Inflation: Simple Idea, Complex Measurement

 

Inflation, the general increase in prices, is a well-known concept. We read about it in the press, we feel its impact in our pocketbooks, and the Fed watches it intensely in determining monetary policy. But it can also be confusing.

 

In March, the consumer price index (CPI) increased 0.6%, or 7.5% at an annual rate, but the change over the last year was only 2.8%. If that isn't confusing enough, the core CPI, net of food and energy, increased only 0.1%, or 0.7% at an annual rate, but it rose 2.5% over the last year. The personal consumption expenditures (PCE) deflator, another measure of inflation, rose 0.4%, or 5.4% at an annual rate, but 2.4% over the last year. But the core PCE increased 0.0% for March, or 0.6% at an annual rate, and 2.1% for the year. So what's the real story? Is inflation high or low, and is that good or bad for the economy?

 

 

Let's take a closer look at some of these measures and see if we can sort through the morass. The most commonly used measure of inflation in the United States is the CPI. The Bureau of Labor Statistics (BLS), a division of the Department of Labor, computes the index by recording the prices of a basket of goods and services consumed by a representative household, weighting each component in the basket and then aggregating the components into the index number. Inflation is then calculated as the change in the index over the desired time period such as a month, quarter, or year. The choice of which particular goods and services to include in the basket, as well as the weights attributed to each component, can have a profound impact on the subsequent calculation of inflation.

 

The CPI is calculated using fixed weights for the contents of its basket. This means that as the relative prices of similar goods change, the index assumes that consumers do not substitute the cheaper good for the more expensive. For example, if the price of butter increased relative to the price of margarine, we would expect that households would consume less butter and more margarine. The CPI calculation, though, assumes no such substitution. Because this procedure introduces an upward bias into the CPI, BLS has attempted to improve its measure of inflation by revising the weights every two years. However, even with the adjustment, there can be significant mismeasurement within a two-year interval.

 

In 2000, BLS, aware of this bias in the CPI, began producing the chained consumer price index (C-CPI). The weights in the C-CPI change each month to take into account product substitution given price changes. Although the C-CPI is probably a better measure of changes in the cost of living conceptually, it is not currently used in any federal legislation as an adjustment mechanism for programs such as Social Security. This is likely because it only goes back to 2000.

 

The Bureau of Economic Analysis (BEA), a branch of the Commerce Department, constructs an alternative price index based on personal consumption expenditures data. The BEA uses a similar formula to the C-CPI for its PCE price index. It changes the relative importance of the PCE index components to try to adjust for product substitution as a result of price changes on a monthly basis. 

 

Moreover, BEA takes an extra step to incorporate a number of price proxies for goods and services that households consume but for which they do not pay "out of pocket," such as the cost of health insurance and government benefits. As a result, the total number of items in the PCE index basket is larger than that of the CPI basket. Consequently, there are notable differences in the magnitude of the weights for many goods and services in each index. 

 

For example, the share of housing costs in the CPI is 42% compared with about 23% in the PCE index. Moreover, while health care represents about 6% of the CPI, this category is about 20% of the PCE index basket. Similarly, "other goods and services" comprises about 4% in the CPI while it represents about 13% of the PCE index, in part because the latter includes services provided by not-for-profit organizations and the government that the former does not.

 

Housing is by far the largest component of the overall CPI, with a weight of 42%. Housing is made up of two categories: shelter and other housing. Shelter, with a weight of nearly 33%, is itself composed of rents and owners' equivalent rents. The latter is derived from owners' survey responses about how much they think their homes would cost to rent given current market conditions. 

 

Considering the house price boom of recent years and the corresponding fall in housing affordability, it is not surprising to see a rise in the CPI's shelter component over the past year. This is because the high cost of purchasing a home propelled the demand for rental units. 

 

In fact, the year-over-year percent change in shelter costs actually accelerated from 2.0% in September 2005 to 4.2% in September 2006. Since then, this rate has moderated to 4.0% in March 2007. And as the housing and rental markets come into balance (a greater supply of rental units has resulted from the recession in the for-sale housing market), it is likely that rental costs will moderate over the course of 2007.

 

While housing costs are a major component of the index, the changes in housing costs are actually relatively small if compared to fluctuations in other components such as energy and food. There are times when one is more interested in the underlying trends in inflation rather than short-run fluctuations and indexes that include highly volatile components, which can be hard to interpret. For these times, we can strip out such volatile components as food and energy and designate these new measures as "core" indexes.

 

To illustrate, chart 3 (bottom) demonstrates the volatility in the energy component. The CPI's energy component rose 14.5% between December 2005 and August 2006, then dropped 13.7% between August 2006 and November 2006, and finally rose 9.6% between November 2006 and March 2007. During the same periods, the overall CPI changes were 3.0%, -0.9%, and 1.6%, respectively. The core CPI, however, was much more stable with changes of 2.0%, 0.4%, and 0.7%, respectively.

 

The core measures of inflation are also important because they tend to be watched more closely by the Federal Reserve, and the Fed's target for inflation is most often thought of in terms of core inflation, specifically the core PCE deflator. The Fed's focus on core rates of inflation stems from its desire to achieve long-run price stability. The core measures of inflation that remove the highly volatile components tend to better reflect the underlying trends and root causes of inflation.

 

For example, when economic growth is strong, the economy employs a larger percentage of its available labor supply and puts upward pressure on wage rates. Higher wage rates, in turn, put upward pressure on unit labor costs that eventually result in higher prices for goods and services. In such circumstances, the Fed tightens monetary policy by raising interest rates and slows the economy to remove the pressure.

 

With the unemployment rate currently below 5%, the tight labor market has indeed pushed up wages and unit labor costs, thus triggering the Fed's recent tightening. With economic growth slowing over the last year, we expect the unemployment rate to rise modestly, wage pressures to ease a bit, and core rates of inflation to gradually decline. If we are correct, the Fed could begin to ease monetary policy toward the end of the year.