edgerton what has british science policy really been

Much discussion of the relationship between history and policy assumes that policymakers are not already exposed to history. In the case of research policy, history is already central to the policy discourse. The problem is that it is largely the wrong sort of history. The quality of the historical and policy discourse around science policy is notably lower than for say economic or defence policy. There is no standard set of historical and policy debates in ‘science policy’ comparable to those in other policy areas (e.g. Keynesianism versus monetarism) to stimulate discussion. A second problem is that while defence, or health, or economic, policy have reasonably robust common sense meanings, what ‘science policy’ refers to is not stable, and this matters a good deal. A third specific problem is that, in contrast to defence and economic policy, science policy too often assumes that the UK is an island unto itself, that what applies to the whole world applies to the UK, that the UK is a little world in miniature, a serious error in this area. Any proper application of history to policy has to be based on an accurate rendering of what actually happened in the past, as well as of what the policy objectives were. This is a tall order for science policy as these questions are only recently being systematically explored, and much of the older literature suffers from systematic flaws which make it unhelpful for policymakers. In this paper I first outline some of the key problems with the older literature,49 and then tell a brief story of research policy drawing on the work of historians who have transformed our understanding over the last twenty years or so.50 By ‘science policy’, professionals mean policy for scientific research, not all science: it is not about policy for scientific knowledge as a whole (which would include education). They mean policy 49 The best of the old – research council focussed histories are Hilary Rose and Steven Rose, Science and Society (London: Allen Lane, The Penguin Press, 1969); Philip Gummett, Scientists in Whitehall (Manchester: Manchester University Press, 1980); Peter Alter, The reluctant patron: science and the state in Britain, 1850-1920 (Oxford: Berg, 1987); and Tom Wilkie, British Science and Politics since 1945 (Oxford: Basil Blackwell, 1991). 50 The essence of the change is in recognising the importance of 1) private business and 2) military and departmental research of the state. See, for the encapsulation of the new picture with lots of statistics, David Edgerton, Science, Technology and the British Industrial ‘Decline’ ca. 1870-1970 (Cambridge: CUP/Economic History Society, 1996). Central Points • We need to beware of bad history – much is indeed bunk. • Research policies have been undertaken by multiple agencies and their scope and ambition have changed radically over time. • Research policies need to be understood in the context of national defence, industrial, economic, and agricultural and other policies. • In the past, research policies have been seen as a substitute for radical policies that governments did not favour. • Today perhaps the most important lesson is not to indulge in delusions of grandeur about the quality and significance of British science. Its weight in the world has changed drastically, as has that of British business, and this needs to be understood for the development of an effective policy. 32 Science Policy History for research, but not all research. In practice, ‘science policy’ is policy for research funded by government agencies that are concerned with civil work of an academic character that is largely taking place in universities. It is the policy of ‘research councils’. Now this might be a perfectly reasonable approach, but problems arise when research council research is identified with all government research, or all research, which happens regularly in the policy and the older historical literature. It is particularly inappropriate looking backwards, because things were very different then. The whole was not the same as the minor part the histories focus on. To put it more generally – the most important agents of research policy have not had ‘science’ or ‘research’ in their titles. Historically most government funded research was funded by mainline departments, not research councils. The greatest state laboratories, the greatest research programmes – whether in aeronautics, or electronics, or nuclear power, or food, had (with few exceptions) belonged to ministries whose names are now obscure. If you want to know about the history of research policy, you need to know about the policies of these ministries. However, histories of ‘science policy’, which were written from the 1960s, focussed on the research supported by research councils, often as if no other government research existed. It was often suggested, for example, that the Department of Scientific and Industrial Research (DSIR), the first research council, pioneered the funding of research by the British state (during the First World War). But research was funded long before this (and gave us radio, for example), and during the war the DSIR was a very minor research player. Indeed there are many other similar category errors routinely made in histories of ‘science policy’. A key historical example which has been central to science policy discourse recently is the ‘Haldane Principle’ of 1918. It has been assumed that there was a single science policy governed by a single principle, though one never clearly defined, along the lines that scientists determined