Management in the 1980'sby

Harold J. Leavitt and Thomas L. Whisler

CARNEGIE INSTIT UTE OF TECHNOLOGY

Pittsburgh 13, Pennsylvania

**) GRADUATE SCHOOL OF INDUSTRIAL ADMINISTRATION

p7. "7/ 77I

i!

Management in the 1980's1

By HAROLD J. LEAVITT and THOMAS L. WHISLERi

REPRINTED FROM

HARVARD BUSINESS REVIEW«

NOVEMBER-DECEMBER 19584

i

{

( Reprints of Harvard Business Review articles are available( at the following prices:

( Single reprint $1.00 Four to 99 $ 0.40 each( Two reprints 1.30 First 100 39-7°( Three reprints 1.50 Additional 10o's 10.00 each

( Harvard Business Review( SOLDIERS FIELD, BOSTON 63, MASSACHUSETTS

\

t

Edited under the direction of theFaculty ofThe Graduate School of Business

Administration,Harvard University

BUSINESS REVIEW BOARDStanley F.

Teele, Dean,

Ex OfficioNeil H.

Bobdbn,

ChairmanVernon R. Alden Stephen H. FullerRobert W. Austin Lincoln GordonJames R. Bright Leonard C. R. LangerEdward C Bursk Ralph W. HidyJ. Keith Butters Thomas J. Raymond

E. Raymond Corey Robert O. SchlaiferBertrand Fox Ross G. Walker

EDITOREdward C. Bursk

Associate Editor: John F. ChapmanAssistant Editor: David W. Ewing

Assistant Editor: Dan H. Fenn, Jr.Assistant Editor: Rollie Tillman, Jr.Managing Editor: Virginia B. FalesAdvertising Director: John Nimmo

Industrial Programs: Jacob F. WeintzBusiness Manager: T. K. Worthington

Circulation Director: Ernest D. FrawleyCirculation Manager: Agnes M. Jordan

Address all communications, including manu-scripts and change of address, to HarvardBusiness Review, Soldiers Field, Boston 63,Mass. Telephone: Klrkland 7-9800.

Reprints of all articles and features in thisissue, as well as most previous issues, areavailable. Information on request from Busi-ness Office at above address.

A Volume Index to the Review is publishedannually with the November-December issue.

The contents are currently indexed inTheIndustrial Arts Index and in the Bulletin ofthe Public Affairs Information Service.

Harvard Business Review is publishedbimonthlyby the Graduate School of Business Administra-tion, Harvard University. Nathan M. Pusey,President; Paul C.

Cabot,

Treasurer; David W.Bailey, Secretary.

$8.00

per year,

$2.00

per copy — UnitedStates & Possessions, Canada.

$10.00 per year,

$2.50

per copy — Pan-American Postal Union.

$15.00

per year,

$3-75

Per copy — elsewhere.Make checks payable to the Harvard Business

Review.Second-class postage paid at Boston, Massa-

chusetts and at additional mailing offices.Copyright, 1958, by the President and Fellows

of Harvard College.

Printed in U.S.A.

November -December 1958

Management in the 1980'sHarold J. Leavitt and Thomas L. Whisler

Sharper Focus for the Corporate ImagePierre Martineau

How to Evaluate New Capital InvestmentsJohn G. McLean

Let's Export Marketing Know-HowWoodruff J. Emlen

New Threat in State Business TaxationPaul Studenski and Gerald J. Glasser

Common Stocks and Pension Fund InvestingPaul L. Howell

The Organization : What Makes It Healthy?

Tougher Program for Management TrainingF. Gordon Barry and C. G. Coleman, Jr.

Fashion Theory and Product Design

In This Issue: Articles and AuthorsThe Editors

From the Thoughtful Businessman: LettersReview Readers

Thinking Ahead: Why Is Education Obsolete?

Looking Around: What Makes Research Sterile?

HARVARD

BusinessReview

Vol. 36, No. 6

ARTICLES41

49

59

70

77

92

107Chris Argyris

117

126Dwight E. Robinson

FEATURES

5

15

23Margaret Mead

149Gerald Morrell

The Harvard Business Review does not assumeresponsibility forthe points of view or opinions of its contributors. It does acceptresponsibility forgivingthem an opportunityto express such viewsand opinionsin its columns.

i

1

From In This Issue

Harold J. Leavitt andThomas L. Whisler

The alert manager likes to planahead. While sharing the univer-sal curiosity of man regarding thefuture, the action-oriented execu-tive also wants to know, "Whatdoes all this mean in terms ofpractical application?" Mindfulof swift advances in technology,management can ignore only atperil to itself the implications oftheresulting pressures on its func-tion. Of particular importance iswhat Harold J. Leavitt and Thom-as L. Whisler call the new "in-formation technology."

