Blackett the Father of OR

by Richard Ormerod

Blackett, who during the second world war played a key role in initiating the application of operational research (OR) to military issues, is widely regarded as the father of OR. From these early military beginnings the subject of OR has diffused into and taken root in government departments, industrial organisations, commercial enterprises and academic establishments across the world. Blackett gave birth to a flourishing professional and academic discipline. He would be delighted at the range of application areas, perhaps disappointed at the operational rather than strategic emphasis of the subject and no doubt surprised at some of the recent directions the profession has taken. Few appreciate that OR is routinely deployed each time you book an airline seat, seek a financial loan or remove an item from a supermarket shelf. OR now makes an important contribution to society: Blackett’s vision for his child has been fulfilled, at least in part.

-oo0oo-

A major event in the calendar of the UK Operational Research Society is the annual Blackett Lecture given by a leading public figure, the text of which is subsequently published in the Journal of the Operational Research Society. Blackett’s role in setting up the Society and initiating the Journal (originally called the Operational Research Quarterly) is thus remembered and he is celebrated as the founding father of OR. Blackett’s activities in seeding new OR groups in Britain’s war machine during the second world war have been amply recorded. In the only paper in the first issue of the first volume of the Operational Research Quarterly, Blackett himself says "The main outlines of the growth of operational research in the armed services during the second world war have been described in numerous articles and books and are certainly sufficiently well known not to need repetition here." He goes on to suggest that others know more about "the actual practical results attained since the war by application of these methods to the great task of increasing the efficiency of our social system and the well-being of our population" than he did. He concludes "Leaving aside, therefore, both its history and its present achievements, I wish to touch on some points relating to its methodology and its organisation." (Blackett, 1950). In my paper, taking the lead from Blackett himself, I will pass over his wartime achievements and concentrate on the evolution and development of the profession and discipline that arose out of his efforts and that of other pioneers of the time. Like Blackett I will touch on some points relating to its methodology and organisation. Unlike Blackett I will also comment on some of the practical results.

In his paper Blackett poses three questions about operational research. Is it new? If so, in what way? Is it scientific? He addresses the last question first concluding that the answer must be yes because most definitions of OR include some such phrase as "the application of the scientific method". Reflecting on what is meant by scientific method he continues as follows: "A broad but eminently reasonable view is that scientific method consists of a systematic method of learning by experience (Jeffreys). In more detail, scientific method may be defined as that combination of observation, experiment and reasoning (both deductive and inductive) which scientists are in the habit of using in their investigations (Yates)." (Blackett, 1950)

In answer to his first two questions Blackett points out that other statistical, economical and social scientific analysis applies the scientific method to the complex data of human society. Despite this, in his view, the difference is that the other subjects were aimed at political action to influence policy whereas OR’s appreciable degree of novelty lies "in the level at which work is done, in the comparative freedom of the investigators to seek out their own problems, and in the direct relation of the work to the possibilities of executive action. Dr. Kittel’s well-known definition of operational research as ‘a scientific method for providing executives with a quantitative basis for decisions,’ expressed this clearly, or, as another writer has put it, operational research is social science done in collaboration with and on behalf of executives." (Blackett, 1950). For the next 20 to 30 years Blackett’s view that OR is a scientific activity conducted for executives in relative freedom would be typical of the views held by most in the UK OR fraternity. However, later these views became the subject of debate and controversy.

The growth of OR

Blackett’s choice of subject for the first paper in the new journal is not surprising: the question of what it is that defines a discipline or profession is crucial at its birth. Immediately after the war the government set up a Committee on Industrial Productivity (CIP) which produced a draft report on ‘The Principles and Practices of Operational Research’, which strongly advocated the application OR in a wide set of peacetime settings. However, a subsequent covering note to a shortened redraft produced by the Treasury’s Economic Information Unit included the comment:

"...the more we read the literature which has been circulated about it, the closer we come to the conclusion that operational research is merely a term covering a whole range of sensible activity (already known, studied and applied under other names), arbitrarily, and to no purpose, differentiated from other sensible activity". (quoted in Kirby and Capey, 1998)."

These issues are seldom satisfactorily resolved at birth and the question of definition remains important later in the life of the profession or discipline as debates continue about the domain of activities, the relationship with neighbouring disciplines and the rules for membership. Much is, of course, simply determined by what those people, who choose to operate under the banner of OR, do in practice. It also depends on how the actual and potential clients of the services perceive the tasks that can best be given to those that operate under the banner of OR. The military successes of OR led in the UK to enthusiasm for the application of OR to civil problems after the war. Initial efforts to establish OR in the government sphere were faltering but the groups established in the newly nationalised coal industry and the steel industry flourished. Early studies included the use of statistical techniques, queuing theory, inventory models, and simulation (initially by hand, later on the computer). Of course, many of the problems could be satisfactorily resolved without sophisticated techniques at all, the solution being apparent once the situation had been appraised and the data gathered. Simulation stands out as the technique that proved most valuable in diverse industrial situations. However, most commentators agreed that it was not the techniques that define OR but its conduct by scientists who applied the scientific method.

Much excitement was generated by two subsequent developments. First, efficient algorithms were developed in the US for the solution of linear programmes. As computer power grew it also became possible to tackle integer, quadratic and goal programming. Collectively these mathematical programming techniques could be used to tackle a wide variety of problems including resource allocation, transportation and production scheduling. The international oil companies in particular, who at that time managed their operations as planned activities from well-head to petrol pump, used the new techniques extensively. Second, the development of critical path analysis and its use on the US Polaris construction project led to its obligatory application for large, governmental construction projects on both sides of the Atlantic.

OR was at its height in the 1960s with industrial OR groups flourishing, growing interest in the commercial and service sectors, gradual take up in the UK civil departments of government and energetic attempts to apply the techniques to governmental planning and budgeting. Various surveys track the take up of OR which despite difficulties of comparison demonstrated the very widespread utilisation of OR in the UK. OR had thus been established as an activity of practical value. Mercer in 196? identified 766 OR groups (defined by the existence of members of the OR Society), 87% of which contained 3 or less people. By then some professional trappings had been established in the form of a society with an annual conference and a journal. Blackett was one of the four founders of the Operational Research Club in 1948 which was later renamed the Operational Research Society. It was the first OR society in the world with the Operations Research Society of America following in 1952. As noted above, Blackett was the first author to publish a paper in the Operational Research Quarterly, which was the first OR journal in the world to be published. The first issue was 15 pages long and Blackett’s paper was less than four pages of what we today call A5 paper. I particularly like the first sentence of the Editorial Notes: ‘To justify burdening the scientific world with yet another journal, two conditions must be rigorously fulfilled: undoubted utility and the utmost brevity.’ Academic life today would be very different if these standards had been upheld. The first issues could be obtained for three shillings a copy or through an annual subscription of ten shillings (post free). By the sixties both the Society and the journal had become established and respected worldwide. Despite this the Society did not introduce a mechanism to restrict entry or maintain professional standards: young OR workers could not obtain a professional qualification in OR. The Society remained essentially an open club of people who wanted to be associated together. Whether or not this was the right strategy to adopt, it had a profound effect on the nature of the Society, its activities, and its membership.

In the sixties OR started to be established in University engineering, economics and business administration departments. Later they would become established in mathematics and management departments and business schools. The first dedicated chair in OR was established at Lancaster University in 1964. OR was thus becoming established as an academic discipline as well as a professional activity. However, University training was not a prerequisite for membership of the OR Society nor for practising OR.

Some doubts set in

OR held out the hope that a scientific approach would bring a more professional, rigorous approach to planning and other managerial problems. Many companies were at this time adopting divisionalised structures to gain better control of increasingly complex national and international operations. OR was part of the technocratic infrastructure used to apply rationality and control. During the seventies there was evidence that the management agenda was moving on and confidence that OR was the answer to management problems was starting to wane. For instance, the number of articles in the Harvard Business Review that featured OR declined. This was, in part, a natural consequence of the absorption of the new ideas by management, in part because OR was being challenged by other management disciplines such as marketing, strategy and organisational behaviour, and, in the UK at least, in part because management itself was starting to become more professional, requiring less support. It may also have been the case that OR approaches were better fitted to resource allocation and control than to the more entrepreneurial approach demanded by the rapidly developing consumers markets. Externally, the scientific approach to management, within which OR was positioned, came under attack. Internally, concern started to be expressed about the direction OR was taking.

There were a number of dimensions to this concern. At a time when the nature and scope of science itself was being hotly debated, OR’s status as a scientific discipline that aided management decision making was questioned. OR’s emphasis on quantitative analysis was said to be ill-suited to many managerial problems where qualitative aspects were important. There was a danger of concentrating on that which could be measured rather than on that which was important, of giving weight to objective, numerical data rather than subjective, qualitative factors, of concentrating on the parts rather than the whole. This unease was ironically given impetus by the great success of the mathematical programming and other optimising techniques. These techniques worked on delimited areas of interest, required reliable information and usually assumed a single objective such as profit maximisation. The computer derived ‘optimum’ with its unchallengable logic offended those who felt that, in the process of formulating the problem in such a way that the computer could perform its trick, many of the important dimensions were left out or marginalised.