science policy. In fact, there was no such 1918 Haldane principle, nor could Lord Haldane ever have defined one principle for science policy. He understood very clearly that most research was done by departments, but he wanted some research to be done in a semi-independent way by what we now call research councils. He had his model in the DSIR, and he followed this precedent in recommending a similar structure for the Medical Research Council (MRC), a structure that was later adopted for other research councils. Haldane gave an intelligent set of reasons for having research councils alongside departmental research, envisioning each doing different sorts of things under different kinds of control. The ‘Haldane principle’ of 1918 of the policy discourse was an invention of the 1960s and reflected a poor understanding of actual research policies and practices even then.51 The ‘Haldane Principle’ isn’t the only fanciful history of science policy that policymakers argue with. In discussions of science policy one will hear that science in Britain is on tap, not on top; that there has long been a deep division between ‘two cultures’; that the civil service has been dominated by classicists, perhaps even historians; that apart from the world wars and Harold Wilson’s White Heat of the Technological Revolution, science was ignored; that Britain has been good at inventing but bad at developing; that politics is short term so government has not been able to make the long term commitment innovation needed. We also know that British universities were until recently ivory towers dominated by arts faculties. All this history explained why, except in emergencies, the nation systematically and disastrously failed to invest enough in R&D, and why the R&D that was undertaken was misdirected, mismanaged, wasted. In the more specialised literature, we learn that in the past the directors of science policy were deluded folk who believed in something called the linear model of innovation. This needs to be understood as left-overs from the claims of self-interested parties who have sought to promote science, and their science, by using stories of failure, indifference to science, and all the rest. Indeed, a useful rule of thumb is that expenditures, influence, and impact correlate positively with the strength of the arguments that claim they are low. Thus, complaints about lack of R&D funding peaked at the moment in British history when R&D funding was at its highest as a proportion of GDP (the early 1960s).52 51 See my talk to the Institute of Government conference on Haldane - https://www.instituteforgovernment.org.uk/events/haldane-reportnext-100-years and my ‘The ‘Haldane Principle’ and other invented traditions in science policy’, History and Policy 88 (2009). 52 For a description of the declinist literature, and a comprehensive rebuttal, see David Edgerton, Science, Technology and the British Industrial ‘Decline’ ca. 1870-1970 (Cambridge: CUP/Economic History Society, 1996). 33 Science Policy History Another set of historical discourse directly informs the present and is also very present in science policy discussion. This is the one that tells us confidently, authoritatively even, that we are living through a revolution comparable to past revolutions. Experts in innovation policy used to talk of these revolutions as long waves of technical innovation, and they often still do, though this language is now less common. The talk today is of our being in a ‘fourth industrial revolution’, preceded of course by other revolutions dating back to the first in the late eighteenth century. We find that each of these great transformations was brought about by one, two, or three key technologies, which between them changed everything. These claims are not based on serious historical research, but rather propagandistic histories that are part of a claim for more public money. From 1900 to 1939 What kind of research policies did the United Kingdom pursue before the Great War? The United Kingdom was the greatest free trading power on earth – it imported half its food, and many manufactured materials as well as raw materials. Some wanted to protect national agriculture and national industry, and to grant preferences to other parts of the British empire. It was not to happen for years, but these political pressures had to be responded to. Economic liberalism did not imply indifference to research. Indeed, research was promoted as an alternative to protection. This is what David Lloyd George did, through what was called the Development Commission, whose remit was to do something for agriculture. It pumped money into agricultural research and would continue to do so until its functions were taken over by other government agencies in the interwar years. Research as an alternative to an active policy would be a recurring theme. On the military and naval side, the state promoted new devices like nobody else – the Navy had co-invented the radio, and the army had a research station developing aviation along scientific lines (under the leadership of Haldane, no less). It pursued a distinctive liberal militarism, which had new machines at its heart, different from Prussian militarism. When it came to the great mass of British industry, the state was not especially concerned – its economic liberalism meant it was not really worried about what was made or developed in the UK. It was not concerned that cars or dyes were predominantly imported; it was neither pleased nor displeased when businesses in the UK adopted techniques of overseas origin. What happened during the Great War is what