No Orwellian fantasy on lifein the 1980*5, Management in the19 80s is a reasoned prediction ofthe adjustments business organi-zation must make in the comingera. When asked how they cametobe interested in this subject, theauthors told the Editors:

"We had heard so much recent-ly of the bankruptcy of personnelmanagementas a staff function thatwe started to explore further. Fromthere we passed quickly to the moreexciting problem of the changingnature of management in general.

"Our interest in the new tech-nology, however, is not purely aby-product of this other interest.Leavitt's research interests have longcentered around experimental stud-ies of communication and organiza-tional design, and Whisler has beeninterested in the larger implicationof job specialization and also intherole of the 'odd' characters in theorganization."

Harold J. Leavitt is Professorof Industrial Administration andPsychology in the Graduate Schoolof Industrial Administration atCarnegie Institute of Technology.He has served as Consultant tothe European Productivity Agen-cy, and was Vice President ofNejelski & Company. His latestbook is entitled Managerial Psy-chology (Chicago, University ofChicago Press, 1958).

Thomas L. Whisler has been acontributor to several professionaljournals and to the EncyclopaediaBritannica. He is active in indus-trial consulting and is AssociateProfessor of Industrial Relations,School of Business, University ofChicago.

— The Editors

I

i

iI

41

Harvard Business ReviewNovember-December 1958

New information flows cutnew organization channels.

MANAGEMENTin the 1980's

By Harold J. Leavitt andThomas L. Whisler

Over the last decade a new technology hasbegun to take hold in American business, oneso new that its significance is still difficult toevaluate. While many aspects of this technol-ogy are uncertain, it seems clear that it willmove into the managerial scene rapidly, withdefinite and far-reaching impact on managerialorganization. In this article we would like tospeculate about these effects, especially as theyapply to medium-size and large business firmsof the future.

The new technology does not yet have a single^established name. We shall call it informationtechnology. It is composed of several relatedparts. One includes techniques for processinglarge amounts of information rapidly, and it isepitomized by the high-speed computer. A sec-ond part centers around the application of sta-tistical and mathematical methods to decision-making problems; it is represented by techniqueslike mathematical programing, and by method-ologies like operations research. A third part is

in the offing, though its applications have not yetemerged very clearly; it consists of the simula-tion of higher-order thinking through computerprograms.

Information technology is likely to have itsgreatest impact on middle and top management.In many instances it will lead to opposite con-clusions from those dictated by the currentlypopular philosophy of "participative" manage-ment. Broadly, our prognostications are alongthe following lines:

(1) Information technology should move theboundary between planning and performance up-ward. Just as planning was taken from the hourlyworker and given to the industrial engineer, wenow expect it to be taken from a number of mid-dle managers and given to as yet largely nonexist-ent specialists: "operations researchers," perhaps,or "organizational analysts." Jobs at today's middle-management level will become highly structured.Much more of the work will be programed, i.e.,covered by sets of operating rules governing theday-to-day decisions that are made.

(2) Correlatively, we predict that large indus-trial organizations will recentralize, that top man-agers will take on an even larger proportion of the

I

Harvard Business Review42

1

II

i

1

innovating, planning, and other "creative" func-tions than they have now.

(3) A radical reorganization of middle-manage-ment levels should occur, with certain classes ofmiddle-management jobsmoving downward in sta-tus and compensation (because they will requireless autonomy and skill), while other classes moveupward into the top-management group.

(4) We suggest, too, that the line separating thetop from the middle of the organization will bedrawn more clearly and impenetrably than ever,much like the line drawn in the last few decadesbetween hourly workers and first-line supervisors.

The New TechnologyInformation technology has diverse roots —

with contributions from such disparate groupsas sociologists and electrical engineers. Workingindependently, people from many disciplineshave been worrying about problems that haveturned out to be closely related and cross-fertil-izing. Cases in point are the engineers' develop-ment of servomechanisms and the related de-velopments of general cybernetics and informa-tion theory. These ideas from the "hard" sci-ences all had a direct bearing on problems ofprocessing information — in particular, the de-velopment of techniques for conceptualizing andmeasuring information.

Related ideas have also emerged from otherdisciplines. The mathematical economist camealong with game theory, a means of orderingand permitting analysis of strategies and tacticsin purely competitive "think-" type games. Op-erations research fits in here, too; OR peoplemade use of evolving mathematical concepts,or devised their own, for solving multivariateproblems without necessarily worrying aboutthe particular context of the variables. And fromsocial psychology ideas about communicationstructures in groups began to emerge, followedby ideas about thinking and general problem-solving processes.

