The logic of management consulting, parts 1 and 2

Canback, Staffan. 1998. The Logic ofManagement Consulting, Part 1. Journalof Management Consulting 10 (2): 3-11A searchable text version of the article is found at the back of this document

THE LOGIC OFMANAGEMENTCONSULTINGPart OneStaffan Canback1998© 1998 Journal of Management ConsultingReprinted with permission from the Journal of Management Consulting

2Editor’s Note: In this the first of twoarticles in our series, “The Logic ofManagement Consulting,” StaffanCanback, a consultant at MonitorCompany, traces the history and trajectoryof the management consultingindustry and introduces transactioncost theory. In the second articlein this series, Canback will usetransaction cost theory—originallydeveloped in the 1930s by Ronald H.Coase—to help explain the existenceand phenomenal growth of this industry.Transaction cost theory has severalapplications in economics and management.One of the most importantis to help explain the boundaries offirms—why certain activities, products,or services are carried out internallyin firms—while others arebought and sold in the market place.As such it is a useful framework forthinking about management consultingservices. Why after all do executiveshire consultants when theymight do the work themselves?By using transaction cost theory asits intellectual foundation, the articleanswers two questions: 1) why domanagement consultants exist; and2) why do they organize in independentfirms?Despite current popularity and astoundinggrowth rates, managementconsulting remains one of the leastresearched and written about industries(Gagnon 1984). We take forgranted that the industry should existand function in the way it does. Yetthe tremendous growth of the managementconsulting industry over thelast 20 years cannot be easily explained.As one “Bernie Ramsbottom”put it in the Financial Times(April 11, 1981):Of all the businesses, by farConsultancy’s the most bizarre.For to the penetrating eye,There’s no apparent reason why,With no more assets than a pen,This group of personable menCan sell to clients more than twiceThe same ridiculous advice,Or find, in such a rich profusion,Problems to fit their own solution.For the purposes of this article, wewill define management consultantsas those who provide general managementadvice within a strategic,organizational or operational context,and who are institutionally organizedin firms. It excludes other types ofconsulting such as human resource,information technology, and actuarialconsulting which have little in commonwith management consultingexcept for the project nature of work.It also excludes management consultantswho are not institutionally organized.My estimate is that the chosensegment of the consulting marketaccounts for around 30 to 40percent of total consulting revenues,and 80 percent of management consultingrevenues.What is management consulting?According to Greiner and Metzger(1983): "management consulting isan advisory service contracted forand provided to organizations byspecially trained and qualified per-

3sons who assist, in an objective andindependent manner, the client organizationto identify managementproblems, analyze such problems,recommend solutions to these problems,and help, when requested, inthe implementation of solutions."There are a few key words in this definition.Advisory service indicatesthat the consultants are responsiblefor the quality of their advice, butthey do not substitute for managersand have no formal authority. Objectiveand independent indicates financial,administrative, political, andemotional independence from theclient (Kubr 1996). Trained and qualifiedshows that a consultant is morethan the individual and his or herpersonal experience. As we will see,these characteristics sometimes contributeto the demand for externalconsulting services, and sometimesdetract from it.Within the context of the definitionabove, management consulting hasa long history (e.g. Moore 1982;Kubr 1996; UNCTAD 1993). The firstmanagement consultants appearedaround the turn of the century andincluded individuals such as FrederickTaylor, Henry Gantt, Arthur D.Little, and Harrington Emerson, all ofwhom are still famous for their contributionsto the science of management.Little and Emerson alsostarted two of the first institutionalconsulting firms. These pioneerswere mainly concerned with operationalefficiency issues such as Taylor'stime-and-motion theory.Between 1910 and 1940 a secondgeneration of consultants expandedthe concept of management consulting.Edwin Booz started offering"business research services" in1914, and James O. McKinseystarted McKinsey & Company in1926. In Europe, Lyndon Urwick andCharles Bedeaux were pioneers whocontributed extensively to definingmanagement consulting in the1920s. These consultants pioneeredor implemented techniques such asbudgeting processes, the divisionalizedorganization, merit-based compensationschemes, and forecastingtechniques.During the early post-war years andin many cases growing out of wartimeexperience, consulting experienceda big surge, with formation ofsuch firms as Cresap, McCormick &Paget, William E. Hill, Bruce Payne &Associates, Hay Associates, andTowers Perrin.Three major developments tookplace in the 1960s. First, Bruce Hendersonmoved from Arthur D. Little,Inc. to start the Boston ConsultingGroup in 1963 and more or less single-handedlyoperationalized theconcepts of strategy and strategyconsulting. Out of this sprang asecond generation of strategy specialistssuch as Bain & Company,Strategic Planning Associates, BraxtonAssociates, LEK Partnership,and Monitor Company. Second, themajor accounting firms started respondingto the growth of managementconsulting and created managementadvisory service groups to

4augment their core accounting practices.Today the consulting practicesof Andersen Worldwide, PricewaterhouseCoopers,Deloitte & Touche,and Ernst & Young often rival theaccounting activities of these firms insize.Also starting in the 1960s with theemergence of Cambridge ResearchInstitute and Management AnalysisCenter (today, both history), firmsinstitutionalizing the combined consultingpractices of leading academicsand practitioners began to maketheir presence known.Yet as late as 1980, despite a growingproliferation of consulting specialties,management consulting was stillan industry in its infancy with perhapsaround 18,000 practicing managementconsultants worldwide, andonly around thirty to forty percent ofthese employed in the large, institutionallyorganized firms of the typementioned above 1 (ConsultantsNews 1982–1997; Payne 1986).Even the largest consulting firm inthose days, Booz•Allen & Hamilton,had revenues of only around $150million. The industry as a whole hadrevenues of $1.2 billion in the U.S.and worldwide perhaps $2 billion.Over the next 17 years, the managementconsulting industry grew toaround $35 billion globally. The annualgrowth rate has been more than20 percent. Today, there are approx-1 The numbers presented in this section are theauthor’s reconciliation of several sources. Theyare broadly in line with most observers.imately 140,000 consultants worldwide(a considerable fraction of thismore recent growth and peoplecount is accounted for by informationtechnology projects manned less bymanagement consultants than bysystems integration specialists).This growth is impressive, but thetrue importance of the industry’s evolutionis the accumulation of institutionalknowledge. In 1980 there wereless than five consulting firms withmore than 1,000 consultants, todaythere are more than 30. If the experiencecurve applies in consultingservices, then it may be noteworthythat approximately 80 percent of allconsulting experience was generatedin the last 17 years, and only 20 percentin the period from 1886 (whenArthur D. Little started the first consultingfirm) to 1980. As we will see,this has had profound implicationsfor the division of labor and the balanceof power between consultantsand clients.MANAGEMENTCONSULTING'SIMPORTANCEMore than just a growth industry,management consulting in and ofitself is one of the most importantand enduring management techniquesdeveloped over the last 50years. A secondary effect of this inventionhas been the rapid disseminationof new frameworks, tools, andtechniques in large companies.

