Conservative Institutions, Radical Circumstances

Following Matt Yglesias to his new digs, I see he hasn’t changed either his views on monetary policy or his views on correcting typos before he posts:

[I]t’s encouraging that the Federal Reserve’s Open Market Committee considered [NGDP level targeting] at their meeting early this month. Unfortunately, according to the minutes they rejected it for reasons that seem pretty shallow. . .

The claim that this would “heighten uncertainty” seems to me to be just flat-out wrong. Normally there’s very little uncertainty about the Fed’s attitude. Normally we’re close to full employment and close to the Fed’s inflation target. . . . Today, however, there’s a great deal of uncertainty. We’re close to the Fed’s inflation target but nowhere near full employment. If millions of currently jobless people return to daily commuting, that’s all but certain to push gasoline prices up. If millions of currently jobless people living with parents or siblings get jobs and move into places of their own, that’s all but certain to push rents up. If that happens will the Fed tighten money to curb the inflation, or will it tolerate the inflation as passing pain needed to return to full employment? Nobdoy really knows, so nobody dares expect growth. One of the great virtues of a move toward NGPD level targeting is precisely to provide certainty on this point. The Fed will tighten if and only if total economy-wide spending increases so fast as to push us above the desired trend level.

True – but trivial. If “uncertainty” in this trivial sense were the problem, the Fed could always eliminate it by simply saying: we will do “X” no matter what, whatever “X” might be. “We won’t tighten until measured inflation is above 3% for two consecutive quarters,” is clear. So is, “we won’t tighten until unemployment goes below 7%.” So is, “we are standing pat until inflation expectations as measured by TIPS spreads drop below 2%, at which point we will do additional quantitative easing until expectations go above 2%.” Any fixed, inflexible rule would eliminate uncertainty.

And it would also eliminate any policy flexibility. Indeed, it eliminates the need for having a Fed – you could just have a computer program setting monetary policy. The very fact that you have a group of people in a room deciding what the rule is going to be means that there can never, ever be a totally credible and inflexible rule of the sort that Yglesias imagines he wants. “We are now moving to NGDP level targeting” sounds pretty clear – but what if inflation expectations jump to 4% overnight and employment growth doesn’t show up immediately? Won’t that put a lot of pressure on the Fed to reverse course? Doesn’t everybody know that it would put a lot of pressure on the Fed to reverse course? Heck, if NGDP level targeting worked it should lead to a sharp rise in nominal interest rates on Federal debt – that would be the market pricing in an expectation of much higher nominal growth. But the immediate result of such a spike would be a sharp increase in the cost of rolling over Federal debt, which might well create strong political pressure for greater short-term budgetary discipline to avoid locking in a huge interest overhang for years to come. Which, in turn, would be contractionary if it actually happened, and didn’t just force the Fed to change policy directions.

The Fed is a big, important, highly conservative institution. Quite properly, it doesn’t change its policy framework easily. If it did, that would lead to a widespread loss of confidence in the Fed as an institution, which would mean any rule it adopted wouldn’t be terribly effective.

Now, the current circumstances are extraordinary. The Fed has responded to those extraordinary circumstances by signalling that it will do pretty much anything to avoid us tripping into a double-dip recession. They’ve declined to be more daring than that, and have basically begged the fiscal side to do more to help them out. That, in turn, implies that they would welcome fiscal action that improved the long-term real growth prospects of the economy. That could mean tax reform that reduced economically distorting deductions; it could mean a shift in spending priorities away from defense and health-care and towards investments in physical infrastructure and human capital; it could mean dozens of other things that the Federal government isn’t doing, that would give either monetary or fiscal policy more running room for straight-out Keynesian expansion without igniting fears of an inflationary spiral.

I find it really hard to blame the Fed for being minimally conservative – not being eager to change its entire framework for thinking about monetary policy – when the fiscal side can’t seem to do anything at all.

Now, none of that means that the Fed shouldn’t consider an NGDP level target. But I’d expect it to want to see a bit more of a “developing consensus” in the profession before making the jump, precisely because you can’t change frameworks very often and remain credible. And if a new consensus is developing, it’s in the very early stages of doing so.

Finally, just to recap, here are the important points I think NGDP targeting needs to address:

- I understand why one implication of an NGDP-level targeting is that a negative supply shock should lead to looser, not tighter, money, and that a positive supply shock should lead to tighter, not looser, money. That makes sense to me. If oil prices spike, that’s not going to trigger an inflationary spiral – it’s going to trigger an economic contraction. So we should loosen. If oil prices crater, that’s going to temporarily goose growth, but there’s no underlying improvement in productivity to make that sustainable; the right response isn’t to loosen in response to lower inflation, but to tighten in response to rising nominal growth. This is not the way the Fed reacts currently – in particular, it’s not how it reacted at the beginning of the financial crisis, when commodity prices spiked even as we started to slip into recession – and this is an area that, I think, deserves very close scrutiny by the Fed, and potentially a declaration of their intent in the event of a future spike in commodity prices.

- But I don’t understand why a sustainable increase in productivity should lead to tighter money. I don’t know how the Fed can know whether productivity “should” increase at 3% or 2% per year over a given decade. And I don’t understand why, effectively, lowering the inflation target when productivity growth is high and raising the inflation target when productivity growth is low makes any sense. I’m not saying outright that it doesn’t make sense. I’m saying I don’t understand the logic of it. Put another way: how do we know that the American economy “should” grow at 5% per year? Where did that 5% come from? Presumably from our historic experience of what stable growth looked like – low inflation, modest population growth, productivity growth in the 2%-3% range. But if either of the latter two variables change, it seems to me that the NGDP level target should change, because they are the “real” drivers of the “real” growth rate of the economy.

- I continue to worry that NGDP has been significantly more volatile than inflation expectations have been, and that therefore an NGDP level target would imply a much more volatile Fed policy. If the Fed is supposed to use medium-term NGDP expectations to guide policy, that means the Fed would loosen preemptively whenever the market predicted a recession. The best instrument for measuring future expectations of nominal GDP growth is probably yields on the 10-year bond. I’m pretty sure, though not certain, that the 10-year bond yield is more volatile than the TIPS spread. But that might reflect the fact that the Fed is actually trying to fix inflation expectations, and is not trying to fix the 10-year bond yield. In any event, this is another area that would deserve real analysis before the Fed made such a shift.