2022学霸代写怎么样指导美国留学生经济学ESSAY

2022学霸代写怎么样指导美国留学生经济学ESSAY

美国留学生经济学ESSAY指导Production, Information Costs, and Economic OrganizationArmen A. Alchian; Harold DemsetzThe American Economic Review, Vol. 62, No. 5. (Dec., 1972), pp. 777-795.Stable URL:The American Economic Review is currently published by American Economic Association.Your use of the JSTOR archive indicates your acceptance of JSTOR’s Terms and Conditions of Use, available at JSTOR’s Terms and Conditions of Use provides, in part, that unless you have obtained指导ESSAY prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content inthe JSTOR archive only for your personal, non-commercial use.Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained atEach copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academicjournals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community takeadvantage of advances in technology. For more information regarding JSTOR, please contact Thu Nov 15 10:46:02 2007Production, Information Costs, andEconomic OrganizationThe mark of a capitalistic society is thatresources are owned and allocated by suchnongovernmental organizations as firms,households, and markets. Resource ownersincrease productivity through cooperativespecialization and this leads to the demandfor economic organizations which facilitatecooperation. \$’hen a lumber millemploys a cabinetmaker, cooperation betweenspecialists is achieved within a firm,and when a cabinetmaker purchases woodfrom a lumberman, the cooperation takesplace across markets (or between firms).Two important problems face a theory ofeconomic organization-to explain theconditions that determine whether thegains from spkcialization and cooperativeproduction can better be obtained withinan organization like the firm, or acrossmarkets, and to explain the structure ofthe organization.I t is common to see the firm characterizedby the power to settle issues byfiat, by authority, or by disciplinary actionsuperior to that available in the conventionalmarket. This is delusion. The firmdoes not own all its inputs. I t has nopower of fiat, no authority, no disciplinaryaction any different in the slightest degreefrom ordinary market contracting betweenany two people. I can "punish" youonly by withholding future business or byseeking redress in the courts for any failureto honor our exchange agreement. That isexactly all that any employer can do. He* Professors of economics at the Fniversit~ of California,#p#分页标题#e#1.0s Xngeles .\ckno~\ledgment is made for financia1aid from the F. I,ill> Endowment, Inc, grant tocCL.4 for research in the behavioral effects of propertyrightscan fire or sue, just as I can fire my grocerby stopping purchases from him or suehim for delivering faulty products. Whatthen is the content of the presumed powerto manage and assign workers to varioustasks? Exactly the same as one little consumer’spower to manage and assign hisgrocer to various tasks. The single consumercan assign his grocer to the task ofobtaining whatever the customer can inducethe grocer to provide at a price acceptableto both parties. That is preciselyall that an employer can do to an employee.To speak of managing, directing,or assigning workers to various tasks is adeceptive way of noting that the employercontinually is involved in renegotiation ofcontracts on terms that must be acceptableto both parties. Telling an employee totype this letter rather than to file thatdocument is like my telling a grocer tosell me this brand of tuna rather than thatbrand of bread. I have no contract to continueto purchase from the grocer andneither the employer nor the employee isbound by any contractual obligations tocontinue their relationship. Long-termcontracts between employer and employeeare not the essence of the organizationwe call a firm. My grocer can counton my returning day after day and purchasinghis services and goods even withthe prices not always marked on the goods-because I know what they are-and headapts his activity to conform to mydirections to him as to what I want eachday . . . he is not my employee.iyherein then is the relationship betweena grocer and his differentfrom that between a grocer and his cus- tomers? It is in a team use of inputs and acentralized position of some party in thecontractual arrangements of all other inputs.It is the centralized contractz~al agentin a team ProductiCLle process-not somesuperior authoritarian directive or disciplinarypower. Exactly what is a teamprocess and why does it induce the contractualform, called the firm? These problemsmotivate the inquiry of this paper.I. The Metering Problem’The economic organization throughwhich input owners cooperate will makebetter use of their comparative advantagesto the extent that it facilitates the paymentof rewards in accord with productivity.If rewards were random, and withoutregard to productive effort, no incentiveto productive effort would be providedby the organization; and if rewardswere negatively correlated with productivitythe organization would be subjectto sabotage. Two key demands are placedon an economic organization-meteringinput productivity and metering rewards.’Metering problems sometimes can beresolved well through the exchange ofproducts across competitive markets, becausein many situations markets yield ahigh correlation between rewards andproductivity. If a farmer increases his outputof wheat by 10 percent at the prevailingmarket price, his receipts also increaseby 10 percent. This method of organizingeconomic activity meters theoutput directly, reveals the marginal productand apportions the rewards to resourceowners in accord with that directmeasurement of their outputs. The successof this decentralized, market exchange inpromoting productive specialization requiresthat changes in market rewards fallMeter means to measure and also to apportion. Onecan meter (measure) output and one can also meter(control) the output. LVe use the word to denote both;the context should indicate which.on those responsible for changes in oz~tpz~t.~’I’he classic relationship in economicsthat runs from marginal productivity tothe distribution of income implicitly assumes#p#分页标题#e#the existence of an organization, beit the market or the firm, that allocatesrewards to resources in accord with theirproductivity. The problem of economicorganization, the economical means ofmetering productivity and rewards, is notconfronted directly in the classical analysisof production and distribution. Instead,that analysis tends to assume sufficientlyeconomic-or zero cost-means,as if productivity automatically createdits reward. ‘IVe conjecture the direction ofcausation is the reverse-the specific sys-A producer’s wealth would be reduced by the presentcapitalized value of the future income lost by loss ofreputation. Reputation, i.e., credibility, is an asset,which is another way of saying that reliable informationabout expected performance is both a costly and avaluable good. For acts of Cod that interfere with contractperformance, both parties have incentives toreach a settlement akin to that which would have beenreached if such events had been covered by specific contingencyclauses. The reason, again, is that a reputationfor "honest" dealings-i.e., for actions similar to thosethat would probably have been reached had the contractprovided this contingency-is wealth.Almost every contract is open-ended in that manycontingencies are uncovered. For example, if a firedelays production of a promised product by A to B,and if B contends that A has not fulfilled the contract,how is the dispute settled and what recompense, if any,does A grant to B? A person uninitiated in such questionsmay be surprised by the extent to which contractspermit either party to escape performance or to nullifythe contract. In fact, it is hard to imagine any contract,which, when taken solely in terms of its stipulations,could not be evaded by one ofthe parties. Yet that isthe ruling, viable type of contract. LL:hy? Undoubtedlyhttp://www.ukassignment.org/daixieEssay/meiguoessaydaixie/the best discussion that we have seen on this question iby Stewart Macaulay.There are means not only of detecting or preventingcheating, but also for deciding how to allocate the lossesor gains of unpredictable events or quality of itemsexchanged. Sales contracts contain warranties, guarantees,collateral, return privileges and penalty clauses forspecific nonperformance. These are means of assignmentof risks of losses of cheating. A lower price without warranty-an "as is" purchase-places more of the risk onthe buyer while the seller buys insurance against lossesof his "cheating." On the other hand, a warranty orreturn privilege or service contract places more risk onthe seller with insurance being bought by the buyer..ALCHT-AX AND DEMSETZ: ECONOMIC ORGANIZ.4TION 779tem of rewarding which is relied uponstimulates a particular productivity response.If the economic organizationmeters poorly, with rewards and productivityonly loosely correlated, then productivitywill be smaller; but if the economicorganization meters well productivitywill be greater. What makes meteringdifficult and hence induces means ofeconomizing on metering costs?11. Team ProductionTwo men jointly lift heavy cargo intotrucks. Solely by observing the total#p#分页标题#e#weight loaded per day, it is impossible todetermine each person’s marginal productivity.