.svg)
Rental listing data for public policy research
Rental listings record what a landlord is asking, for which unit, in which month, across markets where official rent statistics arrive once a year and carry wide margins of error below the metropolitan level. That timing lets a rent movement be dated to the policy that may have caused it, and it makes the advertised stock itself observable: how many units are on the market, how long they sit, and what kind of unit is being offered. The trade is that listings capture the marketed segment rather than the occupied one, which bounds what can be said about the rents sitting tenants actually pay.
How policy researchers use rental listing data
Tracking asking rents between survey releases
Listing level rents give a monthly or weekly series for a ZIP code or neighbourhood, letting researchers date a rent movement to the month an ordinance, a disaster or a subsidy change took effect rather than waiting for an annual survey estimate that carries wide margins of error at that grain. The measure is the asking rent on units that were marketed, not the contract rent paid by a sitting tenant, so it moves earlier and further than household rent burden and says nothing at all about units that never reach a listing platform. Post disaster housing work leans on this asymmetry directly, since the displaced households competing for units are bidding on exactly the advertised segment the data covers.
Evaluating tenant protections and rent regulation
Evaluations of rent stabilization, just cause eviction rules or source of income protections pair the effective date with listing rents and listing counts on either side of a jurisdiction boundary, using the boundary to hold local labour market and amenity conditions roughly fixed. What breaks is composition: regulated or subsidised units may be advertised through different channels, or withdrawn from open marketing entirely, so a change in the mix of what gets listed can read as a price effect. Listing volume and time on market are worth carrying alongside rent, because a supply side response often shows up in how many units are offered and how long they sit before it shows in the asking price.
Short term rental rules and the long term stock
Where a city caps or licenses short term rentals, the policy question is whether restricted units return to the long term market, which researchers test by comparing long term listing volumes and asking rents in affected areas against comparable areas without the rule. Listings describe the advertised stock only, so a unit held vacant, taken back by an owner, or let through an informal channel leaves the series without a trace, which puts a hard floor under how much of any conversion effect can be attributed. Joining listings to parcel and ownership records by address makes a unit level test possible, at the cost of restricting the sample to addresses both sources record.
Rental listing datasets available on Dewey
Dwellsy
US rental listings at the individual unit level, which policy researchers use to build asking rent and listing count series for geographies and unit types that annual survey estimates report only with wide margins of error.
View datasetRentHub
US residential rental and property listings, useful for observing how the advertised stock in a jurisdiction changes in volume, price and unit mix after a tenant protection or land use ordinance takes effect.
View datasetATTOM
Residential property records covering ownership, transactions and assessed value, which join to listings by address so that a rent series can be split by landlord type or ownership concentration rather than by geography alone.
View datasetGreg C. Wright · Working Paper · 2026
Jack Kappelman, Diana Silver, Jin Yung Bae, Kevin Butler, Tanvi Shinkre, Falco J. Bargagli Stoffi, James Macinko · Social Science & Medicine · 2026
Hitanshu Pandit · Applied Economics Letters · 2026
.svg)