The Washington PostDemocracy Dies in Darkness

Michael Woodford may have written the year’s most important academic paper. Here’s why.

Everyone who's anyone in monetary policy is in Jackson Hole, Wyo., this long weekend for the annual conference on central banking held there in late summer. The big story every year is the speech by the Federal Reserve chairman, which is often used to signal changes in Fed policy going forward. Sure enough, Ben Bernanke used his platform to defend past quantitative easing measures the bank adopted to fight the economic downturn and to lay the groundwork for similar actions going forward.

But more important than Bernanke's speech may be an 87-page, highly technical paper released as part of the conference by Columbia University's Michael Woodford. Although not a household name by any means, Woodford is widely regarded as the greatest monetary economist of his generation. He was in the first class of MacArthur genius fellows, before even finishing his PhD, and his book Interests and Prices, in which he laid out in detail his views on monetary policy, was hailed upon publication as a "landmark treatise" by Harvard professor and former White House chief economist Greg Mankiw and as a "classic" by MIT's Olivier Blanchard, now chief economist at the International Monetary Fund.

Woodford's paper is an extended argument for the Federal Reserve to stop targeting a certain level of inflation (about 2 percent annually) and to start targeting a certain level of nominal (that is, not adjusted for inflation) gross domestic product. Woodford is hardly the first person to endorse this idea, commonly known by the not-particularly catchy name of "NGDP level targeting". Bentley University's Scott Sumner popularized the idea first on his blog, The Money Illusion, and then in a long essay in National Affairs. It has since been endorsed by Christina Romer, who was President Obama's head economist earlier in his term, Paul Krugman and the economic team at Goldman Sachs.

The idea is that NGDP encompasses both the rate of inflation and the rate of real economic growth. So, if either inflation grows too large or economic growth is too slow, NGDP targeting would recommend action to correct that. By contrast, inflation targeting such as that engaged in by the Federal Reserve in recent years does not react as aggressively to recessions. NGDP targeting also recommends targeting "levels" rather than rates of growth. To see how this works, consider the recent recession. Over the 2000s, NGDP grew at a rate of about 4.5 to 5 percent annually. But for much of the recovery, it fell well below that mark, reaching only 3.4 percent growth in the start of 2011. Following periods like that, a 5 percent NGDP-level target would have recommended NGDP growth of 6.6 percent, becuase 6.6 and 3.4 percent average to 5 percent. The idea is to make up for lackluster growth with much faster than normal growth, so that the economy stays on track.

Woodford's endorsement adds additional respectability to the NGDP-level targeting proposal, but the paper does much more than that. It answers an important question that the proposal's critics often bring up: What sort of action, exactly, can the Fed take when its fancy new NGDP target says it should act? When interest rates are at zero, as they are today, you can't push them much lower. You can print money and use it buy up bonds in hope of increasing in the money supply (quantitative easing). This all misses the point, Woodford contends. The simple act of the Federal Reserve chairman saying he's going to target NGDP is action enough.

It's all about expectations. The main way the Fed can boost the economy, Woodford says, is by changing where businesses and consumers expect the economy to be in the future. Suppose, to use Matt Yglesias's example, you're a business considering borrowing money to build an apartment building. Interest rates are low, so borrowing is really cheap. So far, so good. But you also want people to have enough money to rent your apartments. And if their incomes increase, then that probably means the economy is doing well in general, and so the interest rate on your loan is going to go up, as interest rates tend to in good times. So it doesn't really matter if interest rates are low now: If you're going to have customers who can pay your rents, your interest rate is going to go up and it'll be a wash.

But suppose the Fed says: "We'll throw you a bone here. Even if people start making more money, we're going to keep interest rates really low. Your loan is going to be cheap no matter what. Build your apartment, count on people making more money and everything's going to be fine." The result is that more people take out loans and build more apartment buildings (or fund other business activities), and the economy grows.

If the Fed said tomorrow that it would start targeting NGDP, Woodford explains, it's basically saying when, as now, NGDP is well below target, the Fed isn't going to do anything that raises interest rates until it stop being far below target. It's the Fed saying to businesses "we got you." This is also, Woodford says, why quantitative easing has worked in the past. What's most effective isn't the actual buying up of bonds; it's the message that action sends to businesses and the effect it has on their planning. It is a credible signal that the Fed is going to keep lending cheap.

But it would be more credible, Woodford contends, if it were paired with a clear explanation of when exactly the Fed would consider raising interest rates. That's where NGDP comes in. "A more logical policy," he writes, "would rely on a combination of commitment to a clear target criterion to guide future decisions about interest-rate policy with immediate policy actions that should stimulate spending immediately without relying too much on expectational channels." A clear target criterion = NGDP targeting. Immediate policy actions = quantitative easing. And easing is even more credible when the assets bought are not federal bonds, as is traditional, but more exotic assets such as mortgage-backed securities, where the effect of the purchases on interest rates (in that case, for mortgages) is more immediate.

Interestingly, Woodford ends with an argument for the other policy he thinks could do a lot to boost growth now: fiscal stimulus, such as tax cuts or spending increases. And pairing NGDP targeting with more stimulus would be especially effective, he says, as it would prevent any "crowding out" of private-sector activities or inflation that could result from increased government spending. NGDP targeting, in short, adds more "bang for the buck" to stimulus spending.

Obviously, the Federal Reserve can't undertake fiscal stimulus. But there are some indications it's taking NGDP targeting more seriously. Chicago Fed president Charles Evans has proposed a "7-3 rule": The Fed declares it won't raise interest rates until unemployment is below 7 percent or inflation is above 3 percent. That's close to an NGDP target, but, Woodford argues, inferior. At its last meeting, the Fed discussed nontraditional policies it could adopt, and it's likely NGDP targeting was among them. But Woodford's paper is a major event in itself, one that more than any other recent development could push the Fed toward the policy. In a way, it's doing what he wants the Fed to do: changing the world of monetary policy just by saying something.