one would expect. First, the armed services ramped up their research and development, growing their existing facilities (for example, Farnborough and the Woolwich Research Department) and creating new ones (for example, Porton Down). Secondly, there was new need to produce and research materials previously imported from a blocked off continental Europe. The state intervened to create new industries and firms – most notably a new British synthetic dye industry, but not only that. A new industrial policy went along with a new research policy. And it all succeeded – the UK ended the war with extraordinary military industrial capacity and new civilian industries too. What did the DSIR have to do with this (recalling that the standard story is that the DSIR was the agency created once the government had realised how disastrous its neglect of science had been)? The answer is nearly nothing at all. The DSIR was there for the postwar era, to support research in universities, in government laboratories, and in what became the industrial research associations. What then of the interwar years? What were the research policies of the British state? The context here was broadly a return to liberalism, with crucial exceptions, and then a move to protectionism and imperialism in the 1930s. The military continued with the development of new methods and continued to dominate state R&D spending. The service ministries all appointed directors of research, and they came up with lots of important new devices, such as radar, jet engines, and ASDIC (sonar), to meet the operational requirements of users. ‘User control’ was the central policy – what was later called the ‘customer-contractor’ principle. In 1939 it would have been hard to argue that there were any significant general deficiencies in the quality of British arms. 34 Science Policy History In civil research policies, one can see a number of strands: a continuing liberal strand, together with emergent national and imperial strands. The liberal strands can be seen in many of the policies of the DSIR. As Sabine Clarke has shown, the DSIR developed a very particular notion of ‘fundamental research’, which was consistent with its limited interventionist stance, while also developing university research.53 For example, the industrial research associations existed for entire industries, not particular firms. The new Empire Marketing Board (EMB) funded research as an alternative to a thoroughgoing imperial protectionism, which government rejected until the 1930s (when the EMB was dissolved). One programme the EMB developed, working with the DSIR, was new refrigeration techniques to improve and increase imports of food from the empire. In the 1930s there was also a decisive attempt to shift to imperial food supply, for example through using new refrigeration techniques to bring beef from Australia. Imperial concerns were clearly there in the state-led development of civil aviation, which included an Imperial airship plan, and a huge flying boat programme of the 1930s – by the late 1930s there were regular flights (involving many stops) to Australia and South Africa. There were also important national dimensions. Infant industry safeguarding promoted private industrial research in the chemicals industry and the electrical industry. There was also a partial move to a national policy in the case of oil. The DSIR’s largest research programme, oil-from-coal, was in the 1920s, and was a programme that nationalists loved. It was put into operation by Imperial Chemical Industries in the 1930s. It was not a British invention, but one of many examples of the UK importing and adapting technology, in this case from Germany. It should be clear by now that the DSIR was much more directly involved with development than research councils would later be. The Second World War As one would expect, the research councils had a minor role in the Second World War. Research activity was directed, on a huge scale, by ministries supplying the armed services. These were the Ministry of Supply, created in 1939 from the War Office (Army) procurement divisions, the Ministry of Aircraft Production, created in 1940 by hiving off procurement and research from the Air Ministry, and the Admiralty (the navy ministry), which kept control of research and procurement. Their main efforts went into the development of old and new weapons. They were responsible for radar and radio, atomic bombs, aviation, including the new jet engines, and much more besides.54 They were also responsible for the development of medicines, including penicillin. Academic scientists were drafted into the research laboratories of the military. The great R&D programmes were overseen by men and institutions unknown to most science policy discourse. A notable wartime development was the internationalisation of certain programmes. The UK operated as part of an empire, with a coordinated imperial military research system. There were important links with France to 1940. Most importantly, there were links with the United States from 1940. These links led to not just research sharing, but joint development of some key innovations, notably the cavity magnetron, the jet engine, and, above all, the atomic bomb, which shifted entirely to the USA (and Canada). The US had taken the lead in proximity fuses and fire-control, in the largest, most powerful aero-engines, and in the largest aircraft too. Were the policies pursued successful? It could certainly be argued that the interwar investment in new machines of war paid off in many important cases. Most of the new weapons, not least radar, jet engines, ASDIC and more, were in development by the military long before the war. Furthermore, technical development was pursued during the war on generally sensible lines, avoiding commitment to large rockets (which did the Nazis no good) or independent development of the atomic bomb. On the other hand, some famous developments proved much less significant than claimed, including the PLUTO pipeline, the Mulberry Harbours, and the Bouncing bomb. 53 Sabine Clarke, ‘Pure science with a practical aim: The meanings of fundamental research in Britain, circa 1916–1950.” Isis 101 (2010): 285-311. 