All of these developments, and many othersfrom even more diverse sources, have in com-mon a concern about the systematic manipula-tion of information in individuals, groups, ormachines. The relationships among the ideasare not yet clear, nor has the wheat been ade-quately separated from the chaff. It is hard totell who started what, what preceded what, andwhich is method and which theory. But, char-

1 Two examples of current developments are discussedin "Putting Arma Back on Its Feet," Business Week, Feb-

acteristically, application has not, and probablywill not in the future, wait on completion ofbasic research.

Distinctive FeaturesWe call information technology "new" be-

cause one did not see much use of it until WorldWar 11, and it did not become clearly visible inindustry until a decade later. It is new, also,in that it can be differentiated from at least twoearlier industrial technologies:

(1) In the first two decades of this century,Frederick W. Taylor's scientific management con-stituted a new and influential technology — onethat took a large part in shaping the design ofindustrial organizations.

(2) Largely after World War II a second dis-tinct technology, participative management, seri-ously overtook — and even partially displaced —scientific management. Notions about decentrali-zation, morale, and human relations modified andsometimes reversed earlier applications of scien-tific management. Individual incentives, for ex-ample, were treated first as simple applications ofTaylorism, but they have more recendy been re-vised in the light of "participative" ideas.

The scientific and participative varieties bothsurvived. One reason is that scientific manage-ment concentrated on the hourly worker, whileparticipative management has generally aimedone level higher, at middle managers, so theyhave not conflicted. But what will happen now?The new information technology has direct im-plications for middle management as well as topmanagement.

Current PictureThe inroads made by this technology are al-

ready apparent, so that our predictions are moreextrapolations than derivations.1 But the signifi-cance of the new trends has been obscured bythe wave of interest in participative manage-ment and decentralization. Information tech-nology seems now to show itself mostly in theperiphery of management. Its applications ap-pear to be independent of central organizationalissues like communication and creativity. Wehave tended until now to use little pieces of thenew technology to generate information, or tolay down limits for subtasks that can then beused within the old structural framework.

Some of this sparing use of information tech-ruary

i,

1958,p. 84; and "Two-Way Overhaul RebuildsRaytheon," Business Week, February

22,

1958, p. 91.

Management in the 1980's 43

i

1

I

i

i

nology may be due to the fact that those of uswith a large commitment to participative man-agement have cause to resist the central impli-cations of the new techniques. But the implica-tions are becoming harder to deny. Many busi-ness decisions once made judgmentally now canbe made better by following some simple rou-tines devised by a staff man whose company ex-perience is slight, whose position on the organi-zation chart is still unclear, and whose skill (ifany) in human relations was picked up on theplayground. For example:

We have heard recently of an electric utilitywhich is considering a move to take away fromgenerating-station managers virtually all responsi-bility for deciding when to use stand-by generat-ing capacity. A typical decision facing such man-agers develops on hot summer afternoons. In an-ticipation of heavy home air-conditioning demandat the close of working hours, the manager mayput on extra capacity in late afternoon. This re-sults in additional costs, such as overtime premi-ums. In this particular geographical area, rapidlymoving cold fronts are frequent. Should such afront arrive after the commitment to added capac-ity is made, losses are substantial. If the frontfails to arrive and capacity has not been added,power must be purchased from an adjacent systemat penalty rates — again resulting in losses.

Such decisions may soon be made centrally byindividuals whose technical skills are in mathe-matics and computer programing, with absolutelyno experience in generating stations.

Rapid SpreadWe believe that information technology will

spread rapidly. One important reason for ex-pecting fast changes in current practices is thatinformation technology will make centralizationmuch easier. By permitting more information tobe organized more simply and processed morerapidly it will, in effect, extend the thinkingrange of individuals. It will allow the top levelof management intelligently to categorize, di-gest, and act on a wider range of problems.Moreover, by quantifying more information itwill extend top management's control over thedecision processes of subordinates.

If centralization becomes easier to implement,managers will probably revert to it. Decentrali-zation has, after all, been largely negativelymotivated. Top managers have backed into itbecause they have been unable to keep up withsize and technology. They could not design andmaintain the huge and complex communication

systems that their large, centralized organizationsneeded. Information technology should makerecentralization possible. It may also obviateother major reasons for decentralization. Forexample, speed and flexibility will be possibledespite large size, and top executives will be lessdependent on subordinates because there willbe fewer "experience" and "judgment" areasin which the junior.men have more workingknowledge. In addition, more efficient informa-tion-processing techniques can be expected toshorten radically the feedback loop that tests theaccuracy of original observations and decisions.

