For most property owners and landlords, screening potential renters is a routine part of the application process, particularly in tight housing markets. More than 90 percent (PDF) of landlords report using tenant screening software, and the number of companies offering reports has increased to meet this demand.

As people’s data—including credit scores, criminal and eviction history, and income history—become more available, and as companies try to set their screening reports apart, many companies are developing algorithmic models that assign tenant risk scores or recommendations. Early studies suggest that landlords often defer to these scores or recommendations, creating high barriers for people who have old, low-severity records, such as misdemeanors or nonconvictions, or inaccurate records on their screening reports. 

Little public evidence shows that risk scores accurately predict rental outcomes, and even less provides a clear mechanism for how certain histories are linked to lease violations. On top of this, these risk scores are unregulated and appear to include incorrect, outdated, or minor records that can prevent applicants from accessing housing.

How can policymakers reduce the negative effects of tenant screening risk scores?

Some states and cities have passed laws to regulate the data included in tenant screening reports, but data are needed to help changemakers understand how certain policy components, such as look-back periods, may affect residents. 

To understand how these policies might affect prospective tenants’ ability to find stable housing, a new tool from the Urban Institute allows users to simulate the effects of different tenant screening regulations on individuals’ risk of being screened out of rental housing applications. Users can choose from a range of policies and look-back periods, such as those restricting access to older court and eviction records that did not result in a conviction.

Using the tool highlights several policies that can help expand access to housing.

Policy 1. Limit the use of nonconvictions, nonevictions, and decriminalized records in screening reports

Depending on the look-back period, sealing these less severe records significantly reduces risk for individuals who are scored as being “moderate risk” under the baseline scenario. As the look-back period is reduced, inferred risk scores continue to decline.

Because these records reflect nonconvictions and nonevictions, or activity no longer criminalized in many states, limiting their use in tenant screening can expand housing access at little cost.

States can implement this policy in a variety of ways. Some states and localities specifically target tenant screening companies’ ability to access these records by regulating what records these companies or housing providers can access. For example, Colorado’s Rental Application Fairness Act prohibits consideration of criminal history older than five years while screening tenants.

Policy 2. Reduce look-back periods to align with perceived severity or relevance of the record

One consideration for local advocates and policymakers is ensuring policies are focused on the lowest-severity histories. In addition to sealing records that didn’t result in criminal charges, policymakers can expand protections to include misdemeanors and eviction filings to reduce how low-severity histories affect risk scores. Using a look-back period of five years for these categories shifts 45 percent of those labeled as “moderate-risk” under the baseline scenario to “low-risk” while a three-year look-back period shifts 63 percent of those scored as a “moderate risk” to “low risk”.

States have shown that it’s possible to apply different look-back periods or sealing timelines to different records. Minneapolis’s ordinance regulating tenant screening applications restricts the use of criminal background data to three years for misdemeanor convictions, seven years for felony convictions, and three years for eviction cases ruled in favor of the landlord. While differing look-back periods aren’t modeled in the tool, reducing look-back periods to align with the perceived severity of a record can greatly reduce barriers to accessing stable housing.

Policy 3. Align look-back periods with federal guidelines on other types of records

Though these policies focus on reducing the effect of low-severity or not-guilty verdicts, policymakers may also want to limit the negative effect of old records that persistently follow people and limit housing access long after they occur. Federal laws on consumer credit screening have specified look-back periods to ensure old histories don’t follow people forever. Policies can be changed to ensure eviction and criminal histories have the same look-back periods.

To reduce barriers for renters with older records, a lower-lift policy change could be updating criminal convictions that don’t have defined look-back periods to align with current federal law. Applying the seven-year look-back period included in standard credit reporting could reduce the risk of prospective renters being screened out.

Washington, DC’s Fair Criminal Record Screening for Housing law  prohibits housing providers from considering arrests that did not result in a conviction. It also limits consideration of certain convictions, including arson, assault, fraud (or attempted fraud), sexual abuse, forgery, and some drug offenses, to those that occurred within the past seven years. Other jurisdictions, like Alaska, simply expunge these records or seal them from public view.

Policy 4. Limit or eliminate look-back periods for eviction filings

The tool shows that targeting eviction filings effectively reduces barriers for individuals at “high-risk” of being screened out of housing applications. In simulating policies where only eviction-related records are selected, the greatest shift occurs in those who were labeled “high-risk” under baseline moving to the “low-risk” category. For those with older records or eviction cases where the landlord did not prevail, this policy can greatly reduce barriers to accessing rental housing.

Landlords often argue that eviction filings are critical inputs to understanding whether an applicant will be a good tenant, but eviction filings can be misleading and incomplete. Many landlords engage in serial eviction filings without an intent to evict, and that eviction filing data itself is often incomplete, inaccurate, and misleading.

States can implement this policy in a few ways. California, in addition to automatically sealing records for eviction cases that were ruled in favor of the tenant, implements a look-back period of seven years for eviction records in which the landlord prevailed. Oregon’s law is similar but applies a five-year look-back period to eviction records in which the landlord prevailed. Other states implement a short look-back period for nonevictions, but we see the greatest effect when look-back periods are applied to successful eviction records.

States and localities can focus on how data sources themselves may create inequitable outcomes

Beyond these specific recommendations, policymakers can consider which data sources are included in tenant screening and better understand the drawbacks and inequalities within these datasets, Crime data, for example, reflect inequity within the criminal legal system through higher rates of arrest and conviction for certain populations. Eviction filings are also more likely to be imposed against single mothers and Black women.

This built-in bias makes it more likely that Black people will be screened out of rental housing opportunities overall. We find that policy levers controlling the data included in tenant screening algorithms show little difference in the racial and ethnic distribution of each risk group before and after policy implementation. This may be because bias shows up earlier within the criminal-legal system, such as in arrests and convictions. For example, in our data from the Administrative Office of Pennsylvania Courts, Black people make up an estimated 25 percent of people, while Pennsylvania’s population is only 10 percent Black.

To address inequities overall in tenant screening processes, policymakers may need to more effectively examine where bias is manifesting within the criminal legal system. For example, states and localities can implement policies to reduce disproportionate arrest and conviction, such as limiting nonsafety and pretextual traffic stops or investing in restorative justice programs.