Relationship Among Different Price Indices

The consumer price index (CPI), the producer price index (PPI), the export and the import price/unit indices are all input in the compilation of national accounts at constant prices. Thus, the implicit GDP deflator is a weighted average of the above mentioned deflators.

Should a unified picture about price changes be expected when observing the different indices? Not necessarily, but in most situations there will be a a strong correlation between the different price indicators. However, for instance in situations were export prices changes rapidly - due to major changes in demand/supply on the world market and/or big exchange rate fluctuations - only limited influence on the CPI should be expected, resulting in rather large deviations between the development of the two indices.

As seen by the examples below, which price index is chosen as a measure of inflation makes a difference. In the first example, the CPI indicates that the price level is rising by 2,2 percent while the implicit GDP deflator indicates it is decreasing by half a percentage. The difference is due to dissimilar coverage and formulas. While the implicit GDP deflator reflects the change in the export unit value, the negative change in the export price is not represented in the CPI. Furthermore, while the CPI is a Laspeyres index – which represents the upper boundary of the price change, the implicit GDP deflator is a Paasche index – representing the lower. The same effects are seen in the second example as well. However, the difference between the GDP-deflator and the CPI is smaller. This is because the export prices are falling by six rather than ten percentage, the GDP deflator is a Fisher index rather than a Paasche – thus, not representing the lower boundary, and export has less weight in this country compared to the one in the first example.

In some cases, large differences in the development of the indices are due to mistakes, in addition to differences in formulas and coverage. Thus, it is important to check the underlying statistical material in order to figure out why differences occur.