Agent Opportunism and the Role of Company Culture - CiteSeerX

Agent Opportunism and the Role of Company Culture - CiteSeerX

Agent Opportunism and the Role of Company Culture: The Hudson's Bay and Royal African Companies Compared Ann M. Carlos • Universityof Colorado Concu...

617KB Sizes 1 Downloads 8 Views

Agent Opportunism and the Role of Company Culture: The Hudson's Bay and Royal African Companies Compared Ann M. Carlos •

Universityof Colorado

Concurrentwith growth of long-distancetrade in Britain in the seventeenth centurywasthe useof the charteredcompanyasthe mechanism by which such trade with Africa, Asia, Russia, and North America was conducted. Indeed, these trading companieswere analogsof modern multinationalswith factoriesand agentslocatedoverseas.Becauseof their hierarchicalstructures,these companiesfaced potential principal/agent problems--they hadto ensurethattheiroverseas agentsoperatedin thebest interestsof the company.The secondary literatureis almostunanimous in its verdict that companyagentswere, in fact, the companies'worst enemies. Keith Davies,the historianof the Royal African Company,wrote that the companywasboundto fail and did fail because"toomuchhad to be left to the discretionof employees abroadwhofor the mostpart followedtheir own concernsto the detrimentof the Company's"[7, p. 165]. This statement definesa principal/agentproblem. E. E. Rich, the historianof the Hudson's Bay Company,in describinga letter of recallfor one manager,wrote that "[Bayly]is to handoverall tradegoods... and to 'bee assistante' to Lydall [new governor]in discovering and preventingprivatetrade. Alwaysprivate trade"[21, p. 101]. Againthisis a statementaboutmanagerialopportunism. That the secondary literatureshouldfocuson agentopportunismas a seriousproblemfacingtheseearly tradingcompaniesis easilyunderstood when one considers the time and distances involved in communication.

Communicationbetween head office and managersabroad could take anywherefrom three monthsto two years[4, p. 407]. In sucha situation, local managerswould have the ability to operateindependently of the companyand in their ownself-interest.Yet despitethisstateof affairs,little archival work has been done to ascertain whether managers were the companies'worst enemies,whetherthesecompaniestook stepsto control managers,and whethertheywere successful in thoseendeavors.As a start in answeringthesequestions, alongwith StephenNicholas,I examinedthe situationfacingthe Hudson'sBay Company.We showedthat the Hudson's Bay Companydid indeedtry to controlits overseas agentsand arguedthat

lI wouldlike to thankthe SocialScienceHumanitiesand ResearchCouncilof Canadafor the fundingwhich made this paper possible.

BUSINESS AND ECONOMIC HISTORY, SecondSeries,VolumeTwenty,1991. Copyright(c) 1991by the Businessllistory Conference.ISSN 0849-6825. 142

143

theydid so successfully [5]. Currently,I am examiningthe strategies usedby the RoyalAfrican Companyand arguethat while the companydid try to controlits agents,it was unsuccessful [3]. This paper,basedon the more extensive documentation presentedin the individualcasestudies,showsthat thoughthesetwocompanies usedsimilarstrategies, the environment in which eachoperatedand the resultingcompanycultureled to oppositeresults. Hudson's Bay and Royal African Company Trade

This sectionfurnishesa contextfor the subsequentdiscussionof managerialstrategiesandcompanyculture. (Detaileddiscussion is presented