54 The story of the British jet engine has long been told (not least by Sir James Dyson) in an extraordinary declinist version bearing little relation to reality. For the astonishing real story of Britain and the jet, see Hermione Giffard, Making Jet Engines in World War II: Britain, Germany, and the United States (Chicago: University of Chicago Press, 2016). 35 Science Policy History The belief that research had been central to British success in the war gained ground. In later work, this was associated with some well-known scientists who were engaged in advice, not the development of weapons. An important argument from the scientific left (though not the right) was that wartime research showed that research could be planned for the national interest. This stimulated a debate on planning versus freedom in research (a debate that had started in the late 1930s), but this debate had virtually no impact on research policy, and was really an ideological struggle between left and right. There was, however, a wide-ranging consensus that more research would need to be done than before the war, for the military, and for industrial reconstruction. The age of techno-nationalism The research policy of the British state after the Second World War can only be understood in the context of a new economic nationalism that was central to politics into at least the 1970s. The state developed the capacity to develop new things, and to put them into operation through its control of major industries and through its ability to control what was imported. One important general aim of research policy was to replace imports with home and imperial substitutes. An example is the development of processes to make sulphuric acid from local anhydride rather than imported sulphur. The most important import substitution programme, and the longest lasting, was in agriculture. The UK went from importing half its food to near self-sufficiency in the 1980s. A very large civil nuclear programme – the largest in the world – would substitute for imported oil. The national impulse was also there in weapons development. For example, the British state decided to develop a national atomic bomb, and it deployed such a bomb in the early 1960s. There was also an imperial aspect of some importance. As Sabine Clarke has shown, for example, the Colonial Research Council (which has never counted as a research council in the literature because it was departmental – dependent on the Colonial Office) spent more than the MRC or ARC in the 1940s. Central to state research policy was the provision of new British equipment to British nationalised industries. Thus, the enormous mission-oriented programmes in aviation, atomic energy, the military, and in electronics. The key agencies involved were the wartime procurement ministries, which were merged in 1946. The Ministry of Supply directed the civil and military aircraft programmes, the civil and the nuclear atomic programmes, and the electronics programmes too. There was some dispersal of functions with the creation of the Ministry of Aviation (in 1959) and the United Kingdom Atomic Energy Authority (UKAEA) (in 1954) out of the Ministry of Supply. There was another agency that would become reasonably important, which came under the Board of Trade – the National Research Development Corporation. This was there to finance the exploitation of inventions made in the public sector (including universities) laboratories (and other laboratories too). It put money into such projects as the fuel cell (a favourite for policymakers in the 1960s) and the Hovercraft. Meanwhile the research councils, while they remained relatively small, expanded. The DSIR, by far the largest, with interests in research in many fields, built up its own laboratories and created new and expanded Research Associations. The MRC and ARC were joined by Nature Conservancy in the 1940s. The early 1960s saw very important administrative realignments that were debated in a context of declinism – the belief that the UK was not doing as well as it should have been, that it was failing, for national reasons. One crucial element was thought to be the lack of, and misdirection of, R&D. Another part of the context was the UK pulling out of large scale projects – large missiles in particular, but also large independent aircraft projects, where it discovered it could not compete with the USA. This was the context of Harold Wilson’s famous declinist speech, which spoke of the ‘White Heat’ of the ‘Scientific Revolution’, which was the cliché of the time.55 The scientific 55 Wilson did not speak here of the ‘technological revolution’ – ‘scientific revolution’ was the term of the era, used, for example, in the title of The Two Cultures and the Scientific Revolution by C.P.Snow. In the Soviet World the ‘Scientific-Technical Revolution’ became a term of art in the 1960s. 36 Science Policy History revolution he had in mind was uncannily similar to the present fourth industrial revolution rhetoric – it was focussed