Recently my main concern has been to reread the most important new classical texts. The texts which can be regarded as the most important items from a methodological point of view. Now I am working on a paper in which I will look into the debate between the realist and antirealist camps and, which is more important, I will study whether new classical macroeconomics has really been an instrumentalist movement.
For me, monetarism, under which new classical macroeconomics as a radical wing can also be sorted, has never been a homogenous theoretical system as far as the methodological principles are considered. This is problem is too difficult to solve easily. In the literature, instrumentalism is identified typically with the famous methodological paper of Friedman (F53), but even Milton Friedman himself never seemed to be able to insist consistently on these guidelines. However instrumentalist he is, he seemed to have some other, realist efforts. For me, the main point of F53 is the example of the model set up for estimating the density of leaves on a tree: this is a clear instrumentalist manifesto. However, his theory on the permanent income with its idea of the economic man trying to optimize on longer time horizons is something which was set up in order to grab the very essence of actual (everyday) optimizing behaviour. This is realism. New classicals, however, with their clearly defined idea of the rational agents always and clearly made up a realist movement in my interpretation.
One can find interesting methodological papers which use the new classical models to introduce and describe the instrumentalist methodology of theorizing. This is a serious mistake. So, getting back to the first paragraph, I am making efforts to prove: although new classical macroeconomics and the Friedmanian orthodox monetarism can (and should) be sorted under the same label (‘monetarism’), because their main economic policy recommendations were common and their efforts to place macroeconomics on microfoundations were also shared (following different lines, though), in methodological terms new classical macroeconomics has been a far cry from orthodox monetarism. In order to draw this distinction and to prove that new classical macroeconomics has been a realist program, I am using the well-known realist-instrumentalist opposition. Traditionally, realism, as a rule of thumb, is said to tend to provide theoretically sound, but empirically poor models, while instrumentalism puts emphasis only on the empirical performance. In my opinion, the realist efforts of new classicals were directed at setting this opposition aside. They tried to set up models which were realistic in that they were aimed at grabbing the essence of reality, i.e. the fundamental economic laws (or some of them, at least), while, for new classicals, empirical performance was always a similarly important purpose, supported by the variety of complex mathematical and econometrical methods. In the case of new classical macroeconomics, realism was not accompanied by poor empirical performance, so one was no longer forced to choose between realism and instrumentalism.