in [7, 23].) The Hudson'sBay Company(HBC) and the RoyalAfrican Company(RAC) receivedtheir royal charterswithin two yearsof one another,the HBC in 1670 and the RAC in 1672. The former company receivedthe monopolyright to trade for furswithinthe drainagebasinof HudsonBay,whilethe RAC receivedthe monopolyof the Britishslavetrade alongthewestcoastof Africa. Althougheachreceivedmonopolyprivilegein theory,neithercompanyhad a monopolyin fact. Each facedcompetition from both Europeancompetitorsand other Britishfirms operatingillegally outsidethe charter. Each companyhad its head officein London,and each conducted a bartertradein its tradingregion. The HBC exchanged European tradegoodsfor fursaroundthe coastof HudsonBay. This wasundertaken by salariedmanagers, whosejob it wasto tradewith theIndiansfor the furs broughtinto the posts. The Indiansbroughtin aboutsixteendifferenttypes of fursandexchanged themfor a largerangeof tradegoods.Thesefursthen had to be baled and shippedback to Londonfor saleat the company's fur auction. In turn, the headofficehad to purchasethe requiredtrade goods and get them shipped,alongwith the food and equipmentnecessaryto supportits personnel, duringthe time thatHudsonStraitwasopen,a period of aboutsixweeks[14, 18, 23]. The RAC conducted a bartertradealongthewestcoastof Africa.This trade had two distinctparts. First, there was the triangularslavetrade in whichgoodssentfrom Englandwere exchanged for slaves.The slaveswere shippedto the WestIndies,wheretheyweresoldandthe proceeds, eitherin billsof exchange or sugar,weresentto London. Second,therewasthe trade in commodities.Trade goodswere exchanged for redwood(a dye),ivory, gold,wax,andmalaguetta(a pepper),whichwereshippeddirectlyto London for sale. Here again,the head office had to organizethe purchaseand shippingof tradegoodsand their subsequent sale[7, 13, 17,24]. The basic problem facing the directorsof both companieswas ascertaining whethertheywerein factreceivingthe maximumoutputfor the trade goodssent. Becausethe actualtrade was takingplacein regionsfar removed from the head office, the companieswere very vulnerableto opportunistic behaviorbytheir managers.Thus,the companies hadto devise strategiesand solutionsto guardagainstsuchbehavior.

144

Opportunistic Behavior:Strategiesand Solutions

The long-distancetrades required that managershave some discretionary power;at the sametime, the companyhad to ensurethat such power was used to the best advantageof the companyand not in the managers' self-interest. Thisisthe"economics-of-agency" problem.Situations in whichmanagers canoperatein theirownself-interest occurin thepresence of asymmetrical information.There are essentially three differentclassesof asymmetrical informationthat are relevantto the issuefacingthe chartered companies andtheirmanagers: moralhazardwithhiddenaction,moralhazard with hiddeninformation,and adverseselection.The latter two are sometimes considered the same.

The headoffice'sproblems wererelatedto itsinabilityto seetheeffort levelof itsmanagers; it onlyknewthe outputit received.It thenhadto assess whetherthat amountwasthe maximumlevelof output. This is an example of moralhazardwithhiddenaction;the headofficeknowssomething about themanager's abilitybutnothiseffortlevel,whichmaybe eitherhighor low. It is also possiblethat neitherplayerknowsthe worker'sability,whichis discovered onlyafter the workeracceptsa contract. If a workerknowshis abilityfrom the start,but the headofficedoesnot, then this is a problemof adverseselection.Indeed,the argumentis sometimes madethat onlythose peoplewho intendedto cheatto maketheir fortunesbecamemanagersin thesecharteredcompanies[16, chs6-8]. In an idealworld it is alwayspossibleto devisean optimalcontract-one that will call forth the requiredlevelof effort--butwe haveto examine actualcontractstructuresand assessto what extentthey might inducethe "correct"response.To do so the theoreticalliteraturesuggests focusingon featuresof the compensation package.The firstis the levelof wages.The higherwagesare,relativeto whatcanbe earnedin otherpositions, themore costlyit is for a workerto losehisjob. This is the notionof an efficiency wage.Alsoimportantis the timepathof wages.If a managergetspaidless than his marginalproductearlyin his careerbut is rewardedwith higher salarieslater, there is also a high opportunitycost to losing one'sjob. Gratuitiesand bonusescan also be tied to output to increasethe level of effort.

At the sametime, unlessa managerknowsthat he is likely to get caughtandlosehisjob, he canstillbehaveopportunistically. Thus,firmsmust monitor. Monitoringincreases informationflowsbywhichthe firm canassess performance.Monitoringcanbe donedirectly,by searching shipsor reading letters,or it may be done indirectly,by generatingaccurateaccounting proceduresor by running tournamentsamong the managers. In a tournament,the outputlevelsof individualmanagersare comparedto the groupasa whole,andmanagers are rewardedaccordingly [16,pp. 167-68]. Thus, there are waysin whicha firm can controlthe opportunistic conductof its overseasmanagers;to do so, however,is not costless. Incentives, suchashighsalaries,gratuitiesor bonuses, andmonitoringare all direct coststo the firm. What intereststhe firm, therefore, is the cost effectiveness of suchmeasures.If it is verycostlyto monitoror if managers

