How do you convert ability into income? This is a question promoted by Ann Bauer's claim that novelists often need subsidies from their family. Her point broadens. Many artists struggle to get by: I suspect Jolie Holland speaks for many when she says she barely makes a living.

A novelist, academic and CEO might have very similar intellect and skill levels, but their income could differ by factors of thousands - and, as Will points out, academics' working conditions are deteriorating. Why the difference?

The conventional neoclassical answer is that wages equal marginal product, and that CEOs have a higher marginal product than others. This is a just-so story which glosses over a lot.

For one thing, what matters is that one's product be monetizable and appropriable. The great writer or musician creates an enormous amount of consumer surplus, but she cannot capture this for herself. Quite the opposite; as Gillian Welch sang*, she is under pressure to give away her work. Similarly, if you believe human capital theory, academics - at least the better ones - create billions of pounds of value. But they don't see much of it. By contrast, the CEO's output is more monetizable.

There are at least five reasons for this.

1. Ideology. Hiring committees' wishful thinking and overconfidence lead them to believe that they can hire a great CEO who can add millions of pounds of value to a business - so they pay a premium for what might turn out to be mediocre performance or worse. Demand, remember, can be an ideological construct.

2. Supply constraints. Thousands of people think they can write or sing. Their entry into the market makes it harder for the more able artists to get themselves heard. By contrast, a manager can only show his ability by having worked with expensive assets. This creates a lack of supply of managerial talent. The upshot, as Marko Tervio has shown, is high wages for modest abilities.

3. Rent-sharing. Joanne Lindley and Steven McIntosh show that workers in the finance industry earn more than others, even controlling for skill. This, they say, is because of rent-sharing. Bankers go home with big money for the same reason zookeepers go home with shit on their boots: if there's a lot of stuff around, some of it will stick to you. This also explains why footballers earn more than they did in the 60s; it's not because they have more ability, but because there's more money in the game. There are fewer rents in academia - hence lower pay.

4. Efficiency wages. Banks' traders and CEOs generally must be bribed in order to not sell off the firms' assets cheaply. Nobody feels the need to bribe academics to stop them teaching badly.

5. Reciprocity. The mere act of communicating with people leads them to treat us more generously. And conversely, if people are out of sight we are apt to treat them meanly. This gives CEOs an advantage.They are often on personal terms with members of the remuneration committee (if only because they often attend its meetings). By contrast, most of the rest of us have our pay set by people less close to us.

My point here is that Ed Walker is right. Marginal product theory is inadequate at explaining inequality. If you want to understand it, you must look beyond the claim that wages equal marginal productivity. We must ask: what is marginal product measuring, and what isn't it? And: through what mechanisms do wages equal (or not) marginal product?

* Everything worth knowing can be learned from country music and football.