Answers are grounded in this paper and cite their sources.
The cost of living index of the United States Bureau of Labor Statistics is in constant use to indicate changes in purchasing power of money, and hence in real wages and real earnings; it has therefore a function of considerable importance. The index is often said to be not so representative as is desirable because of the limited data, both on consumption and on prices, with which the Bureau is provided, and a plea is made for further data. The purpose of this note is to show an evident distortion present in the index, preeminently for the period 1916 to 1921, and to indicate an improvement which can be made now. The Bureau constructs its index on the usual hypothesis that the relative consumption of different commodities remains unaltered month by month and year by year, although changes in price may cause relative expenditures on different commodities to change. That is, the index adopted is a series which compares, at successive dates, the cost of a standard budget of constant commodities and quantities. The construction procedure is to combine commodity price indices based on 1913 by weights taken to represent proportional expenditure on the respective commodities in 1913, and as the only considerable body of expenditure data obtained for any time since 1901 is that of the Bureau's cost of living investigation of 1917-1919, it is natural to use these data to provide expenditure weights for the series. This the Bureau did, but unfortunately in such a way as to cause considerable distortion. To proceed consistently with its own investigation results, the Bureau should have used expenditure weights which assumed as constant, over the time covered by the index series, the consumption habits found by investigation in 1917-1919. Instead, it assumed as constant over the period consumption habits which differ markedly from those found in 1917-1919 and for which there is no evidence at any date. This assumption, apparently generally unnoticed, is implicit in the adoption, unadjusted, of the 1917-1919 expenditure distribution to represent expenditure distribution in 1913. Table I shows what considerable contrast exists, especially for rent, between the expenditure distribution actually found and that imputed by the Bureau's index. This table represents also the contrast between
No takes yet. Share an insight, caveat, or question.
Margaret H. Hogg (1931) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: