Jan Tinbergen was an early pioneer in developing the field of econometrics and building macro-econometric models that could describe business cyclical behaviour as well as stabilization policy and long-term economic planning.

Born on 12 April 1903 in The Hague, Jan Tinbergen studied mathematics and theoretical physics at Leiden University, where he completed a doctorate under the supervision of Paul Ehrenfest in 1929.

The core of his thesis was on minimization problems in mathematics, but he provided an appendix where he also considered economic problems, which was added because of his desire to combine mathematics with his political views on social democracy. Tinbergen was Statistician for Business Cycle Research at the Central Bureau of Statistics (CBS) from 1929 to 1945, and from 1933 he was Professor of Economics at The Netherlands School of Economics, Rotterdam, until his retirement in 1973; but he also served as an expert to the League of Nations from 1936 to 1938 and Director of the Central Planning Bureau (CPB) from 1945 to 1951. In 1969, he shared the first Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel with Ragnar Frisch for “having developed and applied dynamic models for the analysis of economic processes”. Tinbergen died on 9 June 1994 in The Hague, Netherlands.

Tinbergen made several seminal contributions to the mathematical theory of the business cycle and macro-econometric model building while doing business cycle research at the CBS. The CBS was a good place to be as he had access to large amounts of data and was able to test and develop different theoretical models. In his early research, Tinbergen focused on developing an endogenous theory of the business cycle, which included an application of Arthur Hanau’s cobweb-dynamics to agricultural product cycles, and an extension of this analysis to the shipbuilding industry, in which gestation lags in the investment process imply that the dynamics be represented by a mixed difference-differential equation. The method had an important influence on Frisch and Michal Kalecki, who both used it in their later models of the national business cycle.

An invitation by the Royal Economic Association in 1936 to discuss the deteriorating economic conditions and policy problems in the Netherlands prompted Tinbergen (1959) to build and estimate the first large-scale macro-econometric business cycle model. The model was highly novel, containing 24 linear equations, six of which contained lagged endogenous variables that generated dynamics, and a number of non-estimated coefficients. Tinbergen explored several policy scenarios, one of which supported the initiative to abandon the gold parity of the guilder, effectively allowing the guilder to devalue by 20 per cent. Several policy implications were also incorporated into the alternative policy agenda of the Social Democratic Labour Party.

In the same year, Tinbergen was commissioned by the League of Nations to carry out statistical tests of existing business cycle theories. This research, published in two volumes in 1939, was novel in its extensive use of testing procedures (Morgan 1990). The first volume contained an exposition of his econometric method, essentially classical multiple regression analysis combined with Frisch’s confluence analysis, and then applied this method to fluctuations in investment, in general, and in residential building and in railway rolling-stock, in particular. In the second volume, Tinbergen set up a large-scale macro-econometric model of business cycles in the United States that contained 71 variables and 48 equations and covered the period from 1919 to 1932.

A precursor of the giant computer models, the dynamic model was novel in that it could be used to better understand past business cycles as unified, single phenomenon, forecast future trends, and identify policy instruments that can change the direction of the cycle. It was proposed in the preface to carry out a parallel study of the United Kingdom, which Tinbergen carried out in the 1940s and published in 1951.

While these models represent a fundamental contribution to macro-econometric model building, it also sparked a lively debate with several economists, including John Maynard Keynes.

Keynes was very sceptical of the methodology underlying the League of Nations study and published a long review of it in the September 1939 issue of the Economic Journal. Keynes’s main concern was the application of multiple regression analysis to non-homogeneous series in real time, and the consequent problem of independence and misspecification of the economic model. He believed that the complexity and variability observed over the cycle meant that some influences on the cycle could not be reduced to statistical form, such as long-term profit expectations. Keynes elicited a response from Tinbergen who maintained that his model was comprehensive and contained the most important variables, other influences that changed over the course of the cycle were included in the residual variable, and that it is possible to base expectations on the past and thus be extrapolated. He also maintained that explanatory factors only needed to be uncorrelated in statistical relations rather than independent in an economic sense. Overall, Tinbergen (1940: 154) rejected Keynes’s sceptical view of econometrics, not because he considered his criticism irrelevant, but because he believed that the method “promises - and actually yields - much more than Mr. Keynes thinks”. Tinbergen received a life subscription to the journal for his contribution, for which he later commented, “I’ve never been paid that generously” (Magnus and Morgan 1987: 129).

