Reporting inflation

• Article
• Comments ()
• Click-2-Listen
• Videos

By

When the Labor Department reports on consumer inflation every month, do you think the data should count items that consumers don't buy or only items they do buy? If you say the former, you agree with how the press reports the data. If you say the latter, you're right.

The featured price number that makes the monthly headlines includes items that some consumers choose not to buy or to buy less of because of changes in relative prices. For example, if steak gets pricey, some shoppers will buy cheaper chicken instead. But such short-term shifts in buying patterns are not reflected in the featured data.

The Labor Department also publishes in the same consumer price press release an alternative series called the Chained Consumer Price Index, which measures what people actually purchase, i.e., the index uses weights that reflect substitution buying in response to changes in relative prices.

Consequently, as is well known by economists, the featured unchained price index is upward biased, whereas the chained index paints a more accurate picture of consumer inflation.

On Jan. 18 the Labor Department reported in the opening paragraph of its press release that the Consumer Price Index for All Urban Consumers in December was 3.4 percent higher than a year earlier. This compared with a 3.3 percent rise the previous year and smaller increases in the three prior years.

The inflation story was widely reported in the media. Here's a sample of newspaper headlines: "Prices Surged in '05" (The Cincinnati Post); "Consumer Prices Rose by Largest Rate in 5 Years in 2005" (New York Times); "Inflation Hit Five-Year High of 3.4% Last Year" (The Washington Post).

By comparison, the more accurate Chained Consumer Price Index for All Urban Consumers in December 2005 was up by 2.8 percent over a year earlier, a whopping 0.6 percentage point less than the featured number. Although a 2.8 percent inflation rate is less than gratifying and above the Federal Reserve's comfort range, its implications are a far cry from the misleading message of alarmist headlines. You would be hard put to find any mention of the chained index in the media inflation stories.

Nor was the preferred chained measure of consumer inflation, unlike the unchained number, higher than in the previous year. The 2.8 percent year-over-year chained inflation rate in December 2005 was slightly less than the 2.9 percent rate for the year before, again inconsistent with the five-year-high inflation message of some major media headlines.

For the core inflation rate (total less food and energy), the unchained series, at 2.2 percent, was the same year-over-year in December 2005 as in December 2004, whereas the chained index increased more slowly in 2005, by 1.7 percent December over December compared with 2.0 percent in 2004.

The media have a responsibility to learn about the economic data they report and interpret. Perhaps reporters should lunch more often with economists. The reporting of economic statistics has consequences. It influences people's behavior and expectations. It can affect work, spending and saving decisions.

For example, if people are led to believe that inflation is greater than it is and getting worse, they may try to beat expected future price increases by spending more now. An advanced buying splurge could push up demand thereby driving up prices and exacerbating inflation. Such behavior could become a vicious cycle with escalating prices eventually ending in an economic bust.

The Federal Reserve's worst nightmare is that price expectations will become unanchored, leading to accelerating inflation. In that event the Fed would have to raise interest rates sharply to fight the inflation, thereby choking off economic growth and inviting recession. What the Fed doesn't need, nor do the rest of us, is the major media sounding dissonant alarm bells about current inflation based on second-best data.

The Labor Department also has the responsibility of pointing out to the press the differences between the price indexes it publishes, in particular their relative strengths and weaknesses. Similarly, White House economists and the Fed need to be more vocal on the issue.

Fed Chairman Alan Greenspan has talked about the superiority of chained prices indexes over unchained measures, but not loudly enough. One hopes that Fed Chairman-designate, Ben S. Bernanke, currently chairman of the President's Council of Economic Advisers, will emphasize the point in the council's forthcoming annual report. The Congress should also be asking questions, out loud.

If prominent government figures and the issuing statistical agency lead the way, the press will follow.

Alfred Tella is a former Georgetown University research professor of economics.