The news that MPs think they deserve a ''20,000pa pay rise is neither surprising nor relevant. Not surprising, because we all think we're underpaid. And not relevant, because pay does not and should not depend upon desert. Instead, the question is: if we paid MPs more, would we get better governance? I'm not sure.

There are three ways in which higher pay can improve labour supply. And whilst they might apply to some politicians sometimes, I'm not sure they apply to MPs now.

1. "Higher pay would attract better candidates." Richard Murphy says that doctors, headteachers and other professionals have to take a big pay cut if they want to become an MP, and this deters talented people.

There are three problems with this. First, as Paul says, the job of being an MP doesn't require huge skill. You need to be a genius to obey the whips' orders. Secondly, do we really want good doctors and headteachers leaving their jobs to become humdrum backbenchers? Can't they do more good in their current jobs? Thirdly, there's a danger that paying MPs more will crowd out the intrinsic motive of "public service." Do we really want MPs who are only motivated by cold hard cash?

2. "Higher pay would make MPs work harder, as the pain of losing their job would be greater."

But there are few cases of MPs not trying as hard as possible to hold their seats now - not least because of non-monetary motives to be an MP. The exceptions to this hardly support the case for higher MPs pay; would political life really be better if Louise Mensch had remained MP for Corby? And even if MPs did work harder, it's not clear this would be a good thing. I suspect that, in recent years, we've needed less legislative activism and more scrutiny of legislation. Would higher pay really have elicited this?

3. "Higher pay generates good will, and so buys off fraudulent behaviour."

Generally speaking, this is an under-rated reason for high pay. But I'm not sure it's relevant now. I suspect that British MPs are, by global standards, relatively honest already; fiddling a few grand in housing expenses is low-level stuff. And cross-country comparisons suggest little negative correlation between politicians' pay and honesty; MPs are paid more (pdf) in Ireland and Italy than in the Nordic countries, for example.

So much for my hypotheses. What of the historical evidence? My chart shows MPs pay (pdf) relative to average earnings. This shows that their present pay - at 2.7x average earnings - is in line with the historic average. Except for a brief period when MPs' pay was introduced in 1911, the best time financially to be an MP was the 1960s. But were we really well-governed then? Given that MPs included Maudling, Stonehouse and Driberg then, it's not at all obvious.