My new working paper “Economic Science and Political Influence” is available as a working paper at CEPR, TSE, IZA and PSE. It is the text of my Malaga Schumpeter lecture and extends the work I have done over the last two years on ideological bias in macroeconomic modelling. In particular, I discuss:
- How the resources that the government will want to spend on measurement depend on the theory itself, and how this may lead to conflicting views between the government and interested experts about the optimal extent of measurement. Also, there exists a measurement multiplier: the more the theory says a variable is important, the more the government wants to react to it, and the more precisely it wants it to be measured. But the more precisely the variable is measured, the more it is optimal to react to it. Hence the additional feedback from the adjustment of measurement.
- How competition between different models in affecting outcomes may lead to polarized views over some parameters, and consensus over other parameters (and in this case consensus arises because intellectual competition elicits the true value of the parameter).
- How the attempts to maintain political influence may lead to “degenerative research programmes”: in the face of new contradictory evidence, the interested economist prefers to increase the complexity of the model so as to preserve the perceived value of some parameters that are key in delivering the policies he likes, rather than discarding the model and replacing it with another one.