Openness in policy-making

On Twitter, Ben Chu has asked for a counter-argument to Chris Giles' call for the Bank of England to follow the Fed in publishing full transcripts of MPC meetings. I think such an argument exists, but that it draws attention to a general failure of public debate.

The argument - as Chris acknowledges - is that it might stifle debate. I find this plausible. The economy is a complex, emergent process which is inherently unpredictable. What's more, the economy isn't merely risky, but uncertain. We don't know where the big dangers will come from; for example in the mid-00s everyone told us that current account imbalances and hedge funds were a threat to financial stability, but few told us that banks were.

This means that, after the passage of time, many contemporaneous statements about the economy will look wrong. Not only will forecasts be wrong, but we'll be worrying about risks that don't materialize and not worrying about ones that do.

With hindsight, everyone looks daft. The claim that publishing transcripts will make some MPC members reluctant to speak, for fear of looking silly in a few years' time, is therefore one I find plausible.

But its strength rests upon two aspects of our culture which are questionable.

The first is an excessive faith in experts. If you believe the economy is predictable and controllable by men of good judgment, then errors are something to be excoriated. Such a view will stifle debate.

But this is the wrong frame. Errors are inevitable. We should therefore judge people not by whether they happen to be right, but rather by whether their beliefs were rational - that is, proportioned to the evidence as it existed at the time. From this perspective, it would be wrong - and possibly irrational - to decry people for being wrong. If this were so, MPC members would feel less constrained by the threat of future publicity.

The second problematic aspect is the presumption that the role of experts is to express their personal opinion.

But it needn't be so. An expert isn't merely some guy with an opinion. It is someone who is capable of giving the strongest possible argument and counter-argument for a case - who knows all the perspectives on an issue and the evidence and counter-evidence. (This is what I was trying to do in yesterday's post).

This points us to another way of organizing MPC meetings. Someone could be designated an "inflation nutter" and tasked with providing the best argument for raising rates; someone else could be the "dove"; someone else the worryguts about financial risk, and so on. (I'm thinking here of a variant of De Bono's six hats). If any MPC member can't fill such roles, he's probably not qualified to be on the committee.  

Such a structure would create rigorous constructive debate whilst avoiding the twin evils of groupthink or a conflict of egos. It would also free MPC members from the fear of future criticism for being wrong, because the views they expressed would be those of their roles not their selves.

Of course, members would have to step out of their roles to vote on policy. But this doesn't bear upon the issue of publishing transcripts, simply because members' votes are recorded in the minutes of meetings.

I say all this not to say that Chris is right or wrong to call for publication of transcripts. Rather, it's to suggest that, if he is wrong, the problem lies not so much with his argument as in the defective nature of the way policy is organized and debated.

Comments

[Dave Timoney]

The "different hats" approach has merit as a way of rigorously testing evidence. This, after all, is the model employed in a court of law (though the recent tale of the American guy who spent 30 years on Death Row due to an incompetent defence shows that this doesn't always work in practice).

Arguably, the adversarial nature of politics also embodies this role-playing model, but you can now see a problem: what is to prevent the frame of reference being structurally biased? Arguing for more of X or less of X isn't helpful if it excludes consideration of Y.

Formalising roles on the MPC such as "hawk" and "dove" might institutionalise a particular policy paradigm. This in turn would potentially give too much influence to the media (who are largely responsible for the taxonomy), allowing them to gradually shift the terms of debate.

[Phil H]

I think also the hats suggestion ignores the way relatively open cultures actually work. Whereas in a closed culture, there are many positions which literally no-one will publicly advocate for (whether they believe them privately or not), in an open culture there is always someone who believes almost any given proposition. Because of that, we no longer bother with the hats. We just find people who really hold certain kinds of beliefs, and let them be the hats.

[Rob Angell]

The assigning of roles in the MPC will lead to the classic polorisation of argument / debate. And it is the definition of debate - to pursuade / advocate for a position to convince the other side they are wrong - that gives a hint to the problem. For me the solution is to change the frame to dialogue - to deliberate, consider and decide. If the MPC were to be seen to be doing this, on behalf of the interests of the economy then showing what was said seems a whole lot less scary.