145

can engageeasilyin privatetrade, then the level of incentives neededto reduceopportunistic behaviormightbe too highto be costefficient. Thus firm profitabilityand successmay be fairly sensitiveto the exogenous environment[5 p. 859]. The nature of the interactionbetweenthe managerand firm is also potentiallyimportant.Possibletheoreticalsolutionsrevolvearoundideasof reputationand repetition. Therefore,in a multi-periodgame,the firm can encouragestrategiesthatwill resultin the futuregrowthof the firm [11]. Yet the natureof the interactionbetweenthe firm andmanagercanbe influenced in a lessobviousway. The firm cangeneratea companyculturethat,through the creationof an internalizedsetof moresamongemployees, leadsto lower levelsof cheating. This,in turn,meanslowercostsfor thefirm throughlower salariesand lower monitoringlevels. Company Conduct and Culture: Employment Contracts and Monitoring

Withinthirtyyearsof receivingroyalchartersthe RAC essentially had failed, while the HBC is still in existence. The failure of the RAC has been

attributedlargelyto itsinabilityto controlmanagerial opportunism, whilethe success of the HBC

has been attributed

to the fact that it was a small

company.Companysuccess or failure,however,seemsto be a functionof workingenvironment and companyculture. The employmentcontractis centralto the firm's abilityto attenuate opportunismand to call forth the relevantlevel of effort. A feature of employment contracts of thisperiodnot discussed in the theoreticalliterature is that all employeeshad to take an oath to work in the company'sbest interests.Althoughthiswasrequiredbyboththe HBC andthe RAC, neither relied on suchoathsalone [5, p. 863;23b]. The precedingsectionarguedthat to decreasecheating,wagesshould be highand/or risingovertime. The quantitative evidence for the HBC and the RAC suggeststhat these companiesfollowedsuch strategies. In a comparisonwith British wages,managerswere paid high salaries,which tendedto rise overthe yearsof the contract[25]. CharlesBayly,the first governorat HudsonBay,wasinitiallygivena salaryof œ50per annum,but thatwasquicklyraisedto œ200per annum.A successor, JohnNixon,hadan initial salaryof œ100,whichwasraisedto œ200in hissecondyearof office[5 p. 862].

Salariesfor thoseat the top of the RAC hierarchywereevenhigher. In 1676,the agent-general receivedœ400,andby 1680thishad risento close to œ1000.Thisincreaseoccurredbecause the company hadmadeall personal tradeillegal. In additionto the œ600increase, the agent-general wasgivena œ200gratuityat the end of the three-yearcontractin lieu of privatetrade. This wouldsuggest that personaltradewasfinanciallyrewarding.Thoseat the entrylevelof the managerialclassreceivedconsiderably lowerbut still highsalaries,and thosesalariesincreasedoverthe yearsof the contract.In 1688,William Ward washired as a factorfor Cape CorsoCastleat œ40,œ40, andœ50,respectively, for eachof histhreeyears.Again,thisis suggestive of an efficiency wagestructure.Middle-levelmanagers receivedœ200per annum

