Philosophies of economic policymaking - a comment which growed

Think I kind of disagree with this victory lap on the part of Paul Krugman over the legacy of Milton Friedman.  Friedman himself is a bit absent, but sometimes people disappear simply because their victory has been so total that it's not necessary to argue for it any more.  And although k% M2 has turned out to be an embarrassing relic, the general concept of rules-based policy rather than discretionary is more or less hegemonic - when the Taylor Rule broke down, we had a short period of trying to make policy based on judgement, but we were very quick to try and find another rules-based framework.  Adam Posen's criticising the Bank of England's forward guidance strategy for being insufficiently rules-based as if this was a desideratum in and of itself.

And (via Prescott & Kydland of course, although everyone including Taylor would trace the lineage of the Taylor Rule back to Friedman), the impetus for rules-based policy is precisely the tension in Friedmanism that Krugman identifies in his last few paragraphs as the reason for his disappearance; the tension between a free-market ideology and the need for aggregate macroeconomic policy.  As PK correctly notes, Friedman's response to this was to try and package the whole of policymaking into a simple feedback rule, but although k% disappeared, the general project of making macroeconomic policy a rules-based system is definitely not gone - it's the water that we all swim in.

I think that the real Keynes versus Friedman philosophical distinction is whether you believe in actual, active policymaking, in the sense of someone taking a decision about the economy and saying "this is my decision, and as the planner responsible for this aspect of macroeconomic conditions, I am implementing it".  If you believe that someone in the system has to do this - that the buck has to stop somewhere - you're philosophically Keynesian.  If not, then even if your modelling is pure Hicks-Keynes (or even pure Keynes-Keynes - iirc Paul Davidson is basically opposed to active management), you are philosophically a Friedmanite.

And (I think I'm making my own position clear here) I think this is why Friedmanism fails.  Because actually, the buck does have to stop somewhere, and pretending that you can manage a complex system via a simple rule is basically impossible (it falls foul of Stafford Beer's Principle Of Sufficient Variability).  In practice, in a system based on a Taylor Rule, an Evans Rule or even an NGDP target, the buck stops with whoever it is that is responsible for maintaining the model which generates the forecasts of the control parameter.  And this person is always going to deny that he's making activist policy and claim that he's a technocrat who simply goes where the data takes him.  Friedmanism in economic policy, in the general sense I'm talking about here, is nothing more nor less than a distributed responsibility avoidance system.