To enhance local affordability. To foster inclusive communities.
On this page

Unlocking the power of local housing data


In the course of carrying out basic functions like permitting new development or administering public programs, city and county agencies collect information about people, properties, and transactions. Community-based organizations and businesses also gather data about the people they serve.
Image of data is shown to illustrate the importance of data for housing.


These administrative data can be very useful for understanding local housing conditions and needs and evaluating the success of housing programs and policies.

They add nuance, specificity, and timeliness to estimates drawn from the U.S. Census Bureau’s American Community Survey or other national housing datasets (Local Housing Solutions’ Housing Needs Assessment Tool relies on nationally available data to help localities understand their housing needs).

But because local administrative data are gathered for programmatic purposes, such as tax collection, permitting, and code enforcement, policymakers, practitioners, or researchers must do some work to access and use them for research and analysis. Further, to answer some housing policy questions, local officials may need to undertake new data collection efforts, such as a community property inventory or resident survey. 

Using concrete examples, this brief explores the uses, benefits, and challenges of using local administrative housing data. It also outlines strategies for accessing, harnessing, and collecting local data as well as building local data capacity.

What do we mean by local housing data?

Local administrative processes capture a vast variety of information that is often more up-to-date than national housing data sources (see Table 1 below). Property assessments for tax purposes record the value, building size, and certain features of the current housing stock, while building permits tell us about the pace, location, and types of new residential construction. 

In some cases, localities maintain rental registries that contain detailed information about the ownership, rents, and sometimes even the occupancy of rental properties. Local records of code violations, 311 complaints, and demolition permits can provide important insights about housing quality, while flood maps speak to future risks to housing conditions. 

Tax delinquencies recorded by the assessor’s office, court records for mortgage foreclosure and eviction cases, and records of water, gas, and electricity shutoffs from local utilities all speak to housing affordability and stability. Then there are data collected in housing assistance programs, such as demographic information about applicant households or their reasons for seeking assistance, which shed light on housing needs. Finally, there are all kinds of administrative data that aren’t about housing per se but can be linked to housing data and help us understand, for example, how housing affects exposure to crime or access to good schools.

Uses of local data

Broadly, local housing data have three uses:

  • They can illuminate existing conditions and trends in the local housing ecosystem. If home sales prices have increased, it indicates that the local housing market is heating up and it may be time to find ways that developers can be encouraged to build more homes and to include more affordable units in their projects. 
  • They can reveal needs. High eviction rates due to nonpayment of rent, for example, signal unsustainable rent burdens and the need for tools to enhance renters’ housing stability and create more affordable housing. Demographic information about service users and applicants can reveal who is experiencing housing needs.
  • They can show disparities. For example, the City of Arlington uncovered stark differences in housing density across neighborhoods technically zoned for the same number of homes per acre. It also found that zones that historically permitted only single-family homes had much lower shares of residents of color than zones that allowed two-family, townhouse, and multifamily development. 
  • Localities can evaluate new policies and measure overall progress toward their housing goals by tracking certain local data points.

We can put local data to use in answering a wide variety of housing policy and research questions. To give just a few examples:

  • Are we building enough affordable housing? Why or why not? An analysis of local zoning in Minneapolis found that three-quarters of the city’s population lived in neighborhoods zoned primarily for single-family homes. Tax data also showed that many of these homes were assessed at values far above what low-income families could afford. In response, Minneapolis eliminated single-family zoning to encourage denser, more affordable housing types.
  • Are we losing important sources of affordable housing? The Institute for Housing Studies at Chicago’s DePaul University used county assessor data to track the fate of 2-4-unit properties, an important source of lower-cost, family-sized rental units in the city. Their finding that these buildings are being lost to foreclosure and demolition is informing a new strategy to prevent displacement and preserve “missing middle” housing in the city.  
  • How can we help families access homeownership? Researchers at the University of Michigan analyzed Detroit’s Make It Home Program, which helps tenants purchase the properties their landlords are losing to property tax foreclosure. They found that despite having extremely low incomes, 85 percent of new homeowners had successfully sustained homeownership four years after the intervention.
  • How can we prevent eviction? Boston’s Office of Housing Stability invited the city’s largest property managers to join an Eviction Prevention Task Force, which used court records to identify its top-evicting landlords. When one of the participants, WinnCompanies, learned about its own high eviction rate, it implemented a new program to find alternatives for tenants behind on rent.
  • How can we reduce homelessness? A unique survey conducted by the Los Angeles Homeless Service Authority (LAHSA) helped researchers analyze the characteristics of people experiencing homelessness, especially those living in their cars. They found that the “vehicular homeless” are a distinct population that is more likely to include women with children, and that safe parking programs paired with other interventions can help them transition to stable housing.

Combatting High Rates of Investor Purchasing in Newark, New Jersey

In May 2022, two scholars at Rutgers University’s Center on Law, Inequality and Metropolitan Equity (CLiME) analyzed thousands of residential property sales transactions in Newark between 1989 and 2020 using local deeds data from the New Jersey County Tax Boards Association. They uncovered a worrying trend: after the foreclosure crisis, large-scale institutional investors began buying up large amounts of property in the city, including owner-occupied homes. By 2020, almost half of residential sales were to institutional buyers—a threefold increase since 2010. These investor purchases were associated with rising rents and decreasing homeownership. Alarmed by these findings, Mayor Baraka and the Newark Municipal Council created a Homeownership Revitalization Program to sell city-owned properties to longtime residents at low cost. The city is also considering expanding rent registry requirements and fining landlords who raise rent more than five percent every two years, with the proceeds going towards new affordable rental and homeownership opportunities.

The benefits of local data

Given the quality and variety of nationally available data on housing (such as the U.S. Census Bureau’s American Community Survey, the American Housing Survey, and the U.S. Department of Housing and Urban Development’s (HUD) Picture of Subsidized Households, to name a few), why use local data? There are many reasons:

  • Local data are spatially specific and granular. Whereas national data products usually offer data aggregated to the census tract or higher geographic scales, local data sources are more likely to be at the property level and to align with local planning geographies like neighborhoods, wards, or boroughs. 
  • For small geographies or population groups, national estimates are based on survey data from a sample of properties and have large margins of error. In these cases, local data are more accurate.
  • National datasets are often time-lagged; for example, American Community Survey data are not released until late the following year, and the American Housing Survey is only conducted every other year. But housing markets and policies can change quickly. Local data may offer more up-to-date information.
  • Local data may cover topics that national datasets do not. For instance, it is clear that America has a severe housing shortage. Yet only local data can tell us how many new developments are proposed, how long the approval process takes, and at which point(s) proposals for affordable housing are defeated. Another important example is rental registries, which record rental property ownership–information that is not available nationally. Housing quality is another area where very little good national-level data exists. And while the Eviction Lab has built a national dataset on eviction filings, we don’t know who goes to court, who has legal representation, or who ultimately receives an eviction judgment, except at the local level. 
  • Finally, gathering and sharing local data can be a way of building trust and social capital. Collaborative data projects can raise awareness about important issues, create new partnerships across government agencies and institutions, and engage residents and stakeholders.

Analyzing Eviction Data in Allegheny County, Pennsylvania

Allegheny County, which is home to Pittsburgh, is well known for its data-driven approach to policy. The county’s Office of Analytics, Technology, and Planning (ATP) maintains “Allegheny County Analytics,” a hub of interactive dashboards, reports, and local data infographics. One dashboard displays information about cases in landlord-tenant court from 2012 to the present. An interactive chart shows the number of cases filed and the amounts landlords claimed before, during, and after a local COVID-19 eviction moratorium was in effect. Other panels track access to legal representation and case outcomes, case timelines, and the average monthly rent of defendants. This dashboard has helped the county and its partners understand how lifting the moratorium and phasing out the emergency rental assistance programs affected eviction filings and to craft post-pandemic policies to reduce evictions.

