There's no doubt that Jonathan Haskel is an excellent appointment to the MPC. What we should be asking is rather: isn't he over-qualified?

What I mean is that not much hangs upon interest rate decisions. The Bank's own work estimates that a quarter-point rise reduces output only by around 0.15%. Granted, it's possible that the economy might now be more sensitive than this to rate changes now. But even so, the cost of interest rate mistakes are small.

Put it this way. For a monetary policy error to have had the same cost as fiscal austerity or Brexit, interest rates would have had to rise to well over 10%. It doesn't require somebody of Haskel's intellect to see that that would be a stupid idea.

I suspect that the ratio of intellectual input to the stakes of the decision is greater for monetary policy than it is for any other decision in public life.

Which poses the question: why does the MPC need someone of Haskel's expertise?

The question gains force from two considerations. One is that a lot of the job of setting interest rates requires you to interpret noisy high-frequency data and, arguably, to have market feel. Academic economists don't have a comparative advantage here. The other is opportunity cost. If Haskel is thinking about whether to change rates by a quarter-point or not, he's not doing other work, where he might well do more good.

There is an answer here. It starts from the fact that the MPC lacks direct democratic legitimacy. Of course, the inflation target is set by governments (and there's been surprisingly little political debate about it) but the MPC itself is unelected.

And yet it has some power over us. Even the most modest rate rise destroys some jobs; that's how it reduces inflation. And on the other hand, the lack of a rate rise means millions of savers forego some interest income. What legitimacy does it have to take such decisions? Governments have a licence to wreck the economy because that's what voters want. The MPC has no such justification.

Which is why it needs experts. The job of MPC members is not merely to make interest rate decisions. It is to give legitimacy to the MPC by having the ability to explain monetary policy thinking to a sceptical audience of other economists, and to appear to the public as wise technocrats. Paul Tucker calls this "legitimacy through credibility".

It was for this reason that Ben Broadbent's remark about the "menopausal" economy attracted such attention. Technocrats acquire legitimacy by being grey and sober; ill-judged language detracts from this.

A decent trading-floor economist might well be able to take interest rate decisions as well as Haskel. But he wouldn't have the authority with outsiders that Haskel has, nor - no doubt - the ability to construct genuinely interesting economic arguments.

Which brings us to a problem. Even the most genuine expertise, such as Haskel has, is not the only potential source of legitimacy. It's also the case sometimes that people want decision-makers to be representative of those whose lives they affect - hence the complaints that the MPC is too male (and we might add, too white and too posh).

Now, I'm honestly unsure how much force these complaints have. What is clear, though, is that unelected policy-makers need some form of legitimacy; this is true not just of the MPC but of judges, quangos and so on. Expertise is one source of such legitimacy - maybe in most cases the best - but it's not necessarily the only one. Unelected policy-makers should be chosen not just by their ability to take decisions, but by whether they enhance the authority of their office.