J. Opl Res. Soc., Vol. 31, pp. 17 to 28 Pergamon Press Ltd 1980. Printed in Great Britain © Operational Research Society Ltd
0 160-5682/80/0 10 1-0017$02.00/0
The Role of Management Science SAMUEL EILON Department of Management Science, Imperial College of Science and Technology Many of the arguments that are found in the literature on the theme "O.R-JM.S. in crisis" stem not just from alternative definitions of O.R-JM.S., but from differences in strongly held views about its scope and role in organisations, and these are then reflected in the vehemence with which certain definitions are defended or criticised. It is when the hopes and aspirations of O.R./ M.S. are compared with what is being achieved in practice, that we realise the degree to which O.R./M.S. has failed to live up to its original promise. In most organisations, O.R. analysts are expected and are happy to act as technicians and not as advisers, so that technique orientation and concern with tactical problems are constantly reinforced. In addition, managers in various functions have become aware of the potential contribution of analytical modelling in their own spheres and have begun to recruit analysts direct, resulting in the possible fragmentation of O.R. and the loss of its unique identity. The responsibilities of O.R-JM.S. obviously relate not only to its organisational status, but also to an examination of and an identification with organisational goals. Many O.R. analysts are plainly more comfortable when these goals are clearly defined for them, an attitude which people from other professions will readily endorse, but if O.R./M.S. has any aspirations to get involved in strategic problems, it will have to pose questions about the validity and appropriateness of organisational goals and try to influence the formulation of problems accordingly.
INTRODUCTION IF I HAD to nominate the fastest growing subject in the Operational Research world, it would be "O.R. in crisis", a theme seized upon not only by outside observers who look at O.R. and find it wanting, but also by O.R. people themselves, who, for one reason or another, have become disenchanted with the progress of their profession and its standing. You only have to examine the literature, or scan the topics that adorn the conference programmes of T.I.M.S. and various O.R. societies, to realise that raising nagging questions about the present state of O .R. and its future is becoming a frequent occurrence, and this must be due to a deep-rooted anxiety that all is not well in our midst. "The future of O.R. is past", blasted Russ Ackoff at the O.R. Society Conference last year, 1 followed by a sequel on how to resurrect it. 2 Hall and Hess addressed themselves to the question of whether O.R./ M.S. was dead or dying and why 3 and even in the sixties, AI Rubinstein and his colleagues at Northwestern University in Chicago were examining birth and death processes of O.R. groups in industry, so that death and resurrection have been a major preoccupation for some time now, although it has perhaps become a more pressing issue of late, and only last May the Society devoted a National Events Day to a discussion of Ackoff's two papers and their implications. With all of this lengthy debate, it is difficult to see how anything can usefully be added to what has, in one form or another, already been said. And yet, in order to develop the theme on the roles and responsibilities of the management scientist, it is inevitable that I shall refer to the perceived crisis in O.R. It is natural that when you take part in a discussion about the place and expected contribution of O.R./M.S., you are greatly influenced by your role, following the wellknown maxim that where you stand depends on where you sit. In my own case, the views that I have formed about the strengths and weaknesses of O.R.jM.S. as it is currently practised may well seem somewhat schizophrenic. As a consultant, I am keen to see O.R.jM.S. groups maintain an image of high intellectual quality, able to tackle both tactical and strategic problems and report at the very top echelons of their organisations; as a company director, I realise how many central issues are only amenable to qualitative analysis and hence I know the limitations inherent in the studies that the O.R.jM.S. group can conduct and the advice that it can offer; as an academic I am anxious to see the subject advance and new avenues explored. Although there are bound to be differ17
Operational Research Society is collaborating with JSTOR to digitize, preserve, and extend access to Journal of the Operational Research Society. ® www.jstor.org
Journal of the Operational Research Society Vol. 31, No.1
ences in emphasis in these outlooks, I hope that readers will regard this amalgam as a useful contribution to the discussion. There are two ways in which one can proceed. The first is to start with alternative explicit definitions of operational research and management science, argue about the efficacy of these definitions, and then draw conclusions about the way this activity should be organised in practice. I shall not follow this course, except by implication; in the past two decades we have had many sterile discussions about definitions, and, apart from the fact that there are genuine differences in the approaches and attitudes of the many protagonists, I am beginning to doubt whether any proposed definition can do more than provide a general indication of our areas of interest. May I add that this difficulty is not unique to O.R. If you examine the definitions of other disciplines, like physics, mathematics, or mechanical engineering, you find statements that are too embracing and too vague; but these disciplines have been around for such a long time that most people are no longer concerned about their definitions. I wish we felt that O.R. was reasonably mature and sufficiently secure to argue that O.R. is what O.R. people do, but I suspect that arguments about definitions will continue to be a favourite pastime, and that this paper will do nothing to reduce the temperature of this debate. ASPIRATION AND REALITY An alternative way to develop our discussion is to take the various competing definitions as read and to brush aside-at least for a while-their possible implications, even to follow many of our American colleagues who have given up the attempt to distinguish between operational research and management science and use the umbrella acronym O.R./M.S. to