Richard Stone and the Department of Applied Economics

The idea for a DAE was strongly supported by Keynes in November 1939, but it was only established after the Second World War in 1945. There had earlier been an initiative known as the Cambridge Research Scheme, of which the chairman of the steering committee was Keynes and with funding from the

National Institute of Economic and Social Research. Stone was involved with this Scheme, as were a number of other Cambridge economists. In November 1944, the Directorship of the DAE was offered to Stone, who took up his post in May 1945. Initial funding came from the Rockefeller Foundation prompted by Keynes.^[1]

In relation to the programme of research to be carried out by the DAE, Stone said that, ‘My idea was to set up an econometric program which would embrace work on facts, work on theories, and work on econometric and statistical methods needed to analyse the facts in the light of the theories’ (Stone quoted in Stone and Pesaran 1991: 100). This was an ambitious programme and the extent to which the DAE carried it out under Stone is very impressive. Part of the success was Stone’s wide range of international contacts, which made it easy for him to invite them to spend time in Cambridge (see Smith 1998 for details). Among those who visited were Afriat, Anderson, Brumberg, Cochrane and Orcutt, Cramer, Duesenberry, Farrell, Geary, Houthakker, Prais, Prest, Tintner, Tobin, and Watson.

Earlier sections in this chapter discussed Cambridge contributions to data collection and theory, but did not tell the full story of the contribution to quantitative analysis. This may be divided into extensions and improvements in quantitative techniques and the carrying out of applied quantitative studies.

The extensive use being made of time series data led to concerns about problems of autocorrelation, and this was explored initially by Cochrane and Orcutt in work that led to an early correction for first-order autocorrelation (AR(1)) (see Cochrane and Orcutt 1949). This work was then extended by Durbin and Watson (1950, 1951), who derived a bounded test for AR(1) that became the standard for many years. The increasing use of variables defined in terms of logarithms led to work by Aitcheson and Brown (1957) on the lognormal distribution. In addition, Stone et al. (1949) did work on sampling and Roy (1950, 1951) examined the distribution of income.^[2] On the applied side, there was important work by Prais and Houthakker (1955) analysing family budgets, and work continued on the Growth Project.

In 1955, Stone returned from a sabbatical at Johns Hopkins University to a new position, the PD. Leake Chair of Finance and Accounting and an end of his Directorship of the DAE. Deaton (1992: 486) suggests this was a painful event:

However, the Cambridge Keynesians (not Keynes, who had died in 1946) arranged matters so that Stone had to give up the DAE in order to take up the chair, though he made it a condition that he would retain a research group within the department. In his good-humoured way, Stone would later say that this was all for the best — his research was not restricted, and he was relieved of administrative burdens — but there is no doubt that losing the DAE was a source of much sadness.^[3]

The new Director was Brian Reddaway, and he was less interested in econometric research. Smith (1998: 99-100) suggests that ‘He shared the general Cambridge doubts about the applicability of econometrics to complex economic processes, where numerous factors, many unmeasurable, interacted and relationships were likely to change through time’. While applied econometrics continued within the Growth Project, Reddaway’s interests led to important work on foreign investment and its effect on the domestic economy and on the Selective Employment Tax that had been proposed by Nicholas Kaldor.

The material presented so far shows that researchers working at Cambridge made major contributions to the development of all three components of quantitative analysis. Smith (ibid.: 88) suggested that there was a growing interest in the history of econometrics and that:

In this history the DAE plays a central role. Epstein (1987, p. 142) says that ‘the influence of this English work was comparable to that of the Cowles Commission in establishing directions for econometric research.’ The English work was initially done at the DAE, but Epstein, like Gilbert (1986), sees leadership of the English school of econometrics leaving Cambridge for London in the 1950s. Like Cowles, after a period of developing econometric methods the DAE turned its attention to other topics.^[4]

• [1] See Stone and Pesaran (1991: 96—105).
• [2] Some of the research was made possible by work going on at Cambridge to develop digital electroniccalculators that began to make large-scale number crunching exercises possible. See Smith (1998: 98—99and the references given therein) for more details.
• [3] However, when interviewed by Deaton and asked about how giving up his Directorship of the DAEhad affected his work, Stone replied: ‘Very little: it really worked out for the best from my point of view.At first I was a little put out as I had come to think of the Department as mine’ (Stone quoted in Stoneand Pesaran 1991: 105).
• [4] This is a convenient spot at which to end this historical survey of the many contributions to econometrics made by Cambridge. What followed the move to London will be explored in a chapter ‘LSE andeconometrics’ to be included in the forthcoming Palgrave companion to LSE economics.