Questions For Scott Sumner

Apropos of this post.

If I understand his preferred policy correctly, the Fed should target nominal GDP growth, trying to keep that in the 5-6% range. If nominal growth is too strong, whether because of rising inflation or because of accelerating real growth, you raise interest rates to keep it in target; if nominal growth is too weak, whether because of too-low inflation or anemic real growth, you lower rates or, if rates are already at the zero bound, engage in quantitative easing.

Since nominal GDP growth is relatively easy to measure (certainly compared to inflation, the precise calculation of which we never stop arguing about), and since businesses are really making their investment decisions based on nominal, not real growth (if spending drops, they’ll sell less, and if it goes up, they’ll sell more – it doesn’t really matter if the change in spending was due to changes in inflation expectations or changes in real growth expectations), stable nominal GDP growth should make business decisions pleasantly predictable, which, in turn, should increase real growth (which is what we’re all aiming for).

But while it’s easy to measure in retrospect, it’s no easier to measure in advance than any other economic variable. Indeed, it’s arguably harder than some, precisely because we don’t really know what drives real growth moment to moment (if we did, economic policy-making would be easy). But the Fed isn’t going to go into completely reactive mode, and only move rates once we know we’re out of target – among other things, if it did that, by definition it would fail at its task of delivering “stable” NGDP. So if the Fed is going to target nominal GDP growth, it’s going to be targeting a forecast, whether produced by a model or derived from market instruments or (most likely) a combination of both approaches.

Okay. In 2008, we faced the first tremors of the financial crisis. The decline in housing was well underway, but the collapse of Bear Stearns was the first indication that Wall Street might have serious problems. The Fed acted quickly to shore up the banking system, opening the window to investment banks that previously couldn’t borrow there (because they weren’t regulated by the Fed). Then we had a summer lull – the calm before the storm that broke in September with the collapse of Lehman and AIG.

If I remember correctly, that same summer lull was the period when commodity prices entered the blow-off phase. Oil hit an all-time high. Gold was going crazy. Every commodity you could name was surging to unprecedented levels. The markets were full of fears that the Fed was being too loose in response to the housing-and-banking crisis, and that this was going to cause emerging markets (China, Latin America) that were linked to the dollar to overheat, and result in inflation at home. Investors ran to the perceived safety of physical commodities, the classic inflation hedge.

But it was right after this point that real GDP fell off a cliff, to be followed closely by nominal GDP as inflation expectations collapsed. In 2009-2010, inflation was negligible and we had negative real growth. We had entered the Great Recession.

So – again, if I understand him correctly – Sumner’s preferred policy would have required the Fed to ignore the market’s expectations of high inflation, and properly forecast the collapse in nominal GDP, so that they would know that just as the market was freaking out about rising inflation, what was really needed was much looser monetary policy.

Is that right?

I admit, to me it feels not very different from saying “if you can see a huge recession is coming, ease, and keep easing until you’re sure it’s not coming anymore.” Isn’t the whole game inside that “if”?

Sumner talks a lot about the need for a futures market to properly predict NGDP. But futures markets can be hugely volatile, even when they are deep. We have a reasonable proxy for NGDP growth right now – it’s called the stock market. Stocks’ value is derived from future dividends; those are paid out of future profits; and profits, in turn, are driven primarily by growth in nominal GDP. The old line is that the stock market predicted 12 of the last 5 recessions, but in August of 2008 the stock market was holding pretty steady. That’s August, 2008. A month before the beginning of the end of the world.

I’m intrigued by the contrarianism of some of Sumner’s advice. In particular – again, if I understand him correctly – his view is that in response to a supply-shock driven spike in prices, the Fed should lower rates, rather than raise them. Why? Because while these shocks do cause a surge in inflation, they hurt real GDP more – and, therefore, lower nominal GDP growth. So if we’re targeting nominal GDP growth, we should ease, and certainly not tighten, whenever oil or food prices spike sharply. That’s interesting!

But it still seems to me that a Fed policy based on nominal GDP targeting would be extremely volatile. The simple fact is that inflation expectations have been – for some time now – relatively stable. From 1984 to 1992, inflation didn’t budge very much from 4%. Then, in the Clinton years, it drifted down, gently, from 4% to 2%. Then, in the Bush years, it drifted gently upward to 3%, before dropping to 1% in the wake of the financial crisis. By contrast, nominal GDP flops all over the place. The deep recession we just went through was not a period of sharply falling prices. We didn’t have deflation. We had modest disinflation. What cratered was real GDP growth. So, again, it seems to me an NGDP targeting policy amounts to saying: ease if you know a recession is coming. But if we knew that . . .

Moreover, precisely because actual measured nominal GDP growth is so volatile, expectations for nominal GDP growth would also be expected to be highly volatile. If the Fed is going to target nominal GDP, it’ll have to be much smarter than the market (yeah, right), or I rather suspect it’ll be hunting a lot of wolves – both predicted booms and predicted busts – that the market is crying, but that aren’t really there.

But I welcome any corrections (if I have misunderstood him) or explanations why doing what he thinks we should be doing won’t be much harder than it sounds to me.