5Surprisingly, however, not much hasbeen written about this phenomenon.In part, this must be because few areinterested in the topic—it is still seenas an admission of failure by manymanagers to use consultants, andwho wants to read about failure? Inpart it is because the managementconsulting firms are highly secretive,and thus difficult to analyze and understand.A few facts and observations dospeak for themselves. Managementconsultants today employ around 25percent of the graduates from theleading business schools, and thosegraduates are usually among the topperformers in their class. Some traditionalcompanies have essentiallygiven up recruiting at these schoolssince consulting firms and investmentbanks can offer what is perceivedas more career opportunity,better pay and a more stimulatingenvironment than traditional companiesin manufacturing or services.Another aspect is that today thereare approximately 70,000 managementconsultants in the UnitedStates, while there are around150,000 executives of the type consultantsnormally interact with atfirms governed through “complex”management (Granovetter 1984).That is, for each executive there are0.5 consultants who advise, full time.In 1980, this ratio was approximately0.1. Clearly, and without inferringany judgement on the relative contributionof executives and consultants,the balance of influence is shiftingdramatically.Finally, several industry observers,including Payne (1986), argue thatinnovation in fields such as strategyis dominated by management consultants,and not by managers oracademics. The same is probablytrue for other management disciplines.Take, for example, reengineeringin its various incarnations.Consequently, management consultantshave had a large impact on thestate of management due to both thequantity and quality of contributions.Yet, this does not explain why managementconsultants exist. It is notclear why managers would want togive away so much of their companies’intellectual agenda to outsiders.It is not obvious why it is more costeffective to hire experts from the outsidethan to do the same work internallyin companies. And even if it is,why is this happening on a massivescale now, and not 60 years ago?Why is it happening in the UnitedStates but only to a limited extent inJapan?Before addressing these issues, thenext three sections build a platformof understanding of the task of managementconsultants, and the basicsof transaction cost theory, by reviewingthe relevant literature.

6MANAGEMENTCONSULTANTS’ ROLES ANDTASKSSchein (1988) categorizes managementconsultants with respect to therole they play in their interaction withclients. He distinguishes betweenthree models of consultation: 1) purchaseof expertise; 2) doctor–patient,and 3) process consultation.The purchase of expertise model isused by clients who require the consultantto bring their own independentperspective on the industry andthe issues at hand. In its purest form,the consultant is not expected to interactextensively with the client butrather to provide his or her expertisein a hands-off relationship.In the doctor–patient model, the consultantemphasizes his or her diagnosticcapability by carefully analyzingthe client organization’s problems.Using the consultant’s oftenunique experience base and diagnosticskill, the consultant quickly assessesstrategic and organizationalblockages. This model leads to anintimate and often trust-based relationshipbetween the consultant andthe client.The process consultation modelbuilds on the notion that the consultantis the facilitator, while the clientcontributes the expertise. Thus,there is a clear division of roles andtasks. The client ultimately chooseswhat to do about a problem. Theconsultant, on the other hand, providesa methodology for defining theproblem and finding the best possiblesolutions. The similarity to psychologicalanalysis methods is notcoincidental.Schein’s classification reflects arange of roles from the consultant asa content provider, to the consultantas a process provider. A similarsegmentation is suggested by Neesand Greiner (1985), who dividestrategy consultants into five categories.The "mental adventurer" analysestruly intransigent problemssuch as long term scenarios forcountry development, by applyingrigorous economic methods and leveraginghis or her experience base.The "strategic navigator" bases hisor her contribution on a rich quantitativeunderstanding of the market andcompetitive dynamics, and then recommendscourses of action withouttoo much regard of the client’s ownperspective. The "management physician"derives their recommendationsfrom a deep understanding ofthe internal dynamics of the clientorganization, often willingly sacrificingsome objectivity to gain a realisticperspective on what is achievable.The "system architect" impacts his orher clients by helping redesignprocesses, routines, and systems—always in close cooperation with theclient. Finally, the "friendly co-pilot"counsels senior managers as a facilitatorrather than as an expert, andhas no ambition to provide newknowledge to the client.The mental adventurer broadly correspondsto Schein’s expert model,

7the strategic navigator, managementphysician, and system thinker correspondto his doctor–patient model,and the friendly co-pilot is similar tothe process consultation model.Nees and Greiner further show thatinstitutionally organized strategyconsultants are found primarily in thestrategic navigator and managementphysician segments. The BostonConsulting Group, Bain & Companyand Monitor Company are examplesof the former, and McKinsey & Companyof the latter. Clearly, the role ofthe consultant in both segments requiresa relationship between clientand consultant which goes beyond acontractually specified arms-lengthrelationship.Turner (1982) uses a hierarchy oftasks to demonstrate the extent of aconsultant’s involvement with aclient. He argues that up until thelate 1970s, the consultant oftenworked as a supplier to the client,but that the relationship increasinglyis built on a partnership of mutualrespect aimed at fundamentally improvingthe client’s effectiveness.Turner uses eight task categories todelineate management consultingapproaches. The first five correspondto the traditional arms-lengthsupplier status, the last three arenewer, evolving tasks:1. Providing information to a client4. Making recommendations basedon the diagnosis5. Assisting with implementation ofrecommended actions6. Building a consensus and commitmentaround corrective action7. Facilitating client learning8. Permanently improving organizationaleffectiveness.Most management consulting firmstoday aspire to work on the highervalue added activities at the lowerend of the list. Thus, it is once againclear that a management consultants’relationship with their client isbecoming increasingly complicated,and that it relies more and more onsophisticated contractual arrangementsof primarily informal nature,such as trust. However, researchhas also shown (Leontiades andAhmet 1989) that management consultantsstill have a long way to gobefore they exert major influence onthe core issues of their clients. Achief executive is more likely to beinfluenced first by his or her own instinctsand thinking on a particularsubject, followed by the planningstaff, the board of directors, and investmentbankers, than by the consultants.Thus, it is unclear how fardown the task hierarchy managementconsultants have really moved.2. Solving a client’s problem3. Making a diagnosis, which maynecessitate redefinition of theproblem

8PRACTITIONERS’ VIEWSMarvin Bower (1982), the drivingforce behind McKinsey & Companyover almost half a century, suggestssix reasons why hiring external consultantsmakes sense in many situations:1) they provide competencenot available internally, 2) they havevaried experience outside the client,3) they have time to study the problems,4) they are professionals, 5)they are independent, and 6) theyhave the ability to create actionbased on their recommendations.However, he does not make clearwhy most of these statements shouldbe true.In large companies, the core marketfor management consultants, most ofthe skills provided by consultantsshould ostensibly be available internallysince large companies haveencountered most classes of problems.Creating the time to study aproblem should simply be a matter ofpriority-setting. That the degree ofprofessionalism is automaticallyhigher within a consulting firm is notobvious. Furthermore, there are argumentsboth for and against theproposition that consultants are moreindependent than internal managersand experts. Finally, the superiorability to create action, attributed toconsultants by Bower, appears to bea matter of training and methods andnot intrinsic to the consulting capability.Thus, only the second statement—thatconsultants have variedexperience outside the client—appears to be correct prima facie.Implicit in Bower’s argument, however,is the belief that consultants workprimarily with Schein’s first two models,the expert and doctor–patientmodels, since the consultant is expectedto provide an independentperspective on the substantive issuesat hand. In Turner’s hierarchy,this corresponds to the lower levels.Bower appears to see the consultantas a partner to the client in solvingunstructured, difficult problems, ratherthan as a supplier of packagedmethods and approaches.Bruce Henderson, the force behindthe Boston Consulting Group formany years, has a similar perspective(Hagedorn 1982). He arguesthat consultants add significant valueto society (through their clients) byreducing the problem solving cycletime. Exactly why management consultantshave more of this capabilitythan others is, however, unclear. Butas with Bower, Henderson’s implicitargument is that management consultantswork together with theirclients in a complicated relationshipto jointly solve the problems at hand.Henderson also argues that the consultantneeds to work in a specializedinstitutional environment whichtakes into account that the key resourceis the body of consultants, ahighly mobile resource, and that aconsulting environment is characterizedby instability.Kelley (1979) makes a contrary argumentto Bower and Hendersonbased on interviews with more than200 internal consultants at variouscompanies. Among other things, he

9argues that external consultants aremore expensive than internal consultants,they are not available at theright time, and they lack an understandingof the client’s environment.This reduces the external consultant’seffectiveness. Kelley also predictsthat the bulk of consulting workwill be carried out by internal resourcesin the future and that externalconsultants will be used only forspecial problems and when there is aneed to augment the internal resources.As was quantified earlier inthe article, Kelley has been provenwrong by events, and the managementconsulting industry is todaymany times larger than when hewrote his article. In fact, we will seelater that external management consultantsare both cost effective,available, and adept at understandingtheir client’s problems and circumstances.The above summary of the literaturepoints at a number of propositions:• Management consultants increasinglyaddress critical, long-termissues of their clients’ and are asignificant part of the intellectualagenda of executives (correspondingto Turner’s three lowerlevels).• Consultants add value by addressingboth content andprocess issues based on expertise,methodology, and generalproblem solving skills (correspondingto Schein’s expert anddoctor–patient models).• Management consultants worktogether with their clients in acomplicated and fluid relationshipcharacterized by a high degree ofmutual trust.• Management consultants arebest organized in independent,specialized firms with unique characteristicsand success factors(as argued by Bower and Henderson).TRANSACTION COSTTHEORYThe above perspectives do not shedmuch light on why managementconsultants exist. Transaction costtheory, however, may. The theorydeals with the real costs of allocatingresources in an imperfect world ofmisunderstandings, misalignedgoals, and uncertainty. Since managementconsultants deal with thisvery issue it may be that the theorycan help explain the existence of thisprofession.Transaction cost theory was initiallydeveloped in the 1930s by Ronald H.Coase, to help explain why certainactivities, products, or services arecarried out internally in firms—whileothers are bought and sold in themarket place. His ideas were neglectedfor many years, but around1970 several scholars started expandingon Coase’s ideas. Most notableof these is Oliver E. Williamson,who over the last 25 years has dedi-

10cated his research to transactioncost theory issues.Unfortunately, this massive effort hasnot yielded a good definition of whattransaction costs are, and there hasbeen considerable criticism of thelack of clarity and testability of thetheory. The following is yet anotherimperfect attempt at defining transactioncosts.First, a company’s costs are usefullyclassified in two categories: productioncosts and transaction costs.Production costs are those we aremost familiar with. They are all thecosts that are associated directlywith productive activities (Masten1982) such as manufacturing, logistics,and product development.Transaction costs, on the otherhand, are those costs associatedwith organizing economic activity.They thus vary with organizationalform (Masten 1982). Or as KennethArrow (1983) put it, “The distinctionbetween transaction costs and productioncosts is that the former canbe varied by a change in the mode ofresource allocation, while the latteronly depend on the technology andtastes, and would be the same in alleconomic systems.” It has been estimatedthat at least 45 percent ofthe gross national product in a developedsociety is generated bytransaction costs (Wallis and North1986).Ronald H. Coase (1937) defined theterm transaction costs in his pioneeringwork The Nature of the Firm byasking these fundamental questions:"Why is there any organization?" and"Why isn't all production carried outby one big firm?" His answer wasthat there are transaction costswhich determine what is done in themarket, with price as the regulatingmechanism, and what is done insidethe firm, with bureaucracy as theregulator. Coase pointed out that"the distinguishing mark of the firm isthe supersession of the price mechanism."Within this framework, alltransactions carry a cost, either asan external market transaction costor an internal bureaucratic transactioncost. “The limit to the size of thefirm . . . [is reached] when the costsof organizing additional transactionswithin the firm [exceed] the costs ofcarrying out the same transactionsthrough the market.” (Coase 1993).As we will see later, this is exactlythe issue for management consulting.Why do companies buy this servicethrough a market transactionrather than doing it themselves?According to Coase (1937) the mostimportant market transaction costsare the cost of determining the priceof a product or service, the cost ofnegotiating and creating the contract,and the cost of information failure.The most important internal transactioncosts are associated with theadministrative cost of determiningwhat, when, and how to produce, thecost of resource misallocation (sinceplanning will never be perfect), andthe cost of demotivation (since motivationis lower in large organizations).In any given industry the relativemagnitude of market and internal

11transaction costs will determine whatis done where.Williamson (e.g. 1975; 1985) extendedthe argument by noting thattwo behavioral assumptions are critical.First, individuals in an organizationare boundedly rational. This, inthe words of Herbert Simon (1976)means that “human behavior is intendedlyrational, but only limited so.”This limitation makes it impossible tostructure perfect contracts and anycontract will be incomplete even if allinformation is available. Second, individualsbehave opportunistically.This means that they will act in selfinterestwith guile. While some objectto this strong assumption, a numberof studies have shown that it is validin organizations (Williamson 1993)and it is a well established tenet ofDarwinian zoology (Dawkins 1989).The implication is that promises ofresponsible behavior are only crediblewhen they are supported by enforceablecommitments, since individualsotherwise would break anagreement if it is in their self-interest.With the two assumptions ofbounded rationality and opportunism,Williamson (1975) demonstrated thatthree factors play a fundamental rolein determining if market or bureaucratictransactions are optimal. Thefactors are asset specificity, uncertainty,and frequency of transactions.Under conditions of high asset specificitymarket transactions also becomeexpensive. By asset specificityis meant physical assets, human assets,site, or dedicated assets whichhave a specific usage and cannoteasily be transferred to another use.Under this condition, opportunisticbehavior can be expected if the assetis part of a market transaction.An example is if a supplier invests inspecific tooling equipment dedicatedto one customer (or for that matter ifa consulting firm invests in a clientrelationship). Over time, the customerwill be able to put pressure on thevendor since the vendor has no alternativeuse for its investment andwill be willing to accept a price downto the variable cost of production tocover some fixed cost. This leads toa difficult negotiation where eachparty may try to “cheat” and wherecomplicated safeguards have to beincorporated in the contract. On theother hand, if the customer owns theequipment itself, then the incentiveto cheat disappears and the cost ofcreating safeguard contracts is eliminatedsince the asset is owned bythe same company.High uncertainty such as businesscycle volatility or technological uncertaintywill lead to more bureaucratictransactions since it will be difficult,and prohibitively expensive, to createcontracts which cover all possibleoutcomes. Thus, with higher uncertaintyfirms tend to internalize activities.Finally, if the transactions arefrequent there is once again a tendencyto manage the transactionthrough bureaucracy since the repetitivecontracting cost will be higherthan the bureaucratic cost.