describe the general area of our activity, and then proceed to compare the aspirations of O.R./M.S. with its actual achievements. It is this comparison that has led to so much criticism and soul searching, and it goes something like this. The high hopes for O.R./M.S. stem from the conviction that it can and should make a contribution at the top decision-making stratum of the organisation, that its approach to problem solving is inter-disciplinary and based on scientific modelling, that hence its methodology is unique and complements rather than attempts to supplant the work of conventional functions of management. It has further been argued that the dispassionate analysis of data, which is the trade mark of the true scientist, must elevate O.R./M.S. above the politics of power and make it acceptable to-nay, sought after by-the senior executives in the organisation, and it was therefore expected that O.R./M.S. would not be identified or become an integral part of one particular function. Furthermore, O.R./M.S. was seen not only as the means for evaluating crucial problems facing the organisation, but as the natural mechanism for originating new ideas and promoting innovation. Reality, however, is rather different from this rosy picture. The O.R.jM.S. group reports at a level well below the Board, its work concentrates on tactical problems, often with limited terms of reference, its contribution to the evaluation of strategy is minimal, its approach is technique-orientated and the competition from other functions is real and sometimes threatening; the uniqueness of the O.R. approach is not seen as indispensable, its methodology is challenged, it is regarded as a narrow specialist discipline, a suitable sanctuary for mathematicians, its involvement in implementation is tenuous and its general impact somewhat limited. I am very much aware of the fact that generalisations of this kind do an injustice to the few examples where we are assured that O.R./M.S. has lived up to its promise, and where the gap between aspiration and reaiity is sufficiently narrow to bolster the claim for great success. But such examples are indeed few, and against them other cases can be cited where O.R. has largely been ignored, where studies have culminated in stacks of reports that adorn the O.R. manager's bookshelves but have not resulted in implementation, and where O.R. has never been invited to participate in discussions on major decisions. I am, therefore, not dismissing the success stories; on the contrary, they present us with exciting cases histories to contrast with the more prevalent experience to which I have 18
S. Eilon-The Role of Management Science referred, sb that lessons may be learnt about the future development of the practice of O.R./M.S. We have to ask ourselves, of course, whether we need to get worried about it all. It could be argued that in the first flush of enthusiasm during the forties and fifties expectations were bound to be high, and that as the first generation of pioneers gave way to the current generation of professionals, realism and maturity began to prevail, so that the original aspirations should no longer serve as a yardstick for measuring success. O.R. is what O.R. is, it is what reality has made it to be, it has successes and it has failures, and all we need to do is learn, adjust and improve. Clearly, many of us are not satisfied with such an answer, and the literature that I alluded to is ample evidence of this dissatisfaction. Maybe our memories are too long, and once our places are taken by a new generation, unencumbered by our ambitions and expectations, the problems that worry us will vanish. But in the meantime we are here, the problems need to be aired, and I hope that the next generation is listening.
SOUL SEARCHING First, let me draw attention to some of the divergent views that have been expressed recently about the state of O.R./M.S. Hall and Hess 3 echo the sentiment held by many O.R./M.S. practitioners, namely that it .is all the fault of the academics; they claim that "practitioners could be more effective if more of the academics' new discoveries in O.R./M.S. theory were made truly accessible to them", "if some academics would show less disdain for problems they have 'solved before'", and if they would condescend to take part-time jobs in companies to improve rapport with the hard-bitten practitioners. The implication of this argument is that the practitioners need more tools, that bringing them together with the academics would ensure that their knowledge of intricate techniques was constantly updated, and that the more sophisticated their tool-kit became, the better service they would be able to provide. Concern is also expressed by MullerMerbach, who says 4 : "The journals are not readable, the authors do not write for the readers, there is too much useless theory, the authors are not familiar with real life problems and are not even interested in them, the literature is tool-oriented and not problem-oriented, and practitioners do not write". A rather radical view is taken by Parker, and I can do no better than quote what he said in the 1976 Blackett Memorial Lecture, when he summarised the growing disenchantment on the part of managers as follows 5 : "Our failure, in my opinion, has been our tendency to rely too much on rational methods, quantitative methods if you will, in addressing the increasing complex and urgent problems of management. We have tended to ignore or deny the things we could no.t measure, and to cling, almost as an act of blind faith, to the rational, hard-headed, scientific approaches epitomized in the increasingly sophisticated techniques of the so called 'managerial sciences', which enjoyed such a boom in the immediate post-war period". He went on: "We managers believed, because we wanted to believe, that rigorous scientific analysis, based on objective data and all available information, must yield the 'right' answers and thus enable us to solve the problems. But more often than I care to think about, the event proved us wrong. The answers we got failed to solve the problem-not necessarily because they were wrong, but simply because they were often incomplete, or sometimes quite irrelevant. They missed some key element of the problem we were trying to solve." Ackoff, too, is unequivocal in his criticism of quantitative techniques and argues that "O.R. has been equated by managers to mathematical masturbation and to the absence of any substantive knowledge or understanding of organizations, institutions or their 19
