[Access article in PDF]
Quantifying the Qualitative:
Quality-Adjusted Price Indexes in the United States, 1915–61
While economists had taken up questions of the best way to calculate an index of the “average level of prices” in the nineteenth century, the increased application of these concepts in the twentieth century raised a number of new questions.1 Not the least of these questions was how to use the various index formulae when goods appeared and disappeared from the market. A closely related question was how to account for more subtle changes in the quality of the goods. At the same time, questions were arising about the true purpose of a price index. The history of these related concerns provides an interesting case of otherwise abstract ideas hitting concrete barriers in the form of data unwilling to conform to constructed compartments, a confused public, and political interests.
This paper tracks this history in the United States, examining three snapshots corresponding to important reviews of the government's price indexes and interspersing these snapshots with new ideas imported from Europe. These reviews, in two cases during world war, are explicit examples of measurement as statecraft (see the article by Theodore Porter in this volume): making the indexes more fair and accurate would improve them as tools for negotiating wages and setting other payments. [End Page 345] (See Sandra Peart's and Marcel Boumans's essays in this volume for two other contexts for price indexes.)
I begin with a review by Wesley Clair Mitchell in 1915 of the U.S. wholesale price index and with the work of Irving Fisher of about the same period (1922). Mitchell led a second, more extensive review of the government's price indexes during World War II, when price changes were particularly severe. The third review, completed in 1961, was conducted by a committee of the National Bureau of Economic Research (NBER) chaired by George Stigler, which used work by Zvi Griliches on “hedonic prices” to adjust for quality change. These three episodes, plus two interludes, form the outline of this paper. In each case, the ways these authors characterized the problem of quality change—and hence also the ways they tried to overcome the problem—have paralleled the ways in which they have conceptualized the price index itself.
Mitchell and Fisher
I begin this story with two prominent American economists who left a long-standing mark on the study of index numbers, Wesley Clair Mitchell (1874–1948) and Irving Fisher (1867–1947). Both men brought a broader interest in the quantity theory of money and business cycles to their work on price indexes, which included both calculation of actual index number series and theoretical writings. They were also colleagues and friends.2
Early in his career, Mitchell had investigated issues related to the history of money and inflation, especially during the civil war era (Hirsch 1970). He was also a major contributor to the formation of the official price indexes of the United States. The first important contribution was his review of the U.S. Bureau of Labor's wholesale price index in 1915 (revised 1921), published as “The Making and Using of Index Numbers.” Among other things, Mitchell suggested that the wholesale price index was not an appropriate tool for measuring the general purchasing power of money for consumers (24). This point became important when, during World War I, the Shipbuilding Labor Adjustment Board began to use price indexes to control wages.3 Accordingly, the U.S. Bureau of Labor Statistics developed a new “cost-of-living index” which became [End Page 346] available in 1919. (At the time, the term “cost of living” did not imply a link to welfare economics as it does today, but meant loosely an index of consumer prices for the purpose of wage adjustments.) Mitchell also served as editor of the War Industry Board's History of Prices during the War (1919).
Like Mitchell, Fisher had firsthand...