Challenges of local data

Although the potential benefits of local housing data are vast, these resources come with common challenges: 

  • The data of interest may not exist. For example, a court clerk will record the name of the plaintiff, the name and address of the defendant, and the filing date, result, and claim amount of an eviction proceeding, but usually no further detail–such as whether the defendant withheld rent because the landlord failed to make repairs. In some cases, proprietors of a potential information source may be unaware of the data’s possible use in decision-making. Municipal staff and nonprofit employees also face time and resource constraints that often limit data collection. 
  • The data may be inaccessible. For example, eviction filings data are commonly compiled and held solely by municipal courts. But if this information was available to community programs and government agencies, it could be used for outreach, service provision, or analyzing local disparities. In cases such as these, establishing data-sharing relationships across organizations or departments, or creating a shared or public data platform, can be pivotal.
  • Another challenge is the data’s format. Data users may have to transform information collated in an unwieldy format, such as property tax documents generated as individual PDF files, before it can be useful (see more on data scraping below).
  • Data quality can vary considerably. Quality checks and “cleaning” are often needed due to errors or inconsistencies in data entry. Consider potential issues with Zip Codes in address data: Zip Codes could be entered incorrectly, they may be submitted in either five-digit or ZIP+4 format, or defaults in software like Excel may remove leading zeros (for example, the Zip Code “06106” could appear as “6106”). Another common challenge in administrative data is a lack of reliably unique identifiers for people or properties, making it difficult to tell whether a program served two different households or served the same household twice, for example.
  • When working with any data source, it is essential to consider if it could be incomplete or biased. Municipalities may have access to local rent data from property companies, but these data sources will commonly omit residents living in structures with five units or less, where a large portion of small to midsize city renters may reside.

We explore some strategies for addressing data access and capacity challenges below. 

Timely, Reliable Rent Data in Montgomery County, Maryland

Although more U.S. households are renting their homes than at any point in the past 50 years and the nation faces a severe shortage of affordable rental units, the lack of a critical data point–rents at the neighborhood level–handicaps responses to the crisis. Without current, reliable rent data, localities don’t know where rents are rising the fastest and can’t measure the impact of programs and policies on rental rates. Montgomery County, Maryland–a suburb of Washington, D.C.–has taken on this issue by conducting its own annual survey of rental buildings, collecting information on rent levels and occupancy, turnover, and amenities. It has used these data to determine that there is a surplus of units priced affordably for households at 50-100 percent of the Area Median Income (AMI) and a shortage (particularly of larger units) at lower income levels. It also used the data to propose changes to its inclusionary zoning and local housing voucher program.

Building local data capacity

Using local data requires time, resources, and skill. There are a number of actions localities can take to build capacity with data sources they already possess:

  • Sometimes, there is no easy way for municipal staff in one department or division to learn about or access data housed in another. A first step in breaking down silos may be setting up a cross-agency conversation to discuss data resources and data-sharing goals. 
  • An important part of building these relationships is establishing a clear, efficient, and ethical process for sharing local data. The Urban Institute offers a blueprint for interagency data-sharing. The State Data Sharing Initiative offers a toolkit of webinars, legal guidance, and example data use agreements that can help local and state governments share data while safeguarding their residents’ privacy.
  • In the longer term, localities might consider establishing a shared data repository, hiring a data administrator who works across departments, or working with a partner whose role it is to make data interoperable across agencies (for example, Iowa State University maintains the state’s integrated data system and the Poverty Center at Case Western Reserve University integrates neighborhood- and parcel-level data for all of northeastern Ohio in a database called NEOCANDO).