Journal of the Operational Research Society Vol. 31, No. 1
management" 1, and he accuses O.R. of employing "a Machine-Age approach to the self-control problem. Its methodology is analytic and its models are predominantly of closed mechanical systems, not of open purposeful systems". He further argues that "managers do not solve problems; they manage messes" ("messes" in Ackoff's terminology are "dynamic situations that consist of complex systems of changing problems that interact with each other") and he advocates that for a mess we need synthesis and not analysis. 2 Consider now what the current President of O.R.S.A. (Seth Bodner) and the President of T.I.M.S. (Frank Bass) are preoccupied with. The former laments the fact that "the O.R.jM.S. profession lacks public recognition and even recognition by many of the other scientific professions", in spite of the estimated 30--40,000 practitioners and research workers and over 100 O.R./M.S. educational departments and programmes in the US. He is preoccupied with scientific recognition, which he feels will probably have to come before public recognition, and therefore, he says, we need to "further develop the 'operational science' of our field" and that "active efforts ... to identify research, and teach the unique aspects of operational science and O.R./M.S. practice will serve as a cohering force for the profession. O.R./M.S. developments will continue to be needed and supported".6 In other words, he sees our salvation in more of the same, and has made an approach to the National Science Foundation in the U.S. in his quest for scientific recognition. The President of T.I.M.S., too, is concerned about the scope and influence of management science. In a recent message to the membership of T.I.M.S. 7 he starts off: "Is management science science? The original concept of management science as an activity organized to apply scientific methods in the solution of certain classes of problems tended, early on, to give way to a focus on the development of methodology. Indeed, when many people think of our field, they think of the methodologies associated with it: mathematical programming, decision theory, scheduling algorithms, and the like". He argues for a broad view of management science, and he concludes with the optimistic prediction that "management will become management science as our field continues to develop and achieves recognition as a unique science". I find these perorations rather naive. Apart from the implied, but not fully argued, distinction between operational science and O.R./M.S. and the precise nature of their alleged uniqueness, on what basis do the two presidents hope that an appeal to the scientific community can have the desired impact on the industrial and business environment, whose performance and well-being O.R./M.S. is supposed to enhance? If the raison d' etre of O.R./M.S. is to assist in the decision making process, then credibility in the eyes of decision makers depends solely on the nature and quality of the assistance that they can get, and not on whether other scientists enthuse about elegant solutions to non-existent problems. There is, therefore, no escape from the conclusion that O.R./M.S. must be problem orientated and not technique orientated; techniques are only convenient means by which generalisations can be sought, not ends in themselves. As Blackett said way back in 1943: "One obvious characteristic of operational research, as at present practised, is that it has, or should have, a strictly practical character. Its object is to assist the finding of means to improve the efficiency of ... operations in progress or planned for the future". 8 As for the claim of the two presidents and others that O.R./M.S. is unique, it can either be based on the argument that the techniques employed by the O.R.jM.S. analyst are unique, or that his approach to problem solving is modelled on the scientific method and is therefore special. The first argument goes back to the pioneering days of O.R., and again it was Blackett who first argued that "operational research, like every science, must not copy in detail the technical methods of any other science, but must work out techniques of its own, suited to its own special material and problems", adding the pertinent comment: "these techniques must not remain rigid but must change with the nature of the problems". 8 Over the years the claim for uniqueness of techniques has obviously been eroded. We have been so successful in preaching the gospel to engineers, scientists, economists and business administrators. that courses on mathematical models
20
S. Eilon-The Role of Management Science
and optimisation are not longer confined to university O.R. departments, so that-as Ackoff put it-"this has deprived O.R. of its unique incompetence". 1
THE SCIENTIFIC APPROACH We are then left with the claim that as O.R. people we are unique in our approach, not so much in what tools we use, but in how we determine what to use, how we design and evaluate the alternatives, and how we conclude whether a particular course of action has been or is likely to be successful. In short, we stake our claim on the objectivity and process of enquiry of the scientific method. Here, too, Blackett set the scene by describing in 1941 the role of scientists in the military environment: 8 'To what extent is it useful to do analysis of operations in a more scientific manner than is done normally by specialist officers? Experience over many parts of our war efforts has shown that such analysis can be of the utmost value, and the lack of such analysis can be disastrous. Probably the main reason why this is so, is that many war operations involve considerations with which scientists are specially trained to compete, and in which serving officers are in general not trained" and are "not available for detailed analytic work". Similar arguments were subsequently voiced in the civilian field, with some commentators putting the stress on the uniqueness of the scientific method, which managers in executive positions do not possess; others have been content to concede that managers are not always incapable of developing the necessary analysis, but simply do not have the time to do so and therefore need expert help from O.R. workers. Be that as it may, the uniqueness of the scientific method in the way it relates to O.R. has come under