Association for Public Policy Analysis & Management

In a competitive rental housing market, two otherwise identical units with the same cost of operation but different quality should not fetch for the same price. Contrary to this, using micro data on rents and units from the AHS, I find a very small spread in rents between low and high quality rental units in the Milwaukee market among working-age households. Namely, assuming households sort by income into quality, I regress unit type and size-adjusted rents on household incomes in the Milwakee CBSA to estimate whether, on average, a higher quality unit fetches a higher price. I find that in Milwaukee, a household annually earning $50,000 more will only pay $65 more in monthly rent for higher quality neighborhood and unobserved unit characteristics. To partly explain this phenomenon, I find that households who are more entrenched in the rental market are more likely to pay a premium and that in CBSAs with more expensive ownership markets, the spread in rents along the quality dimension is 2-4 times higher. This suggests that landlords of mid or high quality units in CBSAs such as Milwaukee may respond to low demand with low prices while landlords of low-quality units may charge higher prices due to high demand. Finally, I create Herfindahl-Hirschmann Indexes at the Census block level and find that market concentrations are significantly higher in low-price, low-income neighborhoods. This suggests landlords of low-quality units may hold more market power but direct tests of this await access to restricted AHS data from the Census.