The success of the optimising techniques, which organisations did find useful, and which were eminently teachable, led to a greater emphasis on them. This was fed by the new masters courses in management science and OR at Universities, which were increasingly becoming the main source of recruitment for industrial, commercial and governmental OR groups. As the greater knowledge of, and reliance on, OR techniques moved into practice, concern mounted that the craft of investigation and advice was giving way to the instrumental provision of a technical service. A social as well as a functional point was at issue here. In the UK, OR had become established in a peculiarly British way which reflects the use of graduates in the civil service. Oxford and Cambridge had over the years geared itself to supply the upper echelons of the British establishment in general, and the civil service in particular. As Elliot Krause, an American sociologist, observes: "Civil servants are recruited for general work, on the assumption that brilliant amateurs, with first-class degrees from Oxford and Cambridge, can learn any task: such generalists are preferred to those with scientific and technical backgrounds, whose expertise might be limited by their specialities. This principle of amateurism - that ‘superior men of general ability’ are preferable to technicians - and the belief that all those with science and social science backgrounds are technicians - still prevails; since the 1970’s there have been many more appointments for midlevel science and technical graduates, but not in the administrative class." There is, of course, plenty of evidence that in times of stress, such as wartime, the ‘brilliant amateurs’ were clever enough to recognise the value of science and technology as the stories of, for instance, RV Jones, Zuckerman, and Blackett himself attest. The struggle to be heard was, however, formidable.

OR groups in nationalised industries provided a career route for Oxford and Cambridge mathematics and science graduates who, despite their technical background, established themselves as general providers of advice to senior executives, much as their classics, humanities and history colleagues in the Civil Service provided advice to Ministers. OR workers were able to see themselves as an elite, despite, as Krause puts it "the overwhelming cultural bias of the British upper class, and thus of British institutions generally, against technical expertise, the very possession of which signifies social unacceptability". Any emphasis on technique thus threatened to reduce the status to which OR workers aspired and reduce their standing with their client, the senior executives.

Closely linked to the concern about the dominance of techniques was the worry that OR was becoming increasingly concerned with low level operational and planning issues rather than those with strategic and policy implications. Although it was expressed as a functional concern that best use was not being made of OR’s potential, the underlying issue was the lack of influence on and access to the corridors of power. Paradoxically another stream of criticism was that OR had become the unquestioning servants of powerful interests in large organisations and was failing to provide similar support for the weak who had no access to expensive resources.

In retrospect, a further criticism of OR is that it failed to initiate, or at least strongly associate itself with, some of the important managerial developments that it might have been expected to champion. Notions of quality are now embedded in most companies, but the quality movement was largely ignored by OR. Ideas such as ‘quality is free’ perhaps sat uneasily with OR’s micro-economic notion of trade-off of substitutes. OR’s acknowledged expertise in the area of inventory management did not lead it to run with concepts such as Kanban, just-in-time and SMED (single minute exchange of dies, set-up reduction programmes). Something was amiss. One explanation is that the intellectual calibre of the peace time activists did not match those in the war, though this is unlikely; OR personnel during the war were not uniformly of high quality nor always put to good use and there are plenty of examples of excellent intellects being attracted to OR after the war. My preferred explanation is that OR workers, embedded in internal OR groups, were inward looking, the concern being organisational survival rather than the entrepreneurial activity of developing concepts for general application, across companies and industries. This highlights a weakness in the term operational research, a term which may lead one to expect the research to lead to generally applicable results. However, the research referred to is detailed, grounded research carried out to solve a local problem in a particular context. OR was not so much concerned with scientifically building a body of generalisable knowledge as with the engineering of change suited to local circumstances. Particular, local solutions were therefore sought, and if they satisfied the sponsoring managers that was success enough. Perhaps organisational constraints and assumptions were too easily accepted. OR practice was also inward looking in the sense that it was slow to draw on other disciplines, perhaps with some honourable exceptions in the use of social science by the pioneers of soft OR (see below). For instance, ideas in economics have been mainly taken up and made practical by strategists, and those in psychology and sociology by organisational behaviourists. Perhaps there has simply been too much intellectually demanding material for the limited population of mainly mathematicians, scientists and engineers to grasp and make practical (for this reason, if for no other, OR activists should value the social scientists, economists and other disciplines in their midst). Whatever the explanation, the failure of OR to grasp and champion some of the most successful ideas on its own patch is disappointing.