Some of the psychological reasons for decen-tralization may remain as compelling as ever.For instance, decentralized organizations prob-ably provide a good training ground for the;top manager. They make better use of the whole!man; they encourage more active cooperation.But though interest in these advantages shouldbe very great indeed, it will be counterbalancedby interest in the possibilities of effective top-management control over the work done by themiddle echelons. Here an analogy to Taylorismseems appropriate:

In perspective, and discounting the counter-trends instigated by participative management, theupshot of Taylorism seems to have been the sepa-rating of the hourly worker from the rest of theorganization, and the acceptance by both manage-ment and the worker of the idea that the workerneed not plan and create. Whether it is psycho-logically or socially justifiable or not, his creativityand ingenuity are left largely to be acted out offthe job in his home or his community. One rea-son, then, that we expect top acceptance of in-formation technology is its implicit promise toallow the top to control the middle just as Taylor-ism allowed the middle to control the bottom.

There are other reasons for expecting fastchanges. Information technology promises toallow fewer people to do more work. The moreit can reduce the number of middle managers,the more top managers will be willing to try it.

We have not yet mentioned what may wellbe the most compelling reason of all: the pres-sure on management to cope with increasinglycomplicated engineering, logistics, and market-ing problems. The temporal distance betweenthe discovery of new knowledge and its practi-cal application has been shrinking rapidly, per-haps at a geometric rate. The pressure to re-organize in order to deal with the complicating,speeding world should become very great in the

r*\

44 Harvard Business Review

I

I

next decade. Improvisations and "adjustments"within present organizational frameworks arelikely to prove quite inadequate; radical rethink-ing of organizational ideas is to be expected.

Revolutionary EffectsSpeculating a little more, one can imagine

some radical effects of an accelerating develop-ment of information technology — effects war-ranting the adjective "revolutionary."

Within the organization, for example, manymiddle-management jobs may change in a man-ner reminiscent of (but faster than) the transi-tion from shoemaker to stitcher, from old-timecraftsman to today's hourly worker. As we havedrawn an organizational class line between thehourly worker and the foreman, we may expecta new line to be drawn heavily, though jag-gedly, between "top management" and "middlemanagement," with some vice presidents andmany ambitious suburban juniorexecutives fall-ing on the lower side.

In one respect, the picture we might paint forthe 1980's bears a strong resemblance to theorganizations of certain other societies — e.g., tothe family-dominated organizations of Italy andother parts of Europe, and even to a small num-ber of such firms in our own country. Therewill be many fewer middle managers, and mostof those who remain are likely to be routinetechnicians rather than thinkers. This similaritywill be superficial, of course, for the changes weforecast here will be generated from quite dif-ferent origins.

What organizational and social problems arelikely to come up as by-products of suchchanges? ,One can imagine major psychological problemsarising from the depersonalization of relation-ships within management and the greater dis-tance between people at different levels. Majorresistances should be expected in the processof converting relatively autonomous and unpro-gramed middle-management jobs to highly rou-tinized programs.

These problems may be of the same order as jsome of those that were influential in the de-velopment of American unions and in focusingmiddle management's interest on techniques forovercoming the hourly workers' resistance tochange. This time it will be the top executivewho is directly concerned, and the problems ofresistance to change will occur among thosemiddle managers who are programed out oftheir autonomy, perhaps out of their current sta-

tus in the company, and possibly even out oftheir jobs.

On a broader social scale one can conceiveof large problems outside the firm, that affectmany institutions ancillary to industry. Thus:

" What about education for management? Howdo we educate people for routinized middle-management jobs, especially if the path fromthose jobs up to top management gets muchrockier?

" To what extent do business schools stop train-ing specialists and start training generaliststo move direcdy into top management?

" To what extent do schools start training newkinds of specialists?

" What happens to the traditional apprenticesystem of training within managerial ranks?

" What will happen to American class struc-ture? Do we end up with a new kind ofmanagerial elite? Will technical knowledge bethe major criterion for membership?

" Will technical knowledge become obsolete sofast that managers themselves will become ob-solete within the time span of their industrialcareers?

Middle-Management ChangesSome jobs in industrial organizations are more

programed than others. The job that has beensubjected to micromotion analysis, for instance,has been highly programed; rules about what isto be done, in what order, and by what proc-esses, are all specified.

Characteristically, the jobs of today's hourlyworkers tend to be highly programed — an ef-fect of Taylorism. Conversely, the jobs shownat the tops of organization charts are oftenlargely unprogramed. They are "think" jobs —hard to define and describe operationally. Jobsthat appear in thebig middle area of the organi-zation chart tend to be programed in part, withsome specific rules to be followed, but with vary-ing amounts of room for judgment and auton-omy.2 One major effect of information_echnol-ogy is likely to be intensive programing of manyjobsnow held by middle managers and the con-comitant "deprograming" of others.