\I’ith team production it isdifficult, solely by observing total output,to either define or determine each individual’scontribution to this output of thecooperating inputs. The output is yieldedby a team, by definition, and it is not asum of separable outputs of each of itsmembers. Team production of Z involvesat least two inputs, X, and XI, withd2Z1dX,dX,#0.3 The production functionis not separable into two functionseach involving only inputs X, or only inputsXI. consequently there is no sumof Z of two separable functions to treatas the Z of the team production function.(:In example of a separable case is Z= aX: +b,Y; which is separable into Z,= aX; andZj= blY,L, and Z= Z,+Z,. ‘This is not teamproduction.) ‘There exist production techniquesin which the Z obtained is greaterthan if X, and XI had produced separableZ. ‘l’eam production will be used if ityields an output enough larger than thesum of separable production of Z to coverthe costs of organizing and discipliningteam members-the topics of this paper.4The function is separable into additive functions ifthe cross partial derivative is zero, i.e., if a2Z/aAY,a?i,=0.’ LVith suficient generality of notation and conceptionthis team production function could be formulatedas a case of the generalized production function interpretationgiven by our colleague, E. A. Thompson.Usual explanations of the gains fromcooperative behavior rely on exchangeand production in accord with the comparativeadvantage specialization principlewith separable additive production.However, as suggested above there is asource of gain from cooperative activityinvolving working as a team, wherein individualcooperating inputs do not yieldidentifiable, separate products which canbe sz~mmed to measure the total output.For this cooperative productive activity,here called "team" production, measuringmargiltal productivity and making paymentsin accord therewith is more expensiveby an order of magnitude than forseparable production functions.Team production, to repeat, is productionin which 1) several types of resourcesare used and 2) the product is not a sumof separable outputs of each cooperatingresource. .An additional factor creates ateam organization problem-3) not all resourcesused in team production belong toone person.JI’e do not inquire into why all thejointly used resources are not owned byone person, but instead into the types oforganization, contracts, and informationaland payment procedures used amongowners ofteamed inputs. With respect tothe one-owner case, perhaps it is sufficientmerely to note that (a) slavery is prohibited,(b) one might assume risk aversionas a reason for one person’s not borrowingenough to purchase all the assetsor sources of services rather than rentingthem, and (c) the purchase-resale spreadmay beso large that costs of short-termownership exceed rental costs. Our problemis viewed basically as one of organizationamong different people, not of the#p#分页标题#e#physical goods or services, however muchthere must be selection and choice of combinationof the latter.How can the members of a team be rewardedand induced to work efficiently?780 THE *IMERICAN ECONOMIC REVIE%’In team production, marginal products ofcooperative team members are not sodirectly and separably (i.e., cheaply) observable.IThat a team offers to themarket can be taken as the marginalproduct of the team but not of the teammembers. The costs of metering or ascertainingthe marginal products of the team’smembers is what calls forth new organizationsand procedures. Clues to each input’sproductivity can be secured by observingbehatliar of individual inputs. ‘ClThen liftingcargo into the truck, how rapidly does aman move to the nest piece to be ldaded,how many cigarette breaks does hetake, does the item being lifted tilt downwardtoward his side?If detecting such behavior were costless,neither party would have an incentiveto shirk, because neither could imposethe cost of his shirking on the other (iftheir cooperation was agreed to voluntarily).Hut since costs must be incurredto monitor each other, each input ownerwill have more incentive to shirk when heworks as part of a team, than if his performancecould be monitored easily or ifhe did not work as a team. If there is anet increase in productivity available byteam production, net of the metering costassociated with disciplining the team, thenteam production will be relied upon ratherthan a multitude of bilateral exchange ofseparable individual outputs.Both leisure and higher income enter aperson’s utility functi0n.j Hence, eachperson should adjust his work and realizedreward so as to equate the marginal rate ofsubstitution between leisure and productionof real output to his marginal rate ofsubstitution in consumption. That is, hewould adjust his rate of work to bring hisdemand prices of leisure and output toequality with their true costs. However,llore precisely: "if anything other than pecuniaryincome enters his utility function." Leisure stands for allnonpecuniary income for simplicity of e~position.with detection, policing, monitoring, measuringor metering costs, each person willbe induced to take more leisure, becausethe effect of relaxing on his realized (reward)rate of substitution between outputand leisure will be less than the effect onthe trzbe rate of substitutioil. His realizedcost of leisure will fall more than the truecost of leisure, so he "buys" more leisure(i.e., more nonpecuniary reward).If his relaxation cannot be detected perfectlyat zero cost, part of its effects willbe borne by others in the team, thus makinghis realized cost of relaxation less thanthe true total cost to the team. The difficultyof detecting such actions permitsthe private costs of his actions to be lessthan their full costs. Since each personresponds to his private realizable rate ofsubstitution (in production) rather thanthe true total (i.e., social) rate, and solong as there are costs for other people todetect his shift toward relaxation, it willnot pay (them) to force him to readjustcompletely by making him realize thetrue cost. Only enough efforts will bemade to equate the marginal gains of detectionactivity with the marginal costs of#p#分页标题#e#detection; and that implies a lower rate ofproductive effort and more shirking thanin a costless monitoring, or measuring,world.In a university, the faculty use officetelephones, paper, and mail for personaluses beyond strict university productivity.The university administrators could stopsuch practices by identifying the responsibleperson in each case, but they can do soonly at higher costs than administratorsare willing to incur. The extra costs ofidentifying each party (rather than merelyidentifying the presence of such activity)wouid exceed the savings from din~inishedfaculty "turpitudinal peccadilloes." Sothe faculty is allowed some degree of"privileges, perquisites, or fringe benefits.",2nd the total of the pecuniary wages paidALCHIAN ,AND DEMSETZ: ECONOMIC ORGANIZATION 781is lower because of this irreducible (atacceptable costs) degree of amenity-seizingactivity. Pay is lower in pecuniaryterms and higher in leisure, conveniences,and ease of work. But still every personwould prefer to see detection made moreeffective (if it were somehow possible tomonitor costlessly) so that he, as part ofthe now more effectively producing team,could thereby realize a higher pecuniarypay and less leisure. If everyone could, atzero cost, have his reward-realized ratebrought to the true production possibilityreal rate, all could achieve a more preferredposition. But detection of the responsibleparties is costly; that cost actslike a tax on work rewards6 Viable shirkingis the result.It’hat forms of organizing team productionwill lower the cost of detecting "performance"(i.e., marginal productivity)and bring personally realized rates ofsubstitution closer to true rates of substitution?Market competition, in principle,could monitor some team production.(It already organizes teams.) Inputowners who are not team members canoffer, in return for a smaller share of theteam’s rewards, to replace excessively (i.e.,overpaid) shirking members. Rlarket competitionamong potential team memberswould determine team membership andindividual rewards. There would be noteam leader, manager, organizer, owner,or employer. For such decentralized organizationalcontrol to work, outsiders,possibly after observing each team’s totalDo not assume that the sole result of the cost ofdetecting shirking is one form of payment (more leisureand less take home money). \Vith several members ofthe team, each has an incentive to cheat against eachother by engaging in more than the average amount ofsuch leisure if the employer can not tell at zero costwhich employee is taking more than average. As a#p#分页标题#e#result the total ~roductivitv of the team is lowered.Shirking detection costs thus change the form of paymentand also result in lower total rewards. Because thecross I,artial derivatives are ,,os,tive, shirking reducesother people’s marginal products.output, can speculate about their capabilitiesas team members and, by a marketcompetitive process, revised teams withgreater productive ability will be formedand sustained. Incumbent members will beconstrained by threats of replacement byoutsiders offering services for lower rewardshares or offering greater rewards to theother members of the team. Any teammember who shirked in the expectationthat the reduced output effect would notbe attributed to him will be displaced ifhis activity is detected. Teams of productiveinputs, like business units, wouldevolve in apparent spontaneity in themarket-without any central organizingagent, team manager, or boss.But completely effective control cannotbe expected from individualized marketcompetition for two reasons. First, forthis competition to be completely effective,new challengers for team membershipmust know where, and to what extent,shirking is a serious problem, i.e., knowthey can increase net output as comparedwith the inputs they replace. To the extentthat this is true it is probably possible forexisting fellow team members to reccgnizethe shirking. But, by definition, the detectionof shirking by observing team outputis costly for team production. Secondly, assumethe presence of detection costs, andassume that in order to secure a place onthe team a new input owner must accepta smaller share of rewards (or a promise toproduce more). Then his incentive to shirkwould still be at least as great as the incentivesof the inputs replaced, because hestill bears less than the entire reduction inteam output for which he is responsible.111. The Classical FirmOne method of reducing shirking is forsomeone to specialize as a monitor to checkthe input performance of team members.” \Vhat is meant by performance? Input energy, initiative,work attitude, ~ e r s ~ i r a t i orna,t e of euhaustion?