I was very pleased to be invited to the Universidad Carlos III as visiting professor, to meet all my old Spanish friends again, and to have the opportunity to give this public lecture. My address will have two themes: a brief review of the history and current status of the London School of Economics, and then a discussion of the strong tradition which has developed there of the study of mathematical methods in the social sciences, particularly in economics, by building mathematical models of economic behaviour, and then estimating these models statistically. That briefly is how I define econometrics.

The London School of Economics is a great British and international institution. I will stress the international aspects of LSE later. But great institutions are formed by great men, and academic traditions are formed when a group of scholars are responding to a common vision, successfully carry through their vision, so that other good people are attracted to carry out and maintain the tradition indefinitely. A good teaching group attracts good students; a good research group attracts good faculty. I intend to talk about the people who got together in the 1960s and 1970s to make LSE one of the leading universities in the world for mathematical economics and econometrics, and some of their successors, and, so as not to be exclusive, some of the more interesting faculty in the economics department of the same period. The claim that I just made for a leading position for LSE in econometrics and mathematical economics was supported by American studies of the publications record of different universities since, during the 1980s, LSE was top of both rankings.

Starting my brief history at the beginning, the London School of Economics and Political Science was founded by members of the Fabian Society in 1895, in the belief that academic study of the social sciences could only improve the efficiency of the ways in which we conduct our affairs. The LSE grew fairly slowly in its first 30 years, but in 1925 it had 500 students and it had acquired the lease of land from the Greater London Council and had built its current main building in Houghton Street. It now covers a wide range of social sciences, having departments of philosophy, geography, history, statistics, foreign languages, law, anthropology, social administration, sociology, social psychology, industrial relations, international relations and international history, political thought, government, and economics. In all these last ten subjects, except economics, LSE pioneered the teaching of the subject in British universities. LSE early became a part of London University, which now has 18 separate teaching colleges, of which 5 are large with 80% of the total of over 25,000 students at the University. LSE is among this group of 5 large colleges, and is certainly one of the most important in terms of academic prestige.

During the 1930s the department of statistics was founded under Professor Bowley, who pioneered the use of sampling methods in social surveys, and carried out in the 1930s a large survey of London poverty, which could be compared with an earlier survey by Charles Booth at the beginning of the century. He developed a strong tradition in statistics at the School. He was followed by Roy Allen and in the late 1940s Maurice Kendall succeeded in building up the statistics department so that in terms of the number of faculty it was the largest in any British university.

The economics department had in the late 1930s a strong group of economic theorists led by Lionel Robbins, and with the arrival of Hayek became associated with the Austrian theories of capital and the trade cycle. John Hicks, who taught at LSE in this period, was Britain's first Nobel Prize winner in economics. Willian Beveridge was a leading applied economist who became Director of the School in the late 1930s, and prepared there his Report on Unemployment and the Social Services, which was adopted by the British government as the basis of our welfare state system in 1944. Incidentally, Nicolas Kaldor was at the School and was responsible for some of the financial forecasts contained in the report.

With the resumption of peace, the School grew quickly, reaching 2,000 students in the late 1940s. Our professor of international trade at the time was James Meade, who in due course became the second Englishman to win a Nobel Prize in economics, and other notable economists of that period were Henry Phelps-Brown, and Paish.

I moved to the London School of Economics in 1963 when the school had about 3,300 students, about 2/3 of them undergraduate, and about 2/3 of them British. Throughout the last 27 years the numbers have grown so that last year the total number of students was over 5,000, about half of them undergraduate and half of them British. In those early days about 80% of the LSE revenue came directly from the government, 9% in students' fees, and student fees were £250 per annum for all students whether undergraduate or graduate, British or foreign. Contrast this with last year's figures when government grants were 40% of the total revenue, student fees were 30%, specific research grants were 7% of the total, and student annual fees were £1,775 for British and EC undergraduates, £6,075 for other overseas undergraduate and Ph.D. students, £2,104 for British and EC postgraduates, and £6,270 for M.Sc. students.