As organizations have proliferated in size andspecialization, the problem of control and inte-gration of supervisory and staff levels has be-come increasingly worrisome. The best answer

* See Robert N. McMurry, "The Case for BenevolentAutocracy," HBR January-February 1958, p. 82.

Management in the igBo's 45

I

1I

,

until now has been participative management.But information technology promises better an-swers. It promises to eliminate the risk of lessthan adequate decisions arising from garbledcommunications, from misconceptions of goals,and from unsatisfactory measurement of partialcontributions on the part of dozens of line andstaff specialists.

Good illustrations of this programing processare not common in middle management, butthey do exist, mostly on the production sideof the business. For example, the programmershave had some successes in displacing the judg-ment and experience of production schedulers(although the scheduler is still likely to be thereto act out the routines) and in displacing theweekly scheduling meetings of production, sales,and supply people. Programs are also beingworked out in increasing numbers to yield de-cisions about product mixes, warehousing, capi-tal budgeting, and so forth.3

Predicting the ImpactWe have noted that not all middle-manage-

ment jobswill be affected alikeby the new tech-nology. What kinds of jobs will become moreroutinized, and what kinds less? What factorswill make the difference?

The impact of change is likely to be deter-mined by three criteria:

1. Ease of measurement —It is easier, at thisstage, to apply the new techniques to jobs in andaround production than in, say, labor relations,one reason being that quantitative measurement iseasier in the former realms.

2. Economic pressure — Jobs that call for bigmoney decisions will tend to get earlier invest-ments in exploratory programing than others.

3. Theacceptability of programing by the pres-ent jobholder — For some classes of jobs and ofpeople, the advent of impersonal rules may offerprotection or relief from frustration. We recentlyheard, for example, of efforts to program a main-tenance foreman's decisions by providing rules forallocating priorities in maintenance and emergencyrepairs. The foreman supported this fully. Hewas a harried and much blamed man, and pro-graming promised relief.

Such factors should accelerate the use of pro-graming in certain areas. So should the great

* See the journals, Operations Research and Manage-ment Science.

1 Much of the work in this area is still unpublished.However, for some examples, see Herbert A. Shepard,"Superiors and Subordinates in Research," Journal of

interest and activity in the new techniques nowapparent in academic and research settings.New journals are appearing, and new societiesare springing up, like the Operations ResearchSociety of America (established in 1946), andthe Institute of Management Sciences (estab-lished in 1954), both of which publish journals.

The number of mathematicians and economicanalysts who are being taken into industry isimpressive, as is the development within indus-try, often on the personal staffs of top manage-ment, of individuals or groups with new labelslike "operations researchers," "organization an-alysts," or simply "special assistants for plan-ning." These new people are a cue to the emer-gence of information technology. Just as pro-graming the operations of hourly workers cre-ated the industrial engineer, so should informa-tion technology, as planning is withdrawn frommiddle levels, create new planners with newnames at the top level.

So much for work becoming more routinized.At least two classes of middle jobs should moveupward toward deprogramedness: 7

(1) The programmers themselves, the new in-formation engineers, should move up. They shouldappear increasingly in staff roles close to the top.

(2) We would also expect jobs in research anddevelopment to go in that direction, for innovationand creativity will become increasingly importantto top management as the rate of obsolescence ofthings and of information increases. Applicationof new techniques to scanning and analyzing thebusiness environment is bound to increase therangeand number of possibilities for profitable produc-tion. Competition between firms should centermore and more around their capacities to innovate.

Thus, in effect, we think that the horizontalslice of the current organization chart that wecall middle managementwill break in two, withthe larger portion shrinking and sinking into amore highly programed state and the smallerportion proliferating and rising to a level wheremore creative thinking is needed. There seem,to be signs that such a split is already occurring.The growth of literature on the organization ofresearch activities in industry is one indication.4

Many social scientists and industrial researchmanagers, as well as some general managers,Business of the University of Chicago, October 1956,p. 261; and also Donald C. Pelz, "Some Social Fac-tors Related to Performance in a Research Organiza-tion," Administrative Science Quarterly, December 1956,p. 310.

46 Harvard Business Review

1

are worrying more and more about problems ofcreativity and authority in industrial researchorganizations. Even some highly conservativecompany presidents have been forced to breaktime-honored policies (such as the one relatingsalary and status to organizational rank) in deal-ing with their researchers.

Individual ProblemsAs the programing idea grows, some old hu-

man relations problems may be redefined. Re-definition will not necessarily solve the prob-lems, but it may obviate some and give newpriorities to others.

Thus, the issue of morale versus productivitythat now worries us may pale as programingmoves in. The morale of programed personnelmay be of less central concern because less (orat least a different sort of) productivity will bedemanded of them. The execution of control-lable routine acts does not require great enthusi-asm by the actors.

Another current issue may also take a newform: the debate about the social advantages ordisadvantages of "conformity." The stereotypeof the conforming junior executive, more inter-ested in beingwell liked than in working, shouldbecome far less significant in a highly deper-sonalized, highly programed, and more machine-like middle-management world. Of course, thepressures to conform will in one sense becomemore intense, for the individual will be requiredto stay within the limits of the routines that areset for him. But the constant behavioral pres-sure to be a "good guy," to get along, will haveless reason for existence.

As for individualism, our .suspicion. is... thatthe average middle manager, willJhaveiQ satisfyhis personal needs and aspirations off the job,largely as we have forced the hourly worker todo. In this case, the Park Forest of the futuremay be an even more interesting phenomenonthan it is now.

Changes at the TopIf the new technology tends to split middle

management — thin it, simplify it, program it,and separate a large part of it more rigorouslyfrom the top — what compensatory changesmight one expect within the top group?

This is a much harder question to answer.

' See Herbert A. Simon and Allen Newell, "HeuristicProblem Solving: The Next Advance in Operations Re-

We can guess that the top will focus even moreintensively on "horizon" problems, on problemsof innovation and change. We can forecast, too,that in dealing with such problems the top willcontinue for a while to fly by the seat of itspants, that it will remain largely unprogramed.

But even this is quite uncertain. Currentresearch on the machine simulation of highermental processes suggests that we will be ableto program much of the top jobbefore too manydecades have passed. There is good authorityfor the prediction that within ten years a digitalcomputer will be the world's chess champion,and that another will discover and prove an im-portant new mathematical theorem; and that inthe somewhat more distant future "the wayis open to deal scientifically with ill-structuredproblems — to make the computer coextensivewith the human mind." 5

Meanwhile, we expect top management to be-come more abstract, more search-and-research-oriented and correspondingly less directly in-volved in the making of routine decisions. AllenNewell recently suggested to one of the authorsthat the wave of top-management game playingmay be one manifestation of such change. Topmanagement of the 1980's may indeed spend agood deal of money and time playing games,trying to simulate its own behavior in hypothet-ical future environments.

Room for InnovatorsAs the work of the middle manager is pro-

gramed, the top manager should be freed morethan ever from internal detail. But the top willnot only be released to think; it will be forcedto think. We doubt that many large companiesin the 1980's will be able to survive for evena decade without major changes in products,methods, or internal organization. The rate ofobsolescence and the atmosphere of continuouschange which now characterize industries likechemicals and pharmaceuticals should spreadrapidly to other industries, pressuring them to-ward rapid technical and organizational change.

These ideas lead one to expect that research-ers, or people like researchers, will sit closer tothe top floor of American companies in largernumbers; and that highly creative people willbe more sought after and more highly valuedthan at present. But since researchers may beas interested in technical problems and profes-search," Operations Research, January-February 1958,p. 9.

Management in the igBo's 47sional affiliations as in progress up the organi-zational ladder, we might expect more imper-sonal, problem-oriented behavior at the top, withless emphasis on loyalty to the firm and more onrelatively rational concern with solving difficultproblems.

Again, top staff people may follow their prob-lems from firm to firm much more closely thanthey do now, so that ideas about executive turn-over and compensation may change along withideas about tying people down with pensionplans. Higher turnover at this level may proveadvantageous to companies, for innovators canburn out fast. We may see more brain pickingof the kind which is now supposedly character-istic of Madison Avenue. At this creating andinnovating level, all the current work on organi-zation and communication in research groupsmay find its payoff.

Besides innovators and creators, new top-management bodies will need programmers whowill focus on the internal organization itself.These will be the operations researchers, mathe-matical programmers, computer experts, and thelike. It is not clear where these kinds of peopleare being located on organization charts today,but our guess is that the programmer will finda place close to the top. He will probably re-main relatively free to innovate and to carry outhis own applied research on what and how toprogram (although he may eventually settle intousing some stable repertory of techniques as hasthe industrial engineer).

Innovators and programmers will need to bfesupplemented by "committors." Committors art;people who take on the role of approving oi-vetoing decisions. They will commit the organiczation's resources to a particular course of ac-tion — the course chosen from some alterna-tives provided by innovators and programmers.The current notion that managers .ought to be"coordinators" should flower in the 1980's, butat the top rather than the middle; and the peo-ple to be coordinated will be top staff groups.

Tight Litde OligarchyWe surmise that the "groupthink" which is

frightening some people today will be a com-monplace in top management of the future. Forwhile the innovators and the programmers maymaintain or even increase their autonomy, andwhile the committor may be more independentthan ever of lower-line levels, the interdepend-ence of the top-staff oligarchy should increase

with the increasing complexity of their tasks.The committor may be forced increasingly tohave the top men operate as a committee, whichwould mean that the precise individual locus ofdecision may become even more obscure thanit is today. The small-group psychologists, theresearchers on creativity, the clinicians — allshould find a surfeit of work at that level.

Our references to a small oligarchy at the topmay be misleading. There is no reason to be-lieve that the absolute numbers of creative re-search people or programmers will shrink; ifanything, thereverse will be true. It is the headmen in these areas who will probably operateas a little oligarchy, with subgroups and sub-subgroups of researchers and programmers re-porting to them. But the optimal structuralshape of these unprogramed groups will notnecessarily be pyramidal. It is more likely to beshifting and somewhat amorphous, while theoperating, programed portions of the structureought to be more clearly pyramidal than ever.

The organization chart of the future maylook something like a football balanced uponthe point of a church bell. Within the football(the top staff organization), problems of co-ordination, individual autonomy, group decisionmaking, and so on should arise more intenselythan ever. We expect they will be dealt withquite independently of the bell portion of thecompany, with distinctly different methods ofremuneration, control, and communication.

Changes in PracticesWith the emergence of information technol-

ogy, radical changes in certain administrativepractices may also be expected. Without at-tempting to present the logic for the state-ments, we list a few changes that we foresee:

« With the organization of management intocorps (supervisors, programmers, creators, commit-tors), multiple entry points into the organizationwill become increasingly common.

« Multiple sources of potential managers willdevelop, with training institutions outside the firmspecializing along the lines of the new organiza-tional structure.

C Apprenticeship as a basis for training mana-gers will be used less and less since movement upthrough the line will become increasingly unlikely.

« Top-management training will be taken overincreasingly by universities, with on-the-job train-

48 Harvard Business Review

I

ing done through jobs like that of assistant to asenior executive.

« Appraisal of higher managementperformancewill be handled through some devices little used atpresent, such as evaluation by peers.

fl Appraisal of the new middle managers willbecome much more precise than present ratingtechniques make possible, with the developmentof new methods attaching specific values to input-output parameters.

C Individual compensation for top staff groupswill be more strongly influenced by market forcesthan ever before, given the increased mobility ofall kinds of managers.

C With the new organizational structure newkinds of compensation practices — such as teambonuses — will appear.

Immediate MeasuresIf the probability seems high that some of our

predictions are correct, what can businessmendo to prepare for them? A number of steps areinexpensive and relatively easy. Managers can,for example, explore these areas:

(1) They can locate and work up closer liaisonwith appropriate research organizations, academicand otherwise, just as many companies have profit-

ed from similar relationships in connection withthe physical sciences.

(2) They can re-examine their own organiza-tions for lost information technologists. Many com-panies undoubtedly have such people, but not allof the top executives seem to know it.

(3) They can make an early study and reassess-ment of some of the organizationally fuzzy groupsin their own companies. Operations research de-partments, departments of organization, statisticalanalysis sections, perhaps even personnel depart-ments, and other "odd-ball" staff groups often con-tain people whose knowledge and ideas in thisrealm have not been recognized. Such people pro-vide a potential nucleus for serious major efforts toplan for the inroads of information technology.

Perhaps the biggest step managers need totake is an internal, psychological one. In viewof the fact that information technology willchallenge many long-established practices anddoctrines, we will need to rethink some of theattitudes and values which we have taken forgranted. In particular, we may have to reap-praise our traditional notions about the worthof the individual as opposed to the organization,and about the mobility rights of young men onthe make. This kind of inquiry may be pain-fully difficult, but will be increasingly necessary.

Among the "determinists" in administrative theory — and in this category fallmost of those who call themselves behaviorists — the notion is widely held

that what an individual institution may be able to accomplish is controlled by thegeneral environment in which it operates. I suppose a certain amount of supportmay be secured for this position by an oversimplified reading of biology as well asof psychology and sociology. But the behaviorists are wrong. They are wrong not

because what they assume lacks truth but because they leave out of considerationthe contriving factors in man's intelligence. . . .

