The death of Gary Becker, the father of the "economics of everything" set me wondering: could it be that basic neoclassical economics does a better job of explaining "non-economic" behaviour than it does of economic phenomena?

Take three examples where basic neoclassical is, at best, questionable:

- A theory of distribution. The idea that wages are equal to workers' marginal product is deeply questionable. It seems instead that power, rent-seeking and institutional frameworks also matter enormously; the fact that there's no correlation between patterns of income inequality and of human capital inequality should be awkward for neoclassicals. And the idea of a marginal product of capital is just close to meaningless, which gives neoclassical economics little idea of where profits come from.

- The Solow-Swan growth model found that most economic growth was due to exogenous technical progress, which is pretty much no explanation of growth at all. The difficulty (pdf) of relating growth to human capital suggests that augmented theories do little better. And the fact that governments in developed economies can do little to change long-term growth rates suggests that growth theory hasn't (yet?) had a useful policy influence.

- The neoclassical explanation of unemployment stresses wage inflexibility and "rigidities". But this fails to account for why unemployment was high in the 19th century.

By contrast, Beckerian economics has given us some useful insights into crime, family life (pdf) and racial discrimination.

There might be good reasons for this difference. Matters of individual choice are often more tractable in the "non-economic" than economic arena. The man thinking of committing a crime, for example, doesn't face such severely bounded rationality and knowledge as the company boss making investment decisions. Problems of aggregating different goods are less relevant. And whereas financial capital is now abundant, time is not - and so the neoclassical presumption of scarcity remains relevant to our life choices.

Now, I'm not saying here that neoclassical economics of everything is a complete theory. The assumption of an atomic rational maximizer is challenged by cognitive biases and peer pressures - though Becker was well aware (pdf) of the latter. And it's an open question how far the Beckerian approach can explain the collapse in crime rates recently. It might be, though, that neoclassicals' limitations in these areas are more resolvable than in others.

What I'm saying here might earn me the wrath of both heterodox and orthodox economists, but it is in a sense trivial.Different theories can explain different things. The idea that one theory fits all owes more to fanaticism and tribalism than to the data.