on electronics and automation and nuclear energy. Wilson particularly focussed on computers with judgement. The details of the changes are complex, but in essence two things happened. The first was break-up of the first research council, the DSIR. The ‘Haldane Principle’ was invented in response to this.56 What happened was that the bits that funded academic research became the Science Research Council, and alongside this the other research councils went to a new Ministry - the Ministry of Education and Science. But the bulk of the DSIR – the National Physical Laboratory, the Food research bodies, the old Fuel Research institutions, the Road Research and Building Research, and more – went to a new Ministry of Technology (MinTech), along with the UKAEA (which had come out of the Ministry of Supply). MinTech was a ministry of atomic energy research and development, as well as a ministry for a number of other major state laboratories, with responsibilities for some industries. It also got NRDC. The most important development was the takeover of the Ministry of Aviation in 1967, which made Mintech by far the largest state spender on R&D. Mintech was a comprehensive industry and research ministry – a Ministry for Business, Energy (it took over the Ministry of Power) and Industrial Strategy (which came from DEA), and on top of that (most) Defence Procurement. Furthermore, it had extraordinary more power and influence. It could decide the shape of British energy, major industrial programmes, and support whole industries. Understanding the Harold Wilson - White Heat - Mintech story is crucial to understanding research policy of the era.57 Central to policy was a reduction in defence R&D; the shift from large scale prestige projects to smaller more commercial ones; the creation of large national businesses capable of innovating seriously; the idea that the state should use the methods and expertise of the supply departments to promote British industry. It did all of these things to different degrees. Perhaps the most important development within Mintech and among the key government advisers was the increasing recognition that lack of R&D, even of the right sort, was not the issue. It was noted that national R&D spend did not correlate positively with national rates of growth (this is still the case, for good reasons, and should be known as the first law of research policy). Despite high R&D spending, the economy was not growing as fast as hoped – lack of R&D spend was not the problem. Attention shifted decisively to the questions of management and industrial efficiency more generally. From the late 1960s, a great sense of disillusion with native British technology started to emerge, though on the political fringes. No British nuclear reactors had been sold since two from the first programme, and no new sales looked likely, despite vast expenditures partly justified on the grounds that they would sell. Concorde did not look as if it would be bought by anyone, and other big civil projects had poor sales. Arms sales too were concentrated, for special reasons, on Saudi Arabia. The upshot was much more scrutiny of research plans. It now had to be established much more clearly that a research programme would really lead to outcomes the nation wanted. One way of ensuring this was creating more of a sense of the government as customer for research, and the research establishments as suppliers of research. Formalising this was a way of trying to get the key questions asked. This is how the Rothschild Report (1971)58 needs to be understood – as the extension within government of what he called the customer-contractor principle, to bring research programmes under more control. And this happened.59 Rothschild also argued that a fraction of the research of three research councils, the MRC, ARC, and NERC, should be deemed applied research and should come under the customer-contractor principle (it turned out to be a total of less than 20 56 David Edgerton, ‘The ‘Haldane Principle’ and other invented traditions in science policy’, History and Policy 88 (2009). 57 The following accounts of MinTech is derived from an ESRC-funded project I directed in the late 1980s – for which see my ‘The ‘White heat’ revisited: the British government and technology in the 1960s’, Twentieth Century British History 7 (1996): 53-82 and for the wider context Warfare State: Britain 1920-1970 (Cambridge: Cambridge University Press, 2005). See also work by Richard Coopey out of the project. 58 Lord Rothschild, The Organisation and Management of Government R&D Included in A Framework for Government Research and Development November 1971 Cmnd. 4814. The scope and aims of this report are widely misunderstood so consulting the original report and associated documents is essential. 59 Framework for Government Research and Development July 1972 Cmnd. 5046, the government response, makes clear that the bulk of government applied research already operated under customer- contractor principles. 