Illusionary Inflation

When the March consumer price index showed a 0.6 percent rise for the month and 3.1 percent for the year, many claimed inflation had suddenly become a big problem the Fed will have to battle with a series of interest rate hikes.

Those same people did not, however, argue the Fed should lower rates in January simply because the monthly CPI then rose by only 0.1 percent. Nobody should try to find inflation trends in such monthly wiggles. We should be watching for sustained changes in year-to-year inflation rates. And Fed governors should take greater care than they did in 1990 not to confuse energy prices with a general uptrend in other prices.

Aside from energy, there has been no increase in inflation for at least three years. Take out energy, and CPI inflation over the past 12 months was just 2.4 percent -- lower than the 2.5 percent average of 1998-2002. Yet that 2.4 percent rate still overstates inflation, as the Boskin Commission explained in 1996. So, in 2000 the Fed began to focus on a chain-weighted price index for personal consumption expenditures (PCE). Such a chained index is superior to the fixed-weight CPI because people react to prices. If gas prices soar while food prices slow, people are likely to drive fewer miles and eat at home more often. By assuming they're driving just as much as before and eating out just as often, the CPI overstates the cost of living.

The PCE index covers rural areas, not just urban consumers, and it assigns more weight to medical services than does the CPI, and less to housing and new cars. Yet a newer chain-weighted version of the CPI tells the same low-inflation story as the PCE deflator. Both PCE and CPI figures for "core" inflation exclude food as well as energy. That's misleading right now because food is not why the regular CPI exaggerates inflation. Over the past three months, the CPI for energy rose at a 21.1 percent annual rate, compared with 1.3 percent for food.

For the year ending in February, the "core" PCE deflator was up only 1.6 percent. A comparable but preliminary chained CPI estimate was up only 1.9 percent in March. That latter measure is lower than it was even in December 2001, at the end of the recession (2.2 percent). It slowed to 1.9 percent in March 2002 -- the same as today -- yet energy prices were down by 10.9 percent and are now up by 11.6 percent. If we had a figure for the chained CPI less energy alone (rather than energy and food), that would be up about 2 percent over the past 12 months.

 Aside from energy, inflation over the past 12 months has been lower than the 1998-2002 average -- even though price indexes that exclude direct costs of energy (electricity and fuel) nonetheless remain distorted upward by such oil-driven remaining items as taxi services, shipping charges, etc.

Some argue that increases in the producer price index portend faster price increases ahead for consumers. But the PPI does not predict the CPI, and there's no reason to expect it to. The PPI measures prices businesses charge each other, and it only covers goods. Half of the CPI is services, and many goods in the CPI are imports, unaffected by U.S. business costs.

Aside from energy, prices of imports have also not been noticeably affected by last year's decline in the dollar. Import prices of non-automotive consumer goods were only 1 percent higher in March than a year ago -- cars and parts were 1.3 percent higher. The overall Bureau of Labor Statistics index of all import prices less fuels (which includes business imports) was up just 2.5 percent for the year ending in March. Besides, the dollar is up, not down -- it rose 2.1 percent against major currencies from Dec. 1 to April 22.

Other indicators of incipient inflation are benign. The price of gold, a classic hedge against inflation, is the same as a year ago. The 4.2 percent yield on 10-year Treasury bonds and nearly flat yield curve would be a "conundrum" only if inflation was actually worrisome, which it evidently is not. The gap between that nominal rate and the yield on inflation-protected Treasuries implies expected future inflation (including energy) of about 2.6 percent.

New York Times columnist Paul Krugman recently wrote, "In the 1970s, soaring prices of oil and other commodities led to stagflation." In reality, combining inflationary monetary policy with suffocating tax policy gave us broad-based inflation and three increasingly nasty recessions. At that time, unlike now, oil was a sideshow. Aside from energy, CPI inflation was 6.1 percent in 1970, 9.8 percent in 1974, 8.9 percent in 1975, 10 percent in 1979 and 11.6 percent in 1980. How could anyone rationally compare such figures with today's 1.9 percent non-energy inflation?

Krugman went on to ask: "What's driving inflation? ... Oil prices are a big part of the story, but not all of it. Other commodity prices are also rising." Former Fed Governor Wayne Angell showed otherwise in the Wall Street Journal -- his index of 21 commodities includes oil, yet it has not risen at all over the past 12 months. The Economist index of commodity prices for the year ending April 19 was down 4.3 percent.

Aside from energy prices, which have probably peaked, there has been no acceleration of inflation in the past three years. Overpriced oil matters, not because it's inflationary but because it quickly turns profits into losses in energy-intensive manufacturing. Industrial production has fallen for a whole year in South Korea, Taiwan, Singapore, Italy, Spain, Belgium and Denmark. Fourth quarter economic growth was barely above zero in Europe and Japan. Slumps are infectious. The United States is not necessarily immune, as the three-month drop in durable goods orders might portend.

I have avoided complaining about the Fed being too tight since co-authoring a Wall Street Journal article to that effect in October 2000 (and before that, only in mid-1982 and 1984). If asked today if the Fed should again raise interest rates anytime soon, however, my answer is no. The domestic and global economic risks of raising dollar interest rates are real -- the inflation scare is not.