Last week the Seattle City Council passed the Fair Chance Housing ordinance, which prohibits most Seattle landlords from using criminal background to deny housing to prospective tenants. Yesterday morning, Mayor Ed Murray signed the bill into law. The bill was not without controversy, and it’s worth understanding the rationale behind it as well as the strengths and weaknesses of the arguments.
It’s worth stating up front that this is a complex issue. Resist the urge to latch onto a single statistic or chart as the definitive guidance on how the policy should be crafted. This is a long piece, but I encourage you to read all the way through and consider all the relevant facets and dimensions of the problem.
Along with the ordinance, the City Council posted a clerk file with a collection of documents they used to guide their legislative work. It’s not particularly satisfying; most were penned by advocacy groups (and go figure, they argue for prohibiting the use of criminal backgrounds) and they occasionally take some liberties in how they represent the research they cite and the conclusions they draw. In researching this article, I examined all of the materials that the Council provided, as well as several more from the research community, the federal government, and the state government. Together they provide a much more complete and nuanced picture of the policy questions and implications.
There are five basic questions intrinsic to this policy debate; let’s walk through each.
- Is past criminal history predictive of future criminal behavior?
Answer: sometimes. The Council relied on a highly-cited 2006 paper that looks at a cohort of people born in 1958 in Philadelphia and tracks them over decades. It specifically looked at those in the cohort whose first offense was at age 18, and found that they had a higher incidence of arrest than non-offenders, but the highest incidence was in the first two years and after about five years the difference in arrest rate was almost negligible.
But the recidivism rate isn’t uniform across the cohort: the study found that the number of past offenses correlates highly with the likelihood of re-arrest.
That data is fairly old, but fortunately the United States Sentencing Commission and the Department of Justice have also been tracking recidivism statistics and issue regular reports. A May 2016 report from the US Sentencing Commission looks at people released from federal prisons in 2005 and their recidivism rates over the following eight years. They found similar trends for arrest rate following release; more than half of the re-arrests happened in the first two years.
They also found that those with a more extensive criminal background had a higher recidivism rate.
In addition, age and time of release correlated strongly with recidivism: the older you are, the less likely you are to re-offend.
They also found some other demographic differences in predicting recidivism, including gender, race, and education. The difference in rearrest rate by race reflects the well-documented disparate treatment that people of color receive from the criminal justice system — an issue we will return to later.
A 2014 DOJ report on recidivism among prisoners released in 30 states in 2005 showed similar results (the first two years being key), though higher rates of recidivism than federal prisoners (with no explanation why).
What does this tell us? Using the mere existence of a criminal record as a binary instrument is a poor predictor of likelihood of future re-offending. That said, the person’s age and the extent of their history of offenses are strong predictors. So criminal history is not useless as a predictor, but the details are important and it’s ripe for misuse by people who don’t know what they should be looking for.
And that leads us to the second, more relevant question:
2. Is criminal background predictive of whether someone will be a good tenant?
The answer to that question depends on how you define “being a good tenant.” It certainly includes paying the rent on time and not causing damage to the property. It also means being a “good neighbor,” respecting other tenants (including not threatening, harming, or stealing from them) and not participating in illegal activity (such as illegal drug consumption or dealing) on the property. And I suspect for most landlords it includes not getting arrested and/or jailed again, which is a hassle for the landlord (though clearly not as much as it is for the offender).
The proponents of the Fair Chance Housing bill on the City Council, including Council members Herbold, O’Brien, and Gonzalez, have been very careful in how they answer this question. All three have said variations of “There is no evidence that criminal history is predictive of being a good tenant.” While factually correct, this is at best a half-truth. The truth is that there is little to no research on the general question, so while there’s no evidence that criminal history is predictive of being a “good tenant,” there is equally no evidence that it isn’t. There are studies that look at how people with criminal histories do in supportive housing programs that are tailored for those with behavioral health issues, or in “re-entry” programs designed to assist those who have just been released from prison to reintegrate into society. Those studies found that in those specific contexts criminal history alone was not highly predictive of successful exit from the housing program. But as one of the advocacy papers the Council relies upon admits (buried in a footnote), “findings from supportive housing programs may not be completely generalizable to other housing contexts on account of the unique resources and social services available to residents.”
Further, if “being a good tenant” means “not reoffending,” then within the constraints described above someone’s criminal history can be highly predictive.
Even if we just look at the more immediate question of whether they pose a safety or security risk to other tenants, the 2014 DOJ study showed that in some cases the type of crime committed previously was predictive of future offenses; for example, 54% of people who were released after committing a property crime committed another one within five years.
So even though no one has studied whether released offenders are likely to pay their rent on time, some aspects of criminal background are predictive over parts of the question as to whether someone will be a good tenant. Again, though, it’s not a simple analysis to do and is ripe for misuse by landlords — especially those eager to find a reason to deny tenancy to a potential renter.
3. What is the effect on released offenders who are denied tenancy?
There is little debate that lack of available housing for recently released offenders is a major destabilizing effect for them. In fact, it is probably a major contributor to the high recidivism rate in the first two years after release.
The research backing up these findings is a bit convoluted. Studies on re-entry of released prisoners (including this one from Washington state in 2013) show that the key to successful re-entry is to establish a strong social network of family, friends, co-workers, and social service workers who collectively form the person’s connection back into society and also allow for access to behavioral health and drug treatment programs as necessary. Similar to the “housing first” approach for stabilizing homeless persons trying to access other services, for people re-entering from prison having stable housing isn’t the solution itself but provides a foundation for establishing a social network and access to the critical “wraparound services” they need. This is an important nuance: the social network and services are the critical success factor, not the housing itself, but stable housing significantly increases their success rate. The 2013 study puts this in stark terms: “being released homeless or marginally housed puts ex-offenders in almost immediate risk of failure.” An interesting corollary: should landlords receive assurances that released offenders they are considering for housing are plugged into the social network and services they need to successfully re-enter society?
There are other aspects of the question of impact on released offenders that are worth consideration as well, including that many released prisoners have families –or might have them in the future. Denying housing to someone based on their criminal record means denying housing to their whole family: spouses, children, and elderly parents and grandparents in their care.
This year’s One Night Count in King County found that 5% of the county’s homeless population were in jail or prison immediately prior to becoming homeless. 7% said that incarceration was the cause of their current homelessness. 55% reported having previously been in jail or juvenile detention — almost double the rate for the general population. Clearly there is a relationship between the criminal justice system and the homeless crisis.
4. Does denying tenancy to people with criminal backgrounds improve public safety?
This question, in many ways, cuts to the heart of the policy debate. Again, it depends on how you define “public safety.” On one hand, knowing that some aspects of a person’s criminal history are predictive of future criminal behavior that might place neighbors at risk, denying tenancy might increase the safety in the immediate area. But on the other hand, lack of stable housing places released offenders at high risk of immediately falling back into the criminal justice system and increasing the overall crime rate. So the policy question here is whether to optimize for the safety of individual apartment buildings and neighborhoods, or for the safety of the entire community. And that leads into the fifth question:
5. What are landlords required to do to ensure the safety of their properties, and what are they prohibited from doing to that end?
… and here’s where the lawyers all come rushing into the room.
One of the advocacy documents that the Council relied upon, written by Columbia Legal Services (an advocacy group and registered lobbyist that also provides legal representation and assistance to individuals from underprivileged groups) argues that the written and case law requires landlords to provide a safe environment to the best of their knowledge and to act to address any known or foreseeable risks to safety or property. However, according to CLS the law does not require landlords to check the criminal backgrounds of potential tenants; they do not have a duty to provide protection from the unforeseeable criminal acts of third parties, including other tenants.
Let’s take this one step further, returning to an issue we touched on earlier. It’s well understood that the criminal justice system has a disparate impact on people of color: African-Americans, Hispanics, and Native Americans are over-represented in the system and in our prisons. Courts have already established that with few exceptions using criminal background checks to deny employment is a prohibited form of discrimination because of the disparate impact on persons of color that they impose.
The “disparate impact” theory holds that racial discrimination can be proven either by showing an intent to discriminate, or by showing that an act had a clear disparate impact on a protected class. Eleven of the twelve US Circuit Courts of Appeals have found that disparate impact theory can be applied to the Fair Housing Act, which bans housing discrimination on the basis of race. The upshot of this is that since the criminal justice system is already proven to have a disparate impact, using criminal background checks to deny housing is probably a violation of the Fair Housing Act. By that logic, it’s already banned under federal law. In 2016, the Obama Administration’s Department of Housing and Urban Development (HUD) issued guidance to that extent. While it didn’t rule out the use of criminal background checks entirely, it asserted that the use of the criminal background must be based on a specific, legitimate, justifiable and non-discriminatory purpose, and not an broad generalizations. “To do this, a housing provider must show that its policy accurately distinguishes between criminal conduct that indicates a demonstrable risk to resident safety and/or property and criminal conduct that does not.” It’s important to point out that the Trump administration may reverse that guidance, as it has done in several other instances, particularly as it has taken a “tough on crime” stance. But even if that happens, it will be up to the courts to make the final judgment as to whether criminal background checks are permissible in granting access to housing.
So where does that leave us? Seattle’s Fair Chance Housing bill signed into law yesterday broadly prohibits the use of criminal background checks, even narrowly-tailored uses that might conceivably pass muster under the Fair Housing Act. That will certainly be a benefit to Seattle residents with a criminal record (nationally one in three adults have a criminal record, somewhere between 70 and 100 million people). As some on the City Council have argued, underlying the ordinance is a belief that those who have served their time in prison have paid their debt to society for their misdeeds and should not continue to be punished for their crimes.
The ordinance will trade off potentially increased “hyper-local” safety risks with the promise of a decrease in recidivism (and hopefully crime) in the broader community. It will take some time to understand the net effect of that tradeoff; unfortunately, while the ordinance calls for an evaluation of its implementation, that evaluation is focused on the ability of individuals to obtain housing and does not explicitly include evaluating the effect on public safety:
The Department shall ask the Office of the City Auditor to conduct an evaluation of the Fair Chance Housing Ordinance to determine if the program should be maintained, amended, or repealed. The evaluation should include an analysis of the impact on discrimination based on race and the impact on the ability of persons with criminal records to obtain housing. The highest quality evaluation will be performed based on available resources and data. The Office of the City Auditor, at its discretion, may retain an independent, outside party to conduct the evaluation. The evaluation shall be submitted to City Council by the end of 2019.
The ordinance takes effect 180 days from yesterday, to allow time for the city to work out the details of implementations and enforcement and to educate both landlords and tenants.