146

in 1677. Thesesalarieswerenot negligibleandsuggest that losingone'sjob couldbe costly.In addition,boththe HBC andthe RAC providedroomand board for their managers[7, p 252.]. Salaryis nottheonlyfeatureof theemployment contractthatcanhelp reduceopportunism. Thereareotherwaysthatcompanies canmakeit costly to loseone'sjob. Gratuitiesandbonuses are positiverewardsfor workeffort, while bondsare monetaryinvestments for goodbehavior. Suchstrategies make it more costlyto be caughtcheating. Interestingly, here the two companies diverge.The HBC wasverygenerouswith gratuitiesand bonuses. It believedthat suchpoliciesincreased work effort. In 1690the company wrotethatit had"Resolved on ourpartsnotto faileof giveingencouragement to all whomewee find dilligentand activeto promoteour Interest"[5, p. 862; 19, p. 101]. In anotherletter of the sameyear, the companywrote to a Mr. Walshthat "youmayseeour Resolves to Leavenoemerrittunrewarded for your furtherencouragement in the Preservation of our Rightsand Encrease of the tradewee havevotedyoufifty poundsGratuity,overand aboveyour Fixed salary"[5, p. 862]. In the RAC's correspondence with its African managers,no such statementscan be found. Over the period in questionthe companypaid a gratuityto only one memberof its African staff [23 g]. Instead,the RAC concentrated heavilyon ensuringthat all its employees were bonded. The HBC alsousedbondsbut onlyfor top-levelmanagers.EachRAC employee wasrequiredto posta bondof roughlyten timeshis annualsalarybeforehe couldbe employed.The bondsgenerallycouldnotbe postedby one'sfamily. The company preferredthata groupof non-related peoplestandfor thegood conductof the employee. Thus,factorsearningœ40per annumhad to post a bond of œ400,whereasmerchantsplacedbondsfrom œ800to œ1500[23c;7 p. 256]. The RAC alsopaidcloseattentionto thosewhostoodbond. In a letter sentto Africa in 1690,the companywrotethat "...we find nowthat we haveby 400 poundSecurityfrom Mr. Ronanoneof hissecurityMr. Gregory Wale beingfailed"[3 p. 18;23c,11thNovember1690]. In response to a letter for promotionfromMr. Nightingale, thecompany wrotethattherequest"has occassioned us to 1ookeinto his securitywho we find to be his father in Holland if alivea man of smallsubstance ... maybe if he camehomehe will be able to settle his securitybetter to our likeing".[3 p. 19; 23c, 18th September1688]. Indeed, the companywould not promote any person without that person increasinghis bond. In the same letter to Mr. Nightingale,the directorsstressedthat theycouldnot "prefferanyMan to the greatesttrust without a suitableSecurityand if you gert Friendshere to strengthen youSecurity... we shallbe readyto bid youwelcome"[3 p. 19]. Thus, not only do the choiceof toolsusedby thesetwo companies,differ (outsideof salary),but the languageusedin the letters to expressthe company's relationship withitsmanagers differsaswell. I will arguein later sections that the differingchoiceof toolsis not accidental.The companies' choiceswere constrained by the environmentin whicheachoperated. It is clearthat neitheroaths,large salaries,nor large bondsare likely to performwell if a managerbelievesthathe will notbe caught.By cheating, the managercanincreasehisannualincome.Thusa companymustbe able

147

to monitorits managersand then be seento punishthosewho are caught cheating.Both companies monitoredtheir managersdirectlyand indirectly. A methodof direct monitoringwas to searchshipsfor evidenceof illegal trading. This meantthat all shipsleavingLondonfor Africa or HudsonBay aswell asall returningshipsweresearched for anyillegalmerchandise. Both companies hired"waiters" to searchvessels andpaidthema percentage of the valueof merchandise discovered.An entryin 1688for the RAC documents that on searching the shipSarahBonadventure in the Thamesestuarybefore it left for Africa, the waitersfound"4 peccesSayes,15 peccesPerpetuanes, 3 boxescontaining10 grossof Knivesand i barrell containingfive grossof KnivesoverandabovetheCompany's cargo"[23f]. A searchonthe Unityon its arrivalin Londonfound "by accountof the waiters.Inwardsthere was broughthome in the said ship (viz) 5 Negroe boyes,I Negroe girle, 46 Ellephantsteeth,i caskwax 1 Elephantstooth..."[23f]. A major problemfacingthe RAC wasthe fact that not all of its trade wasconducted betweenLondonandAfrica. Slaveswere shippedfrom Africa to the West Indies. The companyaskedits managersto searchall shipson arrival in Africa, and the agentsin the West Indieswere requiredto search shipson arrivalthere. The obviousproblemwasone of collusionbetween managerand ship captain. Indeed, the companyknew that many of its captainslandedslavesillegallyin theWestIndies,but it wasdifficulteither•to catchthem or to get directevidencethat they had doneso. However,the charterparty agreements for this periodshowthat the companydid try to make it more difficultfor slavesto be smuggledinto the West Indies. In particular,slaveshad to be countedand recordedon boardingand then recountedon deck everyfourteendaysof the voyage. The log had to be signedbyall officersof the crew,andall deathshadto be recordedon signed deathcertificates.But hereagain,collusion wasa difficultproblem,especially sincea captainhad the powerto coercethe morejunior membersof the crew [3]. Evenso,all shipswere searchedon arrivalin the WestIndies. The HBC followdthe samesearchprocedure.The companyminutes for 1679reportthat "CharlesWilmottandJeremyGriffithbeingappointedto goeon boardthe JohnandAlexanderandto staythereto preventall Frauds and Embezelments,tooke their Oathesso to doe and carried with them a Letterto Capt.Walkerto receivethemon board"[5 p. 865]. In fact,the HBC searchednot only the shipsbut alsoall personalbaggageand lettersfrom HudsonBayand sealedall legalconsignments of tradegoodsbeforeputting them on boardship. None of this solvedthe underlyingproblemfacingthe headofficesof both companies:how to tie managerialeffort level to output. Direct monitoring,suchas searchingshipsallowedthe head officeto discoverif a managerwas chcating,but both companiesalso neededto increasethe amountof informationtheyhad aboutthe trade and their managers.Each company requireditsmanagers to keepaccount books,journals,andledgers. The companies soughtto ensurethat it receivedthisinformationby laying downpreciseregulations and rigorousstandards governingthe waythat the accountbooksandjournalswere kept.

148

The HBC reprovedits managersfor not keepingletters"inparagraphs in the approvedmanner." In the letterssent from HudsonBay, the first paragraphstateswhat the directors"willfind here enclosed..." [5, pp. 86769]. Indeed, the HBC accountswere kept in a uniform mannerby all managers. The RAC managersalsowere requiredto keep accountsand journals;in 1688,the headofficeevenincludedthe specificform to be used [3, pp. 22-24]. As an example,onesetof instructions requiredthe useof bills of lading. Instructions to Capt.Potloy,factorat Wiburne, requiredthat he get four bills of ladingfrom everyship: one to be sentto the factorsin the West Indies,anotherto be sentto the head officeby the sameship,a third to be sentto the headofficeby the followingconveyance, anda fourth"itwill be reasonthat you shouldkeepeby you"[23c,26th July1687]. Althoughthe level of complianceby RAC managersdoesnot seemto havebeenas great as for HBC managers,both companieswere able to use the accounting systems to increasethe levelof availableinformationaboutthe performance of anygivenmanager. In essence, suchinformationallowedthe managersto run tournaments.The HBC seemsto havedonethisto a greaterextentthan the RAC, generating termsof tradeindicesfor eachpost. If an HBC postfell belowa certainlevel,thenthemanagercouldbe demotedor fired [5, pp. 87071]. Obviously, it is not enoughfor companies to monitortheir managers; theyalsohaveto be seenpunishing thosewhowerecaughtcheating.During the first decade of the RAC's operation,the agent-generalwas caught defraudingthe company.He wasrecalledand replaced;hisreplacementalso had to be replaced[3, p. 26]. There is no questionthat the companyleft in placemanagersthat it knewto be actingfraudulently.The HBC alsorecalled managers for questioning at the headoffice. Not manymanagers werefired, but some were demoted to better learn the trade, as in the caseof Richard

Norton [5, p. 871]. From the precedingdescriptionit would appearthat the HBC and RAC companies usedmethodsthat the theoreticalliteraturesuggests should reducemanagerialmisconduct.Bothusedhighandrisingsalaries;one used gratuities,the other,bonds. Both monitoredmanagerialactivityand were seen to punishthosewho were caught. Yet the HBC was a successful company, whiletheRAC failed.As mentionedearlier,the failureof theRAC is laid on its inabilityto manageits managers.From the discussion laid out here,however,there is no questionthat the companynot onlyattemptedto manageits managersbut alsousedmethodsconsistent with the theoretical materialon principalagentproblems. The lack of success in this endeavor stemsnot somuchfrom lackof effort but from effortsput in placethat were nevergoingto be costefficient. The environmentin whichthis company operatedgenerateda situationparticularly ill-suitedto the problemit hadto solve.

The African Environment and the Role of Company Culture

The necessity to useclosemonitoringor highsalariesis lessened to the extent that the companycan generatea companyculture to which its

149

managersconform. Suchan ethicalsystemalsohelpsalleviateproblemsof adverseselection,becausethe new manageris expectedto behavein a particularway. The questionis, how or under what circumstances is a companylikelyto generatesucha companyculture?Althoughthisquestion cannotbe answeredspecifically, the generationof sucha systemis relatedto the natureof the gamebeingplayedbetweenmanagerand firm and to the typeof externalenvironment in whichthe firm is operating.It maywell be that the firm doesnot consciously decideon the type of culturegenerated; but, onceit is generated,thisculturewill affectmanagerialbehavior. The literatureon the HBC illustratesthat that companysuccessfully generated the notionof the HBC asfamily[5]. Attemptsto createa "family environment" are particularlyobviousduringthe Frenchwars.The company wroteits chiefagentthat "forthe betterencouragement of yourmen in their Duty Wee doe herebypromiseand assureyouand them,that Wee will take Careby wayof Pensionfor suchof themashaveingbehavedthemselves with Courageand fidelityshalbewoundedand the Wivesand Childrenof those whoe shall happento bee killed in our Services"[5, p. 873]. All the correspondence is repletewith positiveencouragement. The largegratuities paid were often not tied to actualoutput,they were intendedto promote future effort.

