Back to Dashboard

Historical Housing Prices Project


Definitions for terms included in the Historical Housing Prices Project.

City (variable “city”)

The city is defined by the coverage of the housing market from historical newspapers at a point in time, including suburbs and other outlying areas beyond the city’s jurisdictional boundaries. The geographic area associated with most cities in the sample thus expands over the 20th century.

Year (variable “year”)

Newspaper real estate sections were sampled from one to four times a year, with the end of May the most common time, to obtain a targeted number of listings for both homes for sale and for rent.

Housing Price Index

An index that measures the quality-adjusted change in the price of housing over time relative to a base year. To adjust for quality, we control for the number of rooms in the housing unit, the housing unit type, and a measure of location. We also include a flexible set of controls for unobserved quality changes in our hedonic regression framework. Users interested in the specification details can find them in Section 4 of the associated working paper, “The Price of Housing in the United States, 1890–2006.”

City Indices (variable “rl_rpi_b_1948” for rents and “rl_hpi_b_1948” for sales in city indices)

The city-level indices are in real terms (i.e., they account for inflation). The city-level indices are relative to 1948 levels, as not every city in the sample had a dense enough housing market to support a newspaper real estate section before that year.

National Indices (variable “rl_pw_rpi_b_1890” for rents and “rl_pw_hpi_b_1890” for sales in national indices )

The national indices are in real terms and relative to a base year of 1890. The 30 sample city indices are aggregated by population (interpolated across census years) to obtain these national indices.