12Empirical research has shown thatthe three factors above indeed dohave an impact on the choice oftransaction mechanism. For example,Masten (1984) demonstratedthis within the aerospaceindustry, Teece (1981) and Klier(1993) in the automotive industry.The final important aspect oftransaction cost theory pertinentto this article restates an argumentfrom the beginning of thissection. Transaction costs alonedo not explain whether transactionsare carried out in the marketor internally in the firm. DouglassNorth, the 1994 Nobel Prize winnerin economics, has forcefullypointed out that firms try to minimizetotal cost, not only transactioncosts (e.g. North 1987; 1991;North and Wallis 1994). In additionto transaction costs, a firm hasproduction costs. Sometimes, andwe will see this in the example ofmanagement consulting, transactioncosts are not always minimized becausethe resultant improvement inproduction costs can outweigh theincrease in transaction costs.We can now summarize transactioncosts economics in the followingframework:TRANSACTION COST FRAMEWORKMarket (external)transaction costs•Price determination•Negotiation•Long-term deviationAsset specificityBounded rationalityUncertaintyOpportunismProduction costsBureaucratic (internal)transaction costs•Administration•Resource misallocation•DemotivationFinally, two specific applications oftransaction cost theory will be usedin the second part of the article.FrequencyAoki (1990) has identified some ofthe basic differences between Japaneseand American style management,and then used elements oftransaction cost theory to explainthese differences. One of his observationsis that spontaneous and voluntarycoordination is much moreprevalent than in Western firms.Thus the need for explicit performancecontracts is reduced. This isachieved by having a long period ofsocializing between employees—thesystem of life-time employmentcombined with a promotion systembuilt on seniority. A consequence isthat it is critically important to have

13stable hierarchies with clearly definedroles, and it is difficult to injectoutside expertise of temporary nature.Thus, while Japanese firms areadept at using suppliers for standardproducts and services, they find itmuch more difficult to use high value-addedservices from the outside.Englander (1984) applied the theoryto the short-lived practice of insidecontracting which was prevalent inthe early days of the manufacturingera, especially in New England. Underthis system, owners contractedwith suppliers to perform all operationswithin a factory, while providingthe productive assets such as machinery.In essence, the inside contractoragreed on a transfer pricewith the owner, and then had thefreedom to hire workers, developwork methods, and take whateveraction necessary to generate a profit.The practice broke down for fundamentaltransaction cost theoreticalreasons. The high asset specificitybetween owner and contractor (bothphysical, human, and site specificity)made it impossible to design contractsbetween owners and contractorswhich gave a fair share of profitsto both parties. The contractor, havingsuperior knowledge of operations,found ways to improve productivitybeyond the expectation of theowner. Thus, supernormal rents accruedto the contractor. At the sametime, the internal contractor did nothave many proprietary skills and itwas therefore relatively easy for theowner to replace the inside contractorwith his own supervisor and workforce.By the end of the 19 th centurythe inside contracting system hadgiven way to the vertically integratedindustrial firm where all resources,human and physical were under thecontrol of management. One maywonder if management consulting,which has much in common with theinside contractor, will disappear in asimilar way.In Part Two, to appear in the followingissue of this journal, StaffanCanback will deal with these andother implications of transaction costtheory as it relates to managementconsulting. He revisits the two questions:1) why do management consultantsexist; and 2) why do theyorganize in independent firms? Anddraws conclusions about the futureof the industry.This article is a reprint from the Journalof Management Consulting, 1998:Volume 10, Issue 2, pp. 3-11.

14LIST OF REFERENCESAoki, M. 1990. Toward an economic model of the Japanese firm. Journal ofEconomic Literature 28: 1–27.Arrow, K. J. 1983. The organization of economic activity: Issues pertinent to thechoice of market versus nonmarket allocation. In Collected Papers ofKenneth J. Arrow. Cambridge, Mass.: Belknap.Bower, M. 1982. The forces that launched management consulting are still atwork. Journal of Management Consulting 1 (1): 4–6.Coase, R. H. 1937. The nature of the firm. Economica n.s., 4: 386–405.Coase, R. H. 1993. The nature of the firm: Origin, meaning, and influence. In Thenature of the firm: Origins, evolution, and development, edited by O. E.Williamson and S. G. Winter. New York: Oxford University Press.Consultants News. 1982–1997. The world's largest management consultingfirms. Consultants News.Dawkins, R. 1989. The selfish gene. 2d ed. New York: Oxford University Press.Englander, E. J. 1984. An inquiry into the economic theory of the firm:Technology, internal organization, and public policy. Ph.D. diss., Universityof Washington.Gagnon, R. J. 1984. An integrated strategy for increasing management consultingresearch. Academy of Management Proceedings (August): 148–152.Granovetter, M. 1984. Small is bountiful: Labor markets and establishment size.American Sociological Review 49 (3): 323–334.Greiner, L., and R. Metzger. 1983. Consulting to management. Englewood Cliffs,N.J.: Prentice-Hall.Hagedorn, H. J. 1982. The anatomy of ideas behind a successful consulting firm.Journal of Management Consulting 1 (1): 49–59.Kelley, R. E. 1979. Should you have an internal consultant? Harvard BusinessReview 57 (November-December): 110–120.Klier, T. H. 1993. Transaction cost theory and just-in-time manufacturing: A newlook at vertical integration in the united states automobile market. Ph.D.diss., Department of Economics, Michigan State University.

15Kubr, M., ed. 1996. Management consulting: A guide to the profession. 3d ed.Geneva: International Labour Office.Leontiades, M., and A. Ahmet. 1989. CEO's perceptions of strategy consultants.Business Forum (Winter): 51–53.Masten, S. E. 1982. Transaction costs, institutional choice, and the organizationof production. Ph.D. diss., University of Pennsylvania.Masten, S. E. 1984. The organization of production: Evidence from theaerospace industry. Journal of Law and Economics 27 (October):403–417.Moore, G. L. 1982. The politics of management consulting. Ph.D. diss., CityUniversity of New York.Nees, D. B., and L. E. Greiner. 1985. Seeing behind the look-alike managementconsultants. Organizational Dynamics 13 (Winter): 68–79.North, D. C. 1987. Institutions, transaction costs and economic growth. EconomicInquiry 25 (3): 419–428.North, D. C. 1990. Institutions and a transaction-cost theory of exchange. InPerspectives on positive political economy, edited by J. E. Alt and K. A.Shepsle. Cambridge: Cambridge University Press.North, D. C., and J. J. Wallis. 1994. Integrating institutional change and technicalchange in economic history: A transaction cost approach. Journal ofInstitutional and Theoretical Economics (Zeitschrift für die gesamteStaatswissenschaft) 150 (4): 609–24.Payne, A. T. 1986. New trends in the strategy consulting industry. Journal ofBusiness Strategy 7 (1): 43–55.Schein, E. H. 1988. Process consultation: Its role in organization development.2d ed. Vol. 1. Reading, Mass.: Addison-Wesley.Simon, H. A. 1976. Administrative behavior. 3rd ed. New York: Free Press.Teece, D. J. 1981. Internal organization and economic performance: An empiricalanalysis of the profitability of principal firms. Journal of IndustrialEconomics 30 (2): 173–199.Turner, A. N. 1982. Consulting is more than giving advice. Harvard BusinessReview 60 (September-October): 120–129.

16United Nations Conference on Trade and Development. 1993. The managementconsulting industry: An overview. In Management consulting: A survey ofthe industry and its largest firms. New York: United Nations.Wallis, J. J., and D. C. North. 1986. Measuring the transaction sector in theAmerican economy, 1870–1970. In Long-term factors in Americaneconomic growth. Studies in income and wealth series, edited by S. L.Engerman and R. E. Gallman. Chicago: University of Chicago Press.Williamson, O. E. 1975. Markets and hierarchies: Analysis and antitrustimplications. New York: Free Press.Williamson, O. E. 1985. The economic institutions of capitalism. New York: FreePress.Williamson, O. E. 1993. Opportunism and its critics. Managerial and DecisionEconomics 14: 97–107.

