Trygve Haavelmo (1911–99)

Trygve Haavelmo was born in Skedsmo, near Oslo in Norway in 1911. He studied economics at the University of Oslo and was awarded his degree in 1933. The 1969 Nobel Laureate Ragnar Frisch (see entry in the present volume) was Professor of Economics at Oslo at this time, and Haavelmo started working for Frisch after his graduation at the newly established Institute of Economics. In 1938 Haavelmo moved to Denmark, taking up an appointment as Lecturer in Statistics at the University of Århus. The following year he left for the United States on a Fulbright scholarship and stayed there for the duration of the Second World War, both as a student (supported by the Rockefeller Foundation in 1940–41), and as a civil servant employed by the Norwegian government. Haavelmo completed his doctoral thesis at Harvard University in 1941 and submitted it to the University of Oslo after the war (he was awarded a PhD in 1946); in the interim the thesis was published in Econometrica as ‘The Probability Approach in Economics’. This work (Haavelmo, 1944) was instrumental in the Royal Swedish Academy’s decision to award the Nobel Prize to Haavelmo (Nobel Foundation, 2004). In 1946, Haavelmo joined the Cowles Commission – then in Chicago – before returning to Oslo in 1947 to work on plans for the reconstruction of the Norwegian economy. In 1948 he was appointed Professor of Economics and Statistics at the University of Oslo; he retired from this post in 1979.

Haavelmo is remembered by those who knew him as a modest, even shy man who refused many of the honours that were offered to him, although he was president of the Econometric Society in 1957 and he accepted membership of the Danish Academy of Sciences in 1975. His gracious Nobel banquet speech suggests that it is economists in general that are being honoured by his award, rather than him in particular. Haavelmo was awarded the Nobel Memorial Prize in Economics in 1989 ‘for his clarification of the probability theory foundations of econometrics and his analyses of simultaneous economic structures’ (Nobel Foundation, 2004).

Haavelmo’s doctoral thesis provided ‘the foundation of modern econometric methods’ (Nobel Foundation, 2004). This work proposed a new ‘probability approach’ in attempting to resolve some of the key problems in econometrics that had been previously explored by Ragnar Frisch. The most fundamental of these is the identification problem: that is, isolating causal parameters in economic data. A major challenge here is that a range of different economic models may be consistent with the same data (Heckman, 2000). How are appropriate choices to be made across this range? In the 1930s, Frisch had attempted to uncover causal economic relationships by means of ‘confluence analysis’. This involved running regressions on all possible relationships on a set of variables in order to work towards the true causations lurking in the data (Moene and Rodseth, 1991). Haavelmo, while generously attributing many of his ideas to Frisch, proposed an alternative method: the prior specification of an economic model – formulated in probabilistic terms – that could be used to guide the statistical measurement, collection and interpretation of data (ibid.; Nobel Foundation, 2004). Haavelmo was also the first to clearly distinguish the identification and estimation problems in econometrics (Heckman, 1992; see also Morgan, 1990).

Again following Frisch, Haavelmo stressed the importance for policy analysis of delineating highly ‘autonomous’ structural relationships in economic models (Nerlove, 1990). More autonomous relationships hold under a wider variety of circumstances and are therefore particularly useful for the evaluation of alternative economic policies and institutional arrangements (Christiansen and Rodseth, 2000). It has also been suggested that Haavelmo’s emphasis on the significance of autonomy in economic modelling might have obviated the need for the Lucas critique:5 there would have been no temptation to predict the consequences of alternative policy changes within a macroeconomic model as the behaviour of economic agents cannot be modelled at the necessary level of autonomy (Moene and Rodseth, 1991). Although Haavelmo’s Nobel citation focuses on his doctoral thesis, two other papers in which he extended and applied his insights are also highlighted: Haavelmo (1943) and, with M.A. Girshick, Haavelmo (1947).

According to the Nobel Foundation (2004), ‘Haavelmo’s doctoral thesis had a swift and path-breaking influence on the development of econometrics’. In particular it was adopted by the Cowles Commission as its econometric modus operandi, and directly influenced the work of other Nobel Laureates such as Koopmans and Klein (Haavelmo, 1997; Heckman, 1992).

In the 1950s, Haavelmo turned his attentions away from econometrics narrowly defined and concentrated instead on economic theory. His reasons for this switch – mostly to do with the need to improve theory – were outlined in his 1957 presidential address to the Econometric Society and reprised in his Nobel lecture (Haavelmo, 1958; 1997). The practical outcome included six books, two of which have been published in English: A Study in the Theory of Economic Evolution (1954) and A Study in the Theory of Investment (1960). Haavelmo’s name is also commonly associated with the theorem discussed in ‘Multiplier Effects of a Balanced Budget’, published in Econometrica in 1945. Although he was not the first to highlight the non-neutrality of increases in government expenditure balanced by higher taxes, his proof of the phenomenon was sufficient for it to become known as ‘Haavelmo’s theorem’ (Nerlove, 1990).

Main Published Works
(1943), ‘The Statistical Implications of a System of Simultaneous Equations’, Econometrica,
11, January, pp. 1–12.

(1944), ‘The Probability Approach in Economics’, Econometrica, Supplement, 12, July, pp. 1–
118.

(1945), ‘Multiplier Effects of a Balanced Budget’, Econometrica, 13, October, pp. 311–18. 5 Named after Robert Lucas Jr, the 1995 Nobel Laureate – see entry in this volume. 6 For discussions of the limits to the work of the Cowles Commission, see Heckman (1992; 2000).
174

(1947), ‘Statistical Analysis of the Demand for Food: Examples of Simultaneous Estimation of Structural Equations’ (with M.A. Girshick), Econometrica, 15, April, pp. 79–110.
(1954), A Study in the Theory of Economic Evolution, Amsterdam: North-Holland.
(1958), ‘The Role of the Econometrician in the Advancement of Economic Theory’, Econometrica, 26, July, pp. 351–7.
(1960), A Study in the Theory of Investment, Chicago: University of Chicago Press.
(1997), ‘Econometrics and the Welfare State’, American Economic Review, 87, December, pp. 13– 15.