The institutions of economic policy have also failed at the national level. Settled doctrine before the crash of 2008 called for keeping fiscal policy, monetary policy and financial regulation in separate silos. This division rested in part on a view about central banking that worked fine in good and moderately bad times, but fails in a situation like this. Central banks, according to this view, could be set to the simple task of keeping inflation low, and be shielded from politics in discharging it.

Post-crash macroeconomic policy recognizes that the lines between fiscal, monetary and regulatory policies aren’t so clear. For maximum stimulus effect, all three have to work in harness.

Consider the case of outright monetary financing of government spending — or “helicopter money,” as it’s sometimes called. This is fiscal policy: The government sends everybody a check. It’s also monetary policy: The central bank finances the outlay by buying and indefinitely retaining government debt. And it implicates regulatory policy, because it changes the cost of capital across different asset classes.

Quantitative easing isn’t quite helicopter money (because the central bank expects to reverse its debt purchases at some point) but the basic criticism applies. Because it isn’t the purely apolitical intervention that advocates of central-bank independence had in mind, central banks are inhibited in its use. (In Europe, this inhibition is enshrined in law.) The limits imposed by independence also complicate the forward guidance the Fed and other central banks have adopted or are about to adopt. For example, in setting thresholds for unemployment and schedules for getting it down, they’re straying onto political terrain, so they have to be circumspect. Deliberate vagueness muddies the message, adding to the confusion that has lately roiled financial markets.

Fears that aggressive QE might cause financial instability when it’s reversed raise a related point. These concerns should be taken seriously, even if signs of a risky “reaching for yield” are few right now. In principle, though, the answer is not to keep QE below the level indicated by the shortfall in aggregate demand; it’s to adopt financial rules that promote safety in times of financial stress (above all by requiring financial intermediaries to be better capitalized). The IMF’s new report makes this point in passing, but without setting it in the bigger picture.

In a way, post-2008 monetary, fiscal and regulatory policies are all one. The old consensus on compartmentalizing the different instruments has plainly failed, but the institutional apparatus and expectations still mostly persist. The same goes for the global policy architecture. It’s not just a question of what the policy should be. Governments need to reconsider how they decide what the policy should be.

A tall order, I know, but if they don’t get it right, nothing much will be different next time round.

Clive Crook is a Bloomberg View columnist.