These changes have taken place against a background of university growth in Britain which was initially rapid. Since the Robbins report (produced by a government inquiry headed by the long period ex-chairman of LSE's economics department) had suggested a doubling of student numbers and a 50% increase in the number of British universities, an increase of the number of universities had duly taken place, and student numbers had grown apart from a setback in the early 1980s from which we are now recovering. The total number of full time students in universities increased from 126,000 in 1964 to 350,000 in 1989, a change from 3% of the relevant age group to 7% now. The government is about to increase this proportion drastically by allowing our polytechnics, which are higher education colleges controlled by our local government authorities, to become universities. This will increase the number of students in universities by some 75%. The government is intending to “increase the efficiency of teaching in universities” by indeed increasing the number of students without increasing the number of faculty. It has already had some success in doing this by financial pressure, as can be illustrated by the fact that our student/faculty ratio was 10.5 in 1975 and is now 14.9.

The government, which previously gave subsidies to each university proportional to the number of different students, has put a quota on its total subsidy to each university, and has encouraged universities to take more students than the quota, particularly by being more generous with capital grants to universities which will do this. Government think-tanks are also discussing the idea of two year eight-term bachelor degrees, and some universities are planning such degrees, with the incentive that some government grants will be related to the number of degrees granted rather than to the number of students taught. Also the government has suggested introducing a three tier system in which the lowest level of university would be undergraduate-teaching-only, the second level would have some graduate teaching and research, but only the top level would have full access to government research grants. The constitution of the government research councils has been modified so that the government ministries have more control, and the grants go primarily to long period research institutes concentrating on topics which are chosen for their relevance to major problems affecting public policy.

At the same time in pursuit of greater efficiency the government has attempted to concentrate teaching in particular subjects at only a few institutions so that the size of classes will be maximised, and the amount of student choice in each institution in a given subject field should also be maximised. This has meant that over the last ten years and probably over the next ten years rapid change has occurred and will occur in our universities. This has provoked great controversy and a considerable resistance by most academics at most universities, even at Oxford and Cambridge. Fortunately LSE seems well-equipped to survive through these troubled times, since it is specialised to the social sciences, which British students seem to prefer to physical sciences or technology, and has sufficiently large student numbers to provide specialist courses and to maintain reasonably large class sizes, and has sufficient reputation to attract large research funds, both private and public, as will be described in more detail later. On the teaching side a recent innovation at LSE has been the development of an international summer school in economics which is now attracting more than 500 students willing to pay teaching fees of £625 for a three week intensive course.

In earlier periods we had student troubles in the late 1960s, which were on the same scale as at other British universities, but rather exaggerated by our tabloid press. Indeed, LSE has always suffered by comparison with Oxbridge from the belief by some uninformed members of the British public that it is dominated by Marxists, particularly among headmasters directing their school-leavers to Oxbridge. The better informed remember that LSE was the home of Hayek, Robbins, and Popper as well as of Laski, Dalton, and Tawney.

I will now turn to my second theme of the development of econometrics at LSE Jim Durbin and Bill Phillips were my primary reasons for moving to LSE in 1963. Jim Durbin was a fellow student of mine in the same college in Cambridge University in 1942, when we were studying for our B.A. degrees in mathematics. I met him again in the late 1940s when he was working on the research leading to the well-known Durbin–Watson statistic. In 1963 he was already a professor in the statistics department at LSE, teaching various courses in statistical theory and in statistical time series analysis. He was always interested in econometric theory. Back in 1965 he wrote a pioneering paper on new methods of estimating simultaneous equation models, and arranged research funds to program his method on the London University computer. He was throughout his long career in the forefront of devising new statistical tests for the correct properties of estimated relationships, and of approximating more accurately the distributions of complicated econometric estimators. He was an enthusiastic teacher always trying to develop better courses and examinations for our statistics and econometric students.

I knew Bill Phillips early through his famous hydraulic machine, which was an attempt to help students to understand the Keynesian model of the economy by the use of transparent plastic pipes, tanks and colored water propelled by electric pumps. For a short period about 1950 every university in the U.K. and quite a few in the U.S.A. had a Phillips Hydraulic Model. But they were difficult to maintain, and rather limited in what they could model. However in the last year LSE has repaired two models as museum specimens, and one is now on display at LSE and the other at Auckland University in New Zealand. When I arrived at LSE his name had become world-famous for his invention of the Phillips curve, with its recognition that one of the fundamental factors in inflation must be the change in the level of wage rates and its determinants. Phillips was also a pioneer in the study of ARMA models and the application of control theory to public policy decision making. (In the 1970s, a junior minister in the current Labour government was so enthusiastic about this idea that a Royal Commission of enquiry was set up with Jim Ball as chairman, and James Mirrlees, Gwillym Jenkins of Box–Jenkins fame and myself as members. We agreed that with the shaky performance of most large scale macro-economic models and the necessity for democratic participation in public policy decisions, the use of control theory was not feasible to control government financial or fiscal decision-taking). Personally, Phillips was very enthusiastic and friendly, and very persuasive. He and the chairman of LSE's economics department were responsible for persuading Terence Gorman and Frank Hahn to join us in 1966. At the same time they were largely responsible for setting up the conditions under which Harry Johnson was to be half time professor both at LSE and at the University of Chicago. This was a unique arrangement for the University of London. Never before had a full time professor at the University of London held a joint appointment at another university. LSE had him for the first two terms of the academic year; the University of Chicago had him only for the third and fourth quarters, including the summer quarter from July to September when relatively little teaching was done. Certainly, in two terms Harry Johnson was quite capable of taking charge of our Master in Economics course, and of playing a major part in the teaching of International Economics in the department. He was a very prolific writer of economics. For many years he led the American Economic Associations statistics for the number of pages published in a year. Towards the end of this period in the late 1970s he was also responsible for a major research project on international trade and finance. He and my friends and colleagues were pioneers in building monetarist models of inflation and foreign trade.

Harry Johnson was a strong believer in simple market mechanisms, and given to describing over-critical approaches as “nit-picking,” a man who whittled small wooden objects whether talking in seminars, eating or drinking, and the only professor in the economics department who could afford to live near the School, and from his flat to dispense hospitality to staff, students and particularly visitors. He was really responsible for putting our M.Sc. in Economics on a sound basis.

In the econometrics field, Gorman and Hahn were unique in moving to the LSE from Oxford and Cambridge, where they were already well established. Gorman was already professor at Nuffield College, Oxford, in 1966, and Hahn a leading member of Churchill College, Cambridge. They were friends of long standing; they, Alan Walters and John Wise were sometimes referred to as the “Birmingham mafia” as they had all been teachers or students at Birmingham in the early 1950s. Undoubtedly, Hahn and Gorman only moved to LSE to be reunited together in the hope of founding a pioneering school of econometrics and mathematical economics, at a time when both Oxford and Cambridge were uninterested in these fields. They, together with Bill Phillips and myself, provided the critical mass sufficient to attract good students and faculty from all over the world to take and teach our courses, particularly in our new M.Sc. in Econometrics.

Gorman was one of the friendliest academics that one could imagine. He had great charm, and general learning and conscientiousness. He attended every seminar in econometrics and mathematical economics, and was a great spontaneous discussant of any topic. He had a soft Northern Irish accent, of which he was profoundly aware, sometimes apologising for the difficulty his audience had in understanding his rather slurred delivery. But he could apply his special techniques of aggregation, and his dual approach to utility and cost functions, to a wide range of theoretical and empirical problems, and he had an extremely wide range of interest over a wide range of applied topics. His technique of solution of aggregation problems seemed wonderfully simple to Gorman, but not to others. But he was particularly good on appointment and promotion committees, always recognising talent or promise of talent, and very good at drawing out shy people at interview.

Frank Hahn was always confident, vigorous and witty in discussion. He was Socratic in his lecture style, expecting students to answer his questions spontaneously and without preparation. He arrived at a time when LSE was not well equipped with mathematical economists, so much so that I had been teaching courses in Mathematical Economics to our M.Sc. students. Frank was chairman of the departmental committee to consider the structure of the B.Sc. (Economics). This led to a wholesale revision of the B.Sc. (Economics) which up to that point had been a London University degree in the sense that all London colleges that had taught that degree had the same regulations and examinations, and the degree had been rather general and inflexible especially in the first year, in that all students had to take the same courses in economics, in political thought, and in history. After Hahn, we had our own LSE B.Sc. (Economics). All students who intended to take economics as a specialism had in their first year a more advanced course in economics, and a course in mathematics and statistics, and students intending to major in econometrics and mathematical economics were allowed to specialise much more in mathematical and statistical courses. We were able to ensure that all our economics students took a course in elementary econometrics using standard statistical computing routines drawing from a standard data base to carry out small econometric projects.

Hahn certainly reorganised our courses on mathematical economics and contributed decisively to our ability to attract first rate economists such as Michio Morishima, Amartya Sen, Partha Dasgupta and Oliver Hart to LSE. However he was somewhat of a sceptic about the value of econometric modelling. He would often challenge me to name one applied econometric study which had decisively changed our understanding of the working of the economy. However he did later appreciate the papers of David Hendry criticising the work of Milton Friedman.

Michio Morishima came to us in 1970 from the University of Essex, having previously decided to move from Kyoto University. Although an outstanding mathematical economist, he had made some interesting applied econometric studies of the Japanese and American economies before he arrived at LSE and he taught a very good introductory course to first year students intending to specialise in economics for several years. He also taught a rather mathematical course on marxist economics intended for students taking our general M.Sc. (Economics). He contributed greatly to our economic and econometric research by persuading two old school-friends who were also millionaires to contribute in 1978 to the Suntory-Toyota International Centre for Economics and Related Disciplines on such a scale that now more than £1M per year is available for research into underdeveloped countries, on taxation, monetary topics, including the history of the World Bank, Japanese and British industrial relations and management, mathematical economics and economic theory, econometric modelling of world trade and finance, and so on. Upon his recent retirement from his professorship at the School, Michio's part in starting the International Centre was commemorated by having his portrait displayed in the lobby of the Centre, which was painted by his own son.

In the early 1970s we acquired a professor of money and banking who was later to become the best known professor of economics to the general public of the U.K. This was Alan Walters. He came to us from Birmingham University with a background in applied econometrics and transport economics. He was appointed to our chair of monetary economics, never having published in the monetary field, but with a promise that this was a field which he would make the centre of research and teaching. During his stay at LSE his principal research output was in connection with a government commission into the Third London Airport. He was responsible for the research activities of the commission. Alan Walters left LSE for the World Bank and Johns Hopkins University, Baltimore, just before our winter of discontent in 1979 when the Labour government was having serious difficulties in controlling the rate of inflation. But within three years he had returned to be Margaret Thatcher's principal economic adviser, establishing a strong personal relationship with the prime minister, and incidentally establishing a record level of salary for such an adviser. He encouraged her to refuse to join the ERM, and he was strongly critical of the chancellor of the exchequer, of the credit boom which was allowed to develop in 1987 as a consequence of the stimulation applied by the government just before that year's election, and of the subsequent high interest rate policy. He also strongly criticised the high exchange rate at which the government entered the ERM. All these criticisms have made Alan Walters unpopular with the present government and we do not expect to see him in London so frequently as in recent years.

Meghnad Desai joined the faculty of the econometrics subdepartment at LSE almost as soon as it was set up in 1965. He was always the most open and friendly of colleagues. He was interested in an immense range of topics from cliometrics to marxist economics, and taught at all levels in the department from first year economics courses to Ph.D.s in applied econometrics. This, in fact, made it difficult for him to get a promotion since it was difficult to find referees who appreciated his whole publication list. However, in the course of time he became full professor, finished his three year period as departmental chairman in 1989, and was then in 1991 made a life peer as a spokesman for the Labour party on economic matters. Since he is universally popular throughout the School, this was an occasion for great rejoicing, and to celebrate the School threw a street party in Houghton Street. As far as I know, this is is the only occasion that an official street party has been held at LSE. Meghnad has taken to being a lord very enthusiastically, and spends the majority of his afternoons at the House of Lords; at other times he is very busy running his Institute of Development Studies.

The economics department recruited one of its most distinguished members in 1980 when Tony Atkinson joined us as chairman of the Suntory-Toyota Centre from 1981 to 1987 and directed many interesting studies of the determinants of income distribution, of public policy issues, and of microeconomic policy issues. He, in turn, led to the recruitment of our newer group of professors at LSE, namely, Mervyn King, Nicolas Stern, and Kevin Roberts, all leading men in their various fields.

However, I now turn to Richard Layard, who first became involved at LSE back in the early 1960s as a research assistant to the Robbins Commission on Higher Education, and has since then become on an ever greater scale the organiser and inspirer of major research projects at LSE. In the eight years up to 1990 he had a government funded centre for Employment Studies which carried out important applied econometric studies of the labour market and unemployment, notably with Steve Nickel and John Hickman. In 1990 a further 10 year grant for £8.6M led to the establishment of the Centre for Economic Performance, which will study the factors leading to economic success on an international scale, considering both European, American and Japanese data, and particularly looking at the problems of Eastern Europe. Layard is principal adviser to the present Russian government and may find it an onerous position. Personally, he has a firm but modest demeanour and runs his large enterprises with a remarkable combination of energy and flexibility.

At LSE we have contributed to the work of the Econometric Society since the 1960s on a large scale. In fact, since 1970 six of the presidents of the society have been teachers at LSE, and 15 of its fellows have either taught at LSE or been graduates from LSE for six years from 1976 to 1981. I was responsible for arranging the annual European meeting of the Econometric Society held in different European capitals, notably the World Congress of 1980 held in Aix en Provence, when 150 sessions were squeezed into six days.

My own research since 1963 was concerned, on the theoretical side, with autoregressive error structures and model specification problems, and with various problems in approximating the distribution functions for various estimators. On the applied side, I was concerned with problems of inflation, particularly with wage and cost inflation. Over this period I supervised more than 40 Ph.D. students, and taught courses on econometric theory and applied econometrics in our M.Sc. (Econometrics).

I will only mention two of these Ph.D. students. David Hendry was undoubtedly the best of my first set of Ph.D. students supervised in the late 1960s. His subsequent career has shown his ability to develop ingenious methods of specifying economic models so as to combine flexibility with appropriate significance testing in the context of complicated dynamic simultaneous equation models. His methodology has undoubtedly affected applied econometric work throughout the world, and has led to a vast upsurge of related new theoretical developments. Of the remaining students, I will pick out only one further name, that of Peter Phillips, now at Yale University, who is undoubtedly the leading world theoretical econometrician, developing many fields of theory, by particularly the attempt to improve the properties of our estimators or significance tests by developing finite sample theory, or by producing better approximations to the sample distribution functions.

After Jim Durbin and I retired from LSE, Peter Robinson in the economics department and Andrew Harvey in the statistics department worthily upheld the traditions of the School in econometrics, both very productive in research and attractive to good students as teachers and supervisors.

During the last fifteen years, LSE has appeared threatened by the British government's increasing pressure to tighten university budgets and to expand student numbers without more finance. We have been extremely lucky to have two directors, Ralf Dahrendorf and I.G. Patel, both with strong personalities and great sympathy for the true ideals of preserving academic standards and freedom, who have successfully resisted these pressures, have preserved the morale of the School through this difficult financial period, and have strengthened the quality of the faculty, when no expansion of numbers has been possible.

We now have a new director, John Ashworth, with relatively little background in the social sciences, being a natural sciences professor originally, but with a strong background in university administration. He and a considerable group of LSE academics favour a bold policy of expansion for the School, accepting a doubling of numbers over the next 20 years, seeking a new home for the School across the Thames in the South Bank building of the old Greater London Council, modifying the structure of our degrees to make a unified school of social science, stressing the importance of hard as compared with soft methodologies, and the importance of making stronger links with industry and government both in teaching and research. These are rather general concepts, and naturally the School has set up committees trying to put some detail on a plan of growth for the next 30 years. The probability of the School taking over the old Greater London Council building has been improved by the result of the last general election, since the planning authority at Lambeth Borough Council, which is the Labour authority controlling the area where the building is located, will no longer be motivated by the hope that a Labour government might reconstitute a Greater London Council at some future date. But the whole problem of the financing and the logistics of the move remains extremely difficult, even although the Greater London Council immediately before it was abolished gave the land to the LSE, which we had been leasing from the G.L.C., so that we can sell the land and buildings currently occupied by the School to help to finance the move.

In 1995 LSE will be celebrating its centenary by organising a week's conference on present and future developments in all the social sciences, using London's largest conference centre to invite 2,000 social scientists from all over the world to participate. One of the greatest strengths of LSE has been its openness to the whole world, its eagerness to attract good students and faculty from wherever it finds them. The conference will be an opportunity for us to strengthen the ties that bind us to all our international colleagues, and to re-iterate our interest and encouragement for our friends and fellow academicians from all the world, and particularly from Europe, to take part in our studies and to interact with our faculty. As we look back in 1995 at the previous 30 years, we will find that econometrics has been making an ever increasing contribution to our understanding of the economy and to the solution of our economic problems. We have learned a great deal in these years as we have accumulated more experience in building mathematical models of the economy, computing power is still increasing exponentially at a very rapid rate, and the very computerising of business and government administration is leading to the collection of ever more detailed data. The ever deeper analysis that this leads to must lead to greater understanding. Of course, greater understanding does not necessarily lead to better public policy, but a clearer presentation of the options available may lead to better decision-making and less political conflict. In all these developments, I am confident that econometrics will be supremely important and that my colleagues at LSE will play a leading part.