Friedman’s Methodology: A Puzzle and A Proposal for Generating Useful Debates through Causal Comparisons (with a postscript on positive vs. normative theories)

摘要

Milton Friedman’s “The Methodology of Positive Economies” is still one of the most widely read pieces on economic methodology. One reason for this might be Friedman’s attractive proposal that economists use theories and hypotheses as pragmatic devices to summarize data and make predictions over the relevant range of observations. Logically, this should lead to a fair minded comparison among many contending theories. However, Friedman's actual examples and discussion of these examples raise a puzzle. The field of comparison seems unduly narrow from the beginning. In my attempt to resolve this, I consider some logical and ontological problems for Friedman's position. I end up by suggesting a scientific realist approach to testing theories by causal comparisons over a wide field of contending theories.