37 Science Policy History percent of total research council spend). This led to a continuing unfounded story that Rothschild represented an abandoning of the (mythical) Haldane Principle and a radical change in British research policy. If it was a radical change it was not due to Rothschild, and it barely affected the research councils as whole, though was important for ARC.60 If there was a change in the 1970s, it was a radicalisation of the direction of travel of the late 1960s. It can be seen in the more general alignment of research with department responsibilities. MinTech was broken up in 1970, with the defence side going first through aviation supply into the Ministry of Defence, in a new Procurement Executive. Many laboratories that had been under MinTech were shifted to more appropriate departments (e.g. Road Research to Transport, and food research to Agriculture and Food). The general picture that emerged of the techno-nationalist programmes was a negative one. Failure was explained by a supposed overemphasis on defence and prestige projects, which Mintech in fact countered. But were there successes? At one level yes – the material infrastructure of the UK was transformed, from the post office, to agriculture, to the railways and mines.61 The question is really whether national programmes were the most effective way to achieve these ends. But where there is no doubt is that something large was achieved. There was some return to a stronger industrial – technological strategy under Labour in the 1970s, notably with the creation of the National Enterprise Board, and Inmos (a state-funded semiconductor firm). Indeed, it could be argued that the 1970s saw a flowering of inventive activity, leading to new micro-computers (Acorn and its descendants like ARM), for example, and the blockbuster pharmaceuticals that drove the growth of these firms into the 1990s. Liberalisation However, the general drift of understanding from 1979 was that techno-nationalism had failed. From the early years of the Thatcher governments the shift in policy was clear. Although there was a wideranging and ambitious programme of research in computing (the Alvey programme), there was, especially from the late 1980s, a shift away from near-market research. The government retreated from what were called ‘near-market’ activities. That meant the radical cutting back, privatisation, and contractorisation of government laboratories. Departmental research and development was slashed along with the rationale for major state-led programmes. The entrepreneurial state (which did exist) was killed off. The research councils were not cut anything like as much. The wider context saw radical changes. Getting rid of the imperative to buy British in itself had major consequences, made stronger by the privatisation of the nationalised industries. Where once infrastructural technologies had been innovated and manufactured in Britain, that over time ceased to be the case. That had huge consequences for nuclear energy, aviation, railway, telephone equipment and more. More generally, there was a fundamental shift away from a focus on R&D as the motor of growth to markets and entrepreneurs. The state had no business in business. At a time when foreign car makers were encouraged to come into Britain, a national industrial strategy would have made no sense. The R&D:GDP ratio fell from the early 1960s. It continued to fall through the 1980s, falling under 2% of GDP in the early 1990s, where it has remained. Other countries now spend considerably more, whereas in to the late 1960s the UK stood out as the major European R&D player. The result of the shrinking of departmental research and industrial research was that the research councils loomed larger and larger in the government’s R&D spending. The result was that politicians 60 ‘The “Haldane Principle“ has, evidently, little or no bearing on the conduct and management of Government R. & D. in the ’70s’ said Rothschild, p. 19. ‘Haldane Principle’ was in inverted commas because Rothschild was himself clear that there was no single Haldane Principle. 61 For this argument, see my The Rise and Fall of the British Nation Lost: a Twentieth-Century History (Allen Lane, 2018). 38 Science Policy History and others now looked to the research councils to produce inventions that would generate economic growth; and the research councils began to look to argue that they could do this. There was much talk of entrepreneurial universities, and spin-outs, and of course Silicon Valley. It now looked as if the UK had a uniquely strong ‘science base’ that needed to be exploited by more vigorous entrepreneurship. Industry may have declined, but not academic research, was the assumption; ‘Realising our potential’ was the name of the key white paper. The paradox was that continuing government support was based on the idea that only research far from the market should be supported, and that this research would lead to new products and processes. That research policy has been a substitute for an active industrial policy may be illustrated by the case of graphene, discovered in Manchester and made a huge fuss of. The government set out to exploit this discovery with a £50m centre. But if the potential was as great as was made out, £50m was a paltry sum in a world where £50bn bought a short-distance railway. It is also a tiny fraction of what was spent on Concorde, for instance. Since the 1990s, research policy has been a substitute for industrial policy, not a complement to it. A partial exception has been biosciences, where claims were made for unique British strength in both research and in the industry. However, it is worth noting that this sector has notoriously low research productivity. It is with justification that there is now talk of a ‘biomedical bubble’ in research.62 And the record of success in biotechnology is not great.63 Indeed what is lacking, remarkably so, is any serious assessment of the effectiveness of policy over the last thirty or forty years. Where are the businesses created by the exploitation of British inventions? Of course, finding the evidence would be difficult, but surely it is not impossible to find some major plausible examples?