistto recommend anypositiveaction to deal withthiskindof situation.It is, as thetitleofthisarticle

6. Summary and conclusion

"

See Tullock [7]. 0 Ibid., pp. 103-106. " See Rubin [5].

This content downloaded from 134.184.26.108 on Fri, 30 Oct 2015 19:18:51 UTC All use subject to JSTOR Terms and Conditions

THE BELL JOURNAL OF ECONOMICS

/ 677

suggests,a trap. I can recommendvery stronglythat we tryto avoid gettinginto such traps in the future,but what about the ones into which we have already fallen? In those cases where there are efficiencygains from reorganizingthe industry,we could presumablycompensate the presentbeneficiaries;but the politicalpossibilitiesseem to me to be verysmall. In those cases where the excess burden on the necessary tax forcompensation would be as greatas the dead-weightloss, this alternativeis not available even in theory. Granted the omnipresence of institutionsof this sort and theirvery large dead-weightloss, it is conceivable that simultaneouslyabolishing all of them would lead to a net gain foralmost everyone. The individualwould lose his particular privilege, but would gain from the loss of privileges of other people. It is doubtfulthat such a change would be trulyPareto optimal,but it mightcome close. As to its political practicality,I take it I do not have to explain why I thinkit is low. The moral of this, on the whole, depressingtale is that we should tryto avoid gettinginto this kind of trap in the future. Our predecessors have made bad mistakes and we are stuck with them, but we can at least make effortsto prevent our descendants from having even more such dead-weightlosses inflictedupon them. References 1. AUSTER, R. D. "The GPITPC and InstitutionalEntropy." PutblicChoice, Vol. 19 (Fall 1974), pp. 77-83. Durham, N.C.: Duke 2. BORCHERDING, T. E., ED. Budgets and Butreautcrats. Univ. Press, forthcoming. 3. BUCHANAN, J. M. AND TULLOCK, G. 'The Dead Hand of Monopoly." AntitrustLaw and Economic Review, Vol. I (Summer 1968), pp. 85-96. 4. PELTZMAN? S. "An Evaluation of Consumer ProtectionLegislation: The 1962 Drug Amendments." Joiurnal of Political Economy, Vol. 81 (September/October1973). pp. 1049-1091. 5. RIJBIN, P. H. "On the Form of Special Interest Legislation." Plublic Choice, Vol. 21 (Spring 1975). pp. 79-90. 6. STIGLER, G. J. "Capitalism and Monopolistic Competition:I. The Theory of Oligopoly-Monopoly and Oligopoly by Merger." The American Economic Review, Vol. 40, No. 2 (May 1950), pp. 23-37. 7. TULLOCK, G. Towvarda Mathematics of Politics. Ann Arbor: Universityof Michigan Press, 1967.

678 / G. TULLOCK

This content downloaded from 134.184.26.108 on Fri, 30 Oct 2015 19:18:51 UTC All use subject to JSTOR Terms and Conditions