Other strategies can help localities access new data sources and analysis:

  • Often, multiple jurisdictions in a region would benefit from the same data resource or analysis and can better leverage their time and resources by pooling funds and coordinating a joint data purchase or request for proposals. For example, Rapid City, South Dakota, and neighboring communities jointly commissioned a regional housing study to inform their decision-making.
  • They can harness the expertise of research and educational institutions. The National Neighborhood Indicators Partnership (NNIP) network includes more than 30 organizations that focus on assembling and transforming local administrative data. At the Housing Solutions Lab, which is based in the NYU Furman Center, our mission is to help small and midsize cities plan, launch, and evaluate evidenced-based housing policies, including through research partnerships. Collecting or analyzing local data can also be an excellent capstone project for college students. In 2021, for example, Humphrey School of Public Affairs students at the University of Minnesota prepared a detailed report for Hennepin County on strategies for improving its point-in-time homelessness count. In 2018, a student in the Goldman School of Public Policy at UC Berkeley used tax assessor’s data to prepare a detailed assessment of vacant parcels for City of Oakland Councilmember Lynette McElhaney.

Finally, localities can build capacity by training municipal staff and partners:

Boulder County, Colorado, Knows Its Veterans by Name

Boulder County had a bold goal to eliminate homelessness among veterans, but there was a problem: they didn’t always know veterans’ names. The Continuums of Care serving the Denver region used one data system (the Homeless Management Information System) to track individuals experiencing homelessness. Meanwhile, the County used a different database (Boulder County Connect). Veterans who registered with one system did not always appear in the other. After creating new data-sharing procedures in 2022, homelessness service providers across the region can now engage in “case conferencing,” that is, meet to review a joint list of veterans experiencing homelessness and problem-solve for each person. This collaborative data effort has helped a large group of providers come together to match their respective housing resources more efficiently with each veteran’s needs.

Collecting new data

Sometimes, no existing data source provides the necessary information for a locality to act on a particular housing challenge. In the boxed examples below, a legal aid organization in Philadelphia, PA, needed to know about the prevalence of illegal evictions and rent hikes, but there was no empirical, site-specific evidence to draw on. The City of Cleveland, OH, needed precise information about deteriorating housing to target rehab, demolition, and neighborhood investments. At other times, localities need more qualitative data, for example, about residents’ perceptions of a new housing program or developers’ experiences with the development approval process. Localities can respond to these gaps by collecting fresh data. There are a variety of quantitative and qualitative methods to choose from:

Quantitative methods

Inventories, registries, data scraping, and surveys usually focus on gathering numeric data that can be analyzed in charts and graphs and using statistical methods (although there are some exceptions, such as write-in survey questions).  

  • Inventories are useful when collecting more detailed information about place-based entities–such as homes, vacant land or structures, or neighborhood amenities–than already exists in administrative data. Inventories often have a mapping component and might use a rubric or index to categorize each item. Conducting an inventory can be expensive, time-consuming, and logistically complex. This brief on creating and managing vacant property inventories includes practical strategies for funding and organizing these efforts.
  • Registries are similar to inventories but rely primarily on owners to share information about their properties (often to secure a permit or avoid incurring a fine). Rental registries or licensing programs are a foundational tool for code enforcement and rent regulation; short-term rental registries are also increasingly common. The information that such registries gather about who owns a property; whether they comply with taxes, fees, and housing codes; and in some cases, details about tenancy, turnover, and amenities, can be extremely valuable for designing local housing policy.
  • Scraping creates a new dataset from information embedded in web applications or PDF documents. It involves creating a code that can identify, extract, and organize the desired data points into a useful format. Examples include scraping real estate listings platforms like Zillow and Craigslist to track local rents; scraping court documents to create databases on evictions, foreclosures, or other housing-related court cases; and scraping social media sites to learn about resident perceptions of housing issues.
  • Surveys can be extremely valuable for learning about the characteristics, attitudes, and priorities of a large group of individuals, households, or organizations and are especially powerful when linked to administrative data. But because a survey gathers data from a sample of people (unlike a census, which collects data from the entire population), a significant challenge is bias. In other words, the survey sample may not be a good representation of the population as a whole, which makes it difficult to extrapolate the results. Robust outreach and appropriate compensation for survey takers can help. Especially for hard-to-reach groups, embedding the survey in existing interactions with government agencies or trusted nonprofits can increase response rates. On the back end, weighting survey responses by demographic characteristics such as income and race can result in more reliable estimates. The Pew Research Center offers helpful resources on crafting a strong survey questionnaire, and the Office of Institutional Research at Wake Forest University has a useful guide for survey analysis.


