Kevin Williamson in National Review writes Economics Lessons Unlearnt
More likely, the science writer David H. Freedman is correct when he argues that “economic models are always wrong,” that the process of calibrating models to account for historical and current data renders them very, very precise and very, very useless. This is another example of the mapmakers’ dilemma: In order to be manageable, models must, by definition, be simplified versions of the real world, and the more simplified they are, the less accurate they are. There is no economics version of the all-seeing Laplace’s Demon. Professor Krugman’s demand — “Show Me the Model!” — assumes the usefulness of the model, which cannot, in fact, be assumed.
There is no obvious reason to believe that successful economic policies are transferable from country to country. Indeed, it may be that “successful” policies in the main merely coincide with happy economic outcomes rather than causing them. There is not much reason to believe that they are transferable from year to year, much less from generation to generation, in spite of the moanings of our phantom armies of New Deal romantics. Economic conditions change very quickly and in ways that are impossible to predict.
The first issue of evidence we should consider is whether there is any evidence that large, complex economies such as those of Japan or the United States are in fact manageable through the blunt instrument of politics. If there is a political constituency for a 6.8 percent contraction in the Japanese economy, it is not obvious what it is. If there is a political constituency for the current state of the U.S. economy, especially the stagnation of middle-to-low-income wages, that is equally non-obvious. All the best people were convinced that Shinzō Abe had it figured out — and, by coincidence, he was putting forward policies very similar to the ones they themselves prefer. That’s evidence of something, too: the human capacity for self-delusion, one of the few commodities to which the economic concept of scarcity does not apply.