THE HISTORY OF INDEX NUMBERS The honor of inventing the device now commonly used to measure changes in the level of prices probably belongs to an Italian, G. R. Carli. In an investigation into the effect of the discovery of America upon the purchasing power of money, he reduced the prices paid for grain, wine, and oil in 1750 to percentages of change from their prices in 1500, added the percentages together, and divided the sum by three, thus making an exceedingly simple index number. Since his book was first published in 1764, index numbers are over 150 years old.
It was in England, however, where practically the same device had been hit upon by Sir George Schuckburg-Evelyn in 1798, that the theory and practice of index numbers were chiefly developed. The generation that created the classical political economy was deeply interested in the violent price fluctuations that accompanied the Napoleonic wars and the use of an irredeemable paper currency from 1797 to 1821. Several attempts were made to measure these fluctuations, and in 1833 G. Poulett Scrope suggested the establishment of a "tabular standard of value." Interest in the study of price fluctuations lagged somewhat in the forties; but the great rise of prices after the Californian and Australian gold discoveries started fresh investigations. W. S. Jevons in England and Adolf Soetbeer in Germany gave a powerful impetus to the theoretical discussion and the practical computation of index numbers. The problem changed somewhat in form but received even more attention after 1873, when a prolonged fall of prices began. In the sixties the chief aim of investigation had been to discover the relations between the rise of prices and the increased production of gold; in the seventies and eighties the chief aim was to find the relations between the fall of prices and the restrictions placed upon the free coinage of silver. The weightiest theoretical contributions of this period were made by Prof. F. Y. Edgeworth, who served as secretary of a committee appointed by the British Association for the Advancement of Science "for the purpose of investigating the best methods of ascertaining and measuring variations in the value of the monetary standard." The problem of price fluctuations entered upon another phase when the world-wide rise of prices which began in 1896-97 had been under way for several years. After 1900, and more insistently after 1910, complaints about the rising cost of living became common in all civilized countries. Efforts to measure this increase as well as efforts to explain it multi
Index numbers are both troublesome and expensive to compile, yet now in the United States not less than seven wholesale-price series are currently maintained, four of them by financial papers. In England there are four important series; in France one; in Germany, before the beginning of the World War, there were three; while the Governments of Canada, Australia, South Africa, India, Netherlands, and New Zealand now publish official index numbers, and private investigators have made series for Italy, Japan, Belgium, Denmark, Norway, Austria, Spain, and Sweden, although not all of these were kept up during the war period. This list may well be incomplete at present, and is almost certain to require additions within a short time.
Most of the series just mentioned have been established but recently. The oldestóthat of the London Economistówas begun in 1869. Sauerbeck's English series dates from 1886, Conrad's German series from 1887 (though in a sense it continues investigations made by Laspeyres in 1864), and Bradstreet's American series from 1897. Of the remaining index numbers regularly published at present, all date from years since 1899, and the majority from years since 1909.
With this increase in numbers there has come an improvement in quality. The early index numbers were made by private investigators, at irregular intervals, from such price quotations as chance had preserved. As public appreciation of the importance of measuring changes in price levels has developed, the work has more and more been assumed by financial journals and Government bureaus. This shift has produced a greater measure of continuity in the series, as well as greater frequency, regularity, and promptness in the publication of the results. Even more important is the improvement in the character and the scope of the price quotations from which the index numbers are made. Whereas the individual investigator had to take what he could get in the way of data, financial journals and Government bureaus can collect those current prices that are best adapted for statistical treatment, and can give better assurance of the representative value of their quotations and the uniform quality of the commodities included in successive years.