Inflation Estimation Problems

John Vandivier
Inflation is reported by the government in the US as the inflation derived from CPI, not the GDP Deflator. It is calculated by the Bureau of Labor Statistics. Problems:
  • Heterogeneity error
    • This effects GDP Deflator as well
    • Geography
    • Age
    • Many other factors
    • Your personal inflation depends on the price behaviors of the goods you consume.
    • A non-subjective, market-wide estimate depends on non-selective averages.
  • Substitution bias
  • Quality changes (up or down)
  • Goods no longer bought, new goods on the market
  • Selection bias for the basket of goods
    • Political determination of the basket of goods
    • Confirmation bias on the size on inflation
    • Calculation problem from a restricted set of minds estimating the market effects
  • Representative of the \"average consumer.\"
    • This consumer doesn't exist.
    • This consumer doesn't represent a real measure of economic growth. The full distribution of consumption better predicts market behavior. Market calculations are based on full distribution calculations. This is another form of a calculation problem. The GDP Deflator better accounts for this and many of the other errors.
    • The \"average consumer\" is selectively determined and as a result selection bias occurs once again. The \"average consumer\" is constructed by taking various averages of various selected behaviors and consumptions.
In conclusion, a more proper estimate of inflation would be based on GDP, as GDP is far less selectively determined and biased than the CPI, although GDP is a simplification as well. The secondary goods market matters for the economy as well, for example.