Over the weekend, Daniel Kuehn commented on a post I made on Facebook regarding the budget sequester. For anyone who follows me on Facebook, you probably know that my status updates are often intended to ruffle…and this one was no different:
How scary is sequester? Not at all scary, provided monetary accommodation.
No mistakes: the sequester should happen.
I stand behind this statement, but I before I get into the meat of the issue at hand, let me take an aside to make some positions clear. Given the choice of “cut while the economy is weak”, and “cut during a robust expansion”, I would unabashedly choose the latter. In fact, I would be perfectly happy if Congress kicked the can down the road — but if they are simply going to kill the sequester on the grounds that it will have a “significant impact on recovery”, it should happen…and that’s double for the defense portion, because that position is not self-evidently true. And contrary to some in my camp; I do believe that fiscal policy has a role to play in counter-cyclical management…and I think that the welfare state is a legitimate use of resources. These are all interesting issues; we have a Federal government that generates revenue inefficiently, and transfers it incoherently. That’s the subject of a different post, however.
Kuehn takes issue with market monetarism itself, claiming it a tautology:
What worries me about this attitude is that MMers seem to define “monetary accommodation” as “the Fed being successful”, and they don’t worry enough expectations trap type problems that could cause trouble and that actually present an opportunity for fiscal stimulus to give a helping hand.
So here’s a question for you: how could I identify a case where monetary policy IS accommodating but where it fails? What would that look like? If there’s not a good answer to that (and I haven’t heard one yet), then the MM position reduces to “don’t worry – everything will be OK as long as everything is OK”, which needless to say is not reassuring. :
You can do this with fiscal policy. Get an estimate of the output gap, get an estimate of the multiplier, and back it out. Any remaining problems are chalked up to the fact that fiscal policy is not a silver bullet. I have no idea what a comparable exercise would be for monetary policy in the context of an MM outlook (you could probably figure something comparable for a New Keynesian view on monetary policy).
In my view, there are two ways monetary policy can fail (assuming no permanent liqudity trap):
- Higher NGDP will fail to raise RGDP.
- The Fed will face a significant risk-adjusted capital loss on assets purchased.
So there are those. I don’t think either are significant concerns. However I do think that the experience of the Great Depression and Great Recession shows that monetary policy always has the capacity to affect NGDP, and the trend rate of NGDP growth is what market monetarists (and some new Keynesians!) focus on as the proper indicator of the stance of monetary policy.
To answer Kuehn’s question; if the Fed bought up all of planet earth, and NGDP didn’t budge, we could say that the Fed tried and failed…but no one in their right mind would expect that result. Is that too high a bar? I suppose to answer that question you’d have to ask yourself what is the marginal (private and social) cost of unemployed labor and capital? When it comes down to it, that is a subjective proposition. I think it implausible that, in the real world, we would face a situation where the Fed would be unable to hit a realistic nominal target.
This isn’t to say that any slight wiggle in NGDP was the intention of the Fed. It is perfectly natural to expect NGDP to oscillate around a trend because the Fed isn’t perfectly omnipotent, and aggregate demand isn’t a cosmological constant. However — and perhaps market monetarists as a whole have been squishy on the subject — there are reasonable levels at which the MM hypothesis can be tested, much like the original monetarist position of targeting the aggregates (although I have a different reading than most regarding this), where reasonable people would be able to judge the experiment to have been a success or failure. And there are mechanical ways in which the Fed can implement this targeting regime to actually define success or failure.
This was a hasty post, so I don’t expect it to fully satisfy Kuehn’s questions, but I think it’s a good start.
Update: I also want to add that it is possible that the MM position could be proven of little usefulness even if not invalidated.