There were some positive results eminating from all the talk of crisis and the agonising about what OR is or ought to be in the UK. Some initiatives were set away on OR in underdeveloped countries and OR in communities. In addition, a number of approaches under development by various academics, which were rooted in social as opposed to natural science, were identified as representing a new school of thought now usually referred to as ‘soft OR’. The soft OR methods are designed to help groups of people work together on unstructured, ‘messy’ problems where the scope, dimensions and objectives of the issue need to be determined. This was a useful response to telling criticism. However, it had two effects. First, OR is now less securely located within science. This may not in practice matter as science has become a less secure place to be intellectually located, and I am also not convinced that OR was ever a science that through research undertook to build a generally applicable body of knowledge; rather it undertook research into particular, local situations to seek locally applicable understanding and improvements. Second, a wider gap has been opened up with the US which has not gone down this particular avenue.

Just as the forward momentum of OR was beginning to falter, it was given a new lease of life by the introduction of the PC. The PC proved a cheap, flexible, and portable tool on which to test algorithms, build models, and interact with the client. PC’s reduced the OR groups’ dependence on IT support and they provided an opportunity to help client departments make use of their new purchases. More importantly OR workers suddenly became much more productive. Despite all the concerns, OR activity levels in the UK continued unabated throughout the seventies and well into the eighties.

What does OR in the UK look like today?

The activities of OR practitioners today can be grouped under three headings: (i) developing and deploying algorithms embedded in information systems, (ii) helping clients in organisational settings address issues and seek improvements, and (iii) contributing to the debate about the purpose and direction of public institutions and policies. I refer to these three areas of activities as smart bits, helpful ways and things that matter. (Ormerod, 1997).

Smart Bits: With information technology enabling the capture, production and manipulation of data in ever increasing quantities, the need for algorithms and logical routines has never been greater. OR algorithms are used to forecast demand, order stock, schedule production, control maintenance, allocate resources, maximise revenue, design networks and assess risks in organisations ranging from supermarkets to banks and from manufacturing to telecommunications.

Helpful Ways: In many instances OR consultants are less concerned with providing technical expertise on appropriate algorithms than simply helping to address the problem as presented by the client by whatever means are appropriate. This could involve a debate about means and ends, organised as a series of structured workshops bringing together various stakeholders. It may involve an assessment of the feasibility, costs and benefits, and risks of a research programme, or a major investment, or a change in policy. It could involve the specification and design of a decision support or planning system for a senior executive, or an air traffic controller or a post office teller.

Things That Matter: Every so often an OR approach will contribute to a public issue. This may be in the form of an Audit Commission report on some aspect the NHS, or an analysis of AIDS-related policies, or a report on the effect of the electricity pool, or an evaluation of the relative performance of schools. Very often academics are best placed to contribute to such issues.

I would classify most of the frequently quoted second world war examples as helpful ways: there were no sophisticated algorithms and no computer systems to embed them in. I suspect that the challenge of the investigations lay in the craft skills of observation, conceptualisation, and data gathering, and the organisational skills of gaining access to the problem and earning the confidence of the decision makers. I imagine that the mathematics of the solution was well within the capabilities of the scientists employed. It could be argued that the subjects addressed were things that matter, and in a sense that was certainly the case. However, in the first instance I would classify most of the activity as helpful ways as the investigators were generally working to find technical solutions to defined problems rather than working on the implications at the policy level. In other words they were engaged with means rather than ends. These generalisations are based on a weak and distant understanding of wartime OR and I am applying distinctions, which I believe are relevant today, but may not have been then.

On reading accounts of wartime exploits I am struck by the close linkage between technology and operational capability on the one hand and military strategy on the other. This can, of course, be true of civil enterprises as well. Thus, less intrusive surgery reduces the requirement for hospital beds leading to the need to review the balance of investments in equipment, staff and hospital facilities. OR may be involved in exploring the deployment of the new technology and its consequences or it may sometimes be providing the new technology itself. An example of the latter, which originated in the airline industry, is referred to as yield management. Airlines have developed decision making aids with embedded algorithms to enable them to determine at any point of time before take-off whether to sell a seat to a customer on a particular tariff. The capability has enabled airlines to grow the market for low cost seats for the general public and as a consequence provide a frequent service for business customers prepared to pay higher prices. In effect the business traveller cross-subsidises the intrepid traveller in order to gain the convenience of a readily available, frequent service. All parties gain from the arrangement. The OR in the decision support system ensures that, as each request for a seat purchase is dealt with, the right balance is struck between filling the plane up with the low tariff customers who are prepared to book early, and maintaining some seats for late booking, high tariff business travellers. Without the ability to work out the consequences of accepting a booking or not in real-time, the airlines could not ensure that they fill their seats, maximise their revenues, and provide an attractive service to different sets of customers. Yield management is now being applied to hotel bookings, care hire, and an ever wider set of circumstances.

Where does OR’s future Lie?

Today there are 28 countries in the European federation of OR Societies (EURO) and 45 in the International Federation (IFORS). Membership of the UK Society is stable. Nevertheless, in 1988 the American sociologist Abbott cited OR as a classic case of a profession in regression, a state of decline which he says the evidence shows is irreversible. The reality is a little more complex and a little less dramatic. The model of OR as an activity conducted for executives by internal OR groups with a good deal of choice as to which issue to tackle, essentially Blackett’s model, is on the way out. A new model of highly specific investigations and developments conducted by external specialist firms and management consultancies is alive and flourishing. In time, the supply side will sort itself out with the specialist companies dominating the provision of the smart bits, external consultancies dominating the provision of helpful ways, and academics dominating the contribution of OR to things that matter.

This is not an unequivocally happy outcome. There was, and still is, great merit in the sort of grounded, detailed investigation that is the hallmark of good traditional OR. It has been argued that the Japanese were successful in developing lean production because they were atheoretical, lacked sociological awareness and based their thinking on a very detailed knowledge of how things work. Thus the technically efficient, goal oriented, socially aware OR consultants of tomorrow may get things done and satisfy their clients but they may miss the potential implications or the valuable insight as they rush to the next challenge. Only time will tell. I am confident, however, that the more relaxed, reflective times are behind us and the pressures that have resulted in the decline of internal OR groups will continue. I have seen no evidence of a decline in military and governmental OR.

Conclusion

Blackett’s child is alive and well. In some sense it has reached and passed its maturity. This should be no surprise 50 years on. The OR profession was for a time the centre of managerial attention. It had its golden era. It is now a small professional grouping which has found its niche. Few people realise that OR lies behind many every day events, providing the algorithms for airline reservation systems, checking the credit worthiness of loan applicants, and calculating the replenishment quantities required by supermarkets. Despite its important role in today’s society OR lacks pomposity, it includes many, diverse interests and has demonstrated a certain capacity for survival. We would all wish as much for our own children.

Acknowledgement

I would like to thank Norman Lawrie, chair of the ORS Archives Committee, for suggesting I write this paper and for commenting on the first draft. However, all opinions and errors are my own.

For the interested reader

• Abbott A (1988) The System of Professions: An Essay on the Division of Expert Labour, University of Chicago Press, Chicago.
• Blackett, P M S (1950) Operational Research, Operational Research Quarterly 1, 3-6.
• Colcutt R H (1965) The First Twenty Years of Operational Research, Unwin, Old Woking.
• Goodeve C F and Ridley, G R (1953) A survey of OR in Great Britain, Operational Research Quarterly 4, 21-24.
• Jones H G (1992) Early OR in the Steel Company of Wales, Journal of the Operational Research Society 43, 563-567.
• Kirby M and Capey R (1998) The origins and diffusion of operational research in the UK, Journal of the Operational Research Society 49, 307-326.
• Krause E A (1996) Death of the Guilds: Professions, States, and the Advance of Capitalism, 1930 to the Present, Yale University Press, New Haven.
• Lawrie N (1996) Unpublished paper, presented at the History of OR stream of the 1996 UK Annual OR Conference, which quotes from Zuckeerman’s biography.
• McCloskey J F (1987a) The beginnings of operational research 1934-1941, Operations Research 35, 143-152.
• McCloskey J F (1987b) British operational research in World War II, Operations Research 35, 453-470.
• Mercer A (1968) The membership of the operational research society, Operational Research Quarterly 19, 371-376.
• Ormerod R J (1997) The role of OR in shaping the future: smart bits, helpful ways and things that matter, Journal of the Operational Research Society 48, 1045-1056.
• Ormerod R J (1998) Beyond internal OR groups. Journal of the Operational Research Society 49, 420-429,
• Rivett, BHP and Ackoff, RL (1963) A Manager’s Guide to Operational Research. Wiley, Chichester.
• Rosenhead J (1989) Operational research at the crossroads:Cecil Gordanmand the development of Post-War OR, Journal of the Operational Research Society 40, 3-28.
• Tomlinson R C (1971) OR Comes of Age, Tavistock, London.

RICHARD ORMEROD is Professor of Management at Warwick Business School, located in the Operations Research and Systems Group. He is also the School’s Director of Executive Programmes and a Founder Member of the Guild of Management Consultants.

First published to members of the Operational Research Society in OR Insight April - June 1999