Canback, Staffan. 1999. The Logic ofManagement Consulting, Part 2. Journalof Management Consulting 10 (3): 3-12A searchable text version of the article is found at the back of this document

THE LOGIC OFMANAGEMENTCONSULTINGPart TwoStaffan Canback1999© 1999 Journal of Management ConsultingReprinted with permission from the Journal of Management Consulting

2Editor’s note: In this the second oftwo articles in our series, “The Logicof Management Consulting,” StaffanCanback, a consultant at MonitorCompany, draws deeply on the industrybackground and the explanationof transaction costs detailed inthe first article to present a uniquerationale for the industry’s existence.He then develops scenarios for theindustry’s evolution, ending with aperspective on the future of managementconsulting. A future characterizedby continued growth and increasinginfluence of consultants.Consultants and clients alike oftenask why the management consultingindustry has grown so fast over thelast 20 years. Graduating studentssimilarly ask if the growth can besustained and if career opportunitiesin the industry will continue to be excellent.Skeptics, such as O’Sheaand Madigan (1997), argue thatmanagement consultants often donot add real value to their clients andthat the industry is a fad—albeit afad with longevity. Proponents arguethat we live in a free and open economyand if clients did not derive valuefrom consulting services, thenthey would stop using them.Transaction cost theory helps us understandthat there are fundamentalreasons why management consultantsexist and that the industry ismore than a fad. The theory alsohelps make predictions about the future.Under what circumstances willthe industry continue to grow? Is itpossible, and advisable, for clients torecapture some of the activities thatare now performed by consultants?These issues are discussed in thisarticle, ending with the perspectivethat the industry will continue to growand that external management consultantswill continue to increasetheir “problem solving market share.”WHY DO MANAGEMENTCONSULTANTS EXIST?Drucker (1979) argues that “themanagement consultant is an extraordinaryand indeed truly uniquephenomenon.” He argues that thereare two reasons why managementconsultants exist. First, managementis neither a science nor an art, it is apractice learned through exposure toand experience with a wide variety ofcompanies in a wide variety of industries.A typical executive, however,lacks that exposure: As Druckernotes: “He works with the same organization—orat the most, with veryfew. He lacks exposure and cannotgain it. Nor can he simulate it.” Consultants,on the other hand, transcendorganizations and thus gainexposure. Second, Drucker observesthat executives yearn for objectiveinsights into their management problems.Empirical research by Gattikerand Larwood (1985) confirms thatclients first and foremost look forstimulation, expertise, and objectivitywhen they turn to outside consultants.Both these explanations forwhy management consultants existare compelling, but they suffer fromnot being anchored in an underlyingtheory. Transaction cost theory pro-

3vides a rigorous and consistent explanationfor the existence of managementconsulting. To understandthe growth of management consultingwithin a transaction cost economicscontext, two fundamentalquestions need to be answered:• Why is there increasing demandfor the types of services managementconsultants provide?• Why is this demand best filled byexternal consultants who are notdirect employees of the firm—butrather contracted outsiders?along the lines suggested in thetransaction cost framework.While national accounts and censusTRANSACTION COST FRAMEWORKMarket (external)transaction costs•Price determination•Negotiation•Long-term deviationBounded rationalityBureaucratic (internal)transaction costs•Administration•Resource misallocation•DemotivationDemand for managementconsulting servicesAsset specificityUncertaintyOpportunismFrequencyIn Part One (Canback 1998), Greinerand Metzger (1983) defined whatmanagement consultants do: theyhelp solve management problems bygiving objective and independentadvice. Why is there such extraordinarydemand for these types of servicestoday, while the demand wasmuch lower 50 years ago?An answer is provided by Wallis andNorth (1986) who studied changes inthe U.S. economy between 1870 and1970 by dividing the gross nationalproduct into production cost andtransaction cost components. Theyfurther divided transaction costs intomarket transaction costs (i.e., thecosts of buying and selling in themarket place) and bureaucratictransaction costs (i.e., the costs ofcoordinating activities within firms),Production costsdata do not easily conform to thisbreakdown, Wallis and North neverthelessmanaged to show that transactioncosts have become an increasinglyimportant part of the U.S.economy. Their estimate is thattransaction costs have increasedfrom 8 percent to 45 percent of theeconomy between 1870 and 1970,with the highest growth in bureaucratic(internal) transaction costs.Using the same methodology, thisauthor found a continued increase intransaction costs over the past 30years.

4To understand this trend, considerhow the following underlying mechanismmight operate. As companiesstrive to reduce productioncosts by exploiting scale and scopeeconomies, they need to increasespecialization, which in turn leads toa need for internal coordination. Iftransaction costs did not exist, thenthe largest company would also bethe most profitable company in eachmarket, since coordination betweenfunctions could be achieved withouteffort. But with transaction costs, thisdoes not happen. Instead, largecompanies need to deploy considerablecoordination resources to realizethe production scale and scopeeconomies. On balance, this pays offand total productivity increases yearafter year. Reductions in productioncosts are larger than the additionalbureaucratic transaction costs incurred,and value added grows.Thus traditional blue collar jobs aredisappearing as production costs arereduced, while the number of whitecollar jobs aimed at coordination areincreasing. Moreover, more effort isspent on creating the appropriatecontractual mechanisms inside andbetween firms. Witness, for example,the increased use of non-traditionalforms of cooperation between firmsthrough different forms of alliancesand partnerships.As a consequence, senior executivestoday deal primarily with abstractissues relating to transactioncosts, while 50 or 100 years ago themanagement task was more concreteand aimed at production costreduction. Thus, the role of top managementin a large company haschanged beyond recognition. One ofthe most famous books by a chiefexecutive, Alfred P. Sloan, Jr.’s([1963] 1990) description of GeneralMotors under his stewardship, illustratesthe point. The book deals almostexclusively with production costissues in sales, manufacturing, development,and finance, and has aninsignificant amount of abstraction.For example, most of the excerptsfrom executive committee meetingminutes deal with practical issuessuch as forecasting and inventorybuild-up, production schedules,project development issues, andcash management. Other illustrationscan be found in old corporateannual reports. In Asea’s 1 annual reportof 1948 the opening statementconcerns factory utilization. The reportthen continues to discuss manufacturingand product developmentissues, while it totally ignores whatwe today call strategic and organizationalissues.Today’s executives must still manageproduction costs, but an evenlarger challenge lies in optimizingtransaction costs. As Herbert Simon(1976) anticipated: "In the postindustrialsociety, the central problemis not how to organize productionefficiently (although this will alwaysremain an important consideration),but how to organize to makedecisions—that is, to process infor-1 Today part of Asea Brown Boveri (ABB), theSwedish-Swiss electrical engineering conglomerate.