The Keynes-Tinbergen debate marked an important milestone in the development of macro-econometric model building, but the urgent need to reconstruct much of Europe after World War II made it essential to the development of a theory of economic policy. As the first director of the CPB and a key policy adviser to the Social Democrats, Tinbergen was highly influential in shaping economic policy of the Netherlands, but he was also essential to the development of the methodology underlying policy-making. Tinbergen’s (1952, 1956) novel approach to the problem was to combine large-scale macro-econometric models with instrumental analysis.

What he had done was to invert the economic problem so that the desired goal or policy objective becomes the datum of the theory, and the objective analysis is to identify the economic means suitable for the attainment of the desired macro-goal (Lowe 1965). Variables in these types of models can either be controlled (instruments), whose values are chosen directly by the policy-maker given certain constraints, or be uncontrolled, whose values are only influenced indirectly by the policy-maker given their choice of instrumental variables. Some of the variables are defined as the targets, which is what the policy-maker wants influence through the use of policy instruments. Any number of policy instruments can be used within a stated economic policy, but consistent policy requires that the number of independent instruments must equal the number of policy targets.

Policy advice based on the large-scale structural macro-econometric models received much criticism from the new classical economic perspective. The “Lucas critique”, named after Robert Lucas (1976), focuses on the Tinbergen model, particularly on the way it represents expectations, because it does not consider the dependence of individual behaviour on perceived government policy rules, potentially misleading policy-makers as to the effectiveness and desirability of their policy choices. Lucas maintained that these models were not adequately grounded in dynamic economic theory and suggested that a theory of economic policy should always start from the basic neoclassical parameters that govern individual behaviour, namely, preferences, technology and resource endowments. The issue of policy neutrality is at the centre of the Lucas critique, with instrumental analysis its main target, but the solution it offers is to require overly restrictive assumptions about individual behaviour, which may not be empirically relevant, as Tinbergen would point out.

Tinbergen continued to do policy-relevant research for the rest of his life.

After the mid-1950s, most of his research centred on the problem of economic development, developing and applying his modelling methodology to low-income countries. Tinbergen (1968) developed a novel multistage approach to development planning, based on the Harrod-Domar growth model, which involved private and public decision-makers at the national, sectoral and project levels. These models focused not only on the need for physical capital and a transport infrastructure but also on the need for education and skill development. Tinbergen (1959) also made original contributions to the theory of income distribution, focusing on the issue of the distribution of personal income well after retirement. Throughout his life, Tinbergen was concerned with economic and social problems, placing particular importance on empirically based and policy relevant research. For this reason it is difficult to place him in a particular school of economic thought, but instead to consider the important role he played in making economics more empirically relevant.

Mark Knell

Jan Tinbergen was an early pioneer in developing the field of econometrics and building macro-econometric models that could describe business cyclical behaviour as well as sta­bilization policy and long-term economic planning.

More on the topic Jan Tinbergen was an early pioneer in developing the field of econometrics and building macro-econometric models that could describe business cyclical behaviour as well as sta­bilization policy and long-term economic planning.:

Jan Tinbergen was an early pioneer in developing the field of econometrics and building macro-econometric models that could describe business cyclical behaviour as well as stabilization policy and long-term economic planning.

More on the topic Jan Tinbergen was an early pioneer in developing the field of econometrics and building macro-econometric models that could describe business cyclical behaviour as well as stabilization policy and long-term economic planning.: