Quantifying Quality Growth

We introduce an instrumental variables approach to estimate the importance of unmeasured quality growth for a set of 66 durable consumer goods. Our instrument is based on predicting which of these 66 goods will display rapid quality growth. Using pooled cross- relatively sections of households in the 1980 through 1996 U.S. Consumer Expenditure Surveys, we estimate quality Engel curves' for 66 durable consumer goods based on the extent richer households pay more for a good, conditional on purchasing. We use the slopes of these curves to predict the rate of quality-upgrading. Just as if households are ascending these quality Engel curves over time, we find that the average price paid rises faster for goods with steeper quality slopes. BLS prices likewise increase more quickly for goods with steeper quality slopes, suggesting the BLS does not fully net out the impact of quality-upgrading on prices paid. We estimate that quality growth averages about 3.7% per year for our goods, with about 60% of this, or 2.2% per year, showing up as higher inflation rather than higher real growth.