Assessing Housing Quality in Cleveland, Ohio

Cleveland, a rustbelt city that has experienced population decline, faces persistent challenges related to vacant and run-down housing. But in order to act, the city needed more information about the location and conditions of problem properties. In 2015, 2018, and most recently in 2023, the city partnered with a local nonprofit, the Western Reserve Land Conservancy, to conduct a property inventory. In teams of two, 30 surveyors are trained to use a mobile app to grade each property on an A to F scale based on visual inspection, utilities data, and local property data assembled in Case Western Reserve University’s NEOCANDO database. They determine whether the structure is occupied and record information about illegal dumping, sidewalk conditions, and ADA accessibility. This inventory is then used to target demolitions and rehab investments, inform programs to help residents adopt vacant lots as side yards, and most recently, identify likely lead paint hotspots.

Qualitative methods

Other methods, such as interviews and focus groups, deal with words and meanings and are destined for thematic analysis.

  • Interviews provide rich and nuanced data but are especially time-consuming to conduct and analyze. They can be a useful tool for collecting detailed information from a small group of experts (such as organizational leaders, or professionals with a specific housing skill set), or they can complement larger survey or administrative datasets by delving into questions about personal history, perceptions, opinions, and rationale for a small subset of representative individuals. When designing an interview protocol, consider whom to recruit; how to structure the exchange (or leave unstructured); whether and how to record; and how to analyze the interview data.
  • Focus groups, like interviews, are open-ended conversations–but they bring together multiple participants and one or more facilitators. Focus groups are especially valuable for observing group dynamics, bouncing around ideas, and giving members of a marginalized group (such as a linguistic minority) a chance to provide input in a comfortable environment. In the housing context, a focus group could gather feedback about a proposed homelessness strategy from people with lived experience of homelessness, for example. 

Each of these data collection methods has limitations and is usually much more costly than analyzing an existing dataset. But collecting original data also offers great flexibility–survey questions and inventory methods can be shaped to meet local needs–and can shed light on a reality that was previously unknown. Engaging and listening to the community is also a crucial way to foster trust and build support for a new housing strategy or movement.

Surveying Hard-to-Reach Households in Philadelphia, Pennsylvania

Past surveys of housing needs and attitudes in Philadelphia have oversampled homeowners, older adults, and individuals who identify as White. Community Legal Services (CLS)–a nonprofit that provides legal aid to low-income residents–wanted to better understand how issues like rising rents, illegal evictions, and past eviction records affect the population they serve. CLS partnered with the Housing Initiative at Penn (HIP) to survey renters. By embedding the survey in the city’s application for emergency rental assistance and working with neighborhood groups to disseminate the link, CLS and HIP collected over 6,000 responses, making this the largest-ever survey of renters in Philadelphia. The results suggested that informal evictions are strikingly common, that a past eviction has lasting negative effects, and that pests and mold plague the vast majority of low-cost rental units. CLS has used these findings to advocate for stronger tenant protections, eviction prevention efforts, and rental repairs.

Table 1. Local Data Sources

How useful was this page?
This field is for validation purposes and should be left unchanged.

Stay Informed

Stay up to date on the latest research, events and news from the Local Housing Solutions team:

Sign up for LHS newsletter and register for a free My Account which allows you to save LHS resources and Housing Strategy Review Results: