The Missing Middle in Product Price Distribution

The IO literature has typically studied the supply-side factors that determine the price structure of products/services competing in a market. This paper pro-poses that the demand-side demographics could play an important role in shaping the product price structure. In particular, we document a “missing middle” phe-nomenon in both the income and the product price distributions in the U.S., based on the IPUMS ACS dataset (2005–2017) and the Nielsen Retail Scanner Data (2006–2017), for a large set of goods sold in the U.S. at the national, state, or commuting-zone level. We show that the lagged population share of the middle-income class has a positive impact on the market share (in quantity) of middle-priced varieties (and respectively so for the low/high income and price group), after controlling for product category and state (or commuting zone) fixed effects. The impacts are further stronger in commuting zones of higher population density. We then evaluate the cost-of-living implications of the observed missing-middle phe-nomenon, taking into account product entry, exit, and pro-competitive price effects of continuing products, in a framework that allows for non-homothetic preferences across income groups with respective to the price groups. We find that ignoring the non-homothetic demand structure and the missing-middle phenomenon under-states the rise in the cost of living for the period 2006–2017. The downward bias is sizable (as large as 2 percentage points out of 11–13% increase in the cost of living for the period), and particularly noticeable for the middle-income households.