sign in sign up
<<
Home , Economic system, Industrial organization

Emergence of the Theory of , 1890-1990 T.A.B.Corley 1 Universityof Reading

Businesshistory, wrote Louis Galambosin 1966, "triesto achievea synthesisof the characteristicsand effects of decentralized economic decision-makingon both economicchange and economicgrowth." Like otherscholars such as RichardWohl (1954),he wasexercised about what seemedto be businesshistory's chief problem: how to weavethe diverse activitiesof a multitude of independentfirms into a coherentaccount of national, regional, or other developments. While a theoretical structure would clearly strengthenany synthesis,ready-made theories from other disciplines,especially , would fail to reflect the richness of businesshistory case studies. Hence business historians would have to make do with eclecticand partial systemsof generalization.Yet thosehardly would add up to a broad intellectualframework, capable of makingcase studies more meaningful and of testing economic historians'broad conclusions. Despite the of studiessince the 1960s, this fundamental dilemma seemsto have been infrequentlydiscussed in businesshistory literature. Nearly three decadeslater, therefore,it is none too early to reopenthe issue,but this time from the theoreticalside. Is the theoryof corporateactivity still too rigid to help businesshistorians, and if so, could anythingbe done to make the theory a more serviceableinstrument for them? To throwlight on thesequestions, it wouldnot be enoughto examine the current state of theory. Instead, a broader "historyof economic analysis"approach will help to giveperspective to the inquiry. By studying the on the firm by certainmainstream of the past, it should be possibleto see how their observationsevolved into formal

11am very grateful forhelpful comments tothe participants atthe Business History Conferenceat Baltimore,especially Michael Best, LeslieHannah and RichardLanglois, and also to Mark Casson and Michael Waterson of the Universityof Reading. To save space, the very lengthy list of referenceshas had to be omitted. These referencescan be obtainedfrom myself at the Universityof Reading,RG6 2AA, UK, but should be traceable through any university library catalogue and the of Economic Journals/Articles.

BUSINESS AND , SecondSeries, Volume Nineteen, 1990. Copyright(c) 1990 by the BusinessHistory Conference.ISSN 0849-6825.

83 theorieswhich sometimeserred too much toward over-generalityor else over-rigidity. Included in this discussionwill be the writingsof from 1890onward, and the posthumousbut pervasiveinfluence of Augustin Cournot, which in due time led to 's and 'smonopolistic or imperfect competitiontheories of the early 1930s. By con,mingthe firm to an over-restrictivestraitjacket, these latter theoriesdeprived scholars of a flexibleand realisticbody of conceptsjust whenbusiness history studies were beginning to proliferatein the U.S. and Britain after 1945. Economists'subsequent theoretical and empiricalstudies have helpedto illuminatemany facetsof corporatebehavior, but without providingan adequateanalytical substitute for the simple Chainberlin- Robinsonmodels. In conclusion,the questionis raised as to whether economists,as well as businesshistorians, would benefit from a radical overhaulof industrialorgani7ation theory as a meansof overcomingthe dilemma mentioned earlier.

Marshall and His Era

Alfred Marshall was a pioneerin two respects. He was the first to assembleinto somekind of corporatetheory the buildingblocks contributed by many predecessors. Moreover, his unique techniquecombined mathematicalanalysis with a practicalknowledge of how businessand commerce worked. His basic ideas on the firm centeredaround competitionwhich he saw in terms of an activityor a processrather than in modern structural terms as portrayed in static sales or averagerevenue curves. While acknowledgingthe existenceof singleor collectivemonopolies in the private sector,he stronglybelieved in the pressureof competitionthrough free entry into the industrywhich helped to curb tendenciestoward . While aware of the power of combativeadvertising to enhanceartificial productdifferentiation, he thoughtsales curves could be slopingalso as the result of genuineconsumer goodwill. Competitionbred uncertaintyover rivals'responses to a given firm's policies. Whereasthe demandfor a productwould equal supply at the -clearingprice and thereforethe industrywould be in equilibrium,each firm would not be in equilibrium sinceit was in a constantprocess of growthor decline. Marshalltherefore chosea representativefirm to providesome typical supply curves for his valuetheory. Economistshave never yet properlyexamined the empirical questionof whetherfirms are actuallyin equilibrium. Howeverbusiness historiansknow that firmsmaking for stockuse inventories and backlogsto balancesupply and demand. The entrepreneurwas central not merelyto short-termoperations but alsoto the firm's long-termsurvival and prosperity.Marshall was the first economistspecifically to integratethe entrepreneurinto hisanalysis of , addingorganization to the existingfactors of land,labor, and . The factorreward, pure , was not contractualbut residual.While the more successfulentrepreneurs might enjoy a rent of ability,competitive pressures wouldtend to reduceprofits to a normallevel, sometimes allowing for the 85 difficultiesof work doneand the degreeof risk undergone.Since in the real world there was not a perfectmarket in information,entrepreneurs' abilityto winwas helped by superiorknowledge and by experienceacquired over time. To achievesuccess, entrepreneurs constantly had to adapttheir firms to changingcircumstances. Improved machinery would make cheaperand more accurate,while a progressivesubdivision of labor would alsohelp to keepdown costs. Marshall sought to studythe firm'sactivities as a whole,not merelythe pricingand output decisions but alsothe growth and technicalprogress. While manyof his assertionswere subsequently erodedby the analyticaltrends to be discussedbelow, some insights and hints did anticipatelater work.

Cournot's Legacy, 1890-1933

After 1870 the essentiallyliterary subjectof political transformeditself into the mathematically-basedsocial sciencecalled economics.This processrapidly acceleratedafter 1890. Hence Marshall's rich verbal insights,based on detailed personal knowledge,were progressivelyset asidein favorof analyticallyrigorous models concentrating on limited aspectsof the firm's activities. Those modelsowed much to Cournot'swork on value theory,published in 1838 but largelyneglected until the 1870s. Cournotanalyzed firms as creatorsof value. He began with the monopolisticcase and progressivelyextended the number of producersin the marketuntil he reachedthe oppositepole of unlimited .At thispole, eachfirm contributedtoo smalla proportionof the whole to affect the goingindustry . In diagrammaticterms not drawnby him, monopolywas associatedwith a slopingsales curve and unlimitedcompetition was associatedwith a horizontalcurve. Cournot discussedduopoly, suggesting that self-interestwould induce the two rivals concernedto reach a determinateand mutuallyadvantageous solution. However, he failed to analyzethe commonestmarket form in advanced ,namely . Marshall had discussedmonopoly purely in terms of a gas undertaking,with salesand costcurves yielding the uniquepoint of profit maximization.He mayhave derived his diagramfrom DionysiusLardnet, who in RailwayEconomy (1850) had produceda comparablemodel based on total ratherthan averagevalues. The practiceof usingmarginal values camemuch later. Marshalldid not regardprivate-sector as worth analysis. In the U.S., despitethe passingof the ShermanAnti-trust Act of 1890, economistswere either lukewarmor activelyhostile to the act, and consequentlyfelt little interestin pursuingthe analysisof monopolies.John B. Clark in 1899 merely acknowledgedthe conceptualdifficulties in integratingmonopoly into statictheory. Competition,on the otherhand, was the subjectof a lively debate. In 1906Henry L. Moore criticized Marshallfor treatingany deviationfrom the competitivemode as a friction to be disregarded,and stressed the urgentneed to explorethe intermediate area betweenthe two poles. AmongBritish economists, Arthur C. Pigou (1912)implicitly agreed with Moore's view that privatemonopolies or trusts were commonenough to merit seriousanalysis. However, unlike Cournot he believedthat oligopolisticconduct could not easilybe analyzedbecause any solutionswere indeterminate. Soonafter Marshall'sdeath in 1924,the anatomyof the firm began to appearin a shapereco•mai7able to modernscholars, the skeletonbeing the revenueand costschedules, and the marketform beingindicated by the slopeof the salescurve. Economistsportrayed curves as U- shaped,to reflect scaleeconomies in the downwardleg and diminishing returnsto scalethereafter. Concernedas theywere to refine theoryalong the newer mathematicallines, they choseto disregardMarshall's scruples and assumedthat fu'msin general--andnot merely statutoryundertakings- -would aim to be in equilibriumat the level of output where profit was maximized. They were plainly influencedby Marshall's and Lardher's diagramsshowing equilibrium in gas and railway enterprisesrespectively. Whereaspublic monopoliescould, within certain limits, calculatea sales curve in the absenceof closesubstitutes, commercial firms were unable to do so becauseof uncertaintyabout future salesand aboutrivals' reactions. This refining processfor theory steadilyincreased after 1925. Economistsassumed that mostfu-ms were in perfectlycompetitive markets, and hencewould seek to produceat the level where the horizontalsales curvewas at a tangentto the averagecost curve. Yet observationshowed that real life fu'msoperated on the increasingreturns leg of the costcurve, able to reduceaverage costs if they raisedoutput. ,author of a lengthy1925 article in Italian on peripheralmatters, was askedby the editorsof the EconomicJournal to producean Englishversion. Sraffa includeda postscriptdesigned to solvethe increasingreturns problem by suggestingthat fu'ms in general were monopolistsand not perfectly competitive,and thus free to alter their pricesat will withoutwholly losing market shares. Sraffa'sEnglish article, publishedin 1926, seemedto provide an approachthat coveredall market formsshort of perfect(or unlimited) competition.Then in 1928,while offeringthe propositionthat Marshall's representativefirm wouldbe in equilibriumeven if all other fu'msin the industrywere not, Pigou brought back into mainstreamanalysis the geometricalrelationship between marginal and average cost. This relationship,first put forwardby FrancisY. Edgeworthin 1913,showed the one curveintersecting the other at the minimumaverage cost point. By 1932 Richard Kahn had pursuedthe theoreticalimplications of Sraffa's rather denselypacked logic to the pointwhere Joan Robinson was able that year to combinehis analysison the demandside with the Edgeworth-Pigou costcurves into a self-containedmodel, in which-forthe first time-everyfu'm was held to be in equilibrium. Scholars'unease at her many heroic assumptionswere hardlyassuaged when her extendedtreatment appeared in 1933. For instance,she assumed, without empirical justification, that each firm was its own monopoly,so that the all-importantoligopolistic reactions of rival firmscould be neglected.She admittedat the outsetto be offering no more than a box of tools,and was thus content to draw conclusionsfrom 87 geometricaldiagrams rather than arguingto thosediagrams: a practice deploredby Marshalland others. In contrastwith JoanRobinson's ultimately dead-end analysis of 1932- 1933,Edward H. Chambedin'sTheory of MonopolisticCompetition (1933) has been hailed as truly revolutionary.Similar to Lardnet, he arrived at his themethrough studying practical questions about railway rates, but his methodologyowed much to Cournot. That did not deter him from incorporatinginto his analysis such highly relevant topics as oligopoly, selling costs,, and qualityvariation. Robinsonand Kahn madebold attemptsto eliminateentrepreneurship as a productivefactor by claimingthat its marginalvalue product was negative. Chainberlin however was convincedthat entrepreneurshad a genuinefunction in havingto judge the appropriatedegree of interactionwith rivalsin the group,or industry, with adjustmenttaking place over time. Unfortunately,his pioneering attemptto analyzethe intermediatearea of oligopoly,in this presentation of a small-groupmodel, was scotched by RobertTriffin (1940). Trdf'mset out to restatethe core of Chamberlin'sanalysis in termsof Leon Walras's generalequilibrium theory, and in the processdenied that the conceptof the groupor industryhad any usefulness because of productdifferentiation. Chainberlinlater acceptedthese criticisms, and therebyrobbed his analysis of one of its most realistic features. Regrettably, these high-profile treatments by Robinson and Chamberlin diverted economists' attention from some important contributionsto the understandingof the enterprise. Frank H. Knight in 1921 and JosephA. Schumpeterin 1934had assignedcrucial roles to the entrepreneur,such as coordination,decision-making, bearing uninsurable risk, and promotinginnovations. Knight maintained that the entrepreneur was essentialbecause of . In 1934 Adolph A. Bede and GardinerMeans exploredthe consequencesof ownershipbeing separated from controlin joint-stockcompanies, while three yearslater RonaldH. Coasemade an equallypath-breaking distinction between the market and the firm,with the pricemechanism being superseded in the latter. All these ideasstimulated a numberof significantdevelopments in corporatestudies, as will be seen below.

