Comparing new final-demand producer price indexes with other government price indexes
Categorization. The PPI and the CPI categorize a number of goods and services differently within their index structures. Differences in categorization for goods and services are mitigated at high levels of aggregation but can create discrepancies at lower levels. The PPI, for example, classifies utilities, such as electric power and natural gas, as goods, while the CPI categorizes utilities as services. The overall PPI for personal consumption and the CPI both include utilities; however, the PPI for personal consumption services excludes utilities, while the CPI for services includes utilities, making the two services indexes less comparable than the overall indexes.
The PPI and the CPI also differ in their categorization and treatment of trade and transportation. The PPI generally separates the costs of transporting, retailing, and wholesaling goods from the cost of the good itself and classifies trade and transportation as services. In contrast, prices for goods as measured by the CPI typically include the value of the good, the value of transporting the good, and the trade margins associated with the sale of the good.
Other technical differences. Several other technical differences exist between the PPI for personal consumption and the CPI. The PPI and the CPI are both constructed with the use of a modified Laspeyres index formula, but the CPI updates weights every 2 years and the PPI updates weights every 5 years. The CPI also implements a geometric mean formula at the item level that the PPI does not. The geometric calculation reduces substitution bias, leading to lower measures of inflation in periods of price increases. The PPI attempts to collect prices for a specific day of the month (the Tuesday of the week containing the 13th), while the CPI collects prices throughout the month. Finally, prices measured by the CPI include sales and excise taxes, while prices measured by the PPI exclude those taxes.
Historical index comparison. Figures 1 and 2 respectively compare index levels and percent changes of the PPI for personal consumption with index levels and percent changes of the All Items CPI, with neither index seasonally adjusted.7 The figures show that, in spite of the aforesaid methodological differences between the PPI and the CPI, the indexes exhibited similar trends and similar monthly percent changes over the sample period. From January 2010 through June 2012, the all items CPI rose 5.9 percent and the PPI for personal consumption moved up 6.4 percent. The standard deviations of the monthly percent changes in the CPI and PPI were 0.30 and 0.34, respectively, and the correlation between the monthly percent changes was 0.91. Note, however, that although the PPI for personal consumption and the CPI are highly correlated over the timeframe examined, the sample period is small and there was considerable movement in food and energy prices, which the PPI and the CPI measure similarly, over that period. If, instead, there had been high volatility in an area such as owners’ equivalent rent or banking, the indexes may have been far less correlated.
7 For comparison purposes, indexes in all figures were rebased to equal 100 in the first period of the sample.