close scrutiny. I do not propose even to attempt to summarise the numerous contributions that have been made to this debate in the literature, except to note the increasing criticism of the conventional view which tends to regard O.R. as a "hard" science. Bevan and Bryer 9 note some of the misconceptions involved in the attribution of a scientific method to O.R. and speculate on the problem of measuring its contribution to the functioning of social organisations, Boothroyd tries to articulate his interpretation of "O.R. as articulate intervention into articulate programmes" 1 0 while Dando and Sharp are scathing in their criticism of O.R. in practice and dwell on the gap between fact and fiction as they perceive it. 11 It all adds up to a realisation that while the scientific method can greatly contribute to the model building process, the weakness in projecting O.R. as a scientific activity (at least in the hypothetico-deductive sense of Popper/Medawar) lies in our inability always to test conjectured hypotheses, 12 hence the reluctance or the failure of O.R. (depending on your viewpoint) to make a significant contribution to problems in social systems (see Cook 13 ). As an oversimplification, one might say that the difficulty lies in the fact that the analyst feels comfortable when he deals with variables that are measurable, with criteria that are definable, and with conjectures that are testable, so that when he has to face systems that lack these attributes, his contribution as a scientist can be seriously challenged. His reluctance to be drawn into such situations is understandable, but alas this is precisely where his advice-if it is accepted-is likely to have the greatest impact. It is therefore clear, at least to me, why many O.R./M.S. analysts have been content to confine their efforts to the tactical level, and this is by no means a recent phenomenon. Hugh Miser, who served for many years as editor of the Journal of the Operations Research Society of America and as such has left an indelible mark on the O.R. world, asserted in 1963 that "we should never sell operations research as the solver of major strategic problems". 14 The very first example cited by Morse and Kimball in their book on O.R. 15 in 1950 (admittedly they state that the example was "simple to the point of triviality") concerned the O.R. study of soldiers queueing to wash their mess kits, culminating with the proposal that instead of having two tubs for washing and two for rinsing there should be three for washing and one for rinsing. O.R. has been bedevilled by such convincing examples ever since, the literature is full of them, and while they seem to be 21
Journal of the Operational Research Society Vol. 31, No.1
useful frills in selling O.R./M.S. to the uninitiated, they are a far cry from the claim that O.R. can make an intrinsically important contribution at the strategic level. So, what is, or what should be, the role of O.R./M.S. in an industrial, commercial or administrative system? There are two aspects to this question, first the scientific role that the analyst should seek to assume for himself in relation to the problems that he is required to solve, and secondly the type of problems that he should address himself to. I have discussed these issues at some length elsewhere 12 and propose to touch on them only briefly here. The scientific role-apart from what has already been said about the scientific method and its relevance to O.R.-stems from the term "research" which the pioneers in their wisdom decided to include in the O.R. label, a term which has stuck in spite of some misgivings expressed about it from time to time (Littauer of Columbia University, for example, tried even in the fifties to get the name operations analysis accepted, but it never caught on). As I pointed out some time ago, 12 there is no such thing as a typical research worker and several archetypes can be identified, such as the chronicler, the dialectician, the puzzle-solver, the empiricist, the classifier, the iconoclast, and the change-agent. They all serve the cause of research in their own way, but they differ in their approaches and in their interests, so that the same individuals may, in fact, assume different roles at different times, depending on the circumstances of each investigation and on its purpose. The chronicler is pictured as the epitome of the detached observer (except that "the dubious validity of the concept of the completely unbiased observer lies in the proposition that the scientific method requires the researcher to have a theory or hypothesis which he intends to test when he embarks on a data-collection exercise" ;12 ) the dialectician engages in debate and argument, even at the risk of disturbing the system under study, in his endeavour to elicit the truth; the puzzle-solver devotes his efforts to finding a solution to a well-structured, often mathematical or logical, problem; the empiricist engages in designed experiments, from which he hopes to learn about the behaviour of the system; the classifier attempts to put order in place of a plethora of disorganised data; the iconoclast challenges conventional wisdom and beliefs, whether he has alternative theories to propose or not; and the change-agent sees it as his mission not only to gain an understanding of a given system, but also to change it. Clearly O.R./M.S. analysts can, and do, assume any one or several of these roles, but of the seven archetypes I would like to single out two that perhaps best serve to illustrate fundamentally different attitudes to what O.R./M.S. is all about, namely the puzzle-solver and the change agent, and perhaps it would be appropriate if I quoted from what I have written about the two types: 12 "The puzzle-solver is happy to accept the definition and the objectives set by someone else, as for example in the case of mathematical or chess problems. The configuration is given, the rules of what is allowed and is not allowed are prescribed; the problem is to find a solution that satisfies predefined conditions: prove a certain relationship, compute the value of a certain variable, find the best strategy (to win a game, to maximize return, to minimize costs)" ... The change-agent's "prime objective is to change a given system, not by merely studying it and proposing in a consulting role how it could be altered, but by being a part of the system (at least for a while) and helping to change it from within ... even with the intention of changing the structure and organization of the system." Now, most O.R./M.S. workers see themselves as puzzle-solvers (and I distinguish between a puzzle-solver and a problem-solver in that the latter participates in defining the problem that needs to be solved, whereas the former does not), and alas too few are prepared to accept the role of change-agents. It is this divide which has led me to regard the puzzle-solvers as technicians, albeit very good ones, who concentrate on perfecting their tool-kit and who spend all their time on tactical problems, while the change-agents act as advisers, with a commitment to tackle strategic, ill-defined, messy problems, the solution of which is more likely to have a significant and lasting impact on the system
22
S. Eilon-The Role of Management Science than the solution of tactical problems. The distinction between the technician and the adviser is a fundamental one: the technician is a specialist, he has a limited and welldefined role, his techniques are distinct and his main strength is embedded in analysis; the adviser tends to be a generalist, his role is less well-defined, his premises and recommendations are more liable to be questioned, and his approach relies on synthesis. More often than not the technician waits patiently for the problem to come to him and will certainly not embark on a study without being commissioned to do so, whereas the adviser in the role of a change-agent is ever on the lookout to identify important areas in which he can get involved. THE DISPERSION OF O.R. There is nothing new about my plea that more O.R./M.S. people should be prepared to act as advisers. I have already alluded to the many criticisms that O.R./M.S. has failed to make any significant contribution to decision making at the strategic level, although some of the critics realise that by abandoning the sanctity of his tool-kit the adviser becomes vulnerable, in that he loses any claim for the uniqueness of his expertise. As Ackoff once put it, "generalization is not the sum of several specializations", and if we see the adviser as a generalist, then he has to compete with other functional managers for his voice to be heard, and this is not a position that many O.R./M.S. theorists particularly relish. What we must realise, however, is that such competition is becoming inevitable, and that even the seemingly cosy technically orientated O.R. group· now faces competition from other functions of management. Hall and Hess remark on the fact that "O.R./M.S. talent is increasingly being dispersed to the various corporate functions" ;3 last year more than half of my students did not join O.R./M.S. groups, and I suspect that other course organisers have witnessed a similar phenomenon. What does this dispersion of O.R./M.S. mean? I am reminded by what Rivett and Ackoff said in their guide to O.R. some 16 years ago when they asked "Will O.R. take over management? This is a blunt question and it demands a blunt answer. The answer is quite simple-No. Management, in fact, will take over O.R." 16 I used to regard this remark with some scepticism, but the dispersion of O.R./M.S. is perhaps an indication that they were right after all. As I said earlier, one way to view this trend is to suggest that O.R. has been so successful in its application that it has now been warmly embraced by all functions of management to become an integral part of their planning and evaluation procedures, and, viewed iri this light, we should congratulate ourselves on a remarkable achievement. But there is another side to this story. THE CONSTRAINING ENVIRONMENT My remark about most O.R./M.S. people being able and willing to act in the role of technicians rather than advisers may seem an indictment of their poor judgement in choosing the easier and less rewarding option. But people-and O.R. analysts are no exception-have to operate within constraining environments and, by and large, have to play the roles expected of them. True, the profession does often adopt a myopic view, but we have to look for other contributory factors for the current state of affairs, and these lie in the attitudes of organisations that recruit and employ O.R. people. To understand the effect of these attitudes, let us visualise a young O.R./M.S. man in his mid- or early twenties, a graduate (typically in science, mathematics or engineering, possibly in economics), probably with an M.Sc. degree in O.R./M.S.; he may have had a couple of years experience, possibly working as a scientist or an engineer before deciding to take an M.Sc. course in O.R./M.S.; even if he gets a grant to carrie on a course, he makes a financial sacrifice in the hope of better things to come. What is this young man looking for? At the risk of generalising, it appears that he would normally be looking for an opportunity to join an experienced O.R./M.S. group, where he can be exposed to a 23
Journal of the Operational Research Society Vol. 31, No.1
variety of challenging problems, where he can learn how to work as a member ot a team, and where, after a while, he can be entrusted with the responsibility of leading project teams and gaining the confidence of managers in various parts of the organisation. He hopes that after a few years in the O.R./M.S. group he will be able to move out in a managerial capacity to a staff or even line position. Needless to say, there are some who are happy to remain analysts all their lives, who do not seek further responsibilities in the wider managerial hierarchy and are quite content to become managers of O.R., but my observations suggest that such people are in the minority, and that most entrants into O.R.jM.S. see it as an interesting and promising channel through which they can develop their future managerial careers. Now, contrast these expectations with what happens in practice, and perhaps I can quote some examples to illustrate the kind of reality that the young aspirants encounter; (1) One well-known firm in the chemical industry insists on recruiting inexperienced O.R. personnel first as programmers, and after a spell of two or three years they are allowed to become O.R. analysts. Result: no graduates with a master's degree in O.R./M.S. go there, the firm has to rely on a wide intake of new graduates or non-graduates through an inductive course in programming, while the ranks of O.R. analysts are filled with ex-programmers or with a few experienced O.R. workers recruited from other firms. The