“Facts are stubborn things, but statistics are more pliable.” — Mark Twain
Economic statistics are backward looking, not forward; in that sense, they always reflect the past. Moreover, given data collection time lapses, costs, and other issues, the data is always incomplete and more or less out of date. Finally, in a vast and heterogeneous economy, the applicability of one statistic to everyone is impossible. These are the general issues that any statistical service wrestles with. No figures are going to be perfect. At the same time, it’s important to know where the numbers you depend on come from, and what their strengths and weaknesses are.
The Consumer Price Index (CPI) is intended to reflect the cost of living, which is the cost of all items and services that a typical consumer would buy over time. The methodology is based on monthly price samples of 211 item categories in 38 different locations, for a total of more than 8,000 category indices. Category examples include “apples in Chicago” or “eggs in Boston.” The categories are regularly priced, typically on a monthly or bimonthly basis, at outlets selected based on consumer surveys. The Bureau of Labor Statistics publishes three major indices: the CPI for All Urban Consumers (CPI-U), the Chained CPI for All Urban Consumers (C-CPI-U) and the CPI for Urban Wage Earners and Clerical Workers (CPI-W). The CPI-U is used for indexation of tax brackets, personal exemptions, and other tax adjustments. The CPI-W is used for determining social security payments. These figures matter to a lot of people in a very immediate way.
The CPI-U is intended to reflect the buying habits of residents of major metropolitan areas, and the CPI-W is intended to reflect the employed subset of this population. All of the prices used are the same in both series, but the weightings are different. For example, gasoline has a higher weighting in the CPI-W, reflecting the commuting costs associated with employment. The CPI-U includes about 87 percent of the population based on the 1990 census, while the CPI-W only covers about 32 percent of the population as of 1990. The CPI-U includes all households in metropolitan areas in which wages account for 50 percent or more of income, excluding households of professional and salaried workers, part-time workers, the self-employed, and the unemployed, as well as households with no one in the labor force, such as retirees.
So, we have two major consumer inflation indices that are widely used. These price indices, though, reflect either extremely wide populations or may well exclude people like you. Specifically, they exclude most of the mass affluent—salaried professional workers, the self-employed—and the retired. Why does this matter? Because what you are experiencing, and need to plan for, may be significantly different than what is reported.
“Oil prices have fallen lately. We include this news for the benefit of gas stations, which otherwise wouldn’t learn of it for six months.” — William Tammeus
Interestingly, though, you may not see a higher personal inflation rate; it could well be lower. This is because the typical consumer spending weights vary by income level. For example, to use the highest-profile items today, food and gasoline account for 15.3 percent of purchases for the lowest-income quintile and only 9.2 percent of the highest-income quintile. Given the dramatic increases in both categories over the past year, the poorer households are seeing substantially bigger increases in their costs of living than the affluent households.
Other factors that can affect personal inflation rates, both real and perceived, include the fact that costs of frequently purchased items—again, food and gasoline—have been increasing recently. These increases have been offset by cost decreases in TVs and computers, as well as clothing, but the decreases are less obvious as these items are purchased much less frequently.
In aggregate, this explains why my father sees an inflation rate significantly higher than what is reported. For him, and for many consumers, it is. Within these data collection limitations, though, the CPI appears relatively straightforward. You take a basket of goods and compare how the prices change over time. As usual, though, the devil is in the details, and this is where the major criticisms of the CPI methodology come into play.