Era of Controversy, 1933-1951

An unexpectedconsequence of the developmentsup to 1933was the gradualrealization of the existenceof an entirelynew academicsubject, the theoryof the firm. Earlier economists'work on corporatetopics had been undertakenmerely to clarifyaspects of valuetheory. Only by 1942was the new subjectwell enoughestablished for KeunethE. Bouldingto maintain that the theory of the f'wmhad existedsince 1932. Bouldinghimself was not undulyimpressed with the marginalistanalysis, on the groundsthat it simplifiedthe expositionof theorybut at the costof neglectingsome highly importantaspects such as oligopolyand uncertainty. These criticisms foreshadoweda very bitter controversythat soon erupted between the supportersof Robinson'sand Chambedin'smarginal approach and those who rejectedit as essentiallyrigid and unrealistic. In 1939some Oxford economists reported on a researchproject that had involvedquestioning businessmen on how price and output decisions were reached. They found that two paramount concernsfor these businessmenwere uncertaintyand oligopolisticrelations, both--as Boulding had pointedout--neglected by the 1933theorists. Price thereforehad to be fixedaccording to full costs,and as muchas possiblewould be soldat that price. A youngereconomist, Philip W. S. Andrews,later slightlymodified this full-costprinciple to a normal-costone, whichwould reflect also the currentbuoyancy of the market. In the U.S., RichardA. Lester(1946) and Robert A. Gordon (1948) also challengedmarginalist theories on both logical and empirical grounds,only to come under attack from such upholdersof marginalismas Fritz Machlup. Suchcontroversies took up all too muchtime of economistsworking in the field, so that thisperiod can be seenas an interludebefore the pace of constructive work accelerated in the 1950s. Some of that work will be discussedin the followingsection.

