## ABSTRACT

In Chapter 4, we looked at the rise of a new way of carrying out economic research, the building of empirical economic models. To illustrate it, we examined the business-cycle research of the CBS and the work of Jan Tinbergen. We saw a shift from inductive statistics to an approach in which a model is constructed based on conjecture about the underlying mechanism that produces a picture matching the shape of the statistical data collected by the CBS. In his model, Tinbergen tried to give shape to this mechanism with the help of insights provided by economic theory. He then produced the equations in his model based on statistical data. This use of statistical data might be seen as a test for the model, but soon one of the most important matters under debate was precisely the degree to which Tinbergen’s econometric evaluations could be regarded as a test. In this chapter, we shall look at John Maynard Keynes’ criticism of Tinbergen and see that Keynes regarded even Tinbergen’s way of working as inductive statistics. We will examine Keynes’ criticism against the background of his own approach to economics, which was still indebted to that of his nineteenth-century British predecessors. In Chapter 6, we shall take a more detailed look at Milton Friedman’s ideas about the testing of econometric models, and about the testing of economic theory more generally. The vocabulary of evaluation and testing that was introduced with the rise of econometrics changed the playing field and the character of the discipline of economics forever. In earlier chapters, we saw that there had already been moves to identify statistical data as empirical evidence, as illustrated by the work of scientifically orientated economists such as William Stanley Jevons. From this point on, economic theory becomes mathematically formulated and needs to be evaluated and tested against statistical datasets. The playing field of economics as a discipline is described in mathematical, quantitative and statistical terms – anything that cannot be described in such terms no longer belongs in the science of economics.