CPI and PPI biases

As I mentioned in the most recent Portfolio Insights article titled “Inflation FAQs,” there are a few problems with the Consumer Price Index (CPI) and Producer Price Index (PPI) as inflation gauges.

CPI has an upward bias that is estimated to overstate inflation by about 1 percentage point per year. This can be particularly problematic for employment contracts with cost-of-living adjustments based on CPI as well as for a substantial portion of government spending, such as entitlement payments, which automatically increase with CPI.

The most commonly cited biases that tend to overstate the CPI include:

• New goods. New products that replace existing products are often more expensive at first. This biases the index because some of the newly-available goods perform the same function as different lower-priced goods in the base-year market basket.
• Quality changes. If the price of a product increases because the product has improved, the price increase is not due to inflation, but still causes an increase in the price index.
• Substitution effect. Inflation or not, prices of goods relative to each other are always changing. When two goods are substitutes for each other, consumers increase their purchase of the relatively cheaper good and buy less of the relatively more expensive good. Over time, such changes can make the CPI’s fixed basket of goods a less accurate measure of typical household spending. The chained CPI, however, does adjust to the substitution effect in a timely manner as the basket of goods is “chained.”
• Outlet substitution. When consumers shift their purchases toward discount outlets like Wal-Mart and away from convenience outlets like the neighborhood grocer, they reduce their cost of living in a way the CPI does not capture.
  

The biases to Producer Price Indexes (PPI) are not quite as extreme and not as important because firms can absorb costs and they don’t get passed on to the consumer. But for educational purposes, here are some of the minor shortfalls of PPI:

• Industry weightings. PPI uses relative weightings for different industries, but these weightings might not accurately represent their actual proportion to real gross domestic product (GDP). As a result, the weightings are adjusted every several years, but small differences still occur.
• Hedonic adjustments. PPI calculations involve an explicit “quality adjustment method,” called hedonic adjustments, to account for changes that occur in the quality and usefulness of products over time. These adjustments may not effectively separate out quality adjustments from price level changes as intended.
• Volatile elements. Energy and food often skew the data because they are so volatile. As a result, the removal of food and energy prices is almost implicit in most media releases. However, the long-term growth rates should not be ignored if these costs grow faster than the core PPI (or CPI) over time because consumers and eventually GDP will feel the pinch.

--

Peter J. Lazaroff