Will Liam Son

8/8/2019 Will Liam Son

1/19

Journal of Economic LiteratureVol. XXXVIII (September 2000) pp. 595613

Williamson: The New Institutional Economics

The New Institutional Economics:Taking Stock, Looking Ahead

OLIVER E. WILLIAMSON1

1. Introduction

IOPEN MY DISCUSSION of the new institutional economics with a confes-

sion, an assertion, and a recommenda-tion. The confession is that we are still

very ignorant about institutions. The as-sertion is that the past quarter centuryhas witnessed enormous progress in thestudy of institutions. The recommenda-tion is that, awaiting a unified theory, weshould be accepting of pluralism.

Chief among the causes of ignorance

is that institutions are very complex.That neoclassical economics was dismis-sive of institutions and that much of or-ganization theory lacked scientific am-bitions have also been contributingfactors. As to progress, that is whatmost of this paper is about. There beingmany instructive lenses for studyingcomplex institutions, pluralism is whatholds promise for overcoming our

ignorance.Speaking for myself, I subscribe toJon Elsters view that we work predomi-nantly on partial mechanisms ratherthan general theories at this stage of de-

velopment (1994, p. 75). In considera-tion, however, of the splendid plausi-

bility of error to which Lord Acton re-fers,2 we need to sort the sheep fromthe goats. That is accomplished by ask-ing each would-be theory to advance re-futable implications to which the dataare applied.

R. C. O. Matthews, in his presidentialaddress to the Royal Economic Societyin 1986, pronounced that the econom-ics of institutions has become one of theliveliest areas in our discipline (Mat-

thews 1986, p. 903). Such a pronounce-ment was a surprise to most of the pro-fession. Hadnt institutional economicslong since been relegated to the historyof economic thought? Whence the vitalityto which Matthews made reference?

Matthews response was that the newinstitutional economics (NIE) turned ontwo propositions. First, institutions domatter; and second, the determinants

of institutions are susceptible to analy-sis by the tools of economic theory(Matthews 1986, p. 903). The second ofthese is what distinguishes the NIE, itbeing the case that institutional econo-mists of all kindsold and newareunanimous in the view that institutionsmatter.

Indeed, although both the older andnewer styles of institutional economics

595

1 University of California, Berkeley. [email protected]. This paper was first presentedat the third annual meeting of the InternationalSociety for New Institutional Economics in Wash-ington, DC in September 1999 in my capacity aspresident-elect. Helpful comments received thereand from Bengt Holmstrom and John McMillan

are gratefully acknowledged. 2 As quoted in Daniel Boorstin (1998, p. 281).


2/19

subscribe to many of the same good ideas,a progressive research program requiresmore. Kenneth Arrow speaks to thetransformation as follows (1987, p. 734):

Why did the older institutionalist school failso miserably, though it contained such ableanalysts as Thorstein Veblen, J. R. Commons,and W. C. Mitchell? I now think that . . .[one of the answers is in the] important spe-cific analyses . . . of the New InstitutionalEconomics movement. But it does not consistof giving new answers to the traditional ques-tions of economicsresource allocation andthe degree of utilization. Rather, it consistsof answering new questions, why economicinstitutions emerged the way they did andnot otherwise; it merges into economic his-

tory, but brings sharper [microanalytic] . . .reasoning to bear than had been customary.

There is no question that the NIEhas grown in stature and influence overthe fourteen years since Matthews pro-nouncement. Initial skepticism hasgradually given way to respectit beingthe case that economists are very prag-matic people. Tell them something dif-ferent and consequential about phe-

nomena that are of interest to them anddemonstrate that the data are corrobo-rative: that will get their attention. TheNIE has progressed not by advancing anoverarching theory but by uncoveringand explicating the microanalytic featuresto which Arrow refers and by pilingblock upon block until the cumulative

value added cannot be denied.The NIE, moreover, will not stand

still. Even as institutional economics isbeing incorporated within orthodoxy,new opportunities and challenges await.Both unfinished business and new proj-ects yet to be undertaken await the newmillennium.

I begin with a sketch of four levels ofsocial analysis, next turn to some of thegood ideas out of which the NIE works,and then examine some of the applica-tions to which the NIE has been put.

Concluding remarks follow.

2. Four Levels of Social Analysis

It will be useful for purposes of per-spective to consider the four levels ofsocial analysis that are distinguished in

Figure 1.3

The solid arrows that con-nect a higher with a lower level signifythat the higher level imposes con-straints on the level immediately below.The reverse arrows that connect lower

with higher levels are dashed and signalfeedback. Although, in the fullness oftime, the system is fully interconnected,I mainly neglect these feedbacks. TheNIE has been concerned principally

with levels 2 and 3.

The top level is the social embedded-ness level. This is where the norms, cus-toms, mores, traditions, etc. are located.Religion plays a large role at this level.Although Level 1 analysis is undertakenby some economic historians and othersocial scientists (E. C. Banfield 1958;Robert Putnam, Robert Leonardi, andRaffaella Nanetti 1993; Samuel Hunt-ington 1996; and Victor Nee 1998),

Level 1 is taken as given by most insti-tutional economists. Institutions at thislevel change very slowlyon the orderof centuries or millenniawhereuponDouglass North poses the query, Whatis it about informal constraints that givesthem such a pervasive influence uponthe long-run character of economies?(1991, p. 111).

North does not have an answer tothat perplexing question, nor do I. The

concept of embeddedness, both at thelevel of society and in the context of on-going network relations, has been ad-

vanced to help explicate these issues(Granovetter 1985). The vast literatureon culture (Paul DiMaggio 1994) is alsopertinent. Neil Smelser and RichardSwedberg discuss these and related issuesin their introduction to the Handbook

3 This framework was first set out in Williamson

(1998).

596 Journal of Economic Literature, Vol. XXXVIII (September 2000)


3/19

of Economic Sociology, where they ob-serve that different kinds of embedded-nesscognitive, cultural, structural,and politicalshould be distinguished,and conclude that the concept of em-beddedness remains in need of greatertheoretical specification (1994, p. 18).

An identification and explication ofthe mechanisms through which informalinstitutions arise and are maintained

would especially help to understand theslow change in Level 1 institutions. Iconjecture in this connection that many

of these informal institutions have

mainly spontaneous originswhich is tosay that deliberative choice of a calcula-tive kind is minimally implicated. Giventhese evolutionary origins, they areadopted and thereafter display a greatdeal of inertiasome because they arefunctional (as with conventions); otherstake on symbolic value with a coterie oftrue believers; many are pervasivelylinked with complementary institutions(formal and informal), etc. Be that as itmay, the resulting institutions have alasting grip on the way a society con-

ducts itself. Insular societies often take

Williamson: The New Institutional Economics 597


4/19

measures to protect themselves againstalien values.

The second level is referred to as theinstitutional environment. The struc-tures observed here are partly the prod-

uct of evolutionary processes, but de-sign opportunities are also posed. Goingbeyond the informal constraints (sanc-tions, taboos, customs, traditions, andcodes of conduct) of a Level 1 kind, wenow introduce formal rules (constitu-tions, laws, property rights) (North1991, p. 97). This opens up the oppor-tunity for first-order economizing: getthe formal rules of the game right.

Constrained by the shadow of thepast, the design instruments at Level 2include the executive, legislative, judi-cial, and bureaucratic functions of gov-ernment as well as the distribution ofpowers across different levels of gov-ernment (federalism). The definitionand enforcement of property rights andof contract laws are important features.

Although such first-order choices areunarguably important to the economic

productivity of an economy (NathanRosenberg and L. E. Birdzell 1986;Ronald Coase 1992; North 1994; BrianLevy and Pablo Spiller 1994; MancurOlson 1996; Witold Henisz 1998) cumu-lative change of a progressive kind is

very difficult to orchestrate. Massivediscontentcivil wars (the GloriousRevolution; see North and Barry Wein-gast 1989), or occupations (following

World War II), perceived threats (theMeiji Revolution), breakdowns (EasternEurope and the former Soviet Union), amilitary coup (Chile), or a financial cri-sis (New Zealand)will, however, occa-sionally produce a sharp break from es-tablished procedures. Rare windows ofopportunity to effect broad reform arethereby opened. Such defining mo-ments are nevertheless the exceptionrather than the rule. At least partly be-

cause of our primitive understanding,

the response to such opportunities isoften one of failure. Absent such a

window, major changes in the rules ofthe game occur on the order of decadesor centuries. The European Union,

for example, has been in progress forfifty years and is still in early stages ofdevelopment.

What is often referred to as PositivePolitical Theory (PPT) is concerned

with working out the economic and po-litical ramifications of Level 2 features.To be sure, such research also has les-sons for the normative design of betterpolities. Like the NIE of which it is apart, however, PPT is predominantly anexercise in positive analysis. The objectis to better understand how things

workwarts and all. The research prod-uct of PPT scholarship has been noth-ing less than auspicious, which has beengood for both political science and theNIE.

Much of the economics of propertyrights is of a Level 2 kind. Such re-search flourished in the 1960s. A strong

version of the argument is that a private-enterprise system cannot functionproperly unless property rights are cre-ated in resources, and, when this isdone, someone wishing to use a re-source has to pay the owner to obtain it.Chaos disappears; and so does the gov-ernment except that a legal system to

define property rights and to arbitratedisputes is, of course, necessary (Coase1959, p. 12; emphasis added). Onceproperty rights have been defined andtheir enforcement assured, the govern-ment steps aside. Resources are allo-cated to their highest value as the marvelof the market works its wonders.

This compact statement illustratesboth the strength and the weakness ofthe property rights literature. The greatstrength of this literature is that itbrings property rights to the forefront,

where they belong, whereupon novel



5/19

property rights reasoning could bebrought to bear in informative ways(Armen Alchian 1961, 1965; Coase1959, 1960; Harold Demsetz 1967). The

weakness is that it overplayed its hand.

The claim, for example, that the legalsystem will eliminate chaos upon defin-ing and enforcing property rights as-sumes that the definition and enforce-ment of such rights is easy (costless).Plainly, many transactions do not qual-ify (Coase 1960). Going beyond therules of the game (property) to includethe play of the game (contract) wasneeded. That is the opening through

which the governance of contractualrelations walked in during the 1970s.

This brings me to the third level,which is where the institutions of gover-nance are located. Although propertyremains important, a perfectly function-ing legal system for defining contractlaws and enforcing contracts is not con-templated. Costless court ordering be-ing a fiction, much of the contract man-agement and dispute settlement action

is dealt with directly by the partiesthrough private ordering. The need tocome to terms with contract laws (plu-ral), rather than an all-purpose law ofcontract (singular), is posed (ClydeSummers 1969; Ian Macneil 1974). Thegovernance of contractual relationsbecomes the focus of analysis.

John R. Commons prefigured thiswork with his observation that the ulti-mate unit of activity . . . must containin itself the three principles of conflict,mutuality, and order. This unit is atransaction (1932, p. 4). Not only doestransaction cost economics subscribe tothe idea that the transaction is the basicunit of analysis, but governance is an ef-fort to craft order, thereby to mitigateconflict and realize mutual gains .

So conceived, a governance structureobviously reshapes incentives. To focus

entirely on ex ante incentive alignment,

however, is a truncated way to study or-ganizationespecially if all complexcontracts are unavoidably incompleteand if adaptation is the central problemof economic organization (Chester Bar-

nard 1938; Friedrich Hayek 1945). Mov-ing beyond the agency theory tradition ofex ante incentive alignment, transactioncost economics turns its attentionad-ditionally and predominantlyto the expost stage of contract.

This entails four moves: (1) to nameand explicate the principal dimensions

with respect to which transactions differ(thereby to uncover differential adap-tive needs); (2) to name and explicatethe principal attributes for describinggovernance structures (where each isdefined by a distinctive syndrome of re-lated attributes, whence markets, hy-brids, firms, regulation, bureaus, non-profits, etc. differ in discrete structural

ways); (3) to effect a discriminating match,according to which transactions arealigned with governance structures soas to promote adaptation of autonomous

and cooperative kinds; and (4) to ascer-tain whether the predicted alignmentsare corroborated by the data.

The canonical problem for dealingwith these issues is that of vertical inte-gration, which is the issue posed byCoase in his classic 1937 article on TheNature of the Firm. As it turns out,any issue that arises as or can be refor-mulated as a contracting issue can beexamined to advantage in transactioncost economizing terms. A huge num-ber of phenomena turn out to be con-tractual variations on a common theme.

What I refer to as second-order econo-mizingget the governance structuresrightis realized at Level 3. The possi-ble reorganization of transactions amonggovernance structures is re-examinedperiodically, on the order of a year to adecade, often at contract renewal or

equipment renewal intervals.



6/19

Such discrete structural analysis ofgovernance is to be distinguished fromthe fourth level, which is the level at

which neoclassical analysis works. Opti-mality apparatus, often marginal analy-

sis, is employed, and the firm, for thesepurposes, is typically described as a pro-duction function. Adjustments to pricesand output occur more or less continu-ously. Agency theory, which emphasizesex ante incentive alignment and effi-cient risk bearing, rather than ex postgovernance, nonetheless makes provi-sion for nonneoclassical complications,of which multi-tasking is one (BengtHolmstrom and Paul Milgrom 1991).

Indeed, a still earlier (zero level) ofanalysis warrants remark: an evolution-ary level in which the mechanisms ofthe mind take shape (Steven Pinker1997). The application of these ideas toeconomics even now is beginning toreshape our understanding of humanactors. Our evolutionary psychologistand cognitive science colleagues are

vital to the exercise.

Finally, I should call attention totechnology. As compared with techno-logical innovation, the study of orga-nizational innovation has been compara-tively neglected. The NIE has attemptedto rectify thatthe idea being thattruly among mans innovations, the useof organization to accomplish his endsis among both his greatest and his earli-est (Arrow 1971, p. 224). We cannotfail, however, to be awed by the pro-found importance of technological inno-

vation (Robert Fogel 1999). Inasmuchas these two work in tandem, we needto find ways to treat technical and or-ganizational innovation in a combinedmanner.

3. Good Ideas

The new institutional economics had

its origins in good critics of orthodoxy

who believed that institutions wereboth important and susceptible toanalysis. Feeling expansive, I would in-clude six Nobel Laureates among thekey figures: Kenneth Arrow, Friedrich

Hayek, Gunnar Myrdal, Herbert Simon,Ronald Coase, and Douglass Norththe last two being the first two presi-dents of ISNIE. But there are others.Armen Alchian has been an influentialfigure. So too has been research on or-ganization theory, especially at Carnegie(some of it prefigured by earlier workby Chester Barnard)where the namesof Richard Cyert and James March jointhat of Simon. Alfred Chandlers pio-neering work in business history wasalso pathbreaking. Thoughtful contribu-tors from the law, especially contractlaw, include Karl Llewellyn, StewartMacaulay, Lon Fuller, and Ian Macneil.John R. Commons also brought originaland important ideas to the study of in-stitutional economics. The German His-torical School was also concerned withrelated ideas (Erik Furubotn and

Rudolf Richter 1997, pp. 3435).Among the key good ideas that I

associate with the NIE are these:Human Actors. If nothing is more

fundamental in setting our researchagenda and informing our researchmethods than our view of the nature ofthe human beings whose behavior weare studying (Simon 1985, p. 303),then social scientists should be pre-pared to name the key attributes of hu-man actors. Both the condition of cog-nition and self-interestedness need tobe addressed.

There is close to unanimity withinthe NIE on the idea of limited cognitivecompetenceoften referred to asbounded rationality. Mind being ascarce resource, cognitive specializationhas economizing consequences. Also,given cognitive limits, the complex con-

tracts to which I referred earlier are



7/19

unavoidably incomplete. But while thereis near-unanimity that complete contin-gent claims contracting is impossible,the appropriate way to model incom-plete contracts remains controversial.

Lack of agreement on the definitionand operational import of boundedrationality is a major obstacle (ArielRubinstein 1998; David Kreps 1999).

Contractual incompleteness posesadded problems when paired with thecondition of opportunismwhich mani-fests itself as adverse selection, moralhazard, shirking, subgoal pursuit, andother forms of strategic behavior. Be-cause human actors will not reliably dis-close true conditions upon request orself-fulfill all promises, contract asmere promise, unsupported by crediblecommitments, will not be self-enforcing.

But for opportunism, the courts would simply ask witnesses to tell us what you know that is germane to ourdecision. That is not, however, the waythat testimony is taken. Witnesses arerequired to take an oath to tell the

truth, the whole truth, and nothing but thetruth: dont lie; dont conceal; dontmislead. Inasmuch, moreover, as oathsare not self-enforcing, penalties for per-

jury remind witnesses that prevaricationhas consequences.

Still a third attribute of human actorswarrants remark, and that is the capac-ity for conscious foresight. Indeed, asRichard Dawkins observes, it is the ca-pacity to simulate the future in imagina-tion . . . [that saves] us from the worstconsequences of the blind replicators(1976, p. 200). Parties to a contract

who look ahead, recognize potentialhazards, work out the contractual rami-fications, and fold these into the ex antecontractual agreement obviously enjoyadvantages over those who are myopicor take their chances and knock on

wood. The governance of contractual

relationsthe Commons triple of con-

flict, mutuality, and order to which Ireferred earlieris centrally implicated.

Feasibility. Students of the NIE es-chew hypothetical idealswhich workoff of omniscience, benevolence, zero

transaction costs, full credibility, andthe likeand deal instead with feasibleorganizational alternatives, all of whichare flawed. Coase (1964) and Demsetz(1969) were among the first to take ex-ception with the asymmetric standardsthat were once used in the market fail-ure literatureaccording to whichmarkets are beset with failures whereasomniscient, omnipotent, benevolentgovernments (Avinash Dixit 1996, p. 8)

would reliably administer efficaciousremedies. As we all should have recog-nized (but needed to be told), all feasi-ble forms of organizationgovernmentincludedare flawed.4

What I have referred to as the reme-diableness criterion is intended to rec-tify this asymmetric state of affairs. Thiscriterion holds that an extant mode oforganization for which no superior fea-

sible alternative can be described andimplemented with expected net gains ispresumed to be efficient.

To be sure, public policy analysis be-comes more complicated when analystscan no longer condemn extant modesbecause they deviate from a hypotheti-cal ideal, full stop. The remediablenesscriterion presses the public policy ana-lyst to display a superior feasible alter-native. If, moreover, a proposed feasi-ble alternative cannot be costlesslyimplemented, then the costs of imple-mentation are appropriately included inthe net benefit calculuswhich has ma-

jor ramifications for the path depen-dency literature. Finally, grounds forrebutting the efficiency presumptionneed to be addressedwhich brings in

4 Dixit (1996) counsels the older public financetradition to come to terms with government fail-

ures.



8/19

politics (Williamson 1996, 1999). Ab-

sent rebuttal, the remediableness crite-rion stands as a reminder of the obvi-ous: it is impossible to do better thanones best.

Firms and Bureaus. In addition to thenature of the human beings to which Si-mon referred, we need also to be self-conscious about the Nature of the Firm,

which was the title of Coases classic1937 article from which the NIE drawsmuch of its inspiration. Arrow speaks tothe fundamental importance of the the-ory of the firm, and to long-standingmisconceptions thereof, as follows: Anystandard economic theory, not just neo-classical, starts from the existence offirms. Usually, the firm is a point or atany rate a black box. . . . But firms arepalpably not points. They have internalstructure. This internal structure mustarise for some reason (1999, p. vii).

The need was to get beyond the ana-

lytically convenient (and sometimes

adequate) conception of the firm-as-production function (which is a techno-logical construction) to consider thefirm as a governance structure (which isan organizational construction) in whichinternal structure has economic pur-pose and effect. More generally, theneed was to identify and explicate theproperties of alternative modes of gover-nancespot markets, incomplete longterm contracts, firms, bureaus, etc.

which differ in discrete structural ways.Because each generic mode of gover-nance possesses distinctive strengths and

weaknesses, there is a place for each yeteach needs to be kept in its place. Thelogic of discriminating alignment to

which I referred earlier applies.In a heuristic way, the choice of

governance structure moves from mar-ket to hierarchy through the sequence

of moves shown in Figure 2 (where h



9/19

denotes contractual hazards and s de-notes safeguards).5 This can be inter-preted as a move from simple to com-plex. We thus begin with autonomouscontracting, which is the ideal transac-

tion in both law and economics: sharpin by clear agreement; sharp out byclear performance (Macneil 1974, p.738). This discrete transaction para-digm comes under strain as contractualhazards appear. The inability of courts,for example, to verify what is commonknowledge between the parties to an ex-change (Williamson 1975, p. 30) couldinduce a move from interfirm to in-trafirm organization. Other sources ofcontractual hazard include bilateral de-pendency (by reason of nonredeploy-able assets), weak property rights (es-pecially intellectual property rights),undisclosed quality, health, and safetyhazards, failures of probity, and the like.Such hazards compromise contractualintegrity and give rise to contractual im-passes, maladaptations, and investmentdistortions. Here, as elsewhere, ineffi-

ciency invites relief. Cost-effective haz-ard mitigation through added governanceensues.

Moving from less to more complexgovernance entails introducing addedsecurity features, reducing incentive in-tensity, and incurring added bureau-cratic costs. Moving from simple (dis-crete) contracts to complex (incompletelong term) contracts is thus attended bya whole series of features: the length ofthe contract increases, penalties to de-ter breach are introduced, provision ismade for added information disclosureand processing, and specialized disputesettlement mechanisms appear.

Additional mechanisms, to includethe use of hierarchy to effect coordi-nation and decide disputes by fiat, are

introduced when transactions are re-moved from the market and placed un-der unified ownership (the firm). Be-cause added compliance and securityfeatures always come at a cost, more

complex modes of governance are re-served for those transactions for whichcontractual hazards are especially difficult.

The public bureau, in this scheme ofthings, can be thought of as the organi-zation form of last resort: try spot mar-kets, try incomplete long-term con-tracts, try firms, try regulation, andreserve recourse to public bureaus for

when all else fails (comparatively). Notethat the common practice of condemn-ing public bureaus because they havelower-powered incentives, more rulesand regulations, and greater job secu-rity than a counterpart firm completelymisses the point. These features havebeen deliberately crafted into the pub-lic bureau, thereby to make it bettersuited to govern some (especially diffi-cult) transactions.6 Vigilance is none-theless neededlest the public bureau

be overused.If transaction cost economics works

out of variations on a few key themes,then this schema, which was initiallydevised to help explicate the gover-nance of contractual relations in inter-mediate product market transactions,should also apply, with variation, toother classes of transactions. It does.

The four nodes are interpreted withreference to intermediate producttransactions above. Consider final goodsmarkets and government procurementtransactions.

Transactions in final goods markets, where individual consumers are the

5 A variant of Figure 2 originally appeared in mypaper on Public and Private Bureaucracies

(1999).

6 This is a recurrent theme not only of the trans-action cost economics literature but also of partsof the agency theory literature. See especiallyHolmstrom (1989) and Holmstrom and Milgrom(1991, 1994), where the benefits of low-powered

incentives in firms are featured.



10/19

buyers, are similar but different. NodeA transactions are generic and competi-tively organized. Node B transactionsare rare. These correspond to P. T.Barnumtheres a sucker born every

minuteand other fly-by-night transac-tions. Node C is the credible commit-ment node. Branding in combination

with reputation effects and productwarranties appear. Also, for some natu-ral monopoly transactions, public utilityregulation serves credibility purposes.Node D is a nearly empty set. Econo-mies of scale and of specialization areimpediments to own-supply by consum-ers, although collective organization(consumer cooperatives) can be usedto manage some transactions. (Manyhousehold services can be thought of asown-supply, but few fit comfortably

within the schema.)Government procurement transac-

tions are also similar but different.Node A describes generic transactionsto which, often, tedious technical speci-fications apply. Very few government

transactions are of a Node B kind.Credibility mechanisms at Node C in-clude the elaborate machinery of ad-ministered contracting, as with defenseprocurement (which transactions, how-ever, are sometimes compromised bythe shared interests of the governmentagency and the private supplier). AndNode D is the public bureau, where forprobity or political reasons the govern-ment chooses to manage the transactionitself.

Other applications of the schema in-clude the employment relation (JamesBaron and Kreps 1999, ch. 4) and cor-porate finance (the choice betweendebt and equity). Some transactions,such as alliances and joint ventures,pose complications of a disequilibriumcontracting kind (Williamson 1991) thatare beyond the reach of the schema.

Operationalization. Many good ideas

are initially expressed as tautologies, which Coase has wryly defined as aproposition that is clearly right (1988,p. 19). Because good tautologies expandthe mind and are hard to come by, they

deserve respect. Lest, however, we slipinto the speculations to which WesleyMitchell once referred7which is a fatethat beset the older style institutionaleconomics as well as the American Le-gal Realism movementwe need to ask

what are the mechanisms through which a proposed theory operates andwhat are the refutable implications.

The effort to operationalize promis-ing ideas has both theoretical and em-pirical parts. The theoretical often takesthe form of a progression from informalto preformal, semi-formal, and fully for-mal modes of analysisideally acquir-ing value added in the process. Such aneffort helps to sort the sheep from thegoats. Nicholas Georgescu-Roegen hada felicitous way of putting it: althoughthe purpose of science is not predic-tion, but knowledge for its own sake,

prediction is nevertheless the touch-stone of scientific knowledge (1971, p.37). Would-be theories for which pre-dictive content is lacking must eventu-ally step aside (be set aside) for thosefor which the hard work of formalizationand empirical testing are undertaken.

Theory Development. Formalizationis vital to a progressive research agenda,but it sometimes comes at a cost. Thusalthough Simon once argued thatmathematical translation is itself a sub-stantive contribution to theory . . . be-cause it permits clear and rigorous rea-soning about phenomena too complexto be handled in words (1957, p. 89)

7 Speculative systems can be quickly excogi-tated precisely because they do not require theeconomist to collect and analyze masses of data, totest hypotheses for conformity to fact, to discardthose which do not fit, to invent new ones and testthem until, at long last, he has established a factu-

ally valid theory (Mitchell 1945, p. 2).



11/19

and subsequently asserted that thepoverty of mathematics is an honestpoverty that does not parade imaginaryriches before the world (1957, p. 90),provision also needs to be made for the

possibility that core features of the the-ory are left out or obscured by thetranslation. There is, after all, such athing as prematurely formal theory.Kreps speaks to the issues as follows(1999, p. 122):

If Markets and Hierarchies has been transla-ted into game theory using notions of infor-mation economics, it is a very poor transla-tion . . . In particular, mathematics-basedtheory still lacks the language needed to cap-ture essential ideas of bounded rationality,

which are central to . . . transaction costsand contractual form. Anyone who relies onthe translations alone misses large and valuablechunks of the original.

What is referred to as the propertyrights theory of the firm, which had itsorigins with Sanford Grossman andOliver Hart (1986) and has subsequentlybeen developed by Hart and John Moore

(hence the reference to the GHM model),relates to but differs significantly fromthe transaction cost economics setup (aspresented, for example, in Williamson1985, 1991). It is similar in that it deals

with the make-or-buy decision througha setup where contracts are incomplete(by reason of bounded rationality),mere promise cannot be used to over-come noncontractibility (by reason ofopportunism), and parties to the con-tract are bilaterally dependent (by rea-son of asset specificity). These com-monalities notwithstanding, there arealso major differences.

Some of these differences are attrib-utable to simplifications that invariablyattend formal modelling. Ideally, corefeatures of the verbal argument and themechanisms through which they workare made more precise in the process of

formalization. Such a case can be made

for the property rights theory of thefirm, which is a major intellectualachievement that has spawned a grow-ing literature on the formal modellingof incomplete contracts.8 As Kreps sug-

gests, however, valuable chunks aremissing. In the spirit of full disclosure(honest poverty), I focus on these.

The most consequential differencebetween the TCE and GHM setups isthat the former holds that maladapta-tion in the contract execution interval isthe principal source of inefficiency,

whereas GHM vaporize ex postmaladaptation by their assumptions ofcommon knowledge and costless ex postbargaining. The upshot is that all of theinefficiency in GHM is concentrated inthe ex ante investments in human assets(which are conditional on the ownershipof physical assets).9

This shift from ex post maladaptation(the hazards from which vary with thecondition of asset specificity and thedisturbances to which a transaction issubject) to ex ante investment distor-

tions matters. For one thing, GHMmakes very limited contact with thedata10 whereas (as discussed below)TCE is an empirical success story. Re-lated (ex post) governance and (ex ante)investment differences are the following:

8 The January 1999 issue of the Review of Eco- nomic Studies is entirely devoted to recent contri-butions, critiques, responses, and extensions uponthe GHM model.

9 Bounded rationality enters this setup in a very

peculiar way: parties who are unable to write com-plete contracts ex ante are nevertheless able to an-ticipate ex ante what decisions will be taken expost, contingent on state realizations (Kreps 1999,pp. 12325). In effect, the GHM setup is one of

selective unbounded rationality: not everythingthat is logically consistent is credulous (Kreps1999, p. 125).

10 That the data relevant to GHM are so limitedand inaccessible explains why there has been noformal testing of the property rights approach(Hart 1995, p. 49)although the inside contract-ing system (John Buttrick 1952) is an approxima-tion to (and its failures could be interpreted as a

partial contradiction of) GHM.



12/19

(1) The TCE rendition of the make-or-buy decision between succes-sive stages (A and B) asks whetherA and B should be separatelyowned and operated or if the own-

ership and operation of these twostages should be unified. If in-dependent, then each stage appro-priates its net receipts (high-pow-ered incentives obtain) butmaladaptation problems can ariseduring contract execution. If uni-fied, then the two stages are man-aged coordinately through hierar-chy. (Maladaptation problems arethereby relieved; incentives arelower-powered; and added bureau-cratic costs arise.) By contrast,GHM view vertical integration in adirectional way: either A buys B orB buys A, and it matters which waythis is done. That is because com-mon ownership under GHM does

not imply unified management. In-stead, each stage (in all configura-tionsA and B are independent; A

buys B; B buys A) appropriates itsnet receipts. This last is a very un-usual condition, in that unifiedownership is normally thought ofas a means by which to effectcooperation.11

(2) TCE maintains that each genericmode of governancespot market,incomplete long-term contract,firm, bureau, etc.is defined by asyndrome of attributes to whichdistinctive strengths and weak-nesses accrue. Specifically, TCEholds that alternative modes differin incentive intensity, administra-tive controls (to include auditing,accounting, and transfer pricing),access to the courts, and informalorganization (to include politick-

ing). GHM assume that incentiveintensity, administrative controls,and informal organization are un-changed by ownership and thatcourts are irrelevant (because of

costless renegotiation). None ofthe physical asset utilization andtransfer pricing distortions that Iassociate with the impossibility ofselective intervention (Williamson1985, pp. 13540) thus occurunder the GHM setup.

(3) TCE examines a wide range of expost devices for infusing crediblecommitments into contracts andapplies this reasoning to a wide setof transactions. Variations on thistheme include hybrid modes of or-ganization (Scott Masten 1996,Part III), exchange agreementsand other uses of hostages to sup-port exchange, the organization of

work, the organization of labor andhuman resources more generally,corporate governance, regulation(and deregulation), public bu-

reaus, and project financing. Be-cause GHM is a property rightsand property rights only construc-tion (Holmstrom 1999), it relatesto some of these issues not at alland others very selectively (Hart1995; Hart, Andrei Shleifer, andRobert Vishny 1997).

GHM is nonetheless a pathbreakingcontribution and has set the formalmodelling of incomplete contracting inmotion. New formal models of incom-plete contracts which are closer in spiritto TCE include the treatment of pro-curement by Patrick Bajari and StevenTadelis (1999), which focuses on the in-centive and ex post adaptation differ-ences between fixed price and cost pluscontracting. Also, the recent paper bySusheng Wang and Tian Zhu (2000)

employs the idea that alternative modes

11 The inside contracting system referred to in

note 7 supra is in the spirit of GHM organization.



13/19

of governance work out of differentcontract law regimes (Williamson 1991).And Gene Grossman and Elhanan Help-man (1999) appeal to the added bureau-cratic costs of unified as compared with

market governance in their assessmentof alternative modes for producing dif-ferentiated consumer products. More

veridical treatments of incomplete con-tracting are thus in progress and stillmore are in prospect.12

Empirical. Some scoff at prediction,evidently in the belief that prediction iseasy. Also, since everyone knows thatit is easy to lie with statistics, whatuseful purpose is served by empiricaltesting? My experience is different: pre-diction is a demanding standard, whichis why so many would-be theories re-main excogitated speculations; and cor-roboration is difficult, which explains

why few predictions are tested.Because, however, good theories are

rarely fully developed at the outset, thetheory and the evidence are often inter-active. As Alan Newell observes (1990,

p. 14):Theories cumulate. They are refined and re-formulated, corrected and expanded. Thus,

we are not living in the world of Popper. . . .[Theories are not] shot down with a falsifica-tion bullet. . . . Theories are more like gradu-ate studentsonce admitted you try hard toavoid flunking them out. . . . Theories arethings to be nurtured and changed and builtup.

Good but underdeveloped ideas are

evidently like good but underdevelopedminds: both are precious things. Be-cause development is costly, promisingtheories, like promising graduate stu-dents, are admitted only if they cross athreshold. Once admitted, theories (and

graduate students) are progressivelybuilt upmoving from less formal tomore formal stages of development. Fi-nally, as with promising graduate stu-dents, we do not hold on to cherished

theories indefinitely: some do flunk out.Specifically, theories that remain tauto-logical or yield predictions that are con-tradicted by the data must make way fortheories that yield predictions for whichthe data are corroborative.

Empirical applications of transactioncost economics got under way in theU.S. in the 1980s and have grown expo-nentially since: the number of pub-lished studies exceeds 500 and involvessocial scientists in Europe, Japan, India,China, Mexico, South America, Austra-lia, New Zealand, and the list goes on.It could have been otherwise, but thetheory and evidence display a remark-able congruity (Scott Masten 1995, p.xi). Recent empirical surveys includeHoward Shelanski and Peter Klein(1995), Bruce Lyons (1996), KeithCrocker and Masten (1996), and Aric

Rindfleisch and Jan Heide (1997).Not only has this research been

broadly corroborative of the predictionsof transaction cost economics, but theimportance of risk aversion to commer-cial contracting has been placed indoubt (Douglas Allen and Dean Lueck1999). To be sure, transaction cost eco-nomics, like everything else, will bene-fit from more and better empirical

work. I have no hesitation, however, indeclaring that the NIE is an empiricalsuccess story. Paul Joskow concurs:this empirical work is in much bettershape than much of the empirical workin industrial organization generally(1991, p. 81). Those who have done thismodest, slow, molecular, definitive

work deserve enormous credit.1312 For an earlier formal treatment (of a reduced

form kind) akin to Bajari and Tadelis, see MichaelRiordan and Williamson (1985). The recent paperby Hart and Moore, On the Design of Hierar-chies (1999a), also makes express provision for

organization.

13 The recurring regularity is this: more complexmodes of governance appear as contractual haz-

ards build upwhere bilateral dependency, due



14/19

4. Phenomena

The NIE is predominantly concernedwith Levels 2 and 3 of the four levels ofsocial analysis shown in Figure 1. Theseare the levels of the institutional envi-ronment and the institutions of gover-nance, respectively. Between them,they cover a lot of ground.

The formal features of the institu-tional environmentthe laws, polity,

judiciary, bureaucracyare crucial inexamining the development of nationstates (North and Weingast 1989) andfor making intertemporal comparisons

within and cross-national comparisons

between nation states. Indeed, this lasthas come to be a growth industry to

which many economists who are onlyslightly associated with the NIE havemade contributions. It is nonethelessnoteworthy that the NIE has donemuch of the pioneering work in thisarea.

To repeat, any issue that arises as orcan be posed as a contractual issue canbe examined to advantage in transactioncost economizing terms. Examples for

which contractual issues are evident atthe outset include contracts for inter-mediate products, for labor, for finalgoods and services, for the rental orlease or purchase of land, equipment,and buildings, for professional services,for marriage, and the list goes on. Even,moreover, if contractual features arenot immediately evident from the out-

set, many issues can be reformulated soas to disclose their contractual qualities,the choice between debt and equity,

the oligopoly problem,14 and the multi-national corporation being examples(Peter Buckley and Mark Casson 1976;Hubert Gatignon and Erin Anderson1988).

Many public policy issues, moreover,turn jointly on the combined use ofLevel 2 and Level 3 reasoning. In thearea of privatizing telecommunications,for example, Levy and Spiller examinethe institutional environments in fivecountries through a comparative con-tractual lens in which issues of crediblecontracting are featured (1994, 1996).The recent study of reforming urban

water systems by Claude Menard andMary Shirley (1999) likewise makesclear that ownership is not determina-tive but needs to be examined in con-

junction with the support, or the lackthereof, of the mechanisms of gover-nance. Again, issues of credible contract-ing are salient. The same is true of com-mercial contracting in Vietnam (JohnMcMillan and Christopher Woodruff1999).

Broad reach notwithstanding, theNIE is not and does not pretend to bean all-purpose construction, as the re-form of economies of Eastern Europeand the former Soviet Union illustrate.Thus Coase in his Nobel Prize lectureobserved that (1992, p. 714):

The value of including . . . institutional fac-tors in the corpus of mainstream economics ismade clear by recent events in EasternEurope. These ex-communist countries are

advised to move to a market economy, andtheir leaders wish to do so, but without theappropriate institutions no market economyof any significance is possible. If we knewmore about our own economy, we would bein a better position to advise them.

Two years later, North, in his Nobel Prizelecture, expressed similar precautions.

to asset specificity in any of its forms (physical,human, site-specific, dedicated assets, brand, andtemporal), in combination with disturbances thatbeset contracts during the contract execution in-terval are responsible for many of these hazards.

Working, as it does, out of noncontractible humanasset investment distortions, GHM cannot layclaim to these same empirical successes (Michael

Whinston 1997; Holmstrom 1999).

14 Of the various ways in which it can be posed,its contractual nature becomes more evident whenit is posed as a problem of reaching and enforcing

a cartel agreement (Williamson 1975, Chap. 12).



15/19

Thus even if we are confident that poli-ties significantly shape economic perfor-mance because they define and enforcethe economic rules, whereupon an es-sential part of development policy is the

creation of polities that will create andenforce efficient property rights, there isthe further problem that we know verylittle about how to create such polities(North 1994, p. 366).

Real-time events, however, cannot beput on hold. Hard choices have to bemade. Economic reform in Russia is anexample.

The team of Maxim Boycko, AndreiShleifer, and Robert Vishny respondedto the perceived need to give shape tothe reform with the recommendationthat the Russian economy should beprivatized quickly and massively. Con-siderations of both Realpolitik and eco-nomic theory were invoked in supportof this recommendation.

There being widespread agreement thatpolitical influence over economic life

was the fundamental cause of economic

inefficiency [in Russia], Boycko, Shleifer,and Vishny (1995, p. 11) declared that:

. . . the principal objective of reform was . . .to depoliticize economic life. . . . Privatiza-tion fosters depoliticization because it de-prives politicians of the opportunity to allo-cate goods. . . . The goal of privatization wasto sever the links between enterprise manag-ers and politicians. . . . There was no other

way to achieve restructuring and efficientoperation of firms.

The two strategic actors in this re-form program were the official bureauc-racy, which was viewed as the enemyto be fought at all costs and the stake-holdersmanagers, employees, and lo-cal governments. The Boycko et al.team consistently and generously rec-ognized stakeholders claims, and thusensured their eventual support of priva-tization (Boycko, Shleifer, and Vishny

1995, pp. 1314).

This political prescription for massiveand rapid privatization was reinforcedby the economic theory of the firm on

which the Boycko et al. team relied.Specifically, they appealed to the afore-

mentioned work by Grossman and Hart(1986), which views ownership as a sys-tem of control rights and treats the ap-propriate assignment of property rightsas determinative (Boycko, Shleifer, and

Vishny 1995, p. 13). Upon privatizingstate-owned enterprises, therefore, ef-fective restructuring by the new stake-holders would presumably follow (opcit, p. 150). In the confidence that thefuture would take care of itself, themass privatization program that was be-gun in the spring of 1992 had purport-edly reached a triumphant completionin June 1994 (op cit, p. 8), by whichdate two-thirds of Russian industry wasprivately owned.

Had the Boycko et al. team consultedthe new institutional economics, a morecautious and selective program of priva-tization with greater attention to imple-

mentation would have resulted. Considerfirst the literature on franchise biddingfor natural monopoly, where the prop-erty rights approach and the gover-nance approach reach very differentconclusions.

The property rights approach to theproblem of natural monopoly is to con-duct an ex ante bidding competitionand award the right to serve the marketto the group that tenders the best bid(Demsetz 1968; George Stigler 1968;Richard Posner 1972). Very much in thespirit of Boycko et al., the future willtake care of itself once the assets havebeen privatized in this way.

That sanguine view does not with-stand scrutiny if serious ex post imple-mentation problems are in prospect.Under the governance approach, theaward of a monopoly franchise needs to

be assessed comparatively. This entails



16/19

looking ahead and uncovering ex postcontractual hazards, thereafter workingout the ramifications for alternativemodes of governance (Williamson 1976,pp. 7991). Because franchise bidding

works much better for some natural mo-nopoly industries than others (William-son 1976, pp. 102103), the use of fran-chise bidding will be reserved for thoseindustries where comparative net bene-fits can be projectedbut not other-

wise. Privatization, it turns out, is not anall-purpose solution (Victor Goldberg1976; George Priest 1993).

Although privatizing an entire econ-omy is a much more ambitious undertak-ing than privatizing a natural monopolyindustry, the key lessons neverthelesscarry over. Specifically, privatizing needsto go beyond the ex ante award stage toinclude an examination of possible expost implementation problems and, inconsideration of the differential hazards,to proceed selectively.

Recall, moreover, that the NIE oper-ates at two levels. Upon moving from

the level of governance to that of theinstitutional environment, the rules ofthe game come under review. The Levyand Spiller (1994, 1996) study of priva-tizing telecommunications in five coun-tries reveals that the decision to priva-tize and the nature of privatization turncritically on the condition and quality of

judicial independence, the division ofpowers between the executive and leg-islative branches, the competence ofthe regulatory bureaucracy, and con-tractual safeguards. Whether and howto privatize telecommunications shouldtherefore be made conditional on thesefeatures.

As Bernard Black, Reinier Kraakman,and Anna Tarassova detail in their pa-per on Russian Privatization and Cor-porate Governance: What Went Wrong(1999), the triumphant completion of

privatization in Russia was a premature

verdict. Thus, although privatization wasevidently a success for small firms, it

was deeply problematic and attended bymassive corruption in others. But forundue reliance on ex ante property rights

reasoning, some of these problemscould have been anticipated by lookingahead and examining the hazards of expost implementation. Greater apprecia-tion for the shortfalls of the institutionalenvironment in Russia would have led tomore cautious pronouncements (AndersAslund 1995). Whether added respectfor the rules of the game (to include anappreciation for the limited efficacy ofRussian law enforcement) would haveresulted in rule improvements in Russiacould be disputed. Arguably, however,the effort to reform Russia would haveproceeded in a more modest, slow,molecular, deliberative way.

None of this is to suggest that theNIE could have done it all. The Boyckoet al. team made heroic efforts. Myclaim is much more modest: the NIE isinformative and should be included as

part of the reform calculus.

5. Concluding Remarks

The new institutional economics is aboiling cauldron of ideas. Not only arethere many institutional research pro-grams in progress, but there are com-peting ideas within most of them. Withreference to history, for example, we seeNorth (1990) and Avner Greif (1999)pursuing complementary but separateagendas. The institutions of embedded-ness (Level 1) are an important but un-derdeveloped part of the story. Withintransaction cost economics we distin-guish between governance and measure-ment branches. The attributes of mixedownership modes (alliances, joint ven-tures, franchising, and the like) as wellas the mechanisms for supporting cred-

ible contracting between autonomous



17/19

firms are incompletely worked out. In-complete contracting of semi-formaland fully formal kinds differ in conse-quential ways, although the gap hasbeen closing. Evolutionary economics

of selectionist, population ecology, andontogenetic kinds are in progress. Pathdependency is a real and important con-dition, but its interpretation is activelydisputed. The merits of privatizationare real but are not uniform and needto be assessed with reference to boththe rules of the game and the play ofthe game. The firm is variously de-scribed in technological, contractual,and competence/ knowledge-based per-spectives. How best to describe humanactors is still unsettled, although evolu-tionary psychology holds promise. Poli-tics is judged with reference to a hypo-thetical ideal by some (North 1990) andin comparative institutional terms byothers (Williamson 1999). Efficiency ar-guments have mainly prevailed overpower interpretations because the latterare tautological, but power issues refuse

to go away. Bureaucracy remains apoorly understood condition no matter

what lens is brought to bear. Private or-dering approaches to contract havemade progressive headway, but legalrules remain important and their rela-tion to private ordering is incompletely

worked out. Positive political theory hasmade major conceptual advances, butan overarching understanding of poli-ties does not appear imminent. And thelist goes on.

The upshot is that, its many accom-plishments notwithstanding, there is a

vast amount of unfinished businessre-finements, extensions, new applications,more good ideas, more empirical test-ing, more fully formal theory. I con-clude that the new institutional eco-nomics is the little engine that could.Its best days lie ahead. Who could ask

for more?

REFERENCES

Alchian, Armen. 1961. Some Economics of Prop-erty, RAND Corporation D-2316.

Allen, Douglas, and Dean Leuck. 1999. The Roleof Risk in Contract Choice, J. Law Econ. Org.15:3, pp. 70436.

Arrow, Kenneth J. 1971. Essays in the Theory ofRisk-Bearing. Chicago: Markham.. 1987. Reflections on the Essays, in Ar-

row and the Foundations of the Theory of Eco- nomic Policy. George Feiwel, ed. NY: NYUPress, pp. 72734.

. 1999. Foreword in Firms, Markets andHierarchies. Glenn Carroll and David Teece,eds. NY: Oxford U. Press, pp. viiviii.

Aslund, Anders. 1995. How Russia Became a Mar-ket Economy. Washington, DC: Brookings Insti-tution.

Bajari, Patrick and Steven Tadelis. 1999. Incen-tives versus Transaction Costs, unpublished

paper, Stanford U.Banfield, E. C. 1958. The Moral Basis of a Back-ward Society. NY: Free Press.

Barnard, Chester. 1938. The Functions of the Ex-ecutive. Cambridge, MA: Harvard U. Press.

Baron, James and David Kreps. 1999. StrategicHuman Resources. NY: John Wiley & Sons.

Black, Bernard; Reinier Kraakman, and AnnaTarassova. 1999. Russian Privatization andCorporate Governance: What Went Wrong?unpublished manuscript, Stanford Law School.

Boorstin, Daniel. 1998. The Seekers. NY: RandomHouse.

Boycko, Maxim; Andrei Shleifer, and Robert

Vishny. 1995. Privatizing Russia. Cambridge,MA: MIT Press.Buckley, Peter and Mark Casson. 1976. The Fu-

ture of the Multinational Enterprise. London:Holmes and Meier.

Buttrick, John. 1952. The Inside Contracting Sys-tem,J. Econ. Hist. 12:3, pp. 20521.

Coase, Ronald. 1937. The Nature of the Firm,Economica, 4:6, pp. 386405.

. 1959. The Federal CommunicationsCommission, J. Law Econ. 2:2, pp. 140.

. 1960. The Problem of Social Cost, J.Law Econ. 3, pp. 144.

. 1964. The Regulated Industries: Discus-

sion, Amer. Econ. Rev. 54:3, pp. 19497.. 1988. The Firm, the Market, and the Law.

Chicago: U. Chicago Press.. 1992. The Institutional Structure of Pro-

duction, Amer. Econ. Rev. 82:4, pp. 71319.Commons, John R. 193233. The Problems of

Correlating Law, Economics and Ethics, Wisc.Law Rev. 8:1, pp. 326.

Crocker, Keith and Scott Masten. 1996. Regula-tion and Administered Contracts Revisited:Lessons from Transaction Cost Economics forPublic Utility Regulation, J. Regulatory Econ.9:1, pp. 539.

Dawkins, Richard. 1976. The Selfish Gene. NY:

Oxford U. Press.



18/19

Demsetz, Harold. 1967. Toward a Theory ofProperty Rights, Amer. Econ. Rev. 57:2, pp.34759.

. 1968. Why Regulate Utilities? J. LawEcon. 11, pp. 5566.

. 1969. Information and Efficiency: An-other Viewpoint,J. Law Econ. 12:1, pp. 122.

DiMaggio, Paul. 1994. Culture and Economy, inThe Handbook of Economic Sociology. NeilSmelser and Richard Swedberg, eds. Princeton:Princeton U. Press.

Dixit, Avinash. 1996. The Making of Economic Pol- icy: A Transaction Cost Politics Perspective.Cambridge: Cambridge U. Press.

Elster, Jon. 1994. Arguing and Bargaining in TwoConstituent Assemblies, unpublished manu-script.

Fogel, Robert. 1999. Catching Up with the Econ-omy,Amer. Econ. Rev. 89:1, pp. 121.

Furubotn, Erik and Rudolf Richter. 1991. TheNew Institutional Economics: An Assessment,

in The New Institutional Economics. Furubotnand Richter, eds. College Station, TX: TexasA&M U. Press.

. 1997. Institutions and Economic Theory.Ann Arbor: U. Michigan Press.

Gatignon, Hubert and Erin Anderson. 1988. TheMultinational Corporations Degree of Controlover Foreign Subsidiaries: An Empirical Test ofa Transaction Cost Explanation, J. Law Econ.Org . 4:2, pp. 30536.

Georgescu-Roegen, Nicholas. 1971. The Entropy Law and Economic Process. Cambridge, MA:Harvard U. Press.

Goldberg, Victor. 1976. Regulation and Adminis-

tered Contracts, Bell J. Econ. 7:2, pp. 42652.

Granovetter, Mark. 1985. Economic Action andSocial Structure: The Problem of Embedded-ness,Amer. J. Sociology, 91:3, pp. 481510.

Greif, Avner. Impersonal Exchange and the Ori-gin of Markets: From the Community Responsi-bility System to Individual Legal Responsibilityin Pre-Modern Europe, forthcoming in Com-

munities and Markets. M. Aoki and T. Hayami,eds.

Grossman, Gene, and Elhanan Helpman. 1999.Incomplete Contracts and Industrial Organiza-tion, NBER Working Paper 7303.

Grossman, Sanford and Oliver Hart. 1986. TheCosts and Benefits of Ownership: A Theory of

Vertical and Lateral Integration,J. Polit. Econ.94:4, pp. 691719.

Hart, Oliver. 1995. Firms, Contracts, and Finan-cial Structure. NY: Oxford U. Press.

Hart, Oliver and John Moore. 1999a. On the De-sign of Hierarchies, unpublished paper, Har-

vard U.. 1999b. Foundations of Incomplete Con-

tract, Rev. Econ. Stud. 66:1, pp. 11538.Hart, Oliver; Andrei Shleifer, and Robert Vishny.

1997. The Proper Scope of Government: The-ory and Application to Prisons, Quart. J. Econ.

112:4, pp. 112761.

Hayek, Friedrich. 1945. The Use of Knowledgein Society, Amer. Econ. Rev. 35:4, pp. 51930.

Henisz, Witold. 1998. The Institutional Environ-ment for International Investment, unpub-lished Ph.D. dissertation, UC Berkeley.

Holmstrom, Bengt. 1989. Agency Costs and Inno-

vation, J. Econ. Behav. Organ. 12:3, pp. 30527.

. 1999. The Firm as a Subeconomy, J.Law Econ. Org. 15:1, pp. 74102.

Holmstrom, Bengt and Paul Milgrom. 1991.Multi-Task Principal-Agent Analyses: Incen-tive Contracts, Asset Ownership, and Job De-sign, J. Law Econ. Org. 7: Special Issue, pp.2452.

. 1994. The Firm as an Incentive System,Amer. Econ. Rev. 84:4, pp. 97291.

Huntington, Samuel P. 1996. The Clash of Civili- zations and the Remaking of World Order. NY:Simon and Schuster.

Joskow, Paul. 1991. The Role of Transaction CostEconomics in Antitrust and Public Utility Regu-latory Policies, J. Law Econ. Org. 7: SpecialIssue, pp. 5383.

Kreps, David M. 1999. Markets and Hierarchiesand (Mathematical) Economic Theory, inFirms, Markets, and Hierarchies. Glenn Carrolland David Teece, eds. NY: Oxford U. Press, pp.12155.

Levy, Brian and Pablo Spiller. 1994. The Institu-tional Foundations of Regulatory Commit-ment, J. Law Econ. Org. 10:2, pp. 20146.

. 1996. Regulations, Institutions, and Com-mitment: Comparative Studies of Telecommuni-

cations. NY: Cambridge U. Press.Lyons, Bruce. 1996. Empirical Relevance of Effi-

cient Contract Theory: Inter-Firm Contracts,Oxford Rev. Econ. Policy, 12:4, pp. 2752.

Macneil, Ian R. 1974. The Many Futures of Con-tracts, Southern Cal. Law Rev. 47:2, pp. 691816.

Maskin, Eric and Jean Tirole. 1999. UnforeseenContingencies and Incomplete Contracts, Rev.Econ. Stud. 66:1, pp. 83114.

Masten, Scott. 1995. Introduction To Vol. II, inTransaction Cost Economics. Oliver Williamsonand Scott Masten, eds. Brookfield, VT: EdwardElgar.

. 1996. Case Studies in Contracting and Or-ganization. NY: Oxford U. Press.

Matthews, R. C. O. 1986. The Economics of In-stitutions and the Sources Of EconomicGrowth, Econ. J. 96:4, pp. 90318.

McMillan, John and Christopher Woodruff. 1999.Dispute Prevention without Courts in Viet-nam,J. Law Econ. Org. 15:3, pp. 63758.

Menard, Claude and Mary Shirley. 1999. Reform-ing Contractual Arrangements: Lessons fromUrban Water Systems in Six Developing Coun-tries, unpublished manuscript.

Mitchell, W. C. 1945. The National BureausFirst Quarter-Century, 25th Annual Report,

NBER.



19/19

Nee, Victor. 1998. Sources of the New Institu-tionalism, in The New Institutionalism in Soci-ology. Mary Brinton and Victor Nee, eds. NY:Russell Sage, pp. 116.

Newell, Alan. 1990. Unified Theories of Cognition.Cambridge, MA: Harvard U. Press.

North, Douglass. 1990. A Transaction Cost The-

ory of Politics,J. Theor. Politics, 2:4, pp. 35567.. 1991. Institutions, J. Econ. Perspectives,

5:1, pp. 97112.. 1994. Economic Performance through

Time, Amer. Econ. Rev. 84:3, pp. 35968.North, Douglas and Barry Weingast. 1989. Con-

stitutions and Commitment: The Evolution ofInstitutions Governing Public Choice in 17thCentury England, J. Econ. History, 49:4, pp.80332.

Olson, Mancur, Jr. 1996. Big Bills Left on the Side- walk: Why Some Nations Are Rich, and OthersAre Poor,J. Econ. Perspectives, 10:2, pp. 324.

Pinker, Steven. 1997. How The Mind Works. NY:

W. W. Norton.Posner, Richard. 1972. The Appropriate Scope of

Regulation in the Cable Television Industry,Bell J. Econ. 3:1, pp. 98129.

Priest, George. 1993. The Origins of UtilityRegulation and the Theories of Regulation De-bate, J. Law Econ. 36:1, pp. 289323.

Putnam, Robert D.; Robert Leonardi, and Raf-faella Y. Nanetti. 1993. Making DemocracyWork: Civic Traditions in Modern Italy. Prince-ton: Princeton U. Press.

Rindfleisch, A. and J. B. Heide. 1997. Transac-tion Cost Analysis: Past, Present, and FutureApplications, J. Marketing, 61:4, pp. 3054.

Riordan, Michael and Oliver Williamson. 1985.Asset Specificity and Economic Organization,Int. J. Industrial Org. 3:3, pp. 36578.

Rosenberg, Nathan and L. E. Birdzell. 1986. Howthe West Grew Rich: The Transformation of theIndustrial World. NY: Basic Books.

Rubinstein, Ariel. 1998. Review of Herbert Si-mon, An Empirically Based Microeconomics, J.Econ. Lit. 37:4, pp. 171112.

Shelanski, H. A. and P. G. Klein. 1995. EmpiricalResearch in Transaction Cost EconomicsAReview and Assessment, J. Law Econ. Org.11:2, pp. 33561.

Simon, Herbert. 1957. Models Of Man. NY: JohnWiley.

. 1985. Human Nature in Politics: The

Dialogue of Psychology with Political Science,Amer. Polit. Sci. Rev. 79:2, pp. 293304.

Smelser, Neil and Richard Swedberg, eds. 1994.Introduction, in The Handbook of EconomicSociology. Princeton: Princeton U. Press.

Stigler, George. 1968. The Organization of Indus-try. Homewood, IL: Richard D. Irwin.

Summers, Clyde. 1969. Collective Agreementsand the Law of Contracts, Yale Law J. 78:4,pp. 52575.

Wang, Shusheng and Tian Zhu. 2000. ContractLaw and the Boundary of the Firm, unpub-lished manuscript, Hong Kong U. Science andTechnology.

Whinston, Michael. 1997. On the TransactionCost Determinants of Vertical Integration, un-published manuscript, Northwestern U.

Williamson, Oliver E. 1975. Markets and Hierar-chies: Analysis and Antitrust Implications. NY:Free Press.

. 1976. Franchise Bidding for Natural Mo-nopoliesIn General and with Respect toCATV, Bell J. Econ. 7:1, pp. 73104.

. 1985. The Economic Institutions of Capi-talism. NY: Free Press.

. 1991. Comparative Economic Organi-zation: The Analysis of Discrete Structural Al-ternatives, Admin. Sci. Quart. 36:2, pp. 269

96.. 1996. The Mechanisms of Governance.

NY: Oxford U. Press.. 1998. Transaction Cost Economics: How

It Works; Where It Is Headed, De Economist,146:1, pp. 2358.

. 1999. Public and Private Bureaucracies:A Transaction Cost Economics Perspective, J.

Law Econ. Org. 15:1, pp. 30642.


Will Liam Son

Documents

Transcript of Will Liam Son