Oxford Institute of Statistics

The next significant step in the development of Oxford statistics was again by its economists, who were increasingly keen to build economic theory on a foundation of sound data analysis.

The first Director of the Institute of Statistics in 1935 was the econometri­cian Jacob Marschak, who was born in Kiev in 1898 as the son of a Jewish jeweller. Marschak had lived an eventful life in Russia and Germany until coming to Oxford fleeing Hitler. He moved to the USA in 1938 where he had a distinguished career at the Cowles Commission.

During the war years, the Acting Director of the Institute was Sir Arthur Bowley, the distinguished economic statistician who had recently retired from a chair at LSE. Although not primarily a statistician, Michal Kalecki was also housed at the Institute from 1939 to 1945 where he contributed to analysing data on many aspects of the Second World War, publishing in the Bulletin. Hubert Henderson, Acting Director of the Institute at the time, recorded his appreciation for Kalecki when he left: ‘[T]he repute that the Institute has won as a war-time centre of lively, yet scientific and realistic economic study, owes much to your stimulating influence' (Henderson quoted in Toporowski 2018: 141). David Worswick (see Chapter 19 in this volume by Seneca) was at the Institute from 1940 to 1960, but did not regard econometrics favourably, arguing that it made ‘pretend-tools' (Worswick 1972: 79) while trying to achieve Frisch's aims.

The Readership was then filled by David Champernowne, who also became Director of the Institute from 1945 to 1948 and Professor of Statistics from 1948 to 1959, after which he returned to Cambridge where he had read mathematics and then economics, graduating in 1934. Champernowne went on to do research on income distribution, for which he was the first to provide a statistical model. In 1937, this work earned him a Prize Fellowship at King’s College, Cambridge. He continued to work on income distribution for the rest of his academic career (see Boianovsky 2017 for more details).

The Oxford Institute of Statistics then became home to a steady stream of distinguished economic statisticians and econometricians. In roughly chrono­logical order, Frank Burchardt was the Director after Champernowne in 1948, and he helped attract Lawrence Klein, later a Nobel Prize winner. Klein worked at the Institute from 1954 to 1958 during the McCarthy era, and helped develop the first UK macroeconometric model with James Ball, Arthur Hazlewood and Peter Vandome (Klein et al. 1961a). Klein spoke of his asso­ciation with the Institute in its early days in his Nobel Prize autobiography.^[3]

Some of the papers related to Klein’s macroeconomic modelling were pub­lished in the Bulletin of the Oxford Institute of Economics and Statistics, estab­lished in 1939, changing its name in 1973 to the Oxford Bulletin of Economics and Statistics. Ball et al. (1959) published “Econometric Forecasts for 1959” (for the UK) in the February issue of 1959, while the February 1961 issue contained “Re-estimation of the Econometric Model of the UK and Forecasts for 1961” by Klein et al. (1961b). That issue also published “A Post-Mortem on Econometric Forecasts for 1959” by Hazlewood and Vandome.

Next, IES was home to Gerhard Stuvel (see, for example, Stuvel 1965), Christopher Winsten (whose serial correlation correction method in a 1954 Cowles Discussion Paper with Sig Prais became widely cited (Prais and Winsten 1954)), N. Schwartz and John Hammersley (at Oxford from 1961 and whose excellent 1964 book on Monte Carlo methods with David Hanscomb (Hammersley and Handscomb 1964) helped Hendry and Pravin Trivedi develop their 1972 paper: Hendry and Trivedi 1972). They were fol­lowed by a non-econometrician, Teddy Jackson, as Director, then Hendry (who was Director from 1982 to 1984) and Stephen Nickell, who was its final Director from 1984 to 1997.

3.3