The interactive in this post was corrected to account for zip codes that span multiple counties, which are now allocated to the county with which they overlap most. When a zip code has multiple Small Area Fair Market Rents for a given year, the estimates are averaged. (Corrected 8/23/2019).
Finding ways to lower rents is already a key issue in the 2020 presidential race. Democratic candidates Elizabeth Warren, Julian Castro, Kamala Harris, and Cory Booker have all promised to invest in the construction of more affordable rental housing, and rightly so. The share of Americans who rent their home is at a 50-year high, and all US metropolitan areas are facing a severe shortage of affordable rental units.
Though the data show this crisis exists, their ability to guide us to solutions is limited because we don’t have timely, reliable data on current rents at the neighborhood level. This means that no matter what the candidates promise, quantitatively evaluating or comparing how their proposals would affect local decision making in specific neighborhoods is difficult.
Where are rents increasing fastest? How should we allocate affordable housing investments to most effectively reduce displacement? What effects do policies have on local rental markets? To answer these questions, we need to better understand how rents change within different parts of a city.
The interactive below shows what we know about zip code–level rental price changes from publicly available data. But these sources often paint contradictory pictures of how rents are changing in a given area.
What’s lacking in current rental data sources?
Nationwide sources of local rental information can be broken down into two categories: public surveys (specifically, the American Community Survey, or ACS) and proprietary real estate data. The ACS provides more reliable information about rents for large areas, which is how we know is the nation is facing a shortage of affordable rental housing.
But the ACS surveys only a few people in each neighborhood, which means its neighborhood-level information can be very uncertain. Moreover, the Census Bureau must combine five years of data to create viable neighborhood estimates. This creates a delay in the publication of survey results and obfuscates shorter-term trends because the latest data represent a five-year average from 2013 to 2017.
On the other hand, rent estimates from proprietary real estate sources are likely biased. These firms often collect online rental listings opportunistically or survey only large, multifamily buildings. The samples are typically biased toward more expensive listings and larger complexes, so they miss informal rentals, smaller complexes, and units that may be affordable for low income families. Although this “big data” approach solves the ACS’s problem of small sample sizes, not all rental listings are equally likely to be posted online, so we cannot be confident these data are representative of all renters.
Better rent data would help create affordable rental housing
Without high-quality local rent data, we cannot identify, predict, or effectively respond to rapid increases in rents. Local rent data would help increase the supply of affordable housing for three reasons.
First, these data would allow policymakers and advocates to identify gentrifying neighborhoods in real time. Although residents of gentrifying neighborhoods know from lived experience that their homes are becoming less affordable, articulating the scale of the problem can be difficult without data. Additionally, when we can identify which neighborhoods are currently facing displacement pressures, we can begin predicting where gentrification will happen in the future.
Second, because the effects of housing policies may vary between neighborhoods, local data are needed to evaluate interventions. When measuring how housing policies affect the rental market, economists typically use the ACS or private real estate data. But as mentioned, these have shortcomings in timeliness, geographic specificity, or representation.
Third, because of resource constraints and efforts to fight segregation, social programs providing housing assistance use estimates of local rental prices to allocate funds. If these programs make inaccurate estimates of rents in gentrifying neighborhoods, they will not provide residents with the assistance they need.
How can we get better local rental data?
So how do we develop accurate, neighborhood-level data on rental trends? Three strategies could help:
1. Research and quantify sample bias among private rental data sources.
Firms such as Zillow, Axiometrics, CoStar, and YardiMatrix combine online listings with surveys and other relevant sources to produce estimates of local rental prices. But they tend to focus on multifamily buildings or exclude less formal arrangements, so they might be missing important segments of the market.
Moreover, their coverage may shift over time. These firms should increase transparency by publishing their methodologies and allowing independent, public audits of their data to assess which types of rental housing units are over- or underrepresented in their sample. This would aid interpretations of their estimates or allow researchers to make their own adjustments to correct for bias.
2. Explore the feasibility of using tax data to estimate rental prices.
The tax code requires individual landlords, partnerships, and certain small corporations to report their rental income to the Internal Revenue Service. For each property landlords own, they must report the address, whether the property is a single- or multifamily residence, and the amount of rent received over the fiscal year.
This information is not publicly available to analysts, because strict privacy regulations govern tax return data, and the IRS grants private access to specific information on a limited basis. Further, access to the data would lag far behind when the rents were paid. And the IRS may not even store the data in a format that would make this analysis possible. But if this information could be accessed without revealing the identities of landlords and tenants, we could better understand US rental market dynamics.
For single-family residences, analysts could convert annual rental incomes to monthly installments and aggregate the data to a geographic area, correcting for partial years of rent. For multifamily complexes, property records would let researchers first understand how many units a landlord owns in a building and then perform a similar aggregation to calculate typical monthly rents. If tax data are only reliable for single-family units owned by individual landlords, then these multifamily data could be complemented by private-sector surveys of the buildings.
3. Catalog local sources of rental data.
Many municipalities and counties are aware that accurate rental data aren’t available and have responded by conducting their own surveys of tenants. Moreover, when local public housing agencies want to determine whether landlords are charging reasonable rents or challenge the Fair Market Rent assigned by HUD , they must also conduct a local survey. These local surveys, as well as their protocols and methods, should be cataloged. By sharing this information, we can encourage common standards across the nation and understand which estimates are comparable. And if we compile a patchwork of information on rental costs, we could begin to correct or adjust our data sources nationally.
One or more of these approaches would give us better information about current rental prices. Without better neighborhood-level rental data, policymakers will continue to struggle to make effective, up-to-date decisions when tackling the nation’s housing crisis.
Tune in and subscribe today.
The Urban Institute podcast, Evidence in Action, inspires changemakers to lead with evidence and act with equity. Co-hosted by Urban President Sarah Rosen Wartell and Executive Vice President Kimberlyn Leary, every episode features in-depth discussions with experts and leaders on topics ranging from how to advance equity, to designing innovative solutions that achieve community impact, to what it means to practice evidence-based leadership.