promoted O.R. analysts are obviously well-versed in various aspects of programming and computer systems, but their knowledge of model building and project work is limited and their outlook is rather narrow. (2) A clearing bank prides itself on the important contribution made by its O.R. group, but regards O.R. people as specialists, whereas for its management trainee scheme, aimed at the eventual filling of executive positions, the bank prefers generalists. The policy, therefore, is to recruit O.R. analysts on a three year contract, at the end of which they have to leave (except for the very few who stay on as project managers), and furthermore, there is an explicit understanding that they will not join any of their competitors, the argument being that O.R. techniques are equally applicable in other industries and the analysts are therefore not penalised by being expected to leave the banking sector. Result: good O.R. men, who wish to make a career in banking, are discouraged from joining; those who do come for the experience spend about a year learning about the job and about banking, and the last six months of the contract are spent looking for another job, so that the analysts are effective only for about half the time, and even during that period they may not feel sufficiently involved or committed to make a full contribution to the organisation. (3) A large engineering company set up its O.R. group many years ago specifically for the purpose of studying problems in inventory and production control. Many of the original problems have long been solved, though they need constant monitoring and updating, and some new problems do emerge, of course, from time to time; but the organisation structure has not changed, the O.R. group is still part of the production function and is rarely allowed to get involved in other problems in the company. Result: the O.R. group is perceived as a useful channel by those who wish to make a career in production, but not by others. (4) A large organisation made a special effort to set up an O.R. group, to pay its members well and provide them with good facilities. The work of the group is highly appreciated, but because of a long tradition that managers in the company have to work themselves up through a variety of relatively mundane and routine jobs before they can be considered for promotion to the senior echelons, members of the O.R. group find it almost impossible to get transferred to managerial positions elsewhere in the company. Result: many analysts become frustrated, some leave, others stay but their motivation is greatly impaired. What is the common thread in all these examples? It is the fact that O.R./M.S. is treated by the organisations as a specialism. Specialists are expected to examine narrow issues in some depth; they are not considered to be adept at tackling ill-defined prob24
S. Eilon-The Role of Management Science
!ems, they are rarely credited with having a sufficiently wide knowledge and experience of the business as a whole and they are regarded as ill-suited to make a contribution to the formulation of policy. It is no wonder, therefore, that most O.R.jM.S. analysts become technicians. Admittedly, many are technicians by choice: their comparative youth, their limited knowledge of the industries they join, their background in science or engineering, and the captivating quality of elegant mathematical modelling-they all contribute to the adoption of a technician's frame of mind. But some are technicians because the organisations in which they work and the managerial hierarchy expect them to be technicians, as the above examples of recruitment policies clearly illustrate. This state of affairs would not be altogether depressing if young analysts could see reasonable opportunities for career progression in· management after spending a suitable spell in the O.R./M.S. group. But it appears that the recruiting policies that I cited are only precursors to the obstacles that lie ahead. As I have remarked on another occasion, 12 analysts who wish to be considered for managerial positions elsewhere in the organisation encounter many difficulties, of which three major barriers are noteworthy: the salary barrier, which arises from the fact that analysts are relatively well-paid, taking into account their age and experience, often better paid than others in managerial posts for which they could be considered and which are normally filled by older people; the tradition barrier, which requires would-be managers to spend long spells in a variety of routine jobs as a pre-requisite for promotion; and the intellectual arrogance barrier, which the O.R./M.S. analysts create for themselves by constantly proclaiming how superior they are to all other living creatures and thereby antagonising the management ranks whom they wish to join, and whose co-operation they require for the purpose.
FACING UP TO THE CHALLENGE The broad picture, then, is that O.R./M.S. is facing threats from many quarters, and these threats can perhaps be summarised as follows: 1. Competition from related services to management-These services are organised under a
variety of generic terms, such as 0 + M, management services, M.I.S. (management information systems), systems analysis, corporate planning, project sections. They are all in the market for people who could go into O.R./M.S. groups and they frequently engage in work that could be undertaken by O.R./M.S. groups. The proliferation of management services and planning departments (under whatever title) and the widening scope of their work that ensues means that the role of the O.R./M.S. group needs to be periodically redefined and that it has to be constantly on guard to get a fair share of the project work that is going. 2. Dispersion-As observed earlier, various functions of management increasingly recruit ·their own analysts, typically to concentrate on problems of production, finance, personnel, or marketing. As a result, demands on a centrally located O.R./M.S. group decline, access to particular functional areas becomes more difficult, the group feels increasingly isolated and its claim to uniqueness is weakened. 3. Fragmentation-In large companies there are, from time to time, decentralising forces that advocate a great deal of autonomy to divisions and subsidiary companies. This can lead to the break-up of a central O.R./M.S. group into units, organised on product or geographical lines, with reporting relationships to local management; While such development puts the O.R. analysts closer to "the shop-floor", there is a risk that the central group will be considerably weakened and neither able to control the work of the units, nor contribute to strategic planning. 4. Relegation-If the O.R./M.S. group is perceived as a bunch of bright immature young analysts working on narrowly based problems, a vacuum is created for the generalist who will take a broader view and whose advice will be sought on questions of policy and strategy. As soon as such a role is filled, the specialist role of the O.R./M.S group O.R.S. 31/l-C
25
Journal of the Operational Research Society Vol. 31, No. 1
may become petrified, and the chances are that its head will report nowhere near the · director's level. 5. Isolation-As already indicated, the more the specialist role is reinforced and the greater the fragmentation process, the more difficult it is for individuals to move out of the O.R./M.S. group in a natural managerial career progression. Faced with these threats and uncertainties, we should not be surprised if the bright young analyst decides not to join any of the companies that have recruiting policies of the type that I alluded to, and indeed not go to any O.R./M.S. group that is ostensibly weak and poorly rated, but chooses instead another route. If this attitude becomes widespread, then O.R./M.S. will be deprived of many high fliers who are destined to become future senior managers. This has obvious long term consequences for the O.R. Society and for the profession as a whole. What courses of action should we consider to reverse this trend? One possibility is to refrain from action. This would acknowledge the fact that after the initial boom, O.R./M.S. is settling down to being one of many specialisations that help the management process, and that the trends of competition, dispersion and fragmentation are symptoms of the large scale infiltration, under one guise or another, of O.R./M.S. into all parts of the organisation, an achievement-as I said earlier-that we should be proud of. But if O.R./M.S. is to remain a distinct contributor, and not just one of the many passing fads that invade the management scene from time to time, then something positive needs to be done. First, I believe that the mobility of analysts can be greatly enhanced by a programme of secondment of personnel, even for short periods, to other parts of the organisation, preferably to line positions. 12 The instrument of short-term secondment is more palatable to all concerned than a permanent transfer, it helps managers to make a proper first-hand assessment of the potential of analysts, who would then become exposed to the company environment and have an opportunity to learn more about the business that the company is in. Secondly, the move of managers-ex-analysts back into the O.R./M.S. group will bring in some valuable experience and undoubtedly will improve the quality of management within the group and its rapport with other departments. In addition, it may be opportune not just to accept, but to positively cultivate the idea that there are two types of O.R.jM.S., the one at the tactical level and the other concerned with strategy. Instead of bitching about the downgrading of O.R./M.S., we should think of ways in which the super-O.R. man can be developed, the change-agent type, the adviser. Perhaps O.R.jM.S. is the wrong title for what I have in mind and it may be better to call the man in charge a Chief Scientist, or a Policy Co-ordinator, assisted by a small band of "think-tankers". This type will not replace what most O.R./M.S. groups are doing now, and in fact he will be in a position to commission work from them and thereby enhance, and not weaken, their status in the organisation. GOALS AND ETHICS Finally, I must refer to another aspect of the roles and responsibilities of the management scientist, which relates to his degree of involvement and identification with organisational goals. This is perhaps obvious when the analyst works for a government agency and when the pursuit of given policies is not always to his liking, particularly as he is expected not only to acquiesce with objectives that he may even thoroughly disapprove of, but diligently devise means by which these objectives can be attained more efficiently. As Howard says 17 : "Many of our present controversies involve the coercion of one group by another using the government as its agent ... The social issues of the draft, military policy, nuclear power, foreign aid, affirmative action, pollution, medical care, etc., etc. are all ethically sensitive-there is no government decision that is not. The decision analyst who works on any govenment problem without having made ethical peace with himself runs
26
S. Eilon-The Role of Management Science
the risk of realizing some day that he has been making the moral equivalent of assassins' rifles." The problem is no less acute in the business world, and there is a marked difference between two schools of thought, the one (which, incidentally, most technicians seem to subscribe to) regards economic and financial goals as sacrosanct, while the other is anxious to question these goals and to examine their wider ramifications. In this, respect management scientists are in the same boat as members of the managerial hierarchy, and their dilemma stems from their attitude to the concept of economic man, epitomised so well by Milton Friedman/ 8 who says: "Few trends could so thoroughly undermine the very foundations of our free society, as the acceptance by corporate officials of a social responsibility other than to make as much money for their stockholders as possible. This is a fundamentally subversive doctrine. If businessmen do have a social responsibility other than making maximum profits for stockholders, how are they to know what it is?" Contrast this with what Boulding 19 says in his discussion on the theory of the firm: "Even in the most capitalist of societies there are limits on the quest for profits which are imposed by the general ideals of the businessman's role, and the 'pure' profits maximizer, where he exist, pays for his profits heavily in terms of social disapproval". It is this view that has led in recent years to the development of the social audit concept, to include various aspects of consumerism, minority rights, equal opportunity employment, pollution and environmental control, contributions to the locality, and it has even been proposed that a weighted sum of these factors should then be computed to produce an overall social audit score. 20 To all this Friedmanites retort that businessmen and managers have no mandate to determine what is good or bad for society; that is the job of elected politicians, who should seek to pass such laws as are appropriate to protect the interests of society, whereas the sole object of management is to further the interests of their own organisations within the legal framework that circumscribes or encourages the operations of business and industry. Thus, they argue, it is perfectly reasonable for managers to make decisions which may adversely affect the environment, or the welfare of certain sections of the community, or the balance of payments, if such decisions are legal and enhance the financial performance of their companies. Clearly, questions of ethics constantly arise: Do you continue perfecting the manufacture and sales of a product that you know or suspect is injurious to health? Do you refrain from calling back faulty products to save the costs involved? Do you speed up the installation of automatic facilities and thereby cause redundancies? Do you shut down uneconomic operations? Do you improve working conditions (which inevitably cost money) only when you have to? Do you relentlessly pursue the interests of your employer or client, irrespective of the harm or damage caused to others? Do you accelerate the process of depersonalising the decision making activity by widening the scope of automatic-prescriptive decisions in the system? 21 These questions are not merely manifestations of ordinary problems involving multiobjectives that O.R. people often encounter (even in the tactical domain), but go further in highlighting the intricate conflict between the short term and the long term, and more significantly the conflict of interests between groups and individuals, between customers and employees, between the organisation and society. The conventional wisdom is to regard these problems as if they were of the multi-criteria type, list the objectives that need to be pursued and then engage in an exercise that attempts to reconcile the various objectives: rank them, convert some into constraints, employ cost-benefit procedures, develop trade-off functions, consolidate several criteria into a single weighted yardstick, 12 and so on. But all these methods, and some are undoubtedly ingenious, are no more than mechanical procedures designed to maintain a consistency of approach and a modicum of concensus between decision centres. The basic ethical problems facing an individual management scientist remain. 27
Journal of the Operational Research Society Vol. 31, No. 1
Add to that the fact that the management scientist has his own goals and aspirations, which are bound to affect his attitude to the way a given problem should be tackled. In formulating his conclusions and recommendations regarding the alternative solutions that may emerge, he and others must be aware of the possible influences of his subjective judgement and self-interest. Ultimately, the management scientist-or the manger for that matter (since everything that I have said in this respect about the management scientist applies equally well to the manager}-has to square his actions with his own conscience. He has a duty to his organisation and to his superior; he is paid to improve the performance of the system of which he is a part, not to impede it. If he feels strongly that the system is moving in the wrong direction, he should try to contribute to the formulation of organisational goals ;12 if he cannot exert any influence, if he cannot or will not bide his time, and if he is convinced that the actions of the organisation are harmful and morally indefensible, he should quit. POSTSCRIPT Let me end with a story, which is presumably apocryphal, but which aptly describes the current state of the O.R./M.S. scene. An invigilator at an examination opened a sealed envelope to distribute the papers to the assembled examinees and discovered to his dismay that the papers were identical to those of the year before. He hurriedly 'phoned the university authorities to ask what he should do under the circumstances, and he was promptly given a reassuring reply: the questions are all the same, but the answers are different. REFERENCES L. ACKOFF (1979) The future of operational research is past. J. Opl Res. Soc. 30, 93-104. R. L. AcKOFF (1979) Resurrecting the future of operational research. J. Opl Res. Soc. 30, 189-199. 3 J. R. HALL and S. W. HESS (1978) O.R./M.S. dead or dying? RX for survival. Interfaces 8, 42-44. 4 M. MULLER-MERBACH (1979) Letter to the editor. J. Opt Res. Soc. 30, 383-384. 5 H. PARKER (1976) Free enterprise and the wealth of nations. Opl Res. Q. 27, 413-424. 6 S. BoNDER (1979) Letter from the President. O.R./M.S. Today 6, 12. 7 F. M. BAss (1979) Message from the President. O.R./M.S. Today 6, 4 and 6, 10. 8 P. M. S. BLACKETT (1948) Operational research. The advancement of science. J. Br. Assoc. 5, 26-38. 9 R. G. BEVAN and R. A. BRYER (1978) On measuring the contribution of O.R. J. Opt Res. Soc. 29, 409-418. Also, a comment by J. D. SMITH and a rejoinder by the authors (1979) 30, 487-490. 10 H. BooTHROYD (1978) Articulate Intervention. Taylor & Francis. 11 M. R. DANDO and R. G. SHARP (1978) Operational research in the U.K. since 1977: the causes and consequences of a myth. J. Opl Res. Soc. 29, 939-949. 12 S. EILON (1979) Aspects of Management, 2nd edn. Pergamon Press, Oxford. 13 S. L. CooK (1973) Operational research, social well-being and the zero growth concept. Omega 1, 647-667. 14 H. MISER (1963) Operations research in perspective. Opns Res. 11, 669-677. 15 P. M. MoRSE and G. E. KIMBALL (1950) Methods of Operations Research. Technology Press of MIT and Wiley, New York. 16 B. H. P. RIVETT and R. L. AcKOFF (1963) A Manager's Guide to Operational Research. Wiley, New York. 17 R. A. HowARD (1978) An assessment of decision analysis. From TIMS/ORSA meeting, May 1978, New York. To be published in Operations Research Special Issue on Decision Analysis. 18 M. FRmDMAN (1962) Capitalism and Freedom. Chicago Univ. Press. 19 K. E. BoULDING and W. A. SPiVEY (1960) Linear Programming and the Theory of the Firm. Macmillan, New York. 20 B. ScANLAN and J. B. KEYS (1979) Management and Organizational Behaviour. Wiley, New York. 21 S. EILON (1979) Management Control, 2nd edn. Pergamon Press, Oxford. 1 R.
2
28