Foreward RICHARD T. ELY LECTURE The Economics of Governance

Foreward

By OLIVER E. WILLIAMSON*

RICHARD T. ELY LECTURE

selected economics, business/management, sociology/orga-nization, and political science journals were 60 vs. I, 76 vs.4, 79 vs. 18, and 60 vs. 25, where the journals surveyedwere: the American Economic Review, Journal 01 PoliricalEconomy, Quanerly Journal 01Economics, Rand Journal 01Economics, and Journal 01 Economic Perspectives in eco-nomics; Straregic Management Journal, Managemenr Sci-ence, Academy 01 Management Journal, and Academy 01Managemenr Review in business/management; Administra-tive Science Quanerly, Organization Science, AmericanJournal 01 Sociology, American Sociological Review, andAnnual Review 01 Sociology in sociology/organization; andAmerican Polirical Science Review, Political Science Quar-terly, Journal 01 Politics, and Political Research Quarterlyin political science. Combining these four categories, arti-cles using the word governance increased from 48 to 275over this 20-year interval. Dixit (2004 pp. 149-50) reportsthat the number of web pages that turn up under the searchfor "governance" is huge.

choke (prices and output, supply and demand),the economics of governance is a lens of con-tract construction, broadly in tlte spirit of JamesBuchanan's (2001 p. 29) observation that"rnutuality of advantage from voluntary ex-change ... is the most fundamental of all un-derstandings in economics."

The economics of governance, as herein de-scribed, is principally an exercise in bilateralprivate ordering, by which I mean tltat the im-mediate parties to an exchange are actively in-volved in tlte provision of good order andworkable arrangements. To be sure, the need forprivate ordering varies with the rules of thegame as provided by tlte state. Distinctions be-tween lawlessness where the state provides lim-ited or unreliable protection for property andcontract (Dixit, 2004) and lawfulness, where thestate undertakes to protect property and enforcecontracts in a principled way, are pertinent. Thefirst of these applies mainly to primitive andtransition economies. The second is commonlyassociated witlt Western dernocracies.

Recourse to private ordering under condi-tions of lawlessness is altogetlter understand-able: given tlte absence of state support, the

The Economics of Governance

• Walter A. Haas School of Business, University ofCalifornia, Berkeley, CA 94720-1900. The paper has ben-efited from workshop presentations at the University ofCalifornia-Berkeley, the University of Valencia, INSEAD(Fountainebleau), and the 2004 annual conference of theInternationalSociety for New Institutional Economics. Anabbreviated version was also given as the Horst ClausRecktenwaldLeclUre at Nuremburg on 4 November 2004.Cornmentsand suggestions from Fred Balderston, ErnestoDalbo, Avinash Dixit, Rohert Gibbons, Witold Henisz, IanLarkin,Steven Tadelis, and Dean Williamson are especiallyacknowledged.

I Lon FulIer's (1954 p. 477) definition of "eunomics" as"the science, theory, or study of good order and workablearrangements"is very much in the spirit of what I refer to asgovernance.

2 One of the immediate rarnificationsof insistently com-paring feasible alternatives. all of which are fiawed, is thatthe purported inefficiencies that are ascribed to failures toachieve"first best" optimality are not dispositive. but invitethe query "As compared withwhat?" I return to this issue inSection IV.. 3 EXcluding"corporate governance," the nurnbers of ar-

lIcles that used the word "governance" during the period1998-2000 as compared with the period 1977-1979 in

The economics of governance is an effort to.'·1 ent the "study of good order and work-~%Fe:angements," where good order inc!ud~sboth spontaneous order in the mark:t, Wh1Ch1S

venerated tradition in econOffilCS (Adam

Sa ·th 1776' Friedrich Hayek, 1945; Kennethm1, , d'A. Arrow and Gerard Deb~eu, 195~), an mten-tional order, of a "consc10us, dehberate, pur-poseful" kind (Chester Irving Barnard, 1938 p.9).1 Also, I interpret workable m:an~ements to

ean feasible modes of orgarnzatlOn, all of:hich are flawed in comparison with a hypo-tltetical ideal (Avinash Dixit, 1996 pp. 4-9).2The object is to work out tlte effici~ncy logic formanaging transactions by alternative mode~ ofgovemance-principally spot markets, vanouslong-term contracts (hybrids), and hierarchies.

Interest among soda! scientists, economistsincluded, in tlte study and practice of good orderand workable arrangements has been steadilygrowing? In contrast with the orthodox lens of

ics), and David Wise (Harvard, health eConics). I think the program committee didorn•exemplary job, and I want to thank them f,antheir careful reviews, good advice, and time~~ness. I wish to particularly commemorate Ben:eSaffran, who died in December, 2004; he Was amajor contributor to the Association, through biscolumn in the Journal of Economic Perspectivesas weil as his work on this committee.

Also included in this volume is the Ely Lec-ture, delivered by Oliver Williamson, whosecontributions to the study of internal control inorganizations have broken down the treatmentof firms as "black boxes," and shown that theproblems of information, incentives, and mech-anism design in markets have analogs in Com-mand and control structures.

The process of organizing the meetings iscomplex and extensive, and progressedsmoothly because of the talent and experienceof Violet Sikes, Gwyn Loftis, and MarleneHight of the AEA office in Nashviile. Alsocritkal to the process was the handling of cor-respondence by Rowilma Balza, who showedcharacteristic care, intelligence, and efficiency.My thanks to John Rust, who provided com-puter programs that we used to organize andtrack submissions. All of the tasks of thePresident-elect (and ofthe President) are greatlyeased by the efficiency and good sense of JohnSiegfried, to whom the Association is indebted.This volume was edited, under difficult circum-stances, by David Baldwin and Ronald Oaxaca.

DANIEL L. McFADDEN

viii

Organizing the program for the annual meet-ings is by far the most important job of thePresident-elect. The job is like that of a journaleditor, in that the actual content of the programcomes from submissions by members, but theimportance given to various research and policyareas by program committee members and theirreviews of submissions are critical. In additionto organizing the usual program of oral sessionsand selecting sessions for inclusion in this vol-urne, committee members assisted me in settingup aseries of tutorials on research and policytopics that gave the AEA membership over-views of subjects outside their specialties, andposter sessions that gave more economists op-portunities to present their research and meetpeers working on similar problems. My pro-gram committee was broadly representative ofthe profession; its members were Alan Auer-bach (UC Berkeley, public finance), RebeccaBlank (Michigan, labor and family), PhilipCook (Duke, public policy), Francis Diebold(Pennsylvania, macro), Henry Farber (Prince-ton, labor), Robert Frank (Cornell, consumerbehavior), Don Fullerton (Texas, environment),Robert Hall (Stanford, macro), Andrew Lo(MIT, finance), Charles Manski (Northwestern,public policy), John Quigley (UC Berkeley,public policy and housing), Robert Porter(Northwestern, industrial organization), KennethRogoff (Harvard, trade), Alvin Roth (Harvard,experimental economics), Daniel Rubinfeld(UC Berkeley, law and economics), BernardSaffran (Swarthrnore, teaching of economics),Jonathan Skinner (Dartmouth, health econom-

AEA PAPERS AND PROCEEDINGS RICHARD T. ELY LECTURE2

parties have no choice but to do their best tocreate their own bilateral or group mechanismsto support otherwise problematic exchange. Butwherein does the need for private ordering ariseif the state has created and enforces efficaciousrules of law, as assumed by the "legal central-ism" tradition?

Private ordering here resides in the inherentlirnitations oflegal centralism. Albeit an analyt-ical convenience, for both law and economics,to assurne that "disputes require 'access' to aforum extemal to the original setting of thedispute [and that] remedies will be provided insome body of authoritative learning and dis-pensed by experts who operate under the aus-pices of the state" (Marc Galanter, 1981 p. I),the facts disclose otherwise. Most disputes, in-cluding those that under current mIes could bebrought to a court, are resolved by avoidance,self-help, and the like (Galanter, 1981 p. 2).That is because in "many instances the partici-pants can devise more satisfactory solutionsto their disputes than can professionals con-strained to apply general mIes on the basis oflirnited knowledge of the dispute" (Galanter,1981 p. 4).

The upshot is that private ordering is centralto the performance of an economy whatever theconditions of lawfulness. Adaptation is taken tobe the central problem of economic organiza-tion. Conditional on the attributes of the activ-ities (transactions) to be organized, I focus onthe comparative efficacy with which alternativemodes of govemance effect good order (adap-tation) during the ex post contract implementa-tion interval.

Section I sketches the background out ofwhich the econornics of governance works. Sec-tion TI sets out the basic logic of efficient gov-ernance with respect to the puzzle of verticalintegration and more generally. Microanalyticfoundations-from law, economics, and organi-zation theory-on which the economics of gov-ernance rests are examined in Section III.Applications are addressed in Section IV. Issuesof lawlessness are briefly discussed in SectionV. Concluding remarks follow.

I. Background

I begin with a sketch of the governance ofongoing contractual relations. The four concep-tual corners tones out of which the lawfulness

MAY 200S

branch of the economics of governance Wo ksare described next. I then turn to the criSisr.public policy toward business during the 196~nthat gave impetus to a rethinking of the pur~poses .se~ed by complex contract and econOmicorganlzatlOn.

As against simple market exchange, goVer-nance is predominantly concerned with Ongoincontractual relationsfor which continuity of th;relationship is a source of value. Given thatincomplete contracts need to be adapted to dis-turbances for which contractual provision Wasnot made or was incorrectly made at the Outsetcontinuity can and will benefit from a spirit oicooperation. But therein lies the mb: continuitycan be put in jeopardy by defecting from thespirit of cooperation and reverting to the letter.Maladaptation to disturbances is where themain costs of govemance reside.4

Taken by itself, the possibility of defection isbad news. But there is also an upside: contrac-tual hazards, like other costs, invite mitigation-which is where the real challenge and analyticalimport of potential breakdown resides. Uponlooking ahead and recognizing that possiblebreakdowns are in prospect, cost-effective pri-vate ordering mechanisms that have the purposeand effect of mitigating contractual hazards willbe devised, thereby better to assure that mutualgains from trade are realized.

The 32-year coal supply contract between theNevada Power Company and the NorthwestTrading Company is an example. Out of aware-ness that disturbances could lead to conflict andpossible breakdown, this contract provided inpart that "In the event an inequitable conditionoccurs which adversely affects one Party, itshall then be the joint and equal responsibilityof both parties to act promptly and in good faithto determine the action required to cure or ad-just for the inequity and effectively implementsuch action." It furthermore instructed "TheParty claiming the inequity shall include in itsclaim such information and data as may be

4 The economies of information also deals with contrac-tual hazards, but mainly of a different kind than those dealtwith here. Thus, whereas insurance is the paradigm problemfor the economies of information, vertical integration is theparadigm problem for govemance. Also. law, organizationtheory, adaptation, and transaction costs all figure moreprominently in studying the govemance of contractualrelations.

vaL. 95 NO. 2

bly necessary to substantiate the claimreasona fu . h hand shall freely and without delay rms sucother information and data as the other Party

bly may deern relevant and necessary. Ifreasona ·th· .parties cannot reach agreement WI III Slxty~~) days the matter shall be submitted to( ."arbitratlOn.

Plainly, the parties to this contra~t weree that things could get out of alignment

awar . d 'd d fduring contract e~ecullo~ an proVI e a rarne-k for correcllve action. Very general lan-wor . . Ige notwithstandlllg, the part1es were a so

r:d-headed. The party reques~ng .correcti~nswas expected to provide SUppOrtlllglllformallon

d data to substantiate the request. And so as:tter to assure that confiicts would be resolvedknowledgeably by a specialist in the industry,provision was made for arbitration in the eventthe parties could not reach agreement. Althoug~it boggles the mind that "reasonably clever bUSI-nessmen and lawyers cope with problems schol-ars might consider intractable" (Victor P.Goldberg and John R. Erickson, 1987 p. 369),evidently contract practitioners can and do de-sign workable order-preserving mechanisms foradapting to disturbances in the service of mutualgains. Some scholars, moreover, had been forg-ing the relevant concepts. The four conceptualcomerstones out of which the economics ofgovemance works are governance, trans actioncosts, adaptation, and interdisciplinary socialscience.

Govemance.-The srudy of govemance wasprefigured by John R. Commons, who was oneof the leaders of older-style institutional eco-nomics in the United States. Of the many goodideas that originated with Commons, none wasmore important to the economics of governancethan his abiding interest in "going concerns."As against the preoccupation of orthodoxy withsimple market exchange and the resource allo-cation paradigm (M. Reder, 1999), Commonsobserved that the continuity of an exchangerelationship was often important, whereuponthe problem of economic organization was re-formulated as follows: "the ultimate unit ofactivity ... must contain in itself the three prin-ciples of conflict, mutuality, and order. This unitis a transaction" (Commons, 1932 p. 4). Com-mons thereafter (1950 p. 21) recommended that"theories of economics center on transactionsand working mies, on problems of organization,

3

and on the .., [ways] the organization of activityis ... stabilized."

A coherent theory of organization for imple-menting these novel ideas nevertheless eludedCommons and his followers, possibly becausethe concept of transaction cost had yet to sur-face and because of the primitive state of orga-nization theory at the time.5 Such a fatenotwithstanding, the Commons triple of con-flict, mutuality, and order anchors the conceptof govemance as herein employed, in that gov-emance is the means by which to infuse order,thereby to mitigate conflict and realize mutualgains. The trans action, moreover, is made thebasic unit of analysis.

TransactionCosts.-As Ronald Coase (1937)developed in bis article on "The Nature of theFirm," the standard assumption that transactioncosts were zero presented neoclassical econom-ics with a logical lapse. Thus whereas ortho-doxy took the distribution of economic activityacross firm and market organization as given,whereupon attention was focused on "the eco-nomic system as being coordinated by the pricesystem" (Coase, 1937 p. 387), firm and marketare properly regarded as "altemative methodsof coordinating production" (1937 p. 388 [em-phasis added)). Rather than take the distributionof economic activity as given, this should bederived. Coase's 1937 paper thus took as itspurpose "to bridge what appears to be a gap ineconomic theory .... We have to explain thebasis on which, inpractice, this choice betweenalternatives is effected" (Coase, 1937 p. 389[emphasis added)).

Orthodoxy remained unmoved over the next35 years, but pressures were mounting as in-terim developments in the market-failure liter-ature revealed that disregard for positivetransaction costs was responsible for confusionover extemalities and other puzzling practices.Upon reformulating the tort problem (or, more

5 Comrnons tumed instead to W. N. Hohfeld's system of"fundamental legal concepts" to implement his ideas. (Forhis reliance on Hohfeld, see especially the extended foot-note in Comrnons [1968 p. 91].) Tbe resulting effort tointerpret transactions and ongoing concerns with the use ofjuridical reasoning resulted in an elaborate taxonomy (Com-mons, 1968 pp. 90-142), but a predictive theory of contractand organization and a follow-on empirical research agendadid not materialize.

AEA PAPERS AND PROCEEDINGS

/

5

ll. The Basic Logic

8 The quoted language is that of the then head of theAntitrust Division, Donald Turner (see Alan Meese, 2004p. 47). Because durable market power is the exceptionrather than the rule, the economics of govemance, like thecommon law, works out of the rebuttable presurnption thatvoluntary exchange serves affirmative economic purposes.

9 United States v. Von's Grocery Co. 384 V.S. 270, 301(1966) (Stewart J. dissenting).

A stripped down version of the basic logic isset out here, first with reference to vertical in-tegration, which would become the paradigmproblem for the economics of governance, andthereafter by formulating the discriminatingalignment hypothesis. Additional rnicroanalyticsupports on which the arguments rest are takenup in Section III.

The modeling precept "keep it simple" (Rob-ert Solow, 2001 p. 111) has its origins in com-plexity (Herbert Simon, 1957b p. 89; E. Wilson,1999 p. 183). Choosing vertical integration asthe specific phenomenon on which to hone inwas both a simplifying move and the obviousplace to start. Not only was vertical integration(the make-or-buy decision) the puzzle to whichCoase referred in his 1937 article, but it re-mained a puzzle, in a zero-transaction-cost

"not hospitably in the comrnon law tradition,but inhospitably in the tradition of antitrust.',gSuch convoluted reasoning carried over tomergers as wen, where Justice Potter Stewartobserved, in a dissenting opinion in 1966, thatthe "sole consistency that 1 can find in the[merger] litigation under Section 7 Cis that] theGovernment always wins.',9 Failures to con-nect with the real purposes served by contractrestrictions and the inner workings of realfirms could often, evidently, have overreachingconsequences.

Confusion, moreover, also reigned in the reg-ulatory area, in large measure because regula-tory issues were treated in a one-sided way.Thus, whereas there was an extensive literatureon market failure, there was no mention of,much less a corresponding literature on, regu-latory failure (Coase, 1964). Taken together,antitrust and regulatory policies toward busi-ness were careening out of contro!. The need forother perspectives, possibly of a comparativecontractual kind, came knocking.


To these four cornerstones I would add thatthe developing crisis in public policy towardbusiness during the 1960s provided added im-petus and urgency to the economics of gover-nance. Victor Fuchs (1972 p. xv), in hisForeword to Policy Issues and Research Oppor-tunities in Industrial Organization, pronouncedthat "all is not weIl in this once flourishingfield." Various reasons can be advanced, amongwhich is that economic organization is muchmore complex than was widely appreciated.Coase's explanation was that industrial organi-zation had been become an exercise in appliedprice theory (Coase, 1972 pp. 60-62), to whichHarold Demsetz's (1983 p. 377) observationthat the neoclassical theory of the firm is not anall-purpose construction is also pertinent: It is a"rnistake to confuse the firm of [neoclassical]econornic theory with its real-world namesake.The chief mission of neoclassical economics isto understand how the price system coordinatesthe use of resources, not the inner workings ofreal firms."

My prior studies of organization theory andrny experience as Special Economic Assistantto the head of the Antitrust Division of the U.S.Department of Justice during 1966-1967 reso-nated with these concems. Thus, although ap-plied price theory was one usefullens, uncritical~pplication of such reasoning led to a presump-hon that nonstandard and unfamiliar forms ofCOntract and organization had monopoly pur-pose and effect-as witness the inhospitalitytradition in anti trust, according to which non-standard (complex) practices were interpreted

vaL 95 NO. 2

lens of contract, as against the lens of choice, ismade the cutting edge.

Interdisciplinary Social Science.- This is thelast of the four cornerstones that I ass.ociate wi~the economics of govemance. As discussed InSection m, both organization theory and aspectsf the law (especially contract law) bear impor-

~tly on the economics of govemance. I merelyobserve here that doing interdisciplinary socialscience is demanding and that the teaching andresearch prograrn in economics, organization the-ory, and operations research at the GraduateSchool of Industrial Administration, Camegie-MeUon University, during the 1950s and 1960shelped to open the door to a disciplined approachto interdisciplinary social science.

MAY 2005

7 lnterestingly, Jean-Jacques Laffont and David Marti-mort (2002 p. 11) credit Barnard as "the first to define ageneral theory of incentives in management," where theyinterpret Bamard's views as broadly in the spirit of theirown agency theory work. Laffont and Martimort (2002 p.13) also write that "Barnard recognized that incentive con-lracts do not roJe all of the activities within an organiza-tion. "ln particular, "the incompleteness of contracts and thebounded rationality of members of the organization reqmrethat some leaders be given authority," presumably 10 exer·eise ex post governance. Bamard advanced prescient ideasthat were pertinent to both agency theory and the econom1CS

of governance.