Or putting the matter more positively, what we have to assume in administra-tion is that man, being intelligent, is capable of a growing amount of free will.That he can help to determine his environment and not merely be bound by itand conform to it. And if you think that this is an idle notion, how else can youexplain the American tradition, with the self-confidence and daring that our pio-neers exhibited, their lusty proclivity to take chances and to experiment, theirinsolent disregard for convention and for the shibboleths of the Old World whichwould have tied them to the rule of privilege and the powerful? Do we want atthis stage in our national history to turn our backs on all this — the chief virtuethat has given us what success as a nation we have achieved?Marshall E. Dimock, A Philosophy of Administration

NewYork, Harper & Brothers, 1958, pp. 21-22.

Graduate School of Industrial AdministrationWilliam Larimer

Mellon,

Founder

Carnegie Institute of TechnologyPittsburgh 13, Pennsylvania

REPRINT SERIES

1. Management Models and Industrial Applications of Linear Programming, by AbrahamCharnes and William W. Cooper. Management

Science,

October 1957. Additional copies,

$1

each.2. Party legislative Representation as a Function of Election Results, by James G. March.Public Opinion Quarterly, Winter 1957-1958.3. Decision Rules for Allocating Inventory to Lots and Cost Functions for Making Aggregate

Inventory Decisions, by Charles C. Holt. Journal of Industrial Engineering, January-February 1958.

4. HeuristicProblem Solving: The Next Advance inOperations Research,by Herbert A. Simonand Allen Newell. Operations Research, January-February 1958.

5. Queuing with Preemptive Priorities or with Breakdown, by Harrison White and Lee S.Christie. Operations Research, January-February 1958.

6. Inflation in Perspective, by G. L. Bach. HarvardBusiness Review, January-February 1958.7. Forecasting Uses of Anticipatory Data on Investment and Sales, by Franco Modiglianiand

H. M. Weingartner. Quarterly Journal of Economics, February 1958.8. Notes on Railroad Productivity and Efficiency Measures, by Edwin Mansfield and Harold

H. Wein. Land Economics, February 1958.9. Environment as an Influence on ManagerialAutonomy, by William R. Dill. Administrative

Science Quarterly. March 1958.10. Linear Decision Rules and Freight Yard Operations, by Edwin Mansfield and Harold H.

Wein. Journal of Industrial Engineering, March 1958.11. A Regression Control Chart for

Costs,

by Edwin Mansfield and Harold H. Wein. AppliedStatistics, March 1958.

12. Mathematics for Production Scheduling, by Melvin Anshen, Charles C. Holt, Franco Modig-liani, John F. Muth, and Herbert A. Simon. Harvard Business Review, March-April 1958.

13. A Model for theLocation of a Railroad Classification Yard,by Edwin Mansfield and HaroldH. Wein. Management

Science,

April 1958.14. Cost Horizons and Certainty Equivalents: An Approach to StochasticProgramming of Heat-

ing Oil, by Abraham

Charnes,

William W. Cooper, and G. H. Symonds. ManagementScience, April 1958.

15. Selective Perception: A Note on the DepartmentalIdentification of Executives, by DeWittC. Dearborn and Herbert A. Simon. Sociometry, June 1958.

16. Elements ofa Theory of Human Problem Solving,by Allen Newell, J. C. Shaw, and HerbertA. Simon. Psychological Review, May 1958.

17. TheCost of Capital,CorporationFinance and the Theory of Investment,by Franco Modig-liani and Merton H. Miller. American EconomicReview, June 1958.

18. New Developments on the OligopolyFront, by Franco Modigliani. Journal of PoliticalEconomy, June 1958.

19. Some Observations on the Business School of Tomorrow, by G. L. Bach. Management

Science,

Jury 1958.20. The Size Distribution of Business Firms, by Herbert A. Simon and Charles P. Bonini.

American Economic Review, September 1958.21. A Study of Decision-Making Within the Firm, by Edwin Mansfield and Harold H. Wein.

Quarterly Journal of Economics, November 1958.22. A Comment on Market Price and Stabilization Policy, by Allan H. Meltzer. Review of

Economics and

Statistics,

November 1958.(Continued on back cover)

(Continued)23. Managementinthe 1980's, by Harold J. Leavitt and Thomas L. Whisler. Harvard Business

Review, November-December 1858.24. Extremal Principles for Simulating Traffic Flow in a Network, by Abraham Charnes and

William W. Cooper- Proceedingsof the National Academyof

Science,

February 1958.25. Determining the "Best Possible" Inventory Levels, by Kalman Joseph Cohen. Industrial

QualityControl, October 1958.

f

The present series begins with articles written by the faculty of the GraduateSchool of Industrial Administration and published during the 1957-58 academicyear. Single copies may be secured free of charge from: Reprint Editor,G_J JUCarnegieInstitute ofTechnology,Pittsburgh

IS,

Pa. Additional copiesare 25 eentaeach, unlessotherwise noted. y