(Contrnut i i )THE ,AMERICAN ECONOMIC REVIEWBut who will monitor the monitor? Oneconstraint on the monitor is the aforesaidmarket competition offered by othermonitors, but for reasons already given,that is not perfectly effective. Anotherconstraint can be imposed on the monitor:give him title to the net earnings of theteam, net of payments to other inputs.If owners of cooperating inputs agree withthe monitor that he is to receive anyresidual product above prescribed amounts#p#分页标题#e#(hopefully, the marginal value products ofthe other inputs), the monitor will havean added incentive not to shirk as amonitor. Specialization in monitoring plusreliance on a residual claimant status willreduce shirking; but additional links areneeded to forge the firm of classical economictheory. How will the residualclaimant monitor the other inputs?iVe use the term monitor to connoteseveral activities in addition to its disciplinaryconnotation. It connotes measuringoutput performance, apportioningrewards, observing the input behavior ofinputs as means of detecting or estimatingtheir marginal productivity and giving assignmentsor instructions in what to doand how to do it. (It also includes, as weshall show later, authority to terminateor revise contracts.) Perhaps the contrastbetween a football coach and team captainis helpful. The coach selects strategiesand tactics and sends in instructionsabout what plays to utilize. The captainis essentially an observer and reporter ofOr output? It is the latter that is sought–the eject oroutput. But performance is nicely ambiguous because itsuggests both input and output. I t is nicely ambiguousbecause as we shall see, sometimes by inspecting a teammember’s input activity we can better judge his outputeffect, perhaps not with complete accuracy hut betterthan by watching the output of the team. I t is not alwaysthe case that watching input activity is the only or bestmeans of detecting, measuring or monitoring outputeffects of each team member, but in some cases it is auseful way. For the moment the word performanceglosses over these aspects and facilitates concentrationon other issues.the performance at close hand of the members.The latter is an inspector-stewardand the former a supervisor manager.For the present all these activities are includedin the rubric "monitoring." Allthese tasks are, in principle, negotiableacross markets, but we are presuming thatsuch market measurement of marginalproductivities and job reassignments arenot so cheaply performed for team production.And in particular our analysissuggests that it is not so much the costsof spontaneously negotiating contracts inthe markets among groups for team productionas it is the detection of the performanceof individual members of theteam that calls for the organization notedhere.The specialist who receines the residualrewards will be the monitor of the membersof the team (i.e., will manage the useof cooperative inputs). The monitor earnshis residual through the reduction inshirking that he brings about, not only bythe prices that he agrees to pay the owners#p#分页标题#e#of the inputs, but also by observing anddirecting the actions or uses of these inputs.Managing or examining the ways towhich inputs are used in team productionis a method of metering the marginal productivityof individual inputs to the team’soutput.To discipline team members and reduceshirking, the residual claimant must havepower to revise the contract terms and incentivesof indioidual members withouthaving to terminate or alter every otherinput’s contract. Hence, team memberswho seek to increase their productivitywill assign to the monitor not only theresidual claimant right but also the rightto alter individual membership and performanceon the team. Each team member,of course, can terminate his ownmembership (i.e., quit the team), butonly the monitor may unilaterally terminatethe membership of any of theALCHIAS ASD DEMSETZ: ECOSOMIC ORGANTZ=\TTOSother members without necessarily terminatingthe team itself or his associationwith the team; and he alone can expand orreduce membership, alter the mix ofmembership, or sell the right to be theresidual claimant-monitor of the team. Itis this entire bundle of rights: 1) to be aresidual claimant; 2) to observe inputbehavior; 3) to be the central party commonto all contracts with inputs; 4) toalter the membership of the team; and5) to sell these rights, that defines theoelnership (or the employer) of the classical(capitalist, free-enterprise) firm. Thecoalescing of these rights has arisen, ouranalysis asserts, because it resolves theshirking-information problem of teamproduction better than does the noncentralizedcontractual arrangement.The relationship of each team memberto the owner of the firm (i.e., the partycommon to all input contracts and theresidual claimant) is simply a (‘quid proquo" contract. Each makes a purchaseand sale. The employee (‘orders" the ownerof the team to pay him money in the samesense that the employer directs the teammember to perform certain acts. Theemployee can terminate the contract asreadily as can the employer, and longtermcontracts, therefore, are not an essentialattribute of the firm. Nor are((authoritarian," "dictational," or (‘fiat"attributes relevant to the conception of thefirm or its efficiency.In summary, two necessary conditionsexist for the emergence of the firm on theprior assumption that more than pecuniarywealth enter utility functions: 1) It ispossible to increase productivity throughteam-oriented production, a productiontechnique for which it is costly to directlymeasure the marginal outputs of the cooperating#p#分页标题#e#inputs. This makes it moredifficult to restrict shirking through simplemarket exchange between cooperating inputs.2) It is economical to estimate marginalproductivity by observing or specifyinginput behavior. The simultaneous occurrenceof both these preconditions leadsto the contractual organization of inputs,known as the classical capitalist firms with(a) joint input production, (b) several inputowners, (c) one party who is commonto all the contracts of the joint inputs, (d)who has rights to renegotiate any input’scontract independently of contracts withother input owners, (e) who holds theresidual claim, and (f) who has the rightto sell his central contractual residuals t a t u ~ . ~Other Theories of the FirmAt this juncture, as an aside, we brieflyplace this theory of the firm in the contextsof those offered by Ronald Coase andFrank KnighLg Our view of the firm is notnecessarily inconsistent with Coase’s; weattempt to go further and identify refutableimplications. Coase’s penetrating insightis to make more of the fact thatmarkets do not operate costlessly, and herelies on the cost of using markets to formcontracts as his basic explanation for theexistence of firms. IVe do not disagree withthe proposition that, ceteris paribus, thehigher is the cost of transacting acrossmarkets the greater will be the comparativeadvantage of organizing resourceswithin the firm; it is a difficult propositionto disagree with or to refute. LYe couldwith equal ease subscribe to a theory ofthe firm based on the cost of managing,for surely it is true that, ceteris paribus,the lower is the cost of managing thegreater will be the comparative advantageof organizing resources within the firm. TOmove the theory forward, it is necessaryto know what is meant by a firm and toRemoval of (b) converts a capitalist proprietary firmto a socialist firm.9 Recognition must also be made to the seminal inquiriesby Morris Silver and Richard Auster, and byH. R. Malmgren.784 THE AMERIC.43 ECOKOMIC REVIEWexplain the circumstances under whichthe cost of "managing" resources is lowrelative to the cost of allocating resourcesthrough market transaction. The conceptionof and rationale for the classical firmthat we propose takes a step down thepath pointed out by Coase toward thatgoal. Consideration of team production,team organization, difficulty in meteringoutputs, and the problem of shirking areimportant to our explanation but, so faras we can ascertain, not in Coase’s. Coase’sanalysis insofar as it had heretofore beendeveloped would suggest open-ended contracts#p#分页标题#e#but does not appear to imply anythingmore-neither the residual claimantstatus nor the distinction between employeeand subcontractor status (nor anyof the implications indicated below). Andit is not true that employees are generallyemployed on the basis of long-term contractualarrangements any more than on aseries of short-term or indefinite lengthcontracts.The importance of our proposed additionalelements is revealed, for example,by the explanation of why the person towhom the control monitor is responsiblereceives the residual, and also by ourlater discussion of the implications aboutthe corporation, partnerships, and profitsharing. These alternative forms for organizationof the firm are difficult to resolveon the basis of market transactioncosts only. Our exposition also suggests adefinition of the classical firm-somethingcrucial that was heretofore absent.In addition, sometimes a technologicaldevelopment will lower the cost of markettransactions while, at the same time, itexpands the role of the firm. When the"putting out" system was used for weaving,inputs were organized largely throughmarket negotiations. LTith the developmentof efficient central sources of power,it became economical to perform weavingin proximity to the power source and toengage in team production. The bringingin of weavers surely must have resulted ina reduction in the cost of negotiating(forming) contracts. Yet, what we observeis the beginning of the factory systemin which inputs are organized withina firm. \Thy? The weavers did not simplymove to a common source of power thatthey could tap like an electric line, purchasingpower while they used their ownequipment. Now team production in thejoint use of equipment became more important.l’he measurement of marginalproductivity, which now involved interactionsbetween workers, especially throughtheir joint use of machines, became moredifficult though contract negotiating costwas reduced, while managing the behaviorof inputs became easier because of the increasedcentralization of activity. Thefirm as an organization expanded eventhough the cost of transactions was reducedby the advent of centralized power.l’he same could be said for modern assemblylines. Hence the emergence ofcentral power sources expanded the scopeof productive activity in which the firmenjoyed a comparative advantage as anorganizational form.Some economists, following Knight,have identified the bearing of risks ofwealth changes with the director or centralemployer without explaining why that isa viable arrangement. Presumably, the#p#分页标题#e#more risk-averse inputs become employeesrather than owners of the classical firm.Risk averseness and uncertainty ~ l i t hregardto the jirm’s fortulzes have little, ifanything, to do with our explanation althoughit helps to explain why all resourcesin a team are not owned by oneperson. That is, the role of risk taken inthe sense of absorbing the windfalls thatbuffet the firm because of unforeseen competition,technological change, or fluctuationsin demand are not central to ourtheory, although it is true that imperfectknowledge and, therefore, risk, in thissense of risk, underlie the problem ofALCHIAS -4NT) DEJ ISETZ: ECOSOMIC ORGANIZ-4TIOS 785monitoring team behavior. We deduce thesystem of paying the manager with aresidual claim (the equity) from the desireto have efficient means to reduce shirkingso as to make team production economicaland not from the smaller aversion to therisks of enterprise in a dynamic economy.We conjecture that "distribution-of-risk"is not a valid rationale for the existenceand organization of the classical firm.Although we have emphasized teamproduction as creating a costly meteringtask and have treated team production asan essential (necessary?) condition for thefirm, would not other obstacles to cheapmetering also call forth the same kind ofcontractual arrangement here denoted asa firm? For example, suppose a farmerproduces wheat in an easily ascertainedquantity but with subtle and difficult todetect quality variations determined byhow the farmer grew the wheat. -4 verticalintegration could allow a purchaser tocontrol the farmer’s behavior in order tomore economically estimate productivity.But this is not a case of joint or teamproduction, unless "information" can beconsidered part of the product. (While agood case could be made for that broaderconception of production, we shall ignoreit here.) Instead of forming a firm, a buyercan contract to have his inspector on thesite of production, just as home builderscontract with architects to supervise buildingcontracts; that arrangement is not afirm. Still, a firm might be organized inthe production of many products whereinno team production or jointness of use ofseparately owned resources is involved.This possibility rather clearly indicatesa broader, or complementary, approachto that which we have chosen. 1) As we doin this paper, it can be argued that thefirm is the particular policing deviceutilized when joint team production ispresent. If other sources of high policingcosts arise, as in the wheat case just indicated,some other form of contractual arrangement#p#分页标题#e#\+rill be used. Thus to eachsource of informational cost there may bea different type of policing and contractualarrangement. 2) On the other hand, onecan say that where policing is difficultacross markets, various forms of contractualarrangements are devised, but there isno reason for that known as the firm to beuniquely related or even highly correlatedwith team production, as defined here. I tmight be used equally probably and viablyfor other sources of high policing cost. Wehave not intensively analyzed othersources, and we can only note that ourcurrent and readily revisable conjectureis that 1) is valid, and has motivated us inour current endeavor. In any event, thetest of the theory advanced here is to seewhether the conditions we have identifiedare necessary for firms to have long-runviability rather than merely births withhigh infant mortality. Conglomerate firmsor collections of separate production agenciesinto one owning organization can be interpretedas an investment trust or investmentdiversification device-probablyalong the lines that motivatedKnight’s interpretation. A holding companycan be called a firm, because of thecommon association of the word firm withany ownership unit that owns incomesources. The term firm as commonly usedis so turgid of meaning that we can nothope to explain every entity to which thename is attached in common or even technicalliterature. Instead, we seek to identifyand explain a particular contractualarrangement induced by the cost of informationfactors analyzed in this paper.IV. Types of FirmsA. Projit-Sharing FirmsExplicit in our explanation of thecapitalist firm is the assumption that thecost of managing the team’s inputs by acentral monitor, who disciplines himselfbecause he is a residual claimant, is low786 THE AMEKICXK ECONOhf IC REYIEKrelative to the cost of metering the marginaloutputs of team members.If we look within a firm to see whomonitors- -hires, fires, changes, promotes,and renegotiates-we should find him beinga residual claimant or, at least, onewhose pay or reward is more than anyothers correlated with fluctuations in theresidual value of the firm. They morelikely will have options or rights or bonusesthan will inputs with other tasks.An implicit "auxiliary" assumption ofour explanation of the firm is that thecost of team production is increased if theresidual claim is not held entirely by thecentral monitor. That is, we assume thatif profit sharing had to be relied upon forall team members, losses from the resultingincrease in central monitor shirking#p#分页标题#e#woulti exceeti the output gains from theincreased incentives of other team membersnot to shirk. If the optimal team sizeis only two owners of inputs, then anequal division of profits and losses betweenthem will leave each with strongerincentives to reduce shirking than if theoptimal team size is large, for in the lattercase only a smaller percentage of the lossesoccasioned by the shirker will be borne byhim. Incentives to shirk are positively relatedto the optimal size of the team underan equal profit-sharing scheme. loThe preceding does not imply that profitsharing is never viable. Profit sharing toencourage self-policing is more appropriatefor small teams. And, intieeti, where inputowners are free to make whatever contractualarrangements suit them, as generallyis true in capitalist economies, profitsharing seems largely limited to partner-IVhile the degree to which residual clainis are centralizedwill afiect the size of the team, this will be onlyone of many factors that determine team size, so as anapproximation, we can treat team size as exogenouslydetermined. Cnder certain assumptions about theshape of the "typical" utility function, the incentive toavoid shirking with unequal profit-sharing can 1)e measuredhy the Herfindahl index.’0ships with a relatively small number ofuctioe" partners. Another advantage ofsuch arrangements for smaller teams isthat it permits more effective reciprocalmonitoring among inputs. Rlonitoringneed not be entirely specialized.Profit sharing is more viable if smallteam size is associated with situationswhere the cost of specialized managementof inputs is large relative to the increasedproductivity potential in team effort. IVeconjecture that the cost of managing teaminputs increases if the productivity of ateam member is difficult to correlate withhis behavior. In "artistic" or "professional"work, watching a man’s activitiesis not a good clue to what he is actuallythinking or doing with his mind. 1I:hile itis relatively easy to manage or direct theloading of trucks by a team of dockworkers where input activity is so highlyrelated in an obvious way to output, it ismore difficult to manage and direct alawyer in the preparation anti presentationof a case. Dock workers can be directedin detail without the monitor himselfloading the truck, and assembly lineworkers can be monitoreti by varying thespeed of the assembly line, but detaileddirection in the preparation of a law casewoulti require in much greater tiegree thatthe monitor prepare the case himself. Asa result, artistic or professional inputs,#p#分页标题#e#such as lawyers, advertising specialists,and doctors, will be given relatively freerreign with regarti to intiividual behavior.If the management of inputs is relativelycostly, or ineffective, as it would seem tobe in these cases, but, nonetheless if teameffort is more productive than separableproduction with exchange across markets,then there will develop a tendency to useprofit-sharing schemes to provide incentivesto avoid shirking.’"l1 The use of the word active will he clarified in ourdiscussion of the corporation, which follows below,l2 Some sharing contracts, like crop sharing, or rentalAILCHIAX AXD I3EAISETZ: ECONOAIIC ORGASIZATIONB. Socialist Firms\Ye have analyzed the classical proprietorshipand the profit-sharing firms in thecontext of free association and choice ofeconomic organization. Such organizationsneed not be the most viable when politicalconstraints limit the forms of organizationthat can be chosen. I t is one thing to haveprofit sharing when professional or artistictalents are used by small teams. But ifpolitical or tax or subsidy considerationsinduce profit-sharing techniques whenthese are not otherwise economicallyjustified, then additional managementtechniques will be developed to help reducethe degree of shirking.For example, most, if not all, firms inJugoslavia are owned by the employees inthe restricted sense that all share in theresidual. This is true for large firms andfor firms which employ nonartistic, ornonprofessional, workers as well. \Yith adecay of political constraints, most ofthese firms could be expected to rely onpaid wages rather than shares in the residual.This rests on our auxiliary assumptionthat general sharing in the resiclualresults in losses from enhanced shirkingby the monitor that exceed the gains fromreduced shirking by residual-sharing employees.If this were not so, profit sharingwith employees should have occurred morefrequently in IYestern societies where suchorganizations are neither banned norpreferred politically. \There residual sharingby employees is politically imposed,as in Jugoslavia, we are led to expect thatsome management technique will arise toreduce the shirking by the central monitor,a technique that will not be found frequentlyin \Yestern societies since themonitor retains all (or much) of the rep—- — -paJments based on gross sales In retail stores, comeclose to profit sharing Honeber, it is gross output 5haring rather than proht shar~ng If e are unal~le to sl)ecif\the impl~c a t~oonf s t he d~fference\ I e refer the reader toS iT Cheungsidual in the \Test and profit sharing is#p#分页标题#e#largely confined to small, professionalartisticteam production situations. IYe dofind in the larger scale residual-sharingfirms in Jugoslavia that there are employeecommittees that can recommend(to the state) the termination of a manager’scontract (veto his continuance)with the enterprise. IYe conjecture thatthe workers’ committee is given the rightto recommend the termination of themanager’s contract precisely because thegeneral sharing of the residual increasesiiexcessively" the manager’s incentive toshirk.13C. The CorporationAll firms must initially acquire commandover some resources. The corporationdoes so primarily by selling promisesof future returns to those who (as creditorsor owners) provide financial capital. Insome situations resources can be acquiredin advance from consumers by promisesof future delivery (for example, advancesale of a proposed book). Or where thefirm is a few artistic or professional persons,each can "chip in" with time andtalent until the sale of services brings inrevenues. For the most part, capital canbe acquired more cheaply if many (riskaverse)investors contribute small portionsto a large investment. The economiesof raising large sums of equity capital inthis way suggest that modifications in therelationship among corporate inputs arerequired to cope with the shirking probleml3 Incidentally, investment activity will be changed.The inal~ility to capitalize the investment value as"take-home" proviate property ~c*enlt/orf the members ofthe firm means that the benefits of the investment musthe taken as annual income 11y those who are employedat the time of the income. Investment will he confinedmore to those with shorter life and with higher rates orpay-offs if the alternative of investing is paying out thefirm’s income to its employees to take home and use asprivate propert). 1,’or a development of this proposition,see the pal)ers hy Eirik 1:urohotn and SvetozarPejovich, and t ~ yPe jovich.788 THE AMERICAN ECOKOMIC REVIEWthat arises with profit sharing among largenumbers of corporate stockholders. Onemodification is limited liability, especiallyfor firms that are large relative to a stockholder’swealth. I t serves to protect stockholdersfrom large losses no matter howthey are caused.If every stock owner participated ineach decision in a corporation, not onlywould large bureaucratic costs be incurred,but many would shirk the task ofbecoming well informed on the issue to bedecided, since the losses associated withunexpectedly bad decisions will be bornein large part by the many other corporate#p#分页标题#e#shareholders. More effective control ofcorporate activity is achieved for mostpurposes by transferring decision authorityto a smaller group, whose main functionis tonegotiatewith and manage(renegotiatewith) the other inputs of the team. Thecorporate stockholders retain the authorityto revise the membership of the managementgroup and over major decisions thataffect the structure of the corporation orits dissolution.As a result a new modification of partnershipsis induced-the right to sale ofcorporate shares without approval of anyother stockholders. ,Any shareholder canremove his wealth from control by thosewith whom he has differences of opinion.Rather than try to control the decisionsof the management, which is harder to dowith many stockholders than with only afew, unrestricted salability provides amore acceptable escape to each stockholderfrom continued policies with whichhe disagrees.Indeed, the policing of managerialshirking relies on across-market competitionfrom new groups of would-be managersas well as competition from memberswithin the firm who seek to displace existingmanagement. In addition to competitionfrom outside and inside managers,control is facilitated by the temporarycongealing of share votes into voting blocsowned by one or a few contenders. Proxybattles or stock-purchases concentrate thevotes required to displace the existingmanagement or modify managerial policies.But it is more than a change in policy thatis sought by the newly formed financialinterests, whether of new stockholders ornot. I t is the capitalization of expectedfuture benefits into stock prices that concentrateson the innovators the wealthgains of their actions if they own largenumbers of shares. IYithout capitalizationof future benefits, there would be less incentiveto incur the costs required to exertinformed decisive influence on the corporation’spolicies and managing personnel.Temporarily, the structure of ownership isreformed, moving away from diffusedownership into decisive power blocs, andthis is a transient resurgence of the classicalfirm with power again concentratedin those who have title to the residual.In assessing the significance of stockholders’power it is not the usual diffusionof voting power that is significant but insteadthe frequency with which votingcongeals into decisive changes. Even aone-man owned company may have along term with just one manager-continuouslybeing approved by the owner.Similarly a dispersed voting power corporationmay be also characterized by along-lived management. The question is#p#分页标题#e#the probability of replacement of themanagement if it behaves in ways not acceptableto a majority of the stockholders.The unrestricted salability of stock andthe transfer of proxies enhances the probabilityof decisive action in the event currentstockholders or any outsider believesthat management is not doing a good jobwith the corporation. We are not comparingthe corporate responsiveness to thatof a single proprietorship; instead, we areindicating features of the corporate structurethat are induced by the problem of.\LCHIAK ASP) DEMSETZ: ECOSO?tlIC ORGASIZA4TIOKdelegated authority to manager-monitors.’"D. Mutual and Nonprofit FirmsThe benefits obtained by the new managementare greater if the stock can bepurchased and sold, because this enablesof anticipated future im-‘"nstead of thinking of shareholders as joint oztners,we can think of them as investors, like bondholders,e\-cept that the stockholders are more optimistic thanbondholders about the enterprise prospects. Instead ofbuying bonds in the corporation, thus enjoying smallerrisks, shareholders prefer to invest funds with a greaterrealizable return if the firm prospers as expected, butwith smaller (possibly negative) returns if the firm performsin a manner closer to that expected by the morepessimistic investors. The pessimistic investors, inturn, regard only the t~ondsa s likely to pay off.If the entrepreneur-organizer is to raise capital on thebest terms to him, it is to his advantage, as well as thatof ~ ~ r o s ~ ~ e cintvievseto rs, to recognize these differences inespectations. The residual claim on earnings enjoyed byshareholders does not serve the function of enhancingtheir efficiency as monitors in the general situation. Thestockholders are "merely" the less risk-averse or themore optimistic member of the group that finances thefirm. Being more optimistic than the average and seeinqa higher mean value future return, they are willing topay more for a certificate that allows them to realizegain on their expectations. One method of doing so is tobuy claims to the distribution of returns that "they see"while bondholders, who are more pessimistic, purchase aclaim to the distritjution that they see as more likely toemerge. Stockholders are then con1parat)le to warrantholders. They care not about the voting rights (usuallynot attached to warrants) ; they are in the same positionin so far as voting rights are concerned as are hondholders.The only difference is in the proba1)ility distributionof rewards and the terms on which they canplace their bets.If we treat t~ondholders, preferred and convertiblepreferred stockholders, and common stockholders and#p#分页标题#e#warrant holders as simply different classes of investorsdifferingnot only in their risk averseness hut in theirbeliefs about the pro1)ability distritjution of the firm’sfuture earnings, why should stockholders he regarded as"owners" in any sense distinct from the other financialinvestors? The entrepreneur-organizer, who let usassume is the chief operating officer and sole repositoryof control of the corporation, does not find his authorityresiding in common stockholders (except in the case of atake over). Does this type of control make any differencein the way the firm is conducted? Li’ould it makeany difference in the kinds of behavior that ~ r o ~ i lbdetolerated 1)y competing managers and investors (and wehere deliberately refrain from thinking of them asowner-stockholders in the traditional sense)?provements into present .c:lealth of newmanagers who bought stock and createda larger capital by their managementchanges. But in nonprofit corporations,colleges, churches, cou1ltry clubs, mutualsavings banks, mutual insurance companies,and "coops," the future c o n s -quences of improved management are notInvestment old timers recall a significant incidence ofnonvoting common stock, now prohibited in corporationswhose stock is traded on listed exchanges. (Li’hyprohibited?) The entrepreneur in those days could holdvoting shares while investors held nonvoting shares,which in every other respect were identical. Nonvotingshare holders were simply investors devoid of ownershipconnotations. The control and behavior of inside ownersin such corporations has never, so far as we have ascertained,been carefully studied. I’or esample, at thesimplest level of interest, does the evidence indicate thatnonvoting shareholders fared any worse because of nothaving voting rights? Did owners permit the nonvotingholders the normal return availaljle to voting shareholders?Though evidence is prohibitively espensive toobtain, it is remarkable that voting and nonvotingshares sold for essentially identical prices, even duringsome prosy battles. I lo~rever ,o ur casual evidence deservesno more than interest-initiating weight.One more point. The facade is deceptive. Instead ofnonvoting shares, today we have warrants, convertiblepreferred stocks all of which are solely or partly "equity"claims xirithout voting rights, though they could he convertedinto voting shares.In sum, is it the case that the stockholder-investorrelationship is one emanating from the divisioiz ofowrzerslrip among several people, or is it that the collectionof investment funds from people of varying anticipationsis the underlying factor? If the latter, why#p#分页标题#e#should any of them he thought of as the owners in\vhom voting rights, whatever they may signify or howeverexercisable, should reside in order to enhance efficiency?IVhy voting rights in any of the outside, participatinginvestors?Our initial perception of this possibly significant differencein interpretation was precipitated by HenryXIanne. -1reading of his paper makes it clear that it ishard to understand why an investor who wishes to backand "share" in the consequences of some new l~usinessshould necessarily have to acquire voting power (i.e.,power to change the manager-operator) in order toinvest in the venture. In fact, we invest in some venturesin the hope t h a t no other stockholders ~i-ilbl e so"foolish" as to try to toss out the incumbent management.\Ye want him to have the power to stay in office,and for the prospect of sharing in his fortunes we buynonvoting common stock. Our willingness to invest isenhanced 11). the knowledge that we can act leqally viafraud, emt,ez~lement and other laws to help assure thatwe outside investors will not be "milked" beyond ourinitial discounted anticipations.790 THE AMERICAN ECONOMIC REVIEITcapitalized into present wealth of stockholders.(As if to make more difficult thatcompetition by new would-be monitors,mutiple shares of ownership in those enterprisescannot be bought by one person.)One should, therefore, find greater shirkingin nonprofit, mutually owned enterprises.(This suggests that nonprofit enterprisesare especially appropriate inrealms of endeavor where more shirking isdesired and where redirected uses of theenterprise in response to market-revealedvalues is less desired.)Team production in artistic or professionalintellectual skills will more likelybe by partnerships than other types ofteam production. This amounts to marketorganizedteam activity and to a nonemployerstatus. Self-monitoring partnerships,therefore, will be used rather thanemployer-employee contracts, and theseorganizations will he small to prevent anexcessive dilution of efforts through shirking.Also, partnerships are more likely tooccur among relatives or long-standingac~~uaintances,not necessarily becausethey share a common utility function, butalso because each knows better the other’swork characteristics and tendencies toshirk.F. Employee UnionsEmployee unions, whatever else theydo, perform as monitors for employees.Employers monitor employees and similarlyemployees monitor an employer’s performance.Are correct wages paid on timeand in good currency? Usually, this isextremely easy to check. Hut some forms#p#分页标题#e#of employer performance are less easy tometer and are more subject to employershirking. Fringe benefits often are in nonpecuniary,contingent form; medical. hospital,and accident insurance, and retirementpensions are contingent paymentsor performances partly in kind by employersto employees. Each employeecannot judge the character of such paymentsas easily as money wages. Insuranceis a contingent payment-what theemployee will get upon the contingentevent may come as a disappointment. Ifhe could easily determine what otheremployees had gotten upon such contingentevents he could judge more accuratelythe performance by the employer.He could "trust" the employer not toshirk in such fringe contingent payments,but he would prefer an effective and economicmonitor of those payments. We seea specialist monitor- t h e union employees’agent-hired by them and monitoringthose aspects of employer payment mostdifficult for the employees to monitor. Employeesshould be willing to employ aspecialist monitor to administer suchhard-to-detect employer performance,even though their monitor has incentivesto use pension and retirement funds notentirely for the benefit of employees.V. Team Spirit and LoyaltyEvery team member would prefer ateam in which no one, not even himself,shirked. Then the true marginal costs andvalues could be equated to achieve morepreferred positions. If one could enhancea common interest in nonshirking in theguise of a team loyalty or team spirit, theteam would be more efficient. In thosesports where team activity is most clearlyexemplified, the sense of loyalty and teamspirit is most strongly urged. Obviouslythe team is better, with team spirit andloyalty, because of the reduced shirking—not because of some other feature inherentin loyalty or spirit as such.15Is Sports Leag~ies: Professional sports contests amongteams is typically conducted hy a league of teams iVeassume that sports consumers are interested not only inabsolute sporting skill but also in skills relative to otherteams. Being slightly better than o[y)osing teams enablesone to claim a major portion of the receipts; the- -AIXHIAK AKI I I IEMSETZ: ECONOMIC ORGAKIZATIOK 791Corporations and business firms try toinstill a spirit of loyalty. This should not ~ ~be viewed simply as a device to increaseprofits by ocer-working or misleading theemployees, nor as an adolescent urge forbelonging. I t promotes a closer approsimationto the employees’ potentially availabletrue rates of substitution betweenproduction and leisure and enables eachteam member to achieve a more preferred#p#分页标题#e#inferior team does not release resources and reduce costs,since they were expected in the play of contest. Hence,absolute skill is developed beyond the equality of marginalinvestment in sporting skill with its true socialmarginal value product. I t follows there will he a tendencyto overinvest in training athletes and developingteams. ”Reverse shirking" arises, as budding playersare induced to overpractice hyperactively relative to thesocial marginal value of their enhanced skills. To preventoverinvestment, the teams seek an agreement witheach other to restrict practice, size of teams, and evenpay of the team members (which reduces incentives ofyoung people to overinvest in developing skills). Ideally,if all the contestant teams were owned by one owner,overinvestment in sports would be avoided, much asownership of common fisheries or underground oil or指导ESSAY water reserve would prevent overinvestment. Thishyperactivity (to suggest the opposite of shirking) iscontrolled by the league of teams, wherein the leagueadopts a common set of constraints on each team’s behavior.In effect, the teams are no longer really ownedby the team owners but are supervised by them, muchas the franchisers of some product. They are not fullfledgedowners of their business, including the brandname, and can not "do what they wish" as franchises.Comparable to the franchiser, is the league commissioneror conference president, who seeks to restrainhyperactivity, as individual team supervisors competewith each other and cause esternal diseconomies. Suchrestraints are usually regarded as anticompetitive, antisocial,collusive-cartel devices to restrain free open competition,and reduce players’ salaries. However, theinterpretation presented here is premised on an attemptto avoid hyperinvestment in team sports production.Of course, the team operators have an incentive, oncethe league is formed and restraints are placed on hyperinvestmentactivity, to go further and obtain the privatebenefits of monopoly restriction. To what estent overinvestmentis replaced by monopoly restriction is notyet determinable; nor have we seen an empirical test ofthese two competing, but mutually consistent interpretations.(This interpretation of league-sports activitywas proposed by Earl Thompson and formulated byMichael Canes.) Again, athletic teams clearly exemplifythe specialization of monitoring with captains andcoaches; a captain detects shirkers while the coach trainsand selects strategies and tactics. Both functions maybe centralized in one person.situation. The difficulty, of course, is tocreate economically that team spirit and#p#分页标题#e#loyalty. I t can be preached with an auraof moral code of conduct-a morality withliterally the same basis as the ten commandments-to restrict our conduct towardwhat we would choose if we bore ourfull costs.VI. Kinds of Inputs Ownedby the FirmTo this point the discussion has examinedwhy firms, as we have definedthem, exist? That is, why is there anowner-employer who is the commonparty to contracts with other owners ofinputs in team activity? The answer tothat cluestion should also indicate the kindof the jointly used resources likely to beowned by the central-owner-monitor andthe kind likely to be hired from peoplewho are not team-owners. Can we identifycharacteristics or features of various inputsthat lead to their being hired or totheir being owned by the firm?How can residual-claimant, centralemployer-owner demonstrate ability topay the other hired inputs the promisedamount in the event of a loss? He can payin advance or he can commit wealth sufficientto cover negative residuals. Thelatter will take the form of machines, land,buildings, or raw materials committed tothe firm. Commitments of labor-wealth(i.e., human wealth) given the propertyrights in people, is less feasible. These considerationssuggest that ,residual claimants-o\n7ners of the firm-will be investorsof resalable capital equipment in thefirm. The goods or inputs more likely tobe invested, than rented, by the ownersof the enterprise, will have higher resalevalues relative to the initial cost and willhave longer expected use in a firm relativeto the economic life of the good.But beyond these factors are those developedabove to explain the existence ofTHE AMERICAAN ECOKOMIC REVIEITthe institution known as the firm-thecosts of detecting output performance.\Then a durable resource is used it willhave a marginal product and a depreciation.Its use requires payment to cover atleast use-induced depreciation; unless thatuser cost is specifically detectable, paymentfor it will be demanded in accordwith expected depreciation. *And we canascertain circumstances for each. indestructiblehammer with a readily detectablemarginal product has zero usercost. But suppose the hammer were destructibleand that careless (which iseasier than careful) use is more abusiveand causes greater depreciation of thehammer. Suppose in addition the abuseis easier to detect by observing the way itis used than by observing only the hammerafter its use, or by measuring theoutput scored from a hammer by alaborer. If the hammer were rented andused in the absence of the o\n7ner, the depreciation#p#分页标题#e#would be greater than if the usewere observed by the o\n7ner and the usercharged in accord with the imposed depreciation.(Careless use is more likelythan careful use-if one does not pay forthe greater depreciation.) An absenteeowner would therefore ask for a higherrental price because of the higher expecteduser cost than if the item were used by theo\n7ner. The expectation is higher becauseof the greater difficulty of observingspecific user cost, by inspection of thehammer after use. Renting is therefore inthis case more costly than o\n7ner use. Thisis the valid content of the misleading expressionsabout ownership being moreeconomical than renting-ignoring allother factors that may work in the oppositedirection, like tax provision, shorttermoccupancy and capital risk avoidance.Better examples are tools of the trade.iyatch repairers, engineers, and carpenterstend to on7n their own tools especially ifthey are portable. Trucks are more likelyto be employee owned rather than otherequally expensive team inputs because itis relatively cheap for the driver to policethe care taken in using a truck. Policingthe use of trucks by a nondriver owner ismore likely to occur for trucks that arenot specialized to one driver, like publictransit busses.The factor with which we are concernedhere is one related to the costs of monitoringnot only the gross product performanceof an input but also the abuse or depreciationinflicted on the input in the course ofits use. If depreciation or user cost is morecheaply detected when the owner can seeits use than by only seeing the input beforeand after, there is a force towardowner use rather than renting. Resourceswhose user cost is harder to detect whenused by someone else, tend on this countto be owner-used. AAbsentee ownership, inthe lay language, will be less likely. Assumemomentarily that labor service cannotbe performed in the absence of itsowner. The labor owner can more cheaplymonitor any abuse of himself than if somehowlabor-services could be provided withoutthe labor owner observing its mode ofuse or knowing what was happening. Alsohis incentive to abuse himself is increasedif he does not own himself.16’6 Professional athletes in baseball, football, and basketball,where athletes having 5old their source ofservice to the team owners upon entering into sportsactivity, are owned by team owners. Here the teamowners must monitor the athletes’ physical conditionand behavior to protect the team owners’ wealth. Theathlete has less (not, no) incentive to protect or enhancehis athletic prowess since capital value changes have less#p#分页标题#e#impact on his own wealth and more on the team owners.Thus, some athletes sign up for big initial bonuses(representing present capital value of future services).Future salaries are lower by the annuity value of theprepaid "bonus" and hence the athlete hasless to lose bysubsequent abuse of his athletic prowess. Any decline inhis subsequent service value would in part be borne bythe team owner who owns the players’ future service.This does not say these losses of future salaries have noeffect on preservation of athletic talent (we are not makinga "sunk cost" error). Instead, we assert that theAALCHIXN AAND DEMSETZ: ECONOMIC ORGAANIZAATION 793The similarity between the precedinganalysis and the question of absenteelandlordism and of sharecropping arrangementsis no accident. The same factorswhich explain the contractual arrangementsknown as a firm help to explainthe incidence of tenancy, labor hiring orsharecropping."VII. Firms as a Specialized MarketInstitution for Collecting, Collating,and Selling Input InformationThe firm serves as a highly specializedsurrogate market. Any person contemplatinga joint-input activity must searchand detect the qualities of available jointinputs. He could contact an employmentagency, but that agency in a small to\n7nwould have little advantage over a largefirm with many inputs. The employer, byvirtue of monitoring many inputs, acquiresspecial superior information abouttheir productive talents. This aids hisdirective (i.e., market hiring) efficiency. He"sells" his information to employee-inputsas he aids them in ascertaining good inputcombinations for team activity. Thosewho work as employees or who rent servicesto him are using him to discern superiorcombinations of inputs. Not onlypreservation is reduced, not eliminated, because theamount of loss of wealth suffered is smaller. The athletewill spend less to maintain or enhance his prowessthereafter. The effect of this revised incentive system isevidenced in comparisons of the kinds of attention andcare imposed on the athletes at the "eupense of theteam owner" in the case where atheletes’ future serviesare owned by the team owner with that where futurelabor service values are owned by the athlete himself.LVhy athletes’ future athletic services are owned by theteam owners rather than being hired is a question weshould be able to answer. One presumption is cartelizationand monopsony gains to team olvners. Another isexactly the theory being expounded in this paper-costsof monitoring production of athletes; we know not onwhich to rely.l7 The analysis used by Cheung in explaining the#p#分页标题#e#prevalence of sharecropping and land tenancy arrangementsis built squarely on the same factors-the costsof detecting output performance of jointly used inputsin team production and the costs of detecting user costsimposed on the various inputs if owner used or if rented.does the director-employer "decide" whateach input will produce, he also estimateswhich heterogeneous inputs will work togetherjointly more efficiently, and hedoes this in the contest of a privatelyowned market for forming teams. The departmentstore is a firm and is a superiorprivate market. People who shop andwork in one town can as well shop andwork in a privately owned firm.This marketing function is obscured inthe theoretical literature by the assumptionof homogeneous factors. Or it istacitly left for individuals to do themselvesvia personal market search, much as if aperson had to search without benefit ofspecialist retailers. Whether or not thefirm arose because of this efficient informationservice, it gives the directoremployermore knowledge about theproductive talents of the team’s inputs,and a basis for superior decisions aboutefficient or profitable combinations ofthose heterogeneous resources.In other words, opportunities for profitableteam production by inputs alreadywithin the firm may be ascertained moreeconomically and accurately than for resourcesoutside the firm. Superior combinationsof inputs can be more economicallyidentified and formed from resourcesalready used in the organization than byobtaining new resources (and knowledgeof them) from the outside. Promotion andrevision of employee assignments (contracts)will be preferred by a firm to thehiring of new inputs. To the extent thatthis occurs there is reason to expect thefirm to be able to operate as a conglomeraterather than persist in producinga single product. Efficient productionwith heterogeneous resources is a resultnot of having better resources but in knowingmore accurately the relative productiveperformances of those resources. Poorerresources can be paid less in accord withtheir inferiority; greater accuracy ofTHE AMERICAAN ECONOMIC REIrIEIVknowledge of the potential and actual productiveactions of inputs rather than havinghigh productivity resources makes afirm (or an assignment of inputs) profitable.18VIII. SummaryIVhile ordinary contracts facilitate efficientspecialization according to comparativeadvantage, a special class of contractsamong a group of joint inputs to ateam production process is commonlyused for team production. Instead ofmultilateral contracts among all the joint#p#分页标题#e#inputs’ owners, a central common partyto a set of bilateral contracts facilitatesefficient organization of the joint inguts inteam production. The terms of the contractsform the basis of the entity calledthe firm-especially appropriate for organizingteam production processes.Team productive activity is that inwhich a union, or joint use, of inputs yieldsa larger output than the sum of the productsof the separately used inputs. Thisla .lccording to our interpretation, the firm is aspecialized surrogate for a market for team use of inputs;it provides superior (i.e., cheaper) collection and collationof kno\vledge at)out heterogeneous resources. Thegreater the set of inputs ahout which knowledge of performanceis being collated within a firm the greater arethe present costs of the collation activity. Then, thelarger the firm (market) the greater the attenuation ofmonitor control. To counter this force, the firm will 11edivisionalized in \yay that economize on those costs –just as will the market be sl~ecial i~eSdo. far as ive canascertain, other theories of the reasons for firms have nosuch implications.In Japan, employees t)y custom ~vork nearly theirentire lives with one firm. and the firm aqrees to thatexpectation. Firms will tent1 to be large and coriglomerateto enable a ljroader scope of input revision. Eachfirm is, in effect, a small economy engaging in "intranationaland international" trade. ;\nalogously, . h e r -icans expect to sljenrl their whole lives in the UnitedStates. and the bigger the country, in terms of varietyof resources. the easier it is to adjust to changing tastesand circumstances. Japan, with its lifetime employees,should he characterized more by large, conglomeratefirms. Presum;~hly, at some size of the firm. specializedknowletlge about inputs becomes as expensive to transmitacross divisions of the firms as it does acrozs marketsto other firms.team production requires-like all otherproduction processes-an assessment ofmarginal productivities if efficient productionis to be achieved. Nonseparabilityof the products of several differentlyowned joint inputs raises the cost of assessingthe marginal productivities ofthose resources or services of each inputowner. Monitoring or metering the productivitiesto match marginal productivitiesto costs of inputs and thereby toreduce shirking can be achieved moreeconomically (than by across market bilateralnegotiations among inputs) in afirm.The essence of the classical firm isidentified here as a contractual structurewith : 1) joint input production; 2) severalinput owners; 3) one party who is common#p#分页标题#e#to all the contracts of the joint inputs;1)who has rights to renegotiate anyinput’s contract independently of contractswith other input owners; 5) whoholds the residual claim; and 6) who hasthe right to sell his central contractualresidual status. The central agent is calledthe firm’s owner and the employer. S oauthoritarian control is involved; the arrangementis simply a contractual structuresubject to continuous renegotiationwith the central agent. ?’he contractualstructure arises as a means of enhancingefficient organization of team production.In particular, the ability to detect shirkingamong owners of jointly used inputs inteam production is enhanced (detectioncosts are reduced) by this arrangement andthe discipline (by revision of contracts) ofinput owners is made more economic.Testable implications are suggested bythe analysis of different types of organizations–nonprofit, proprietary for profit,unions, cooperatives, partnerships, and bythe kinds of inputs that tend to be ownedby the firm in contrast to those employedby the firm.IVe conclude with n highly conjecturalALCHrAN AND DEMSETZ: ECONOMIC ORGAANIZXTION 795but possibly significant interpretation. Asa consequence of the flow of informationto the central party (employer), the firmtakes on the characteristic of an efficientmarket in that information about the productivecharacteristics of a large set ofspecific inputs is now more cheaply available.Better recombinations or new uses ofresources can be more efficiently ascertainedthan by the conventional searchthrough the general market. In this senseinputs compete with each other within andvia a firm rather than solely across marketsas conventionally conceived. Emphasis oninterfirm competition obscures intrafirmcompetition among inputs. Conceivingcompetition as the revelation and exclzangeof knowledge or information about qualities,potential uses of different inputs indifferent potential applications indicatesthat the firm is a device for enchancingcompetition among sets of input resourcesas well as a device for more efiiciently rewardingthe inputs. In contrast to marketsand cities which can be viewed as publiclyor nonowned market places, the firm canbe considered a privately owned market;if so, we could consider the firm and theordinary market as competing types ofmarkets, competition between privateproprietary markets and public or communalmarkets. Could it be that themarket suffers from the defects of communalproperty rights in organizing andinfluencing uses of valuable resources?REFERENCESM. Canes, "A Model of a Sports League," unpublished#p#分页标题#e#doctoral dissertation, UCL.4 1970.S. N. Cheung, The Theory of Share Tenalzcy,Chicago 1969.R. H. Coase, "The Xature of the Firm," Econonzica,Xov. 1937, 4, 386-405; reprinted inG. J. Stigler and K. Boulding, eds., Readingsin Price Theory, Homewood 1952, 331-51.E. Furobotn and S. Pejovich, "Property Rightsand the Behavior of the Firm in a SocialistState," Zeitschrift fiir 2Vationalokolzonzie,1970,30,431-454.F. H. Knight, Risk, Crl.tcertainty and Profit,Sew York 1965.S. Macaulay, "Non-Contractual Relations inBusiness: A Preliminary Study," Anzer. SociologicalRev., 1968, 28, 55-69.H. B. Malmgren, "Information, Expectationsand the Theory of the Firm," Quart J. Econ.,.LZug. 1961,75,399-421.H. Manne, "Our Two Corporation Systems:Law and Econon~ics," L7irginia L a z ~ Rcv.,Mar. 1967, 53, No. 2, 259-54.S. Pejovich, &’The Firm, Monetary Policy andProperty Rights in a Planned Economy,"Tl’cstern Econ. J., Sept. 1969, 7, 193-200.M. Silver and R. Auster, ‘,Entrepreneurship,Profit, and the Limits on Firm Size," J. Bus.Ulziv. Clzicago, A4pr. 1969, 42, 277-51.E. A. Thompson, "Nonpecuniary Rewards andthe Aggregate Production Function," Rezt.Econ. Statist., Xov. 1970, 52, 395-404.You have printed the following article:Production, Information Costs, and Economic OrganizationArmen A. Alchian; Harold DemsetzThe American Economic Review, Vol. 62, No. 5. (Dec., 1972), pp. 777-795.Stable URL:This article references the following linked citations. If you are trying to access articles from anoff-campus location, you may be required to first logon via your library web site to access JSTOR. Pleasevisit http://www.ukassignment.org/daixieEssay/meiguoessaydaixie/your library’s website or contact a librarian to learn about options for remote access to JSTOR.ReferencesThe Nature of the FirmR. H. CoaseEconomica, New Series, Vol. 4, No. 16. (Nov., 1937), pp. 386-405.指导ESSAY Stable URL:LINKED CITATIONS- Page 1 of 1 –