5mation.” The level of abstraction hasincreased commensurately. Todaywe talk about vision, strategic intent,learning organizations, and virtualcorporations. We find that mostcompanies’ value can not be calculatedby studying the income statementand balance sheet alone, sincemuch of the market value is embeddedin abstractions such as brandimage and intellectual capital.DEFINITIONSTransaction costs: the costs of allocatingresources in an imperfect world ofmisunderstandings, misaligned goals,and uncertainty. External transactioncosts center around the cost of contracting,internal transaction costs are dominatedby the cost of coordination. Transactioncosts are often described as“economic friction.”Asset specificity: The degree that aparticular asset, or set of assets, is dedicatedto a given use. In the context ofthis article, the degree that a consultingfirm's investment in know-how of itsprofessional staff is applicable only to asingle client, i.e., human asset specificity.Uncertainty: Lack of information aboutthe future. In particular, demand volatilityand technological uncertainty are importantto transaction cost theory. Similar torisk.Bounded rationality: The notion thathuman beings strive to be rational inmaking choices, but since the brain cannotprocess infinite amounts of informationthose choices are not always correct.In this world, it is necessary to begood at symbol manipulation (Reich1991): “Symbolic analysts solve,identify, and broker problems by manipulatingsymbols. They simplify realityinto abstract images that can berearranged, juggled, experimentedwith, communicated to other specialists,and then, eventually, transformedback into reality.” The symbolsare often qualitative rather thanquantitative. Examples are the fiveforces framework and the valuechain developed by Michael E. Porter,and the 7-S framework designedby McKinsey & Company. Reich estimatesthat in 1990 close to 20 percentof American jobs were held bysymbolic analysts, while no morethan 8 percent of workers could beclassified as symbolic analysts atmidcentury. Thus, as the transactioncost part of the economy has grown,so has the demand for symbol manipulation.Nature of demandThe transaction cost framework canalso be used to more specifically deducethe nature of this demand.First, bureaucratic (internal) transactioncosts stem principally from thecost of administration, the costs ofresource misallocation, and the negativeimpact of demotivation in largeorganizations. Management techniquesaimed at minimizing thesecan, for example, be found within thefields of organizational design, strategicplanning, and governance. Or-

6ganizational design influences thecost of administration and the levelof motivation significantly. An exampleis the superior performance ofmultidivisional organizations overfunctional organizations (Armour andTeece 1978). Strategic planning reducesresource misallocation bychanneling scarce resources intoareas where the company has acompetitive advantage. The choiceof governance models help improvemotivation through incentives, andreduces organizational slack such asexcessive bureaucracy. These areexactly the kinds of problems managementconsultants solve.Second, market transaction costsderive from the cost of price determination,the contract negotiationcosts, and the risk that there will belong-term deviations from the contractsince all aspects of the futurecan not be anticipated. To reducethese costs in dealing with customers,suppliers, and partners, executivesprimarily need information. As aconsequence, the demand for marketand competitive information andthe intelligent synthesis of this informationhas increased dramaticallyover the last 30 years. Services suchas these are offered by managementconsultants.In sum, the increase in demand formanagement consulting services isexplained by fundamental shifts inthe economy. Today’s complex businessenvironment requires hightransaction costs to function. This inturn leads to an increasing demandfor symbolic analysts—the kinds ofprofessionals found in modern managementconsulting firms. Stryker(1954) identified this trend years agowhen he observed that consultantsused to work on “specialized problems—inplant layout, for example,or in wage-incentive programs,” but“a relatively new kind of consultant—the man or firm that in effect offers toset a company’s basic objectives,policies, structure, and strategies”was emerging.Reasons for using externalmanagement consultantsWhy then is the demand for symbolmanipulation to a significant part satisfiedby external management consultants?After all, corporate executivescould do the symbol manipulationthemselves, or they could useinternal consultants. Instead they oftenuse external resources. As a result,since 1980 management consultinghas grown by 20 percent peryear. It has not always been thatway, however. Once upon a time,the executives did indeed do thework themselves. Chandler (1962)describes how executives at the duPont Company struggled between1917 and 1921 with how to organizethe company. They created workingparties and ad hoc committees, andat the same time worked individuallyon position papers and proposals.No consultants were involved. Similarly,when General Motors faced amajor crisis in 1920, it turned to oneof its senior executives, Alfred P.Sloan, Jr., to diagnose and solve the

7problem. Sloan’s write-up, the OrganizationStudy (1919), soon catapultedhim into the chairmanship ofGeneral Motors—without the help ofconsultants.Over time, though, the do-it-yourselfapproach to solving business problemshas decreased in importancebecause it is inefficient. A senior executivemost likely is not familiar withthe particular problem he or she isfacing and does not know whichproblem solving technique to apply.This is increasingly true as managementbecomes more complex,while executives remain boundedlyrational (Simon 1976) and do nothave the capacity to learn everything.Thus, the choice for the executiveoften is whether to turn to internal orexternal experts for advice. Accordingto transaction cost theory, thischoice hinges on the degree of assetspecificity, uncertainty due to demandvolatility and technological uncertainty,and the frequency oftransactions involved (as explainedin Part One). If these factors are low,then buying the services in the externalmarket will be the better solution(Rubin 1990): “When a competitivemarket exists, this usually offersthe most powerful method of controllingcosts. If a product is made internally,then the firm must spend substantialmanagerial resources monitoringcosts and efficiencies…Thefirst presumption should always befor purchasing inputs on the market.”What then, can be said about thedegree of asset specificity, uncertainty,and frequency of transactionsin management consulting services?The two latter factors have worked infavor of using outsiders, althoughtheir influence probably is weak. Uncertaintyhas decreased over the last50 years, as evidenced by the declinein volatility of the S&P 500 indexand of GDP growth. The frequencyof transactions is usually low,with most problems to be solved beingunique and singular.Asset specificity, which can be brokendown into physical asset specificity,human asset specificity, sitespecificity, and dedicated assets, isthe most important factor. Givingconsulting advice does not usuallyrequire an investment in physical assetsthat are specific to the client,and when it does (such as the purchaseof client-specific software), thecost is usually billed directly to theclient. Site specificity is low since theconsultant rarely moves permanentlyto the client’s location. Dedicated assetsthat cannot be redeployed areuncommon. The only aspect of assetspecificity that truly affects the decisionof using internal or external expertsis human asset specificity. Thatis, to what extent is the knowledge ofthe consultant specific to the client.High human asset specificity exists ifthe consultants need to invest significanttime and effort to understandthe client’s business, or conversely,if the client needs to invest in understandinghow the consultants work.In Turner’s (1982) eight task catego-

8ries described in Part One, there isan increasing degree of human assetspecificity the further down the listthe consultant works. Task 1: Providinginformation to a client usuallydoes not require a client-specific investment,while Task 8: Permanentlyimproving organizational effectivenessdemands that the consultantshave a thorough understanding ofthe idiosyncrasies of the client organization—anunderstanding thatoften takes at least a year to build.If human asset specificity is high,then there is significant risk that theclient or the outside consultant willopportunistically try to take advantageof the other party, a so-calledholdup situation. For example, theclient may try to reduce price or askfor free additional work since itknows that the consulting firm cannoteasily reassign people who have investedin building an understandingof the client organization. Similarly,the consultants know that it will taketime for the client to find, evaluate,and build the knowledge of a newconsultant. In the end, it may be easierfor the client to avoid the hold-upsituation by using internal resourcesrather than to go through a painfulnegotiation with outsiders.Thus, all other things equal, externalconsultants can be expected to workon issues that have low human assetspecificity, while internal experts dealwith issues close to the heart of theorganization. Indeed, this is the waysymbol manipulation was done up tillthe 1970s, with fast-growing internalconsulting staffs (such as those atGeneral Electric and Xerox (Kelley1979)) addressing core issues, andexternal consultants working primarilyon projects with low human assetspecificity.All other things are not equal though.External consultants have been ableto use three other transaction costrelatedfactors to their advantage,while they have tried to minimize thenegative impact of high human assetspecificity.First, the theory holds that opportunisticbehavior can be expected withinand between firms. This opportunismbecomes stronger as specializationto realize scale and scope economiesincreases, since specializationleads to goal conflicts between organizationalunits and individuals: Amanager in marketing may not necessarilyhave the same goal as amanager in manufacturing, eventhough the goal of the company is tomaximize shareholder returns. Thus,the risk of efficiency losses due tomisaligned goals has increased withthe growth of transaction costs. Tooffset this, executives more than everneed objective, detached, advice.Who then can best provide the objectivity?External management consultantshave the benefit of not beingmembers of the organization. Theyusually do not have vested interestsor oblique loyalties. (The counterargumentis that the consultant hasone unique sponsor to whom he orshe will yield if necessary. Research(Gattiker and Larwood 1985), however,suggests that this does not