The abilityto generatethisfamilyenvironment wasfurtherenhanced by the waythe companychoseits managers.In its earlyyears,the company either hired managersin Englandor promotedpeoplethroughthe ranks. Althoughthesepeoplecouldbe verygoodmanagersand traders,the head officebelievedthat performancecouldbe better. To this end, the company startedan apprenticemanagerialclassmadeup of youngboys,takenfrom institutions suchas Christ'sCollege. These"bluecoat"boysusuallybetween the agesof fourteenandtwentyone,learnedtherequiredfur tradeskillsand accountingprocedures. Most of theseboyshad no immediatefamily in England,andthe fur tradingskillstheylearnedwerenot easilytransferable. Thus,not onlydidthe company solveitspotentialadverseselection problem, but it alsogenerateda groupwhowerelikelyto remainwiththe companyfor a longtime. In game-theoretic terms,thecompany waslookingat a repeated game,andfor bothsidesreputation wasclearlyimportant.With itsgratuities andbonuses, theheadofficewasinvesting in a futurerelationship betweenthe directorsand the agents[5, pp. 872-74]. The RAC situationwasverydifferent. While the HBC wasplayinga repeatedgame,theRAC wasengaged in a non-repeated one. Managerswere hiredfor a three-yearterm;onlytheveryoccasional managersignedup for a secondterm. This madeit difficultfor the companyto investin the future. Gratuitiesand bonuseswouldnot lead to futurepay-offsfor that agent. Twobasicproblems facedtheRAC in itsattemptto generatea positive feedbackrelationship withits managers.The first relatedto the factthat the coastof Africawasa veryunhealthyplacefor Europeans.The mortalityrate for managerswas high. The Companyletters or minutesreport such statements as "thedeathof threeAgentssuccessively at the Fort in Gambia" or "thelongillnessof AgentHodgkin"[3 p. 19]. In fact,usingthe Listsof LivingandtheDead,Daviesestimated thatover60% of thosewhofirstlanded

150

in Africa were dead by the end of the first year, althoughthe death rate declinedsubstantially thereafter[12 ch.4]. In thesecircumstances a manager

wasunlikelyto makea long-termcommitment to the firm. Anotherfeature' of thisunhealthyenvironment wasthe ambiguityconcerning the amountof effortexpectfor anygivenlevelof salary.The headofficecouldhavebeen expecting a givenamountof effort,whilesicklymanagers mightonlyhave beenableto operateat a lowerlevelof productivity.Sucha difference in expectations wouldtend to generatea more confrontational relationship betweenthe managerandthe headofficeandmayin part explainthe lackof anypositivereinforcement by the company. The second factor lies in the nature of the outside environment. The

HBC tradingareawasrelativelyclosed.Evenwhenthe managers hadgained expertise, thefirmwaspractically theonlyplacewheretheycoulduseit. The managersof the RAC operatedwithina radicallydifferentenvironment.Any RAC agentoperatingin Africa was in dailycontactwith tradersof other nationsandwith interlopersfrom Britain. A letterwrittenin 1684statesthat interlopersfrom both Englandand the plantations"notonly spoilthe trade but Corrupt our own Servants..."[3 p. 29]. Cheatingwas easy. Yet the problemfacingtheRAC wasevenworsethanthe availability of opportunity to cheat.Becauseof theopenness of theenvironment, managers weremaking contacts aswell asgainingexperience thatwouldbe usefuloutsidethe firm structure. They could work for themselves, which made managersless interestedin theirfuturewiththe firm. Giventhe rapidturnover,the lack of positivereinforcement, andthe easewithwhichoutsideoccupation couldbe obtained,there was little that boundthe managerto the firm. Conclusion

The failure of the RAC and the successof the HBC do not stem from

lackof effortto controlmanagers.Bothcompanies monitoredtheirworkers and usedhighsalaries,bonds,and oathsto reinforcerequiredbehavior.Yet the environmentin which the RAC operatedproved inadequate. The companypaid highsalaries,but it couldnot meet the alternatives available. The companymonitored,but firingas a punishment did not havea large impactif formeremployees couldthenwork for themselves.The company did not use strategiesthat would enhancethe future growthof the firm becauseof the non-repeatednatureof the game. The RAC wasa company for whomthe costsof agencywere too highto makea hierarchicalstructure work. References

1. 2.

K. J. Arrow, "The Economicsof Agency," J. Pratt and R. Zeckhauser,eds., Principalsand Agents.'TheStructureof Business (Boston,1984). Jennifer Brown, Strangersin Blood (Vancouver, 1980).

3. Ann M. Carlos,"AgentOpportunism and the Role of the Environment:Evidence from the Royal AfricanCompany,1672 -1691," workingpaper,Department of Economics,Universityof Colorado,Boulder,(1991).

151

4.

Ann M. CarlosandStephenNicholas,"Giantsof An Earlier Capitalism:The Early Chartered Companiesas Modem Multinationals,"BusinessHistory Review, 62 (1988), 398-419.

5. 6. 7. 8. 9.

10. 11. 12.

13.

14.

15. 16. 17. 18.

19.

, "AgencyProblemsin Early CharteredCompanies: The Caseof the Hudson'sBay Company,"Journal of EconomicHistory, L (1990), 853-75. R. Caves,MultinationalEnterpriseandEconomicAnalysis(Cambridge,MA, 1982). Keith G. Davies, TheRoyalAfrican Company(London,1957). , ed., LettersfrownHudsonBay 1703-1740 (London, 1965). R. Davis, "EnglishForeignTrade 1600 - 1700," in The Growthof EnglishOverseas Trade in the 17th and 18th Centuries,ed., W.E. Michinton(London, 1973). RichardDunn, Sugarand Slaves(ChapelHill, 1972). David W. Galenson,Traders, Planters,and Slaves(Cambridge,1986). Stanley Engermanand Eugene Genovese,eds., Race and Slavery in the Western Hemisphere.'QuantitativeStudies(Princeton,1975). KathleenHagerty, AharonOfer, andDaniel Siegel, "ManagerialCompensation and Incentivesto Engage in Far-SightedBehavior", working paper, Northwestern, University November 1989. Harold A. Innis, The Canadian Fur Trade (Toronto, 1930). SidneyPollard, The Genesisof ModernManagement(Cambridge,MA, 1965). Eric Rasmussen,Gamesand Information (Cambridge, 1989). JamesA. Rawley, The TransatlanticSlave Trade (New York, 1981). Arthur J. Ray and Donald Freeman, "Give Us Good Measure".'An Economic Analysisof RelationsBetweenthe Indiansand the Hudson'sBay CompanyBefore 1763, (Toronto, 1978). , TheHisto .ryof theHudson'sBay Company1670 -1870 (London, 1958).

20.

, ed., Hudson's Bay Copy Booke of Letters, Commissions, Instruction Outward, I688-1696 (London, 1957). 21. E. E. Rich, ed., Minutesof the Hudson'sBay Company,1671-74 (London, 1942).

22.

,ed., MinutesoftheHudson'sBay Company1679-1684,(London,

1954) 23. Royal African Company,PublicRecordOffice, Kew, England a. Extract of Letters sentto the Royal African Companyof Englandfrom West Indies and Africa 1678 to 1681, T70/1.

b. c.

Copiesof An Oath givento Servantof the Company,T70/1506. Copiesof Letterssentby the Royal African Companyof Englandto the Coast

d.

of Africa 1685 to 1698, T70/50. Abstracts for the Committee of Accounts 1678 to 1681, T70/15

e.

Minute Bookof the Generalcourtof the RoyalAfrican Company,No. 2 1678 to 1690, T70/101.

f. Black Book of the Royal African Companyof England,T70/1433. g. Abstractof lettersreceivedby Royal African Companyof Englandas far as relateto the Committeeof Correspondence, T70/10. 24. BarbaraL. Solow and StanleyL. Engerman,British Capitalismand Caribbean Slavery: TheLegacyof Eric Williams(Cambridge,1987). 25. Jeffrey Williamson, "Structureof Pay in Britain, 1710-1911," Research in EconomicHistory, 7 (1982).

26. Oliver Williamson, The EconomicInstitutionsof Capitalism(New York, 1985).