The 1950s,Emergence of Industrial Organization Studies

The modern theory of industrialorgani7ation was born out of a numberof academicprojects in the U.S., alreadyyielding significant results by the early1950s. In Britainthe term "industrialeconomics" was preferred, and importantresearch findings on the subjectwere disseminatedin the Journal of Industn'alEconomics, founded by Andrews in 1952. Of the American contributions,three are examinedhere. The leadingproject was the one at Harvard, where Chamberlinand Edward S. Mason promotedindustry-wide studies, helping to test the hypothesisthat market or industrialstructures determined member firms' conduct and performance. While this structure-conduct-performance relationshipfeatured prominently in industrialorganization studies until the 1970s,its importancedeclined once empirical research yielded only weak practicalresults and scholarsbegan to acceptthat the widespreadmarket form of oligopolymade businessbehavior very difficultto forecast. One breakthroughwas achievedby a pupil of Mason's,Joe S. Bain, who in 1949 suggestedthat pricing theory shouldtake accountof two relativelyneglected factors, namely time and potentialentry. Today'sprice might be governedby tomorrow'sprofit targets, while the threat of competitioncould well be as effectivein determiningbusiness conduct as the currentmarket structure.In his 1956book he both categorizedbarriers to entryand showedhow firmsfacing various heights of barriercould hold above minimum unit costswithout encouragingentry. Despite attractingsome criticism, for instanceregarding definitions of barrierheights and the smallnessof his sample in the empirical sections,his work demonstrated how research of this kind could test and refine theories. The secondproject reflected the emergenceafter 1945 of a new field of study,the economicsof development.A panelof labor economistsfrom Harvard, MIT, Princeton, Berkeley, and Chicago, investigatinglabor problemsin developingcountries, identified the lack of an entrepreneurial (as opposedto a mercantile)cadre as a factorimpeding such countries' 89 economicadvance. That approachhelped to reviveacademic in the entrepreneur,which had been all but struckout of the scenarioduring the 1930s. In 1956 one of the Princetonteam, Frederick Harbison, drew on the lately neglectedviews about the entrepreneurput forward by Marshall, Knight,and Schumpeter.Since the qualityof entrepreneurshipcould dearly affectthe degreeof organizationalefficiency, Harbison suggested that so- called inefficiencycould be due to entrepreneursbehaving rationally in pursuingother goalsthan profit maximization,such as socialadvancement. Efficiency could also be reduced by inadequate knowledgeand by inappropriateorganizational structures which could lead to lossof effective control over subordinates. These important ideas later were developed further by one of the labor economists'panel, Harvey Leibenstein,in his discussionsof organizationalor X-,which he was to showas being of far greatermagnitude than the allocativeinefficiencies resulting from, for example,monopoly or tariffs. In 1957while analyzingdevelopment questions in depth,Leibenstein elaboratedsome of his views on entrepreneurshipand its origins. He stressedthe role of knowledge,which couldbecome an economicresource throughthe entrepreneur'sactivities, although in a poor economythe cost of acquiringand utilizingknowledge could be too highfor an innovationto be exploitedthere. On the supplyof ,Leibenstein believed that entrepreneurswould emerge in the economic at a rate dependingon their anticipationof growthin incomeper head. Thusin a stagnanteconomy, there would be few entrepreneurs,but they would increaseas the economyembarked on its development.His "growthagent expansioncurve" postulated a mutual interactionbetween expected and actual income growth, set off by steadily increasingnumbers of entrepreneurs.He did not at this time draw a demandschedule for entrepreneurs,although in subsequentworks he attemptedto do so. The third researchproject, at JohnsHopkins under Machlup, was on the growthof the firm. In 1952one of that team,Edith Penrose,stressed the biologicalanalogies in the theoryof the firm, favoredby Marshallbut sincediscarded in the drivetoward rigor underthe influenceof Cournot's ideas. In her book of 1959, Penroseproved to be a pioneer in two respects.First, in settingout a generaltheory of corporategrowth, she moveddiscussion away from the scaleof the firm to its growth. Second, shefocused attention on the firm as an organizationin its own right,rather than merelyas a unit in a givenmarket structure. For her the engineof growth,whether in one productor by diversification,was the existenceof unusedmanagerial services. Penrose recognized the valueof goodbusiness historiesand of evidencefrom businessmenfor testinghypotheses on the processof corporategrowth. Sinceit turnedits back on the marginalist theories,her book was attackedby mainstreameconomists for insufficient analyticalrigor. However,it has proveda greater spur to subsequent researchthan the equilibriumgrowth models offered by Baumoland by Robin Marris in the early 1960s. Individualeconomists were busy in devisingmore realisticalternative businessmotives to profit maximization.In 1935John Hicks pointed to the "quietlife" as the bestof all monopolyprofits, and in 1943Tibor Scitovsky suggestedfor entrepreneursthe same kind of work-leisurechoice that theoristsassumed for workersable to vary the hoursor intensityof their labor. Businesshistorians are familiarwith the tendencyof entrepreneurs initiallyto work all hours,but then to easeoff oncetheir firms had been established.In 1958 Baumolput forwardthe salesrevenue maximization hypothesis,since the typicalU.S. corporationsought that goal rather than profit maximizingat all costs,but with a minimumprofit constraint. The motive of securingno more than satisfactoryprofits had been suggestedby Gordon in 1948, and followedup after 1955 by Henry A. Simonwith his psychologicalconcepts of drivesand aspirationlevels. In 1963Oliver E. Williamson'smanagerial discretion model took account of the separationof ownershipand control. Managers,as those effectivelyin chargeof limited companies,would seekto grant themselvesrewards well beyondtheir actualproductivity. By the end of the decade,the ever elusivetopic of oligopolywas onceagain the subjectof attention. Two importantforerunners had been John von Neumannand Oscar Morgenstern'stheory of games (1944), relatedmore specifically to firms'operations by Martin Shubikin 1959,and K. W. Rothschild'sarticle of 1947,which likened oligopolistic rivalry to war, especiallyin the alternateperiods of intenseand oftencostly hostilities, and lengthyspells of inactivity.Paolo SylosLabini, in an Italian work cited by Franco Modiglianibut not translatedinto Englishuntil 1962, analyzed oligopolisticstrategies from the viewpoint of aspiring entrants and establishedfirms. The latter'sforestailing tactics could pay off evenif new firms believedthemselves able to produceat costscomparable with those of existingfirms. Modigliani helpfully discussedSylos Labini's work alongsidethat of Bain. While both contributionswere essentiallystatic, they did providea joint frameworkfor future developmentscapable of yielding operationallytestable propositions.

Recent Developments,1960-1990

Puttingrecent events into perspectiveis alwaysdifficult, and in the presentcase the problemis compoundedby the substantialquantity of industrialorganization studies in this period. When in the early 1970s,the researchimpetus of the previoustwo decadesappeared to slackenthe NationalBureau of EconomicResearch commissioned four leadingscholars to identifycurrent problems and possiblefuture directionsof study. Two areaswhich they specifiedwere the internalorganization of the firm and technicalprogress. Despitebeing a key determinantof U.S. industrial growth,the latter hadbeen inadequately analyzed, probably being regarded as fallingoutside the scopeof traditionalcorporate theory. Thesetopics helpedto influencethinking in the 1970s,and in turn createdan empirical renaissancein industrialorganization studies during the followingdecade. The work of Alfred D. Chandlergreatly boostedthe progressof studieson internal organization. During the 1950sChandler had been researchinginto the topic of how large-scalebusiness evolved in the U.S. Chandlermade a uniquecontribution to businesshistory by a combination 91

of three factors. He had a verybroad and relevanttheme; he took account of certainlarge economicforces such as demographicand technological changes;and he had a thoroughresearch grounding in the historyof the fifty largestU.S. corporationswhich he studied. In 1961 Chandlerand Fritz Redlich, as former colleaguesin the Harvard ResearchCenter in EntrepreneurialHistory, jointly set out to update the theory of the entrepreneur,hitherto assumedto be one individualor a smallteam. Their studytook account.of organizational changesnecessitated by the emergenceof diversifiedgiants in the U.S. The variouscategories of enterprisewhich they described, from the smallsingle- productone to large oneswith manyproducts and functions,provided a usefulparadigm for scholars.After the articleappeared, the journal's editor canvassedeight eminent business historians for their views. That initiative yieldedno seriousdiscussion of the authors'main thesisthat '•usiness historicalChandler'smaterial Strategycan be andused Structurein developing(1962) economic did offertheory." two concepts that

were to be influentialwell beyondbusiness history: that structurefollows strategy,and that largefirms needed to evolvean organizationalform, such as a multidivisionalstructure, to permit effectivecontrol. It was an ,Oliver Williamson,who in 1970 drew on Strategyand Structure for a distinctionbetween U-form or unitary firms and M-form or multidivisionalfirms. In Marketsand (1975), Williamson grafted his expositionof Chandler'sconcepts on to the distinctionby Coase(1937) betweenthe firm andthe market,and the conceptof transactioncosts which had beenlargely neglected by the post-Cournoteconomists. Williamsoh's article of 1981 on "the Modern "was impressivein covering manyfacets of the subjectand identifying, along Galambos's lines, some of the differencesin viewpointbetween business historians and economists. Yet the article did not prove as seminalon the businesshistory side, no doubtbecause Williamson sought to explainthe firm'sworkings chiefly by the overridingneed to economizein transactioncosts, an approachnot very amenableto empiricaltesting. Althoughthe secondtopic of studymentioned by the NBER scholars, namelytechnical progress, subsequently attracted much academic research, it wasconsiderably extended after the early1970s. The traditionalconcept of technological-oftenmechanical-innovations, arising mainly from research and development,proved to be onlyone sourceof corporateadvance, to be supplementedby innovationsin the fields of marketing, fmance, and managerialexpertise. The most progressivefu-ms therefore possessed a substantialstore of knowledgeor information. These broader ideas flourishedin the contextof yet anothernew economicsubject, the theory of the multinationalenterprise (MNE). Normallyarising from a home firm's foreigndirect (FDI), the MNE and its originsreceived virtuallyno theoreticalattention before 1960, when Stephen Hymer drewa definitivedistinction between portfolio and directinvestment, the latterbeing undertakento earn not interestbut profits. Hence, he concluded,MNEs shouldproperly be studiedthrough the theoryof the firm. However,the marginalistmodels of the 1930swere not robustenough to permitHymer to build up a satisfactorytheory of the MNE. Although 92 he and successorssuch as Charles Kindlebergerand Richard Caves attributed FDI mainly to monopolisticmotives, emphasizingprice discriminationthrough separationof markets, that explanationsoon lost ground to more flexible ones. In the 1970sJohn Dunning at Reading proposeda more eclecticapproach that stressedthe synergisticmotives for producingoverseas, with firms perceivingadvantages which might be organizationalor locatiohal. Other economistswere fruitfullyapplying Coase'sideas to the MIXrE. They saw a progression,from arms-length marketingof exportsto Settingup overseasbranches, as the internalization of activitiesformerly undertakenthrough the market. The new MiXrEs wouldenjoy two benefits:profits that had previouslygone to others,and reliablequality control. The head office'sstock of informationcould best be safeguardedby being transferredacross boundaries while remaining under home control. Althoughby the 1980svirtually all largefirms were MNEs, industrial organizationstudies were by no means exclusivelyMNE-oriented. As RichardLanglois has shown,William Lazonickamong others was working towardsa theoryof the giantentrepreneurial firm alongSchumpeterian and ChandlerJanlines. Rather thanbeing suspicious of suchgiants for enjoying marketpower, Schumpeter had stressedtheir abilityto generatesignificant innovations.Chandler in turn had emphasizedthe substantialeconomies of high-volumeproduction and marketingthat the giantswere able to achieve. Yet towardthe end of the decade,specifically MIXrE theory was beginning to convergewith the old implicitlyassumed to be a uninationaland single-plantunit. Both of thesetheories could now be seen as specialcases of a generaltheory of the enterprisein geographicalspace, withmany (but not all) of the considerationsaffecting MNE behaviorbeing applicablealso to multi-regionaland multi-plantfirms.

Conclusion

Despite the inevitableselectiveness of the above account,the substantialprogress made in manyaspects of corporatestudies will be clear. Yet the dilemma,stated at the beginningof this paper,remains. While undoubtedlysuperior to partialgeneralizations for bringingorder to the myriadindividual case studies, theory so far is scarcelyin a state to be reallyuseful to businesshistorians. The problemabout economic theories in thisarea is not so muchthat theyare over--rigid--althoughparts such as the 1933models undoubtedly are-but that the variousbranches of these theoriesare poorlycoordinated and lack an innerlogical core. At the sametime, a holistictheory is likely to coverevery possible category,is likely to be too generaland as suggestedat the outsetcould do violenceto the richnessof businesshistory case studies. Perhapsthe infinitelyvaried corporate experience would be betterreflected in the careful building up of analysisthat highlightedthe essentialsimilarities and differencesof varioustypes of firm, suchas , , and non-profitmaking. It is to be hopedthat the secondcentury of research into industrialorganization will be as fruitful as the first centuryhas been.