9happen often enough to warrantconcern.)In addition to giving impartial adviceon key issues, consultants can alsoperform managerial audits. Traditionally,this was within the domain ofaccountants, but as the complexity oforganizations increased the ability ofaccountants to detect shirking decreased(Rubin 1990). Externalmanagement consultants have to alarge extent filled this void since theydeal with managerial issues ratherthan accounting issues. In transactioncost terms, the external managementconsultant is more likelythan an internal counterpart to lessenthe bureaucratic insularity of topmanagement, and to reduce internaltransaction costs due to misallocationof resources within and betweenfunctions.Second, for those activities that donot carry high human asset specificityvis-à-vis the client, the externalconsultants can build experiencemore effectively than inside consultants.Since they work in organizationsthat essentially are specializedby competence, they will have seensimilar problems before and the costfor leveraging this knowledge basewill be low. In contrast, the internalconsultants are experts in how theirown company works, but they seldomhave the size to create an experiencebase by type of problem.Also, the external consultant oftenhas the opportunity to engage in jointproblem solving with colleagues (Paroush1985). Such joint problemsolving is encouraged by the incentivestructure of the consulting firm.Replicating this type of incentive systemwithin the client organization isoften difficult since most client organizationsare joint stock companieswith very different reward systems.Third, the external consulting firmmost likely has higher productivitythan the internal counterpart. Themain reason is that incentives aremore easily tailored to the needs andperformance of individuals in smallerorganizations, while employees inlarger organizations suffer from bureaucraticallyinduced demotivation(and most consulting firms aresmaller than the their clients). A parallelis found in R&D where smallercompanies have 3 to10 times higherproductivity than larger companies(Cooper 1964; Zenger 1994).The three factors are advantagesheld by the external consultants overthe internal consultants. In addition,consulting firms often manage to offsetthe negative impact of high humanasset specificity through contractualmechanisms. In accordancewith the transaction cost framework,it is in the interest of the externalconsultant to minimize the cost ofprice determination, negotiation, andthe impact of long-term deviationsfrom the agreed upon contract. Pricedetermination is simplified sinceconsulting firms mostly follow thepractice of charging a fixed monthlyfee and the cost to the client is proportionalto the length of the project.Negotiations are possibly burdensome,but are alleviated by the man-

10agement consultant’s propensity touse short and standardized proposals.The risk of deviations from theintended task is usually small sincemost efforts are relatively brief andthere is constant feedback betweenclient and consultant. Projects seldomtake more than one year, andthe norm is three to nine months.Consultants further reduce this riskby providing easy exits for the client,such as agreements that the workcan be terminated without advancenotice and without a stated reason.What is sometimes viewed as lessthan rigorous contracting policy is infact a sophisticated way for the consultantsto lower the threshold for theclient to retain their services.The above logic can be summarizedin the following graph:Part One posed the question why wehave seen an explosion in the demandfor management consulting inthe United States, but not in Japan.The answer is complicated. Part ofthe answer lies in Japanese andAmericans being at different stagesin the management skill developmentcycle. More importantly, theJapanese management traditionplaces so much reliance on longtermpredictability of careers and acommensurate need to carry organizationalknowledge within organizations,that it is difficult for outsiders tobe accepted by large corporations.External consultants’ disruptive effectson clients’ managementprocesses, so far, have outweighedthe benefits of stimulation, expertise,and objectivity.MANAGEMENT CONSULTANTS’ DOMAINTask 8: Permanentlyimproving organizationaleffectivenessTurner’s eighttask categories *Task 1: Providinginformation toa client* As described in Part OneNot efficient to useexternalconsultantsImprovedcontractingEfficient to useexternalconsultantsLowHighClient’s internal (bureaucratic)transaction costIncreasinghumanassetspecificity* * * *

11Management consulting firms existfor good reasons. The nature ofmanagement has changed: Unlike inearlier times, abstract issues embodiedin the transaction cost part ofthe economy demand management’sattention. Consequently, there is amarket for symbol manipulation—amarket which hardly existed 50 or100 years ago. External managementconsultants are well suited tofill this demand. They bring objectivity,experience, and have high productivity.The cost to the client ofworking with outside experts is lowerthan the cost of using internal resourceswhen both direct and indirectcosts are factored in. As we willsee in the next section, this is likelyto hold true in the future as well.HOW WILL THEMANAGEMENT CONSULTINGINDUSTRY EVOLVE?More than forty years ago, managementconsulting was considered“one of the hottest—and most influential—growthindustries” (Stryker1954). Today, management consultingarguably is one of the world’smost rapidly growing industries.Many expect that consultants willcontinue to increase market share inproblem solving on behalf of corporationsand other organizations –and thus continued industry growth.On the other hand, it may be thatclients eventually will reclaim theservices provided by managementconsultants—especially those serviceswith high human asset specificity.This would be akin to the disappearanceof the inside contractingsystem discussed in Part One of thisseries. Under this scenario, the consultingindustry could stagnate oreven decline.Continued growth scenarioRemember that the key obstacle tousing external resources such asmanagement consultants, accordingto transaction cost theory, is the degreeof human asset specificity involved,and that high uncertaintymakes it difficult to use outside contractors.For the growth scenario tomaterialize the following conditionswill have to be true.First, the current trend towards managementconsultants’ deeper involvementin more and more aspectsof solving core problems of theirclients will have to moderate; otherwise,asset specificity will increaseso much that external sourcing ofconsulting services becomes unfeasible.Alternatively, contractual arrangementsbetween client and consultantneed to be refined at a pacethat exceeds the increase in assetspecificity (sophisticated contractscan mitigate the negative effect ofasset specificity; witness for examplethe increasing use of success feeswhich tend to align the objectives ofclients and consultants).Second, client organizations willhave to avoid making significant

12strides in reducing internal bureaucracycosts. If, however, clients canreduce the costs of administration,resource misallocation, and demotivation,then transaction cost theorytells us that it will be relatively moreattractive to do symbol manipulationinternally. Indeed, highly bureaucraticorganizations tend to use moreexternal management consultants dothan lean organizations. (A continuedhigh level of internal bureaucracycosts will stimulate demand forexternal management consultants.)The third condition what would haveto prevail is that uncertainty (in termsof demand volatility or technologicaluncertainty) will not increase significantly,given that high uncertaintyreduces the benefit of buying productsor services from the outside.Were the foregoing growth scenarioto develop more or less as outlined,we could, within 15 to 30 years, seea radically different corporate world.Initially, we would see continued rapidexpansion of the managementconsulting industry. Soon therewould be as many external symbolmanipulators as there are executivesin large companies. Over time, thebalance of power would shift to themanagement consultants. Theywould possess the most knowledgeabout management practice in general,and their clients’ problems specifically.They would own the knowledgenetworks which will be essentialin the global economy. The managementconsulting firms would alsodeplete the stock of young, intelligent,and well educated people formingthe backbone of the future economy.We thus would see a shift inthe balance of influence from thetraditional product and services sectorsto the symbolic analyst sector,just as in the 1800s we saw a shift ofinfluence from the agriculture sectorto the industrial sector.Ultimately, management consultingfirms would move from being advisers,to taking over the managementfunction of their clients. We wouldsee a new corporate configuration inwhich the consultants work as thesymbol manipulators of corporations,and the old corporate structures aredismantled to provide the buildingblocks for those manipulative activities.Consultants would manage highvalue added networks of product designand delivery activities, wherebythey would provide strategic and integrativecapabilities. The old corporationswould provide low valueaddedproducts, subassemblies, andservices to the specification of thenetwork operators —the managementconsultants.Decline scenarioUnder the second scenario, managementconsulting would bedoomed, just as inside contractingonce flourished and then declined(see Part One). How would this“doomsday” scenario come to be?First, the asset specificity of managementconsulting advice wouldneed to be so high that clients find it