Adaptation.-The economies of govemanlocates the basic action in the differential capac~ities of alternative modes of govemance ~effect adaptation. Interestingly, both the econo~mist Friedrich Hayek and the organization the-orist Chester Bamard were in agreement thatadaptation is the central problem of econornicorganization. Hayek (1945 pp. 526-27) focusedon the adaptations of economic actors who ad-just spontaneously to changes in the market.Upon looking "at the price system as '" a mech-anism for comrnunicating information," themarvel of the market resides in "how little theindividual participants need to know to be ableto take the right action." By contrast, Barnard(1938 p. 9) featured coordinated adaptationamong economic actors working through ad-ministration (hierarchy). The latter is accom-plished not spontaneously but in a "consciousdeliberate, purposeful" way with the use ofadministration.

In effect, the adaptations to which Hayekrefers are autonomaus adaptations accom-plished in the market, whereas the adaptationsof concem to Bamard are consciously coordi-nated adaptations accomplished through the useof management within the firm. To the widelycelebrated "marvel of the market" (Hayek) isnow therefore joined the hitherto scomed "mar-vel of hierarchy" (Bamard)? Because efficiencyis the product of adaptive capacities of bothkinds, an understanding and appreciation forboth markets and hierarchies (rather than themore familiar dichotomy between markets orhierarchies) is needed. The firm for these pur-poses is described not as a production function(which is a technological construction), but as agovernance structure (which is an organiza-tional construction). And the market is de-scribed as an organizational alternative. The

4

6 Coase (1937 p. 391) argued that ''The main reason whyit is profitable to establish a firm would seem to be that thereis a cost of using the price mechanism, the most obvious'" [beingl that of discovering what the relevant prices are."Although this sounds plausible, the price discovery burdenthat Coase ascribes to the market does not survive compar-ative institutional scrutiny (Williarnson, 2002 pp. 179-80).

generally, the extemality problem) in contrac-tual terms, Coase (1960) showed in his paper on"The Problem of Social Cost" that the external-ity problem vanished if the logic of zero trans-action costs was taken to completion. Plainly,provision for positive trans action costs wouldthereafter have to be made if extemalities, andthe study of complex contracting more gen-erally, were to be accurately described andassessed.

Kenneth Arrow's (1969) examination of"The Organization of Econornic Activity: Is-sues Pertinent to the Choice of Market versusNon-market Allocation" likewise made a prom-inent place for trans action costs, both in generaland with respect to vertical integration. Thegeneral argument is that "market failure is notabsolute; it is better to consider a broader cate-gory, that of transaction costs, which in generalimpede and in particular cases completely blockthe formation of markets" (Arrow, 1969 p. 48).Arrow's remarks about vertical integration areespecially pertinent: "An incentive for verticalintegration is replacement of the costs of buyingand selling on the market by the costs of intra-firm transfers; the existence of vertical integra-tion may suggest that the costs of operatingcompetitive markets are not zero, as is usuallyassumed by our theoretical analysis" (1969 p.48 [emphasis added]).

The time was ripe for the concerted study ofpositive trans action costs, yet obstacles re-mained. For one thing, the concept of transac-tion costs lacked definition. Being a vague andmalleable concept, transaction costs came to beinvoked as an all-purpose explanation for puz-zling practices, whereupon the concept of trans-action cost acquired a "well-deserved badname" (Stanley Fischer, 1977 p. 322). Relat-edly, transaction cost is an expansive concept.Of the variety of ways in which transactioncosts can manifest themselves (of which searchcost was Coase's candidate),6 where does themain comparative institutional action reside?


10 As discussed herein, it will sirnplify to take the at-tributes of transactions as given and deploy governancestructures in relation to these. In fact, the attributes oftransactions and of govemance structures are chosen simul-taneously (Michael Riordan and Williarnson, 1985). Tbebasic regularities of the sirnplified setup nevertheless carryover, although some nuances also appear, when simultane-ity is introduced.

The upshot is that there is a place for eachgeneric mode of organization, yet each shouldbe kept in its place.

The critical attributes of transactions and thesets of concurrent relations among attributesthat define viable modes of governance are dis-cussed elsewhere (Williamson, 1979, 1988,1991b). I merely offer some summary com-ments on each here.

There is general agreement that asset speei-fieity, uncertainty, and frequency are relevantdimensions for describing transactions. Al-though much of the explanatory power of thetheory turns on asset specificity (Williamson,1971, 1975, 1985; Benjamin Klein et al., 1978),which gives rise to bilateral dependency (or theabsence thereof), bilateral dependency by itselfwould not pose a problem but for maladapta-

7

Herbert Simon (1985 p. 303) advised soeialseientists that "Nothing is more fundamental insetting our research agenda and infonning ourresearch methods than our view of the nature ofthe human beings whose behavior we are study-ing." The two attributes ofhuman actors that areespeeially relevant to the economics of gover-nance are cognition and self-interestedness.

A. Affirmative Supports

all three attributes and works weIl, but not sur-passingly weIl, in both autonomous and coordi-nated adaptation respects. The viability of thehybrid turns cmeially on the efficacy of crediblecommitments (penalties for premature tennina-tion, information-disclosure and verificationmechanisms, speeialized dispute settlement,and the like), the cost effectiveness of whichvaries with the attributes of transactions (Wil-liamson, 1991b; Claude Menard, 2004).

Although such differences among modesmay now appear to be "obvious," it was notalways so (A. Alchian and Demsetz, 1972 p. 777).

11 Arrow (1987 p. 734) likewise emphasizes that "nano-economic reasoning" is a distinguishing fealure of the newinstitutional economics.

m. Microanalytic Supports

The overarching logic of efficient alignmentis as described above. Somewhat more tedious,but vital to the exercise, are the microanalyticmechanisms described here. Indeed, I conjec-ture that David Kreps's (1999 p. 122) remarkthat "game theory ... has more to leam fromtrans action cost economics than it will haveto give, at least initially" tums as much onthe microanalytic mechanisms as on the basiclogic.11 Discriminating alignment is easy to im-plement for those who know the basic logic, butan understanding of economic organization re-quires the student of economic organization toprobe deeper.

Mechanisms of affirmative and contestedkinds are distinguished. Human actors, inter-temporal process transformations, and contractlaw in practice are all affirmative supports, inthat the shape of private ordering governancefor ongoing contractual relations is significantlyinfluenced by each.


. between the parties to an incomplete con-uons .t that are induced by disturbances. Indeed,

thtraCroblem of contracting under fuIly station-ep .' "On! h thconditions is urnnterestIng: y w en eary d d . ..eed to make unprogramme a aptatIons IS m-

~oduced does the market ve~su~,inte~~ orga-nization issue become engagmg (WI1hamson,1971 p. 113). Uncertainty ~s ~e sour~e of dis-turbances to which ~daptatIon IS requrred. ~re-quency is relevant m two respects: reputatIoneffects and setup costs, the net effects of whichwill vary with the particulars.

Asset speeificity in conjunction with distur-bances is where the main predictive action re-sides. Sometimes asset speeificity can be tracedto non-redeployable durable investments that aremade imrnediately upon signing the contract.Hut asset specificity also evolves during con-tract implementation. As discussed in SectionmoA, such transactions undergo a fundamentaltransformation, in that, even though a largenumber of suppliers may have been on a parityat the outset, a bilateral dependency conditionbetween the buyer and initial winning biddersets in during contract implementation and atthe contract renewal interval. Because transac-tion speeific assets can be redeployed to alter-native uses and users only at a loss ofproductive value, continuity for such exchangerelations is important.

Tbe economics of governance makes threebasic govemance structure distinctions: classi-cal markets (simple spot-market exchange), hy-brid contracting (of a long-term kind), andhierarchies (firms, bureaus). The key features ofgovemance (differential incentive intensity, ad-ministrative control, and contract law regime)are postulated to vary among modes in in-ternaIly consistent ways. Different attributecombinations give rise to distinctive adaptivestrengths and weaknesses. Speeifically, the mar-ket mode works out of high-powered incentives,little administrative control, and a legal-mIescontract-law regime, which is weIl suited toimplement autonomous adaptations but poorlysuited to effect cooperative adaptations. The setof complementary attributes that describes hier-archy is antipodal to the market mode (in that~erarchy uses low-powered incentives and con-slderable administrative control, and the courtsare deferential), which reverses these adaptivecapabilities. The hybrid is a compromise modethat is located between market and hierarchy on

differences must be named and their .. k d ranufi_catIons wor e out. The economics of gd b· OVe~

nance respon s y namrng asset spec'fi .(which can take a variety of forms) ~ City

. d fre ' neer-~ty, ~ quenc~ ~ three of the criticaldimensIOns for descnbmg transactions.If the comparative efficacy of differentmodes of governance (market, hybrid hi

h b· , er-ar~. y, pu l~c bureau, etc.) differ, then thecntIcal attnbutes that describe alternativemodes of governance need to be nam dand the internally consistent syndromes e fattributes that define viable modes ne~~o be ~orked. ou~. Rel.evant dimensionsmclude mcentIve mtenslty, administrativecontrol, and contract law regime, the com-plementary relations among which are de-scribed below.A predictive theory of economic organi-zation resides in the hypothesis that trans-actions, which differ in their attributesare aligned with governance stmctures'which differ in their costs and competen~eies, so as to effect a (mainly) transaction-cost-economizing result.lo

(3)

(2)

(1) If some transactions are simple and othersare complex, then the attributes of trans-actions that are responsible for these

6

world, thereafter (Arrow, 1969). Also, verticalintegration was an unsettled issue in antitmstenforcement. As compared with other candidatetransactions (such as the employment relationor final product market transactions), verticalintegration has the advantage of being simpler,in that a variety of complications that arise intransactions between firms and workers or be-tween firms and consumers (such as disparitiesof information, differential access to technicaland legal expertise, differential capacity to bearrisk, and the like) are of lesser importance intransactions between firms, where the special-ization of labor within and between functions isextensive. Accordingly, the action in intermedi-ate product market transactions resides moreassuredly in the attributes of transactions inrelation to the properties of alternative modes ofgovernance.

My 1971 paper on "The Vertical Integrationof Production: Market Failure Considerations"drew on aIl of the foregoing and more. Keyfeatures of this article that would find their wayinto the economics of governance included: (1)focusing on a specific phenomenon in compar-ative contractual terms; (2) taking adaptation todisturbances to be the central problem of eco-nomic organization; (3) ascribing contractualincompleteness to bounds in rationality, anddefection hazards to opportunism; (4) tracingbilateral dependency contractual hazards to in-tertemporal transformations in the exchange re-lationship; and (5) recognizing that markets andhierarchies differ in kind, in that each possessesdistinctive strengths and weaknesses, where hi-erarchy enjoys the advantage for managingcooperative adaptations, and the market for au-tonornous adaptations.

As compared with price-theoretic and tech-nological explanations for vertical integration,the cornparative contractual approach locatesthe action in the attributes of transactions, thedifferential capacities of alternative modes ofgovernance to implement autonomous and co-ordinated adaptations, and the efficient align-ment thereof. This applies not only to verticalintegration, but to economic organization moregenerally. The discriminating alignment hy-pothesis is this:


Simon took early exception with the idea thathuman actors are supremely rational and pro-posed instead that human actors be described asboundedly rational, by which he means thatthey are "intendedly rational, but only lirnitedlyso" (Sirnon, 1957a p. xxiv). Human actors arethus neither nonrational nor irrational, but areattempting effectively to cope.

It is Reinhart Selten's (2001 p. 15) view thatbounded rationality is an encompassing conceptthat "cannot be precisely defined. It is a problemthat needs to be explored." I concur. With ref-erence to complex board games (such as chess),bounds on rationality are relieved by heuristics;in the context of search it is manifested assatisficing; my interest is in contracting, wherethe main lesson of bounded rationality is that altcomplex contracts are unavoidably incomplete.

Contractual incompleteness notwithstanding,the economics of govemance also assurnes thatparties to a long-term contract possess "feasibleforesight," by which I mean that they have thecapacity to look ahead, uncover possible haz-ards, and work out the rarnifications, thereuponto incorporate hazard-mitigating mechanismswithin the ex ante contractual agreement,broadly in the spirit of Robert Michels' (1962)classic study of oligarchy. Thus whereas Mich-els observed that democratic structures can beand are subverted by oligarchy, he did not onthat account give up on democracy. Rather thelesson is that "nothing but aserene and frankexarnination of the oligarchical dangers ofdemocracy will enable us to minimize thesedangers" (Michels, 1962 p. 370)-whereuponhazard mitigation can be introduced in cost-effective degree.

The subversion of contracts and organizationraises the issue of how self-interest is to bedescribed. Simon's (1985 p. 303) candidate is"frailty of motive," which is a relatively benign(nonstrategic) construction. The proposition, forexample, that routines describe the behavior ofmost individuals most of the time contemplatesbenign behavior. But while most people will dowhat they say (and some will do more) most ofthe time, much of what is interesting abouthuman behavior in general and in organizationsin particular has reference not to routines, but toexceptions.

Exceptions pose strains when parties to along-term contract perceive that individual ad-vantages can be realized by defecting from the

MAY 2005

spirit of cooperation and reverting to the lettof the contract. The general proposition here ~rthat when the gains to be had by insistence u~sthe literal enforcement of the contract exce ~the ~scou.nted valu~ of continuing the exchan~erelationship, defection from the spirit of coo _eration can be anticipated (Williamson, 199 fbp. 273). In that event, even if most people willdo what they say (and some will do more) mOstof the time, provision also needs to be madefor outliers, where the stakes are great. Self-interested bargaining of an opportunistic kindthus becomes the exception to which coopera_tion is the rule. Strategic considerations that hadbeen ignored by neoclassical economists from1870 to 1970 now make their appearance (L.Makowski and J. Ostroy, 2001 pp. 481-83490-91). Opportunism takes us into the deepstructure of contract and organization in waysthat frailty of motive does not.

Tbe bilateral dependency to which I referredearlier has its origins often in the FundamentalTransformation. Bureaucratization is also an in-tertemporal phenomenon. Both alter the natureof the contractual relation during the contractimplementation interval.

Tbe Fundamental Transformation applies tothat subset of transactions for which large num-bers of qualified suppliers at the outset are trans-formed into what, in effect, is abilateralexchange relation during contract execution andat the contract renewal interval. Tbis is to becontrasted with the standard presumption thatthe number of qualified suppliers at the outset(large or small) will continue into the future.Tbe key factor in deterrnining whether a large-numbers supply condition will evolve into abilateral exchange relation is the degree towhich the trans action in question is supportedby durable investments in transaction-specificassets- by which I mean assets that can beredeployed to alternative uses and users only ata loss of productive value. Because continuityof the exchange relation matters as asset speci-ficity increases, such transactions elicit addedprivate ordering govemance supports.

Specific investments can take the form ofspecialized physical assets (such as a die forstamping out distinctive metal shapes), special-ized human assets (that arise from firm-specifictraining or leaming by doing), site specificity(specialization by proximity), dedicated as~ets(large discrete investments made in expectaHOn

vaL. 95 NO. 2

f Ontinuing business, the premature terrnina-OC l' d b'. of which would resu t m pro ucts emgHon. .

Id at distress pnces), or brand-name capltal.~ISO, nonredeploya~le "organizationa~ .assets"

( actices, relationships, complementanties) be-pr . ral' .een firms often have mtertempo ongms.

~atever the source, failure to appreciate thatIllany transactions underg? a. ~undamentalTransformation, whereupon Identi~es ~ereaftermatter to which govemance rarrufications ac-rue was responsible for many misconceptions

~bo~t contract and organization in the zero-. 12

transaction-cost econonncs era.Although bureaucratization is a much ig-

nored condition, Oskar Lange (1938 p. 109)described bureaucratization, correctly I think, as"the real danger of socialism." Because, how-ever, he was interested in the pure economictheory of socialism, bureaucratization issueswere set aside-where they remained over thenext 50 years until the economies in EastemEurope and the former Soviet Union collapsed.Even now, bureaucratization is a poorly under-stood intertemporal phenomenon. I will returnto it in my discussion of the impossibility ofreplication/selective intervention below.

As against one all-purpose law of contractthat was enforced in a legalistic way, KarlL1ewellyn adopted a purposive perspective andintroduced the idelK)f"contract as framework."As L1ewellyn (1931 pp. 736-37) put it, the"major importance of legal contract is toprovide ... a framework which never accuratelyrefiects real working relations, but which pro-vides a rough indication around which suchrelations vary, an occasional guide in cases ofdoubt, and a norm of ultimate appeal when therelations cease in fact to work." Tbe object ofcontract, so construed, was not to be legalistic,but to get the job done.

To be sure, the norm of ultimate appeal towhich L1ewellyn refers is important, in that

12 Whatever the source, bilateral dependency has mas-sive public policy ramifkations which did not register in thepre-govemance era. Thus contracrual practices that werepreviously thought to be antkompetitive (as with take-or-pay contracts) are now perceived to serve an efficiencypurpose (S. Masten and K. Crocker, 1985) if the requisite~reconditions are satisfied. Also, the purported efficacy ofex ante franchise bidding" as a solution to the problem of

natural monopoly is deeply problematic in industries thatWIll predictably undergo a fundamental transformation(Williarnson, 1976).

9

recourse to the courts for purposes of ultimateappeal serves to delimit threat positions. But thekey idea is this: the legalistic view of contractthat applies to simple transactions needs tomake way for a more flexible and managerialconception of contract as the preservation ofongoing relations takes on economic impor-tance. The convenient notion of one all-purposelaw of contract gives way to contract laws (plu-ral) in the process.13

Such contract-Iaw differences play an impor-tant role in distinguishing among alternativemodes of governance. Specifically, the econom-ics of governance avers that each generic modeof governance is supported by a distinctive formof contract law. Tbe contract law of simplemarket exchange is that of legal rules,14 where-upon each party goes its own way, and courtsaward money damages in the event of a dispute,there being no interest in continuity for suchtransactions. Tbe hybrid mode is supported bycontract as framework, which is a more elasticconcept of contract and (within limits) prornotescooperative adaptation. If and as those adaptivelimits are exceeded, transactions are organizedby hierarchy. But what then is the contract lawof internal organization?

Tbe argument here is that the implicit con-tract law of internal organization is that of for-bearance (Williamson, 1991b). Thus, whereascourts routinely grant standing to contracts be-tween firms should there be disputes overprices, the damages to be ascribed to delays,failures of quality, and the like, the courts havethe good sense to refuse to hear disputes be-tween one intern al division and another overidentical technical issues. Access to the courtsbeing denied, the parties must resolve their dif-ferences intemally, which is to say that the firm

13 Contract laws (plural) can be thought of as the re-sponse by legal realists to the need to support continuity intrades that deviate from the ideal transaction in both law andeconomics-namely, between large numbers of buyers andseilers whose identity was unimportant. Whereas econo-mists made note of deviations from the ideal that took theform of small numbers (by devising monopoly, monopsony,duopoly models and the like), legal realists, if not contract-law specialists more generally, were alert to the benefits ofcontinuity as identity became important.

14 As Dixit's (2004) recent examination of lawlessnessemphasizes, even simple market exchange can benefi! fromprivate ordering if court ordering is weak or corrupt. SectionV excepted, I assume that court ordering works weil formost simple transactions.

AEA PAPERS AND PROCEED1NGS

18 The "property rights theory of the firm" (S. Grossmanand O. Hart, 1986; Hart. 1995) and recent variants tbereon(G. Baker et al., 2002; Hart and B. Holmstrom, 2002)obviously qualifies. For arecent survey and contribution tothe theory of the firm literature, see Robert Gibbons (2005);also see Tian Zhu (2004).

11

IV. Applications

Vertical integration is the paradigm problemfor the economics of govemance, to whichmany other contractual phenomena turn out tobe variations on a theme. Indeed, any issue thatarises as or can be reformulated as a contractingproblem can be examined to advantage throughthe lens of transaction-cost economizing. WhatI have previously referred to as the simple con-tractual schema displays the basic regularities.

Assurne that a firm can make or buy a com-ponent and assurne further that the componentcan be supplied by either a general-purposetechnology or a special-purpose technology,where k is a measure of asset specificity. Thetransactions in Figure 1 that use the general-purpose technology are ones for which k = O. Inthis case, no specific assets are involved, and theparties are essentially faceless. Those trans ac-tions that use the special-purpose technologyare ones for which k > O. As earlier discussed,bilaterally dependent parties have incentives topromote continuity and safeguard their specific

cem with fully formal theory is the possible lossof contact with the phenornena in question. Asecond concem is that the theory becomes non-testable. Robert Solow (2001 p. 112) speaks tothe first as follows: "A model can be rightin ... [al mechanical sense" yet be "unenlighten-ing because ... [it] obscures the key interactions,instead of spotlighting them." And MichaelWhinston (2003) speaks to the latter.

From the perspective of one who is con-vinced that much of the relevant action residesin the ex post contract implementation stage, Ifind the formal models and related empiricalresearch by Steven Tadelis and his co-authors tobe especially promising (P. Bajari andTadelis,2001; Tadelis, 2002; Bajari et al., 2004; J. Levinand Tadelis, 2004). However, the contract ap-proach to economic organization will continueto benefit from pluralism.18 As discussed inSection V the recent literature on lawlessnessand economics is pertinent.

R1CHARD T. ELY LECTUREvaL. 95 NO. 2

dI user costs be compromised by the ac-an or .' b. . . n? Will selective mterventIOn e com-qU1SItIO. .

. d by failures of contract as pronuse,ProITllse

Pon a need for three-way commonwhereu . b' d .ledge arises? WIll the com me enterpnseknowore subject to politieization? Examiningbe m . d' d . rt dth mieroanalytics IS te IOUSan IS repo e

le where (Williamson, 1985 Ch 6). I merelye~ .' brt here that integratIon expenences pro -asse h' h . th tI in all three respects, w lC IS to say aems ..' t(1) replication and selectJ.vemterventlon cannobe 'mplemented as described, on which account(2/the move from mark:t to ~erar~hy is.alwaysattended by a loss of mcentIve mtensI~y anddded bureaucratic costs, and (3) conung to

~erms with these conditions is ~ital to ~ u~der-standing of real-world econonuc org~mz~tIon. Ifurther aver that it would be weIl mgh ImpOS-sible to uncover the relevant microanalytic f~a-tiIfes without posing the issues in a comparatIvecontractual way.

The purported efficacy o~ reputa~ion e~fect~,often examined as a one-sided pnsoner s dI-lemma game with sequential moves, is widelyinvoked to support efficient trade. But what arethe limits? If fully efficacious, why do we notrely entirely on reputation effects to police tradeacross all technologically separable stages? Andif reputation effects require support, why does itnot suffice to invent a suitable collective-actionmechanism, such as the merchant law system(Paul Milgrom et al., 1990), to perform therequisite information-disclosure and punish-ment functions?

Again, the answers reside in the microanalyt-ics, an examination of whieh discloses that rep-utation effects can be costly and experiencebreakdowns (Kreps, 1990a; Williamson, 1991app. 166-72; Jean Tirole, 1996). Because theefficacy of areputation effect varies with thenature of transactions and with the conditions ofembeddedness (local sanctions and the like),this and other theories of spontaneous orderoften need to be augmented by providing trans-action-specific intentional order of an ex postgovemance kind.

Like many other theories, the economics ofgovemance has moved through a natural pro-gression from informal theory (where the earlyintuitions reside) to pre-formal theory (wherethe basic logic is set out) to semi-formal theory(of a reduced-form kind) to fully formal theory.ldeally, value is added at each step. One con-

17 See Frank Knight' s (1965 p. xxiii) preface in lbereissue of Risk, Uncenainty. and Profit. Also see ](njgbt(1965 p. 286 [footnote Il). Tracy Lewis (1983 p. 1092)alsospeaks to the purported advantages of large size.

MAY 2005

The comparative contracting perspective d'putes t~s. ~l that. is required fc:r the the~~iecononues m questIon to be reallzed is for thiron-making and steel-making stages to be lo~cated in the immediate proximity of each otheCommon ownership is not, without more, i~~plied. Indeed, in a zero-transaction-cost worldinterfirm contracting would costlessly imple~ment autonomous and coordinated adaptationsto disturbances of all kinds. In a positive_transaction-cost world, by contrast, interfirmcontracting between autonomous, site-specificstages poses bilateral dependency hazards thatare relieved by common ownership. Site speci-ficity, moreover, is merely one illustration of thecomparative contractual argument that technol-ogy, by itself, is not determinative of verticalintegration.

I exarnine the argument that bureaucracy canbe ignored (because it is a wash) by restating thelong-standing puzzle of firm size as follows:Can a large firm do everything that a collectionof small firms can do and more?I? I proceed byindirection by postulating two mechanisms, rep-lieation and selective intervention, which wouldanswer this question in the affirmative, if theycould feasibly be implemented.

Thus consider the outsourcing of a good orservice to an independent supplier and assumethat the contract works weIl most of the time butoccasionally breaks down. Suppose that the pur-chaser proposes to acquire the supplier on thefollowing terms: the supplier will continue toappropriate its own net receipts (adjusted foroverhead and user costs) in the post-acquisitioninterval; and the supplier will continue to dobusiness as usual (by replication) except as theacquiring stage selectively intervenes, when-ever there is a prospect of expected net gains.In that event, incentive intensity will be un-changed after acquisition, and the combinedfirm will never do worse (by replieation)and will sometimes do better (by selectiveintervention).

Whether or not tlIis can be implemented turnson answers to the following questions: Will theaccounting system, transfer pricing practices,

B. Contested Mechanisms

15 For a discussion of the limits of the two-way commonknowledgelcostless bargaining argument, see Williarnson(1975 pp. 31-33); for assessments of the claim that a se-quence of short-term contracts can implement an optimallong-term contract, see Kreps (1990b p. 760) and Williarn-son (1991 a); on power, see Williarnson (1996 Ch. 9); and ontrust, see Williarnson (1996 Ch. 10).

16 Note that 1 define a stage as a cluster of nonseparableactivities. Vertical integration thus entails the unified own-ership and hierarchical organization of successive separablestages.

By contested mechanisms, I mean oneswhich, if costlessly operative, would undo theneed for an economics of governance. Theseinclude claims that technology is determinativeof economic organization, that bureaucracy is ofno account, that two-way common knowledgeand costless bargaining eliminate problems ofmal adaptation in contracts, that reputation ef-fects are reliably efficacious, that power ex-plains observed contractual practices, and thattrust obviates the need for credible commit-ments. I exarnine all from a microanalytic, com-parative contracting perspective.

I briefly discuss technology, bureaucracy,and reputation effects here. My exarnination ofthe other contested mechanisms is reportedelsewhere.15

Technological explanations were once heldto be central to the vertical integration of sepa-rable stages of production.16 Indeed, verticalintegration that lacked a "physical or techniealaspect" was believed to be deeply problematic(1. Bain, 1968 p. 381). Thus consider the "das-sie case ... of integrating iron-making and steel-making to effect a saving in fuel costs byeliminating a reheating of the iron before it isfed to a steel furnace" (Bain, 1981 p. 381).Vertical integration of these two stages waspurportedly necessitated by thermal economies.

10

becomes its own court of ultimate appeal. Ineffect, forbearance law authenticates hierarchyby supporting its main purpose, namely, timelyresponsiveness to consequential disturbancesfor which coordinated adaptations are needed.The upshot is that the differential adaptive ef-ficacy of alternative modes of governance arerealized, in part, with the support of comple-mentary contract law regimes.

market and organized under hierarchy (unifiedownership; vertical integration), thereby betterto implement coordinated adaptations. Becauseadded bureaucratic costs accme upon taking atransaction out of the market and organizing itintemally, internal organization is usefullythought of as the organization form of last re-sort: try markets, try hybrids, and have recourseto the firm only when all else fails. Node D, theunified firm, thus comes in only as higher de-grees of asset specificity and added uncertaintyaward priority to coordinated adaptation.

Note that the price that a supplier will bid tosupply under node C (hybrid) conditions will beless than the price that will be bid at node B.That is because added security features (s > 0)serve to reduce the risk at node C, as comparedwith node B, so the contractual hazard premiumwill be reduced. One implication is that suppli-ers do not need to petition buyers to providesafeguards. Because buyers will receive producton better terms (lower price) when added secu-rity is provided, buyers have the incentive tooffer cost-effective credible commitments.

Applications of the schema to phenomenaother than vertical integration inc1ude nonstand-ard contracting practices (customer and territo-rial restrictions, take-or-pay contracts, exchangeagreements, price discrimination, and the like),regulation (and deregulation), labor-market or-ganization, the uses of debt and equity, agn-cultural cooperatives, networks, multinationaleconomic organization, corporate strategy (manymarketing practices inc1uded), and the list goes

A (Unassisted market)

13

... some very important gaps in our under-standing and [gave] us some very mis-leading ideas about the possibilities ofbeneficial policy intervention. Economistsstudying business and industrial organiza-tion have long recognized the inadequacyof the neoclassical view of the firm andhave developed richer paradigms andmodels based on the concepts of variouskinds of transaction costs. Policy analysisalso stands to benefit from ... opening theblack box and exarnining the actual work-ings of the mechanism inside.

Applications of such reasoning to policy analy-sis is not only a theme of Dixit' s 1996 book, buthas taken hold more generally.

orbit and mIes of the game (R. Lipsey, 2001). Asecond possibility is that some theories are trulyfanciful. A third is that the refutable implica-tions of some would-be theories are contra-dicted by the data. A multiplicity of theories,some of which are vacuous, others of which arefanciful, and still others of which are contra-dicted by the evidence, is an embarrassment topragmatically oriented social scientists. Amongthis subset, insistence upon the injunction toderive refutable implications and submit theseto the data has attractions.

The economics of govemance has respondedto the challenge by deriving refutable implica-tions and inviting empirical testing. As of theyear 2000, there were over 600 published em-pirical articles on transaction-cost economicswith exponential growth therein (c. Boemerand J. Macher, 2002). Still, there is no occasionto rest content. All theories of economic orga-nization, the economics of govemance in-c1uded, will benefit from more and betterempirical tests: better data, additional phenom-ena, and better statistical procedures.

Given that the economics of govemance wasstimulated in part by the evolving crisis in an-titrust enforcement and regulation during the1960s, little wonder that many of the public-policy applications of governance have been toantitrust (Joskow, 2002) and regulation/deregu-lation (B. Levy and P. Spiller, 1994). But it isalso noteworthy that the lens-of-contractJgover-nance approach has had influence on public-policy analysis more generally. As Dixit (1996p. 9) remarks, the era of black-box appliedwelfare economics had left


19 Interestingly, the make-or-buy decision refuses to goaway.having recently been renamed as the worrisome prac-rice of "'outsourcing" or, even worse, as "offsourcing,"which entails procurement from a foreign country. Somecriticswould have us believe that domestic vertical integra-tion is the ideal to which market procurement is a deeplyproblematic alternative.

20 As Scott Masten (1995p. xi-xii) observes, "surveys ofthe empirical transaction cost literature attest ... [thatl thetheory .nd evidence have displayedremarkable congruity."Paul Joskow (1991 p. 81) moreover, describes empiricalwork of • tr.nsaction cost kind as "in much better shapethanmuch of the work in industrial organization generally"(also see Whinston, 2003). Interestingly, much of the em-pirical research on the economies of govemance does notrely on published data that have been collected for analtogetherdifferent purpose (e.g., to satisfy census or reg-ulatoryrequirements). Instead, much of the best empiricalresearch on govemance uses primary data that have beencollectedwith the microanalytic needs of the discriminatingalignment hypothesis foremost in mind. Tbose who havedone this modest, slow, molecular, cumulative work de-serve enormous credit.

21 For collections of articlesthat work out of the logic ofgovern.nce, see Williamson and Masten (1995) and ClaudeMenard (2005), especially those articles that Menard clus-ters under headings VII-XIII.

22 For an examination of the growing citations to trans-,clJ~n-cost economicslthe economies of governance, see~dhamson (2005). Tbe uses of such reasorting are espe-~'ally gre.t in business and economies but also extend toIllcludeI.w. sociology (organization theory), and politicalSCJence.

VOL. 95 NO. 2

19 Surveys of empirical applications oftrans-on. 20 th . liaction-cost economics, e ~ow~ng. tealrature

ovemance within the new mstltutlon eco-on gcs 21 and citations to this literature all attestnoffil , .' 22to the wide reach of thls reasomng. .

The economics of govemance subsc~bes ~oth propositions that "the purpose of sClence m

e eral is not prediction, but knowledge for itsgenn sake," yet that prediction is "the touch-o~ne of scientific knowledge" (N. Georgescu-~oegen, 1971 p. 37): Some scof~ a~ pre~iction,vidently in the bebef that predlctlon IS easy.

~lso, since everyone knows that "it is e.asy to liewith statistics," what useful purpose IS servedby empirical testing? My experience is that pre-diction is ademanding standard and that cor-roboration is not easy, but difficult. Takentogether, prediction and empirical testing per-form the vital function of helping to sort thesheep from the goats among riYal theories.

Why then are not more social scientists in-sistent upon deriving refutable implications andsubrnitting these to empirical tests? One pos si-bility is that the world of pure theory has its own

MAY 2005

D (Hierarchy)

C (Hybrid)

B (Unrelieved hazard)


FIGURE 1. THE SIMPLE CONTRAcruAL SCHEMA

I2

investments. Let s denote the magnitude of anysuch safeguards, which include penalties, infor-mation disclosure and verification procedures,specialized dispute resolution (such as arbitra-tion), and in the limit, integration of the twostages under unified ownership. An s = 0 con-dition is one for which no safeguards are pro-vided; adecision to provide safeguards isreflected by an s > 0 result.

Absent anarchy or inept or corrupt courts(which would pose lawlessness issues of thekind exarnined in Section V), node A in Fig-ure 1 corresponds to the ideal transaction in lawand economics: "sharp in by c1ear agreement;sharp out by clear performance" (I. Macneil,1974 p. 734). There being both large numbersand an absence of dependency, governanceis accomplished through competitive marketprices and, in the event of disputes, by court-awarded damages. Node B poses unrelievedcontractual hazards, in that specialized invest-ments are exposed (k > 0) for which no safe-guards (s = 0) have been provided. Suchhazards will be recognized by farsighted play-ers, who will price out the implied risks ofcontractual breakdown.

Added contractual supports (s > 0) are pro-vided at nodes C and D. At node C, interfirmcredible contracting mechanisms serve to sup-port cooperative adaptations across a widerrange of disturbances. Should costly contractualbreakdowns continue in the face of best bilateralefforts to craft cost-effective safeguards at nodeC, the transaction may be taken out of the


Pertinent in this connection is that the prac-tice.of c?mparing a~tual alternatives with hypo-theucalldeals has gIven way to a comparison offeasible alternatives, alt of which are flawed.Lapses into comparisons with omniscient, om-nipotent, benevolent alternatives are avoided by(1) recognizing that it is impossible to do betterthan one' s best, (2) insisting that all of thefinalists in a governance-structure competi-tion meet the test of feasibility, (3) symmetri-cally exposing the strengths and weaknessesof all proposed feasible forms, and (4) des-cribing and costing out the mechanisms ofimplementation.23

V. Lawlessness and Govemance

The econornics of lawlessness focuses on in-stitutional environments where the "govern-ment is unable or unwilling to provide adequateprotection of property rights and enforcement ofcontracts through the machinery of the state"(Dixit, 2004 p. vii). 1 repeat, however, that evenin states that make best efforts to provide pro-tection for property rights and contract enforce-ment, the state' s access to information and thestate's protection and enforcement mechanismsare inherently lirnited. Whether, therefore, therules of the game are well-developed (as in theUnited States) or poorly developed (as in Viet-nam [J. McMillan and C. Woodruff, 1999]),property and contractual hazards invite the useof private ordering to infuse order, thereby tornitigate conflict and realize mutual gains fromtrade.

Circumstances where state law is "verycostly, slow, unreliable, corrupt, weak, or sim-ply absent" (Dixit, 2004 p. 3) nevertheless poseadded private ordering challenges. Dixit's re-cent book on Lawlessness and Economics isnoteworthy for the range of phenomena that headdresses and his imaginative development of a"toolki~". of game-theoretic models, broadly inthe spmt of the pragmatic methodology towhich 1 referred earlier (Dixit, 2004 p. 22).

He begins with "private ordering in theshadow of the law," parts of which track the "con-tract as framework" reasoning of Llewellyn. ButDixit also uncovers a nuance in using the courts

23 I discuss these issues elsewhere with reference to theremediableness criterion (Williamson. 1996 eh. 8).

MAY 2005

for purposes of ultimate appeal: relation btween the parties can be temporarily co s e-

. d'f . fr mpro_mIse I, startmg om a weak state the ." d I . , re IS agra ua Improvement of state law"_ hih I "' .. wchas essons lor tranSItIon econornies (D' .2~ pp. 3~-40). He thereafter exarnines ~~;:tratIon, WhiCh often has verifiability advant

th b· d . agesover e courts, com me With Court-enfarc 'bac~p. The gener~ fin?ing here is that "ar~tratIon based on ItS mformation adva t

ks 11. . . nagewor we m conJunction with the formal 1 als~ste~" (Dixit, 2004 p. 47). Describing e~_bitratlOn as providing superior verifiabili(buttressed perhaps by reputation effects) %nevertheless a truncated statement of the pur-poses served by this mode of governance Fmany transactions, arbitration also provilies Orforum with greater give-and-take, which pro~motes cooperation, continuity, and mutual gain(Fuller, 1963). Dixit's models make no pro/sion for this. 1

Especially interesting are transactions forwhich "profit-motivated contract enforcement"is observed, of which the creation of a mafia toprovide order for otherwise problematic trans-actions is an example (D. Gambetta, 1993 p. 15[as quoted in Dixit, 2004 p. 99]):

When the butcher comes to me to buy ananimal, he knows that 1 want to cheat him[by giving hirn a low-quality animaI]. But1 kno.w that he wants to cheat me [byrenegmg on payment]. Thus we needPeppe [that is, a third party] to make usagree. And we both pay Peppe a percent-age of the deal.

To be sure, organization, like the law, has a lifeof its own-which poses intertemporal trade-offs: "protectors, once enlisted, invariably over-stay their welcome" (Gambetta, 1993 p. 197).Again, lessons for transition economies residetherein (Dixit, 2004 pp. 3, 100, 129).

Issues of "private protection far propertyrights" also pose novel issues, both in devel-oped countries, where common pool resomeeutilization problems elicit collective aetion re-sponses, and even more, in less-developedcountries (Dani Rodrik, 2000) and transitioneconomies (McMillan, 2002). As Dixit (2004p. 125) puts it, "threats to property rights eomefrom ... individuals [who] encroach on one'sproperty ... [or], even worse, the state itself or

nts may engage in extortion of privateage . b"" Th

rty to further their own 0 ~ectIves. ePrope . .balanee between government protectlOn, pnvate

tI'on and efforts made by other people toProtee , .capture one's property turns on the underlymg

ameters of the model.parAlthOugh the contractual hazards with whichVirit is eoncerned are due mainly to weak prop-ertY rights, rather than .bilateral depende~cy,they ean nevertheless be mterpreted as a van~nt

the simple contractual schema of Fig-1.24 Absent relief, property-right hazards

loeate the parties at node B, which is'neffieient if cost-effective private ardering sup-~rts for property can be devised that will m~v.ethe parties to node C. Note, however, that DIXItnever moves beyond node C to include unifiedownership at node D (hierarchy) as a gover-nanee alternative. Possibly this will be remediedin follow-on work of the lawlessness kind-although if, as 1 contend, markets and hierar-chies differ in discrete structural ways, thesedifferences will need to be taken into accountexplicitly. Also, as discussed above, Dixit'streatment of arbitration emphasizes informa-tional benefits to the neglect of informal pro-cess benefits. Be that as it may, there is noquestion but that the study of lawlessness use-fully expands the reach of the economics ofgovemance.

VI. Conclusions

Tbe economics of governance has helped topersuade many economists and other socialscientists that (1) institutions matter and aresusceptible to analysis, (2) adaptation to distur-bances is a key purpose of economic organiza-tion, (3) the action is in the microanalytics, (4)positive trans action costs can be addressed in acomparative way, and (5) public policy towardbusiness needs to be informed by a broad (or-ganizational) understanding of the efficiencypurposes served by eomplex contract and eco-nomic organization. Put in the negative, it is nolonger acceptable to treat governance as some-one else's bailiwick, to slight adaptation, toscant the microanalytics, to assurne trans ac-

. 24 Letting r denote property rights hazards and substimt-lOgr for k (asset specificity hazards) in Figure I, the sameregularities appear except for the absence of node D.

15

tion costs to be zero, or to uncritically ascribemonopoly purposes to nonstandard contractualpractices and organizational structures.

As described herein, the economics of gov-ernance joins three fundamental concepts (ad-aptation, governance, and transaction costs)with the purpose of pouring operational contentinto all three. This is accomplished by makingnot one (or a few) changes in the basic setup--such as a change in the objective function of thefirm, or the introduction of a new constraint, orinvoking barriers to entry, or assuming differ-ential risk aversion, or the like- but by makingaseries of related changes.

The economics of governance treats simplemarket exchange as a special case and featuresongoing transactions for which adaptations (ofboth spontaneous and intentional kinds) areneeded. As compared with most theories ofeconomic organization, the economics of gov-ernance is more interdisciplinary and more mi-croanalytic-as with the efficient alignment oftransactions with governance structures (whichturns on the attributes of transactions in relationto the adaptive properties of alternative modesof governance); in working out hitherto ne-glected features of contract (e.g., tlte Funda-mental Transformation), of bureaucracy (theimpossibility of replicationlselective interven-tion), and of contract laws (plural); and in ex-arnining the efficacy of contested mechanisms.

It is said that theories, like beads, need astring to hold them together. The string thatdraws the foregoing features together is thediscriminating alignment hypothesis, where-upon economizing on transaction costs (whichmainly take the form of maladaptation) is madethe main case. The resulting theory of economicorganization applies not merely to the make-or-buy decision and boundary of the firm issues,but has ramifications for a wide range of con-tractual phenomena-within economics and thesocial sciences more generally. Empirical testshave been both numerous and broadly corrob-orative. The study of good order and work-able arrangements from the economics-of-governance perspective nevertheless entailssetup costs. Is the game worth the candle?

1 venture a two-part answer. First, not every-one will want to or should make the requisiteinvestments. The economics of govemance, af-ter all, is only one of several instructive lensesfor studying the economics of arganization. But

AEA PAPERS AND PROCEEDlNGS R1CHARD T. ELY LECTURE) 17

Laffont, J. and Martimort, D. The theory of in-centives: The principal-agent model. Prince-ton, NJ: Princeton University Press, 2002.

Lange, O. "On the Theory of Economic Social-ism," in Benjamin Lippincott, ed., On theeconomic theory of socialism. Minneapolis:University of Minnesota Press, 1938, pp. 55-143.

Levin, J. and Tadelis, S. "Employment versusContracting in Procurement: Theory and Ev-idence from U.S. Cities." Unpublished manu-script, Stanford University, 2004.

Levy, B. and SpiIler, P. "The Institutional Foun-dations of Regulatory Comrniunent: A Com-parative Analysis of TelecommunicationsRegulation." Journal of Law, Economics andOrganization, 1994, 10(1), pp. 201-46.

Lewis, T. "Preemption, Divestiture, and ForwardContracting in a Market Dominated by a Sin-gle Firm." American Economic Review,1983, 73(5), pp. 1092-110I.

Lipsey, R. "Successes and Failures in the Trans-formation of Econornics." Journal of Eco-nomic Methodology, 2001, 8(2), pp. 169-201.

LIewellyn, K. "Wh at Price Contract? An Essayin Perspective." Yale Law Journal, 1931,40(5), pp. 704-51.

Macneil, I. "The Many Futures of Contracts."Southern California Law Review, 1974,47(5), pp. 691-816.

Makowski, L. and Ostroy, J. "Perfect Competi-tion and the Creativity of the Market." Jour-nal of Economic Literature, 2001, 32(2), pp.479-535.

Masten, S. "Introduction to Vol. II," in O. Wil-liamson and S. Masten, eds., Transactioncost economics. Aldershot, UK: Elgar, 1995.

Masten, S. and Crocker, K. "Efficient Adaptationin Long Term Contracts: Take or Pay Provi-sions for Natural Gas." American EconomicReview, 1985, 75(5), pp. 1085-96.

McMillan, J. Reinventing the bazaar: A naturalhistory ofmarkets. New York: Norton, 2002.

McMillan, J. and Woodruff, C. "Dispute Preven-tion without Courts in Vietnam." Journal ofLaw, Economics and Organization, 1999,15(3), pp. 637-58.

Meese, A. "Intrabrand Restraints and the Theoryof the Firm." North Carolina Law Review,2004,83(1), pp. 5-86.

Menard, C. ''The Econornics of Hybrid Organi-zations." Journal of Institutional and Theo-retical Economics, 2004, 160(2), pp. 345-76.

OL 95 NO 2

R gen N. The entropy law and eco-rgescu. oe , d. process. Cambridge, MA: Harvar

nOmlCU iversity Press, 1971. -.

n R "Four Formal(izable) Theones ofGibbOn

Fs: '" Journal of Economic Behavior

- the Irffi· . )and Organization, 2005 (forth:?ffilng ..dberg, V. and Erickson, J. R. Quantlty and

GoI. AdJ'ustment in Long-Term Contracts: Apoce C k "J 1'+Case Study in Petroleum 0 e. ourna oJLaw and Economics, 1987, 30(2), pp. 369-

98. Sand Hart O. ''The Costs and Ben-Grossman, .' .fit f Ownership: A Theory of Verucal and

~t~r~l Integration." Journal of Political _Economy, 1986, 94(4), pp. 691-71~.

IIart, O. Firms, contracts, and finan~lal struc-ture. New York: Oxford UnivefSlty Press,

1995. fF'Hart. O. and Holmstrom, B. "A Th~ory 0 mn

Scope." Unpublished manuscnpt, HarvardUniversity, 2002.

Hayek, F. A. "The Use of ~owled~e in Soci-ety." American Economlc ReView, 1945,35(4), pp. 519-30. .

Joskow, P. "The Role of Transa~tlOn. <;ost Eco-nomics in Antitrust and Pubhc Utlhty Regu-latory Policies." Journal of Law, E~onomics,and Organization, 1991, 7 (SpeCIal Issue),pp. 53-83. .,

Joskow, P. "Transaction Cost Econoffilcs, Antl-trust Rules, and Remedies." Journal of Law,Economics and Organization, 2002, 18(1),pp. 95-116.

Klein, B.; Crawford, R. and Alchian, A. "VerticalIntegration, Appropriable Rents, and theCompetitive Contracting Process." Journal ofLaw and Economics, 1978, 21(2), pp. 297-326.

Knight, F. Risk, uncertainty, and profit. NewYork: Harper and Row, 1965.

Kreps, D. "Corporate Culture and EconornicTheory," in James Alt and Kenneth Shepsle,eds., Perspectives on positive political econ-omy. New York: Cambridge UniversityPress, 1990a, pp. 90-143.

Kreps, D. A course in microeconomic theory.Princeton, NJ: Princeton University Press,1990b.

Kreps, D. "Markets and Hierarchies and (Math-ematical) Economic Theory," in Glenn Car-roll and David Teece, eds., Firms, markets,and hierarchies. New York: Oxford Univer-sity Press, 1999, pp. 121-55.

Coase, R. "The Nature of the Firm " Eco .. . nomlcQNew Senes, 1937,4(16), pp. 386-405 '

Coase, R. "The Problem of Social Cost " 'J. OUr·nal of Law and Economics, 1960 3

1-44. ' , Pp.Coase, R. "The Regulated Industries' Di

'''A' E .. sCus_S10n. mencan conomlC Review 1964(Papers ,~nd Pro.ceedings),. 54~2),pp. 194-97.

Coase, R. Indusmal OrganlzatlOn: A Propo alfor Research," in V. R. Fuchs ed Pol~. .: ., Icylssues and research Opportumtles in indu _trial organization. New York: Columbia Un~-versity Press, 1972, pp. 59-73.

Commons, J. "The Problem of Correlating LawEconornics, and Ethics." Wisconsin Law Re:view, 1932, 8, pp. 3-26.

Commons, J. The economics of collective action.Madison: University of Wisconsin Press1950. '

Commons, J. Legal foundations of capitalism2nd Ed. Madison: University of Wisconsi~Press, 1968.

Demsetz, H. "The Structure of Ownership andthe Theory of the Firm." Journal of Law andEconomics, 1983,26(2), pp. 375-90.

Dixit, A. The making of economic policy: Atransaction cost politics perspective. Cam-bridge, MA: MIT Press, 1996.

Dixit, A. Lawlessness and economics: Alterna-tive modes of governance (Gorman Lecturesin Economics, University College London).Princeton, NJ: Princeton University Press,2004.

Fischer, S. "Long-Term Contracting, StickyPrices, and Monetary Policy: Comment."Journal of Monetary Economics, 1977,3(3),pp. 317-23.

Fuchs, V., ed. Policy issues and research oppor-tunities in industrial organization. NewYork: Columbia University Press, 1972.

Fuller, L. "American Legal Philosophy at MidCentury." Journal of Legal Education, 1954,6(4), pp. 457-85.

Fuller, L. "Collective Bargaining and the Arbi-trator." Wisconsin Law Review, 1963, (1), pp.3-46.

Galanter, M. "Justice in Many Rooms: Courts,Private Ordering, and Indigenous Law."Journal of Legal Pluralism, 1981, 19(1), pp.1-47.

Gambetta, D. The Sicilian Mafia: The businessof private protection. Cambridge, MA: Har-yard University Press, 1993.

REFERENCES

Alchian, A. and Demsetz, H. "Production, Infor-mation Costs, and Econornic Organization."American Economic Review, 1972,62(5), pp.777-95.

Arrow, K. ''The Organization of Econornic Ac-tivity: Issues Pertinent to the Choice of Mar-ket versus Nonmarket Allocation," in U.S.Joint Economic Committee, 91st Congress,1st Session, The analysis and evaluation ofpublic expenditure: The PPB system, Vol. I.Washington, DC: U.S. Govemment PrintingOffice, 1969, pp. 39-73.

Arrow, K. "Reflections on the Essays," inGeorge Feiwel, ed., Arrow and the founda-tions of the theory of economic policy. NewYork: New York University Press, 1987, pp.727-34.

Arrow, K. J. and Debreu, G. "Existence of anEquilibrium for a Competitive Economy."Econometrica, 1954, 22(3), pp. 265-90.

Bain, J. Industrial organization, 2nd Ed. NewYork: Wiley, 1968.

Bajari, P.; McMillan, R. and Tadelis, S. "Auc-tions versus Negotiations in Procurement: AnEmpirical Analysis." Unpublished manu-script, Duke University, 2004.

Bajari, P. and Tadelis, S. "Incentives versusTransaction Costs: A Theory of ProcurementContracts." Rand Journal of Economics,2001,32(3), pp. 287-307.

Baker, G.; Gibbons, R. and Murphy, K. J. "Re-lational Contracts and the Theory of theFirm." Quarterly Journal of Economics,2002, 117(1), pp. 39-84.

Barnard, C. The functions of the executive.Cambridge, MA: Harvard University Press,1938.

Boemer, C. and Macher, J. "Transaction Cast Eco-nornics: An Assessment of Empirical Researchin the Social Sciences." Unpublished manu-script, University of California-Berkeley, 2002.

Buchanan, J. "Game Theory, Mathematics, andEconornics." Journal of Economic Methodol-ogy, 2001, 8(1), pp. 27-32.

second, I subrnit that our understandings of eco-nornic organization and public policy pertinentthereto have been needlessly impoverished byfailures to pay heed to the lessons of gover-nance. The econornics of govemance is an un-finished project whose time has come.

16

AEA PAPERS AND PROCEEDlNGS

INTERNATIONAL INTRAFIRM PRODUCTION DEC/S/ONSt

By GENE M. GROSSMAN, ELHANAN HELPMAN, AND ADAM SZEIDL *

There are two countries, North and South,and one factor of production, labor. The wagerate is wN in the North and wS in the South, withwN > ws. We normalize prices so that W' = 1.

Consumers have Dixit-Stiglitz preferences fordifferentiated products that generate the inverse

I. Tbe Model

ate inputs can be produced domestically or in alow-wage country and can be produced in-house or outsourced. By assumption, assemblyof final goods takes place within the boundariesof the firm that has developed the product, butwe sometimes allow this activity tobe per-formed abroad.

First, we assurne that assernbly takes place athorne and that intermediate goods can be trans-ported at no cost. We identify conditions underwhich cross-industry variation in the fixed costof outsourcing generates a positive correlationbetween outsourcing and foreign sourcing. Wethen introduce transport costs for intermediateinputs and allow firms to choose where to as-semble their final output. In this case, cross-industry variation in the fixed cost of doingbusiness abroad produces a second complemen-tarity between outsourcing and foreign sourc-ing. The latter finding is in keeping withconditions described in arecent article in theFinaneial Times about problems facing firmsproducing in China (see Peter Marsh, 2004).Companies that cannot find efficient localsources for components in<China are burdenedwith the extra costs of shipping inputs fromhorne. Apparently, FD! often goes hand-in-handwith the ability to find suitable Chinese suppli-ers. The trade-offs between in-house productionand outsourcing and between shipping interme-diate goods and producing them in proximity toassembly operations are the subject of our in-vestigation below.

19

complementarities between Outsourcing and Foreign Sourcing

t Discussants: Stephen Yeaple,University ofPennsylva-Dia;James Markusen, University of Colorado.

* Grossman: Department of Economies, Prineeton Uni-versity,Prineeton, NJ 08544 (e-mail: [email protected]); Helpman: Department of Economies, HarvardUniversity, Cambridge, MA 02138 (e-mail: [email protected]);Szeidl: Department of Eeonomies, Universityof California, Berkeley, CA 94720 (e-mail: [email protected]).We aeknowledge with thanks the support oftheNational Seience Foundation and the U.S.-Israel Bina-tionalScienee Foundation.

IMare Melitz (2003) has developed a model of hetero-geneousfirms that is eommonly used in this literature (seeAntrasand Helpman, 2004; Grossman and Helpman, 2004).

Outsourcing has been growing both domesti-cally and internationally. So has foreign directinvestment (FDI). New models of internationaltrade address these phenomena using recent ad-vances in the economic theory of organizations.'1'he models help us to identify circumstancesunder which firms choose to make their inputsthemselves or buy them from third parties, andwhen they choose to produce their inputs locallyor abroad (see Grossman and Helpman, 2002,2004, 2005; Dalia Marin and Thierry Verdier,2002, 2003; Diego Puga and Daniel Trefler,2002; Pol Antras, 2003; Antras and Helpman,2004). Some authors investigate the organi-zational choices of homogeneous firms in anindustry with some particu1ar characteristics

others exarnine the relative prevalence oforganizational structures in industries

heterogeneous finns.1In this paper, we combine elements from

Antras and Helpman (2004) and Grossman et al.(2oo4a) to study the relationship between out-sourcing and foreign sourcing (or "offshoring").Our analysis focuses on industries with hetero-geneous firms that make intensive use of inter-mediate inputs. Contracting problems limit thetypes of contracts that can be written betweenfinal producers and input suppliers. Intermedi-

MAY 2005

Law, Economics and Organization, 200319(1), pp. 1-23. '

Williamson, o. "The Vertical Integration ofPducti~n: Market Failure Consideration:o.:Amencan Economic Review 1971 (P .. ' apersand Proceedmgs), 61(2), pp. 112-23.

Williamson, O. Markets and hierarchies: A lysis and antitrust implications. New Y~~k~Free Press, 1975. .

Williamson,. O. "~ranchise Bidding for NaturalMonopohes-m General and with Respect tCATV." Bell Journal oi Economics, 1976~7(1), pp. 73-104.

Williamson, o. "Transaction-Cost Economies:The Governance of Contractual Relations."Journal oi Law and Economics, 1979, 22(2),pp. 233-61.

Williamson, O. The economic institutions oi cap-italism. New York: Free Press, 1985.

Wil~a~son:, O. "The Logic of Economic Orga-ruzatIon. Journal oi Law, Economics, antiOrganization, 1988, 4(1), pp. 65-93.

Williamson, o. "Economic Institutions: Sponta-neous and Intentional Governance." Journaloi Law, Economics, and Organization1991a, 7(Special Issue), pp. 159-87. '

Williamson, O. "Comparative Economic Orga-nization: The Analysis of Discrete StructuralAlternatives." Administrative Seience Quar-terly, 1991b, 36(2), pp. 269-96.

Williamson, O. The mechanisms oi governance.New York: Oxford University Press, 1996.

Williamson, O. "The Theory of the Firm as Gov-ernance Structure: From Choice to Contract."Journal oi Economic Perspectives, 2002,16(3), pp. 171-95.

Williamson, o. "Transaction Cost Economiesand Business Administration." ScandinavianJournal oi Management, 2005 (forthcoming).

Williamson, O. and Masten, S. Transaction casteconomics, Vols. I and II. Aldershot, UK:Elgar, 1995.

Wilson, E. Consilience. New York: Knopf,1999.

Zhu, T. "Incentive Intensity, Forbearance Law,and the Governance of Transactions." Indus-trial and Corporate Change, 2004, 13(6),pp.855-66.

18

Menard, C. ed. International library oi the newinstitutional economics. Northarnpton, MA:Elgar, 2005.

Michels, R. Political parties. Glencoe, IL: FreePress, 1962.

Milgrom, P.; North, D. and Weingast, B. "TheRole of Institutions in the Revival of Trade."Economics and Politics, 1990,2(1), pp. 1-24.

Reder, M. The culture oi a controversial sei-ence. Chicago: University of Chicago Press,1999.

Riordan, M. and Williamson, O. "Asset Specific-ity and Econornic Organization." Interna-tional Journal oi Industrial Organization,1985,3(4), pp. 365-78.

Rodrik, D. "Institutions for High QualityGrowth: What They Are and How to AcquireThem." National Bureau of Economic Re-search (Cambridge, MA) Working Paper No.7540,2000.

Selten, R. "What is Bounded Rationality?" inGerd Gigerenzer and Reinhard Selten, eds.,Bounded rationality. Cambridge, MA: MITPress, 2001, pp. 13-36.

Simon, H. Models oi man. New York: Wiley,1957a.

Simon, H. Administrative behavior, 2nd Ed.New York: Macmillan, 1957b.

Simon, H. "Human Nature in Politics: A Dia-10gue of Psychology with Political Science."American Political Science Review, 1985,79(2), pp. 293-304.

Smith, A. The wealth oi nations. London: W.Strahan and T. Cadell, 1776; London: Dent,1992.

Solow, R. "A Native Informant Speaks." Journaloi Economic Methodology, 2001, 8(1), pp.111-12.

Tadelis, S. "Complexity, Flexibility, and theMake-or-Buy Decision." American Eco-nomic Review, 2002 (Papers and Proceed-ings), 92(2), pp. 433-37.

Tirole, J. "A Theory of Collective Reputations,with Applications to the Persistence of Cor-ruption and to Firm Quality." Review oi Eco-nomic Studies, 1996, 63(1), pp. 1-22.

Whinston, M. "On the Transaction Cost Deter-rninants of Vertical Integration." Journal oi

Foreward RICHARD T. ELY LECTURE The Economics of Governance

Documents

Transcript of Foreward RICHARD T. ELY LECTURE The Economics of Governance