What's In A Name?

Op-ed in The International Economy

April 21, 2007

When we sat down 12 years ago at the New York Fed to research the hot new monetary regime of the 1990s, we gave little thought to the resulting book's title.1 The Reserve Bank of New Zealand (RBNZ), the Bank of Canada, and the Bank of England all called what they were doing "inflation targeting;" so, in the interests of transparency, that is what we called the book. We were more concerned with addressing the threats we perceived than the label: the vulnerability of US monetary policy to a radical change in the Fed's mandate (such as the gold standard a US senator proposed in 1994) and to the excessive personalization of policy through the reputation of a long-serving chairman. We thought a public inflation target (IT) would preempt both dangers.

It turns out we should have paid more attention to marketing. In today's discussions of the future of the Fed, there is a strong presumption in many quarters that "inflation targeting" would be an excessively rigid regime-akin to monetary targeting or a hard exchange rate peg, concentrating only on inflation. If it were such a rules-based regime, as the word "targeting" (and the absence of the word "employment") apparently connotes on Capitol Hill, it would be a very bad idea indeed.

But an IT in practice actually enhances a central bank's flexibility in responding to real shocks. If you can anchor inflation expectations through a public commitment, you can cut rates in a recession more aggressively or avoid raising rates after an oil shock, because the central bank's short-term stabilization of the real economy does not raise doubts about its commitment to price stability. The most well-intentioned central bank that cuts rates sharply when unemployment rises, or that fails to raise interest rates soon enough after an oil shock, but does so without such an anchor for inflation expectations as an IT provides, ends up with higher inflation rates and then must induce a recession to get them back down. That is what the 1970s taught us.

This use of an IT is how the United Kingdom kept its inflation rate low following its devaluation of the pound in September 1992 and how Brazil managed a similar challenge when its currency was hit in the crisis of 1997-98. In both cases, monetary policy still avoided sharp tightening for the sake of employment. Italy and Argentina, whose currencies depreciated at the same times without the IT anchor, got the inflation, as well as less output stabilization. This pattern also shows up econometrically: Central banks with inflation targeting have significantly less persistence of inflation after shocks than those without it.

On the other side, an IT prevents central banks from being "inflation nutters," obsessed with achieving low inflation at whatever growth cost, because it makes transparent when a central bank is going nuts. When the RBNZ had too tight and low a target range for inflation in its initial IT regime, the costs became obvious and the framework allowed for the elected government to change it. In contrast, when the IT-less Bank of Japan spent the 1997-2002 period in apparent pursuit of deflation, it ducked accountability in part because of the absence of a transparent medium-term goal.

Thus, an IT's bad rap in parts of official Washington is undeserved, because it would help the Fed pursue-not elude-its dual mandate (of stabilizing growth and prices, while increasing Fed accountability). Still, if one has a product with evident virtues that fails to sell, one must sometimes rebrand it. So let's forget the label "inflation targeting" and instead promote the substantive components that would get the US economy IT benefits without the baggage of its undeserved image.

First, we should have a public discussion about what is the desirable long-run average rate of inflation. Fed Chairman Bernanke began this with his 2006 speech at Princeton University, but the understandable reluctance of the Fed to hammer the economy to get down to 1 percent on the Personal Consumption Expenditures shows the question is unresolved. Second, the Fed should release more information about and take more ownership of its forecast. This would continue the multiyear process of increasing transparency. Third, the Fed should be more explicit in flagging events or circumstances that require a change in policy and accepting the risk of a potentially extended deviation from low inflation. What the Fed did in late 2001 and 2002 in response to deflation is a great example to follow.

None of that sounds so bad, does it? Yet if the Fed were to formally pursue those three steps-the first of which would clearly require open consultation with the US Congress, and the other two informing the public in a more disciplined manner-we would have inflation targeting and its benefits in all but name. That is, we would have a monetary framework for the United States that would be one notch more deliberate, transparent, and accountable than it was before, but still consistent with the dual mandate and far from inflexible. Call it what you like, but it would be an improvement for the long-term.

1. Inflation Targeting: Lessons from the International Experience. 1999. Posen, Bernanke, Laubach, and Mishkin. Princeton, NJ: Princeton University Press.