13difficult to handle the interface betweenthemselves and consultantsand, consequently, decide to internalizesymbol manipulation.Second, large corporations wouldhave to develop their managementpractices to accommodate the needsof different types of employees, bothsymbolic analysts and routine workers.In particular, this would requiredifferentiated approaches to performanceevaluation and the setting ofincentives (a process that has alreadystarted as evidenced by theescalating compensation packageslavished on executives).In a third factor leading to a scenarioof decline, uncertainty would have toincrease to a significantly higher levelthan it is today.Fourth, the types of problems handledby management consultantswould have to become more prevalent.(Remember, as an activity becomesmore frequent there is a tendencyto internalize it.)Should all these things happen, wemay live to see a second version ofthe demise of inside contracting.Clients would initially hire away toptalent from consulting firms to do thesame jobs as before, and with thesame compensation, but now asemployees. The alignment of highasset specificity with internal sourcingwould over time prove more costeffective than buying consulting servicesfrom the outside. Knowledgeaccumulation then would shift towardthe clients, and management consultingfirms would find it increasinglydifficult to provide high value addedadvice. However, since managementconsultants also would be providingan auditing function, and assumingthey provided objective advice, theywould not disappear entirely. Thenature of their work, however, mightwell shift from Schein’s expert anddoctor–patient models to the processconsultation model, one in which theconsultant facilitates and the clientprovides the expertise.Under such a decline scenario, externalmanagement consultantswould work primarily on routine assignments.Yes, they would continueto leverage industry knowledge fromclient to client, much as McKinsey &Company and others do today. Butby its very definition, this knowledgeis most unlikely to add unique valueto the individual client. Furthermore,opportunities to work on core issuessuch as strategy and governancewould be highly limited. In the end,the consulting process would becomesubstantially streamlined andhighly efficient; on the other hand,the industry no longer would be ableto attract the best people. Managementconsulting will cease being“one of the hottest—and most influential—growthindustries.” 22 Stryker (1954)

14The future role of managementconsultantsIn reality, neither of these two scenariosseem very likely to fully evolve.Nevertheless, looking to the next tenor fifteen years, several factors pointto the “continued growth scenario” asthe more likely outcome.So far, the management consultingindustry has been able to largelysurmount the hurdle of asset specificityand, thereby, redefine an appropriatedivision of labor betweenclients and consultants. New formsof collaboration have made it easierfor clients to outsource problem solvingof core issues. An example is thetendency of consulting firms to strivefor long-term relationships withclients as opposed to working on oneproject per client. Another example isthat consultants have been backingaway from the classical model of“consultants analyze and recommend,clients decide and implement.”Collaboration today is muchmore sophisticated than it was amere fifteen years ago, with clientsand consultants now working togetherthroughout the entire changeprocess. This trend can be expectedto continue.Of at least equal significance—withor without reengineering and thelike—there is no indication that internal(bureaucratic) transaction costswithin large corporations will decline.To the contrary, as noted earlier, thetransaction cost part of the economyhas grown steadily since the 1870s.Nor is this trend likely to be disruptedanytime in the foreseeable future.For one thing, the increasingly globaleconomy adds to complexity. Withinlarge corporations, the demand forcoordination continues unabated.New technologies such as artificialintelligence appear unlikely within inthe foreseeable future to change thispicture.Finally, there is scant evidence thatlarge corporations will be able to realigntheir management processessufficiently in order to be able to internalizesymbol manipulation. Stinchcombe(1965) found that the way acompany manages itself to a largedegree is determined by when it wasfounded. Most large companies arefairly old and will continue to be so,even though the information technologyrevolution gradually will changethis picture. In the meantime though,it is unlikely that corporate giants willchange their modus operandi fundamentally.If the above arguments hold true, themanagement consulting industry willcontinue to prosper. Consultants, togetherwith other external advisers,will play an increasingly importantrole in the global economy and mayultimately take on the role of networkmanagers. Relationships betweenclients and consultants will growstronger and symbiotic. Managementconsulting will continue to be a preferredcareer choice for many graduatingstudents at the premier businessschools and universities.This article is a reprint from the Journal ofManagement Consulting, 1999: Volume10, Issue 3, pp. 3–12.

15LIST OF REFERENCESArmour, H. O., and D. J. Teece. 1978. Organizational structure and economicperformance: A test of the multidivisional hypothesis. Bell Journal ofEconomics 9 (1): 106–122.Canback, S. 1998. The logic of management consulting, part 1. Journal ofManagement Consulting 10 (2): 3–11.Chandler, A. D., Jr. 1962. Strategy and structure: Chapters in the history of theAmerican industrial enterprise. Cambridge: MIT Press.Cooper, A. C. 1964. R&D is more efficient in small companies. Harvard BusinessReview 42 (May-June): 75–83.Drucker, P. F. 1979. Why management consultants? In The evolving science ofmanagement, edited by Z. Melvin and R. G. Greenwood. New York:AMACOM.Gattiker, U. E., and L. Larwood. 1985. Why do clients employ managementconsultants? Consultation 4 (2): 119–129.Greiner, L., and R. Metzger. 1983. Consulting to management. Englewood Cliffs,N.J.: Prentice-Hall.Kelley, R. E. 1979. Should you have an internal consultant? Harvard BusinessReview 57 (November-December): 110–120.O'Shea, J., and C. Madigan. 1997. Dangerous company: The consulting powerhousesand the businesses they save and ruin. New York: RandomHouse.Paroush, J. 1985. Notes on partnerships in the service sector. Journal ofEconomic Behavior and Organization 6 (1): 79–87.Reich, R. B. 1991. The work of nations: Preparing ourselves for 21st-century capitalism.New York: Alfred A. Knopf.Rubin, P. H. 1990. Managing business transactions. New York: Free Press.Simon, H. A. 1976. Administrative behavior. 3rd ed. New York: Free Press.Sloan, A. J., Jr. 1919. General Motors Corporation: Organization study. Detroit:General Motors.

16Sloan, A. P., Jr. [1963] 1990. My years with General Motors. Edited byJ. McDonald and C. Stevens. Reprint, with a new introduction by Peter F.Drucker. New York: Currency. Original edition, New York: Doubleday.Stinchcombe, A. L. 1965. Social structure and organizations. In Handbook oforganizations, edited by J. G. March. Chicago: Rand McNally.Stryker, P. 1954. The ambitious consultants. Fortune, May, 82–85.Turner, A. N. 1982. Consulting is more than giving advice. Harvard BusinessReview 60 (September-October): 120–129.Wallis, J. J., and D. C. North. 1986. Measuring the transaction sector in theAmerican economy, 1870–1970. In Long-term factors in Americaneconomic growth. Studies in income and wealth series, edited by S. L.Engerman and R. E. Gallman. Chicago: University of Chicago Press.Zenger, T. R. 1994. Explaining organizational diseconomies of scale in R&D:Agency problems and the allocation of engineering talent, ideas, and effortby firm size. Management Science 40 (6): 708–729.