Three ideas for the Fed: seeking new ways to measure inflation.

As Chairman Bernanke, Governor Kroszner, and Governor Warsh settle into their new positions with the Federal Open Market Committee, much of the debate has centered on whether the Fed will pursue a new strategy of adopting a formal target for inflation. But before moving to a more rigid system, the Fed should first consider whether their favorite definition of inflation is broad enough and whether their informal inflation range is too narrow for our high-productivity world.

High energy prices and high home prices have made mainstream America more skeptical that price stability has been achieved, despite a general view among economists that two decades of disinflation, globalization, and rising credibility at central banks has brought inflation close to zero. There are three ways the new Fed could improve its anti-inflation performance, which would also help bridge the gap between what the Fed sees and what workers often feel. First, the Fed should not treat the lower bound of its inflation reference range of 1 to 2 percent more rigidly than they treat the upper bound. Second, the Fed should adopt a new measure of goods and services inflation that more accurately captures trend inflation. And third, the Fed should incorporate rising asset and commodity prices more actively into its decision-making.

Since the middle of the 1990s, the U.S. economy has been marked by a high rate of productivity growth. Catching this trend before it showed tip in statistics was one of the great successes of the Greenspan Fed. High productivity means that it takes less labor to make the same goods--in other words things gets cheaper. An important question in such an environment is whether the downward pressure on the price of goods should be resisted with extraordinary policies by the central bank.

One argument for such a policy is that deflation provides consumers with an incentive to postpone purchases since the same product will be cheaper in the future. An examination of price data since 1997 shows that the United States has experience with deflation, although it's been concentrated in certain goods. Since the end of 1997, apparel prices have fallen 13 percent in aggregate, but nominal spending on apparel still rose over 34 percent in the same period. Consumer computer prices provide an even starker example. Over the same period, prices fell 89 percent, while nominal spending on computers still rose 56 percent.

If deflation becomes an economy-wide phenomenon, however, there are more serious concerns to worry about. Deflation raises the real cost of borrowing money and makes the real return on cash positive. Both provide an incentive to save more and spend less. So corrosive deflation, as Alan Greenspan once called it, is really a condition where a shift toward thrift happens at an inopportune time and sets off a downward spiral, reducing investment, employment, and output. This happens most frequently after an asset bubble has collapsed. Asset prices drop, lowering household net worth and consumer sentiment enough that aggregate demand drops and a recession ensues.

This is the general condition that started Japan's fifteen-year stagnation and is what leading policymakers in the United States feared, with reason, in the 2001 to early 2003 period. But these unique periods that require a central banker's version of overwhelming force are not a good guide for the long periods of normalcy that fortunately occupy most of our economic history. If the Fed is focused solely on creating positive inflation at a time that both productivity and aggregate demand are high, it will create too much liquidity as a byproduct. In such cases, the Fed could inadvertently contribute to asset bubbles that would be destabilizing over time. This has an important policy ramification: in certain situations the Fed may want to tolerate an inflation rate that deviates below its reference range for goods and service prices just as it occasionally tolerates rates above its range.

A second critical issue is whether the Fed and financial markets are following the best definition of inflation. Persistently higher energy prices have made the Fed's favorite inflation measure--personal consumption expenditures excluding food and energy--seem outdated. Energy prices have been excluded from most past inflation analysis because they are extremely volatile. But in recent years their prices have been marked more by an upward trajectory than stochastic volatility.

A better conceptual measure of inflation lies within the Federal Reserve System. The Dallas Fed produces a trimmed PCE price index that strips out the most volatile components every month, rather than simply excluding food and energy. As the Dallas Fed has explained, the system is similar to Olympic judging where both the high and low score are thrown out to improve accuracy. Over the second half of 2005, this measure has been running almost a half a percentage point higher than the core-PCE index. Importantly, during the deflation scare of 2003 this measure never fell below 1.5 percent on an annual basis--meaning it never dipped into the bottom half of the Fed's preferred 1 to 2 percent inflation range. On a three-month annualized basis, the trimmed measure only fell out of the Fed's range during one month while the traditional measure excluding food and energy fell below the range in three months. While the trimmed measure still strips out gasoline the majority of the time, it also strips out the price of goods that fall the most, like computer prices, leaving a more balanced measure.

The job of measuring inflation accurately becomes more important the closer you get to zero. Paul Volcker didn't need much help figuring out that inflation was too high (though he needed extreme courage to bring it down). If you're making a VFR flight from Asia to Australia, your accuracy doesn't have to be that great as long as you're heading in the fight direction. But if you're hopping from Australia to Howland Island, knowing exactly where you are becomes essential.

The third and most vexing issue facing central banks today is the proper evaluation of asset prices, which have bedeviled monetary policymakers since the start of the Fed. Asset prices can go up for many reasons, especially in a high-productivity, low-interest-rate world, making it quite difficult to deconstruct why their prices changed. But central banks face difficulty on almost all forecasting issues, and it's important to remember that excess money can flow into Toyota stock just as easily as it can flow into a new Camry.

In the last few years, monetary policy at the three main central banks--the Fed, the European Central Bank, and the Bank of Japan--has been extremely accommodative. The policy has coincided with a ran-up in house prices, commodity prices, and financial markets across the globe. Major equity markets in Brazil and Mexico were up over 35 percent last year; Japan, India, and South Korea were up 40 percent or more; and the major markets in Europe rose 20 to 35 percent. Gold even surged 20 percent last year--all from an asset that offers nothing except protection from human error. Fortunately, broad-based goods and services inflation has generally been delayed by globalization and productivity. But it's far from certain that the price rise in assets won't eventually seep into the rest of the economy.

Soaring equity prices surely have a number of causes, but one possible contributor is an asymmetric central bank policy that reserves judgment on forming asset bubbles, but accommodates them once they pop. Traders refer to this bias as the "Greenspan put," but the United States does not have a patent on the idea. The "put" implies that investors will have the ability to get out of their stocks in a crisis because central banks will create liquidity if the markets plunge. The implied corollary is that central banks are less willing to resist markets on the way up.

Global home prices are another indicator that liquidity has been too high. The surge in home prices has been even more widespread than the stock market rally. In 2004 and 2005, $5 trillion of the $10 trillion increase in household assets in the United States was attributable to inflation in land prices or inflation in construction costs. At least traditional measures of goods and services inflation attempt to include a cost of shelter in their calculations, though it is quite a difficult job separating out the investment and user cost components of housing. The United Kingdom and the Australia are examples where the central bank took a broader look at prices and proactively worked to limit their housing bubbles. In both cases, their preemptive action proved successful.

With the benefit of hindsight, it appears that global monetary policy has been too accommodative for too long. Does this mean that the Fed has a long way to go in its tightening campaign? Not necessarily, as financial markets have started to signal some bite from monetary policy in 2006. The yield curve, for instance, has flirted with mild inversion but long rates have come up with recent tightenings. If the bond market were to hold a significant inversion, which in the past has tended to predict a slowdown about a year in advance, it could signal the possibility of entering a boom/bust cycle. Presently, the economy is running too hot, with nominal GDP growth at about 6.5 percent and the unemployment rate at 4.7 percent. But with the housing market beginning to lose some steam, the Fed has to also be careful about chasing its own tail. The FOMC shouldn't try to remove excess liquidity all at once so long as inflation expectations are still in line. The best solution is to put policy in a moderately restrictive stance and show as much patience at the top as was shown at the bottom.

Marc Sumerlin is Managing Director of the Lindsey Group and formerly served as Deputy Director of the National Economic Council under George W. Bush. The views expressed in this article are his own.