Toward a Cost-of-Living Index: Progress and Prospects

摘要

Almost anyone with an interest in consumer price trends wants to know how those trends affect consumers, and, in principle, that effect is best captured by an appropriate cost-of-living index. It therefore seems curious that so few of the statistical offices around the world have adopted the cost-of-living index objective for their consumer price indexes. "Accurately tracking the cost of purchasing a fixed market basket of goods and services" no doubt offers a more comfortable standard for a statistical agency to aspire to than "accurately tracking the cost of living." But if the latter is what one really would like to measure, it seems better just to say so. For some years, the theory of the cost-of-living index has guided Bureau of Labor Statistics (BLS) decisions about the Consumer Price Index (CPI), and in the recent past, the agency has stated more explicitly that the goal for the CPI program is to approximate a "conditional" cost-of-living index―that is, an index that measures changes in the cost of living that are due to changes in prices of goods and services, conditional on other outside influences that affect the standard of living remaining unchanged. It may be that a true cost-of-living index is unattainable as a practical matter. Still, the Consumer Price Index surely can be moved closer to that ultimate objective, and the theory of the cost-of-living index has the considerable advantage of offering a unified framework for the evaluation of the methods used to produce the CPI. In this paper, I consider the improvements the Bureau of Labor Statistics has made to the Consumer Price Index since the mid-1990s. These include changes designed to make the index more representative of current expenditure patterns; to account for consumer substitution in response to relative price change; and to account for changes in the quality of the goods and services that consumers purchase. Drawing on selected recommendations made by the National Academy of Sciences Panel on Conceptual, Measurement, and Other Statistical Issues in Developing Cost-of-Living Indexes recently chaired by Charles Schultze (hereafter, the Schultze panel), I highlight those additional areas for exploration that I believe have the greatest potential and discuss those areas I believe to be less promising.