Thursday, July 10, 2014

Long way around is better

Joseph Epstein wrote a moving tribute to his friend Edward Shils "My Friend Edward" (in his Narcissus Leaves the Pool; unable to link using behind-the-Chinese-Wall connections). Epstein notes, "He [Shils] admired the economists of the University of Chicago, trained under Jacob Viner and Frank Knight; and he spent more than half a century in the company of Milton Friedman and George Stigler and found both to be men of superior intelligence. But he felt the Chicago economists, brilliant though they could be, were insufficiently impressed with the mysteries of life." (p. 309). But Arnold Kling writes "Neoclassical economics is obsessed with the concept of equilibrium, and in turn it pays little attention to innovation. I believe that one of the biggest lessons of economics is the value of trial-and-error learning through entrepreneurial activity." (askblog post, July 9). Now we're starting to get somewhere. Equilibrium is just a modeling convenience. But many smart people take it much too seriously. Google Scholar shows 2170 citations for "benevolent social planner." Many modelers rely on this unicorn to make their model work. In his undergrad text, Greg Mankiw properly uses the mythical BSP to demonstrate to students the idea that markets face a huge task. But this is way different from inventing a BSP to satisfy journal editors and referees that they are seeing a real model. Modeling trial-and-error may seem like the long way around.