Another Post About Monetary Policy

Just to clarify my views a little more, the following policies are not equivalent to each other:

1. There should be no discretionary monetary policy; currency should consist of or be backed explicitly by assets such as precious metals.

2. Monetary policy should exclusively be concerned with combating inflation; recessions are caused by structural changes in the economy, and monetary policy should ignore them.

3. Monetary policy should be concerned with price stability – defined as a low rate of inflation to stay above the zero bound below which conventional monetary policy doesn’t really work. When inflation gets too low, the central bank should create more money; when it gets too high, it should take money out of circulation. Outright deflation, meanwhile, should be fought aggressively with every tool at the central bank’s disposal. This goal – stable, low inflation – is what produces the best results in terms of long term growth, and is therefore also the best way for the central bank to deliver on its mandate to achieve full employment in addition to price stability.

4. Monetary policy should be concerned with price stability, but this should be understood as price stability over a period of time, such as a decade. If inflation has been too low for a period, the central bank should aim for higher inflation to “catch up” with where the price level would have been with stable inflation. Presumably, if inflation has been too high for a period, the central bank should aim for lower than usual inflation to catch up (catch down?) with where the price level would have been in the absence of the inflationary period.

5. Monetary policy should balance the goal of price stability with the goal of maintaining low unemployment. The central bank should keep an eye on both indicators; when unemployment is very low, that should motivate the central bank to have a lower inflation target; and when unemployment is high, that should motivate the central bank to have a higher inflation target. Anything less than full employment should be understood as a prima facie indicator that inflation is too low.

6. Monetary policy should not target inflation or unemployment but rather aggregate spending or nominal GDP. When nominal GDP drops below target, whether due to a drop in real growth or a drop in inflation, the central bank should ease; when nominal GDP rises above the target, whether due to a rise in real growth or a rise in inflation, the central bank should ease. The implications, among other things, are that a rise in productivity growth should make the central bank more vigilant about inflation, and that external shocks (such as spikes in commodity prices) that produce more inflation should prompt the central bank to ease, rather than tighten (since they actually cause a drop in NGDP – the hit to real growth more than offsets the spike in inflation).

In this space, I’ve been arguing for proposition #3.

Proposition #1 is, in my view, a position held only by cranks.

Proposition #2 appears to be the position held by a number of GOP politicians, including Governor Perry. I haven’t heard anybody who is actually involved in monetary policy or who is a professional economist advocate this position. If Austrian school economists believe that deflation isn’t a monetary phenomenon, then I stand corrected. In any event, I reject this proposition completely.

Proposition #3 is what I would call a conventional conservative monetary policy view. I’d count most versions of the Taylor Rule as falling within this rubric, notwithstanding the fact that the output gap is a factor in some versions of said rule. I’d also say that arguments about whether or not to have an explicit inflation target, and whether that target should be 1% or 2% or 3% in general (the goal being to have inflation be as low as possible without significant risk of dropping below zero), also fall within this rubric. I don’t think I’ve really made arguments in favor of or against a particular inflation target, or whether inflation targeting is the best way to achieve the goal of price stability.

Proposition #4 I object to mostly on the grounds that, in practice, it would be asymmetric. You could have catch-up inflation, but never catch-up deflation (the latter would be too economically destructive). Because of this, I suspect that in practice it would result in higher inflation on average than the stated target. This, in turn, would undermine confidence in the ability of the central bank to maintain stable prices, and you could wind up in a rising spiral of inflationary expectations.

Proposition #5 sounds to me like a recipe for a back-to-the-70s inflationary spiral. What Paul Volcker was warning about in his recent Op Ed. In my preferred terms, inflation above some low and stable number negatively impacts real growth. Since the goal is to get nominal growth (aggregate spending, aggregate demand) up, this negative feedback means that you need more inflation than you would think to get the result you want. This higher inflation then becomes permanent – if you tried to return to price stability, you’d hit an economy with already low real growth rates with a recession. So, anyway, I’m strongly opposed to this proposition, just as I am strongly opposed to Propositions 1 and 2.

Proposition #6 is the intriguing hobby horse of Scott Sumner and is certainly worth exploring. I’ve raised my concerns about it with him before – nominal GDP is more volatile than inflation expectations, so this would mean monetary policy would become more volatile; moreover, precisely because it’s more volatile, it’s harder to predict, so you might get it more wrong than traditional monetary policy does; finally, it seems to demand perverse actions like “punishing” higher productivity with higher interest rates and “rewarding” drops in productivity with lower interest rates, and it would mean reading commodity price signals in the opposite way that they are usually read (higher commodity prices would mean a hit to real growth, and so would prompt easing rather than tightening). I guess I really hope some small economy adopts this policy so we could study a real-world example, but it feels kind of a radical experiment for a major economy like the United States or the EU.

I’m laying out all these alternatives because the advocates of greater easing have generally been wildly inconsistent about the rationale for greater easing which, I think, has hurt their case. The strongest case for another round of quantitative easing is that we are at greater risk now of renewed deflation than a rise in inflation. I think a very good case can be made for that proposition. (My main response would be that the major reason why we are at risk of deflation is the situation in the Euro-zone, and that unilateral easing by the United States could well exacerbate that situation. But if we really are at risk of renewed deflation, it wouldn’t be hard to convince me that we need a new round of QE.)

And sometimes advocates of greater easing make that case. Or, sometimes they make the case that we could have looser monetary policy without sparking inflation. I take that to be a version of the “we’re at risk for deflation” case: nominal prices and wages haven’t fallen yet – but they will if the economy doesn’t pick up soon to support them at current levels.

But sometimes they make the case for one of the other propositions above, either 4 or 5 or 6. And I can’t stress enough that these are inconsistent views.

Pundits and politicians have the luxury of saying “we need more inflation” without actually articulating a more general monetary policy. But central banks do not. If the advocates of higher inflation don’t pick an actual policy, and accept its implications for different economic conditions, then effectively they are advocating Proposition #5: let inflation rise until unemployment falls below some target level. Since I believe that’s a policy that’s proved catastrophic before, I have to push back.

In general, it’s a good idea to have as many arguments in your quiver as possible. So I understand the impulse to argue in the alternative: we need higher inflation because a 2% target is too close to the zero bound, or because we need to catch up to make up for past deflation, or because we’re at risk of falling into deflation again, or because we should be targeting NGDP and not inflation in the first place, or because unemployment is a bigger problem now than inflation is. Whatever rationale you like: so long as we get higher inflation.

But precisely because these positions are inconsistent, and would therefore mean different behavior in different economic conditions, even if they all agree right now, the central bank cannot adopt them all. They cannot argue in the alternative. They have to have a rule. And if you’re trying to influence the thinking of the central bank, it might make sense not to argue in the alternative in this way, but rather to argue for a position that the bank could actually adopt.

And this applies in spades to the other side of the aisle. So let me just say that letter GOP congressional leaders wrote to the Fed is simply outrageous. Outrageous, that is, if you think the Fed should actually be independent of political influence. It is as nakedly partisan a move as Nixon’s infamous deal with Arthur Burns#.

I sincerely hope nobody thinks I’ve been expressing skepticism about outright advocacy of higher inflation as the solution to the long employment recession because I want President Obama to fail to be reelected. I voted for him last time, and I’m more likely than not going to vote for him again. And, moreover, it seems to me that President Obama agrees with my view on the very question we’re discussing – namely, whether we need to address the long-term growth prospects of the economy to actually solve the short-term unemployment problem, as opposed to the whole problem being “monetary policy is too tight.” So why wouldn’t I support him?

I just wish he were doing more about those long-term growth prospects.