For many years I have used criminal records histories to screen applicants for my own rental properties, and taught others how to search online for PA criminal records histories of their own rental applicants. I have done this without considering the race, national origin, religion, sex or familial status of the applicants. I have had and helped others develop policies to apply only convictions of significant crimes, such as felonies, crimes involving violence, drug dealing, domestic abuse, forgery, theft by deception and extortion as litmus tests to deny an applicant the opportunity to rent housing. I have always told my clients that one being convicted of a crime does not make one a member of a protected class, and that fair and uniform application of non-discriminatory criteria for tenant screening is appropriate and legal.
I have counselled against using mere arrest records without conviction, except if the arrest is recent and a case involving potential incarceration of the applicant. The exception is based on the risk they will not be able to continue employment and pay the rent, or occupy the property, if incarcerated during the lease term.
Growing up, I remember well when my parents rented to a nice, normal appearing family of four without running a criminal check, only to learn later that the husband was on probation following conviction and incarceration for burglary. They learned this after other apartments in the building were broken into several times, with no apparent forced entry to the building, and missing furniture, TV sets and small appliances were all found in the new tenants’ apartment.
My practices in tenant screening and those I advise clients to follow are changing, based on a developing line of cases and the guidance on fair housing recently issued by HUD’s Office of General Counsel.
The “Disparate Impact” cases find employers and housing providers liable for discrimination without the need for any intent, if the practices they follow which are not otherwise discriminatory cause a disparate impact on a protected class. The logic is that if reading the entrails of sacrificed animals is an important religious practice of Zoroastrians, and you restrict the practice without showing a legally sufficient justification for the restriction, you are liable for discriminating against Zoroastrians.
On April 4th, 2016, the office of General Counsel of the U.S. Department of Housing and Urban Development, HUD’s legal department, issued its “Guidance on Application of Fair Housing Act Standards to the Use of Criminal Records by Providers of Housing and Real Estate Related Transactions”. Basically, this guidance is a ten page explanation of how use of criminal conviction histories has discriminatory effects on African Americans and Hispanics because there are higher rates of convictions and incarceration among those populations then the average for the entire population.
The first step outlined by HUD in its guidance is determining whether the landlord’s criminal history practice has a discriminatory effect. HUD’s guidance quotes studies showing nearly one third of the population of the U.S. has a criminal record. For 2013, HUD cites statistics that African Americans were arrested at a rate more than double the population as a whole, and make up 36% of the prison population, but only 12% of the U.S. population. HUD further cites data that Hispanics make up 22% of the prison population, but only 17% of the total U.S. population. By contrast, non-Hispanic Whites comprised 62% of the population, but only 34% of the prison population. Thus, the imprisonment rate for African American males is almost six times that for White males, and the rate for Hispanic males is over twice that for non-Hispanic White males.
HUD concludes that although these are national statistics and state and local statistics may be relevant to show differing conclusions, in the absence of different statistics, these statistics show reasonable cause to believe blanket use of conviction records would have a disparate impact on African American and Hispanic populations, and therefore have a discriminatory effect.
The second step of HUD’s analysis is evaluating whether the landlord’s criminal records policy is necessary to achieve a substantial, legitimate, nondiscriminatory interest. If HUD has found the discriminatory effect in step one, the burden in step two shifts to the landlord to prove that the challenged policy or practice is justified. That means the landlord has to provide evidence that the landlord has a substantial, legitimate, nondiscriminatory interest in using the criminal records policy it has, AND evidence that the policy is successful in achieving that interest or result. Blanket statements that “we do this to protect other residents’ will not cut it. You have to show evidence that your policy works.
In this part of its analysis, HUD comments that exclusion because of prior arrests not resulting in conviction will certainly fail to meet the landlord’s burden. Further, exclusions based on prior conviction without considering the nature of the conviction, the length of time since the conviction, and what the person has done since will also most likely fail. In order to successfully exclude persons from tenancy based on prior convictions, HUD directs that the landlord must show that the way it uses conviction records accurately distinguishes between criminal conduct that indicates a demonstrable risk to resident safety and/or criminal conduct that does not.
No more rejection of convicted embezzlers, forgers and tax evaders would appear to be available under this standard. As I recall, the only significant thing Al Capone, the notorious Chicago gangster was convicted of was income tax evasion. He never had a conviction for his role in the St. Valentine’s Day massacre. A policy as directed by HUD would appear to mean Al Capone would qualify to rent your property.
The third step is evaluating whether there is a less discriminatory alternative to achieve the same result. Even if a landlord proves its policy of using conviction records actually works to provide a safer place for its tenants to live, HUD gives the prospective tenant an opportunity in its third step to show the landlord’s goal could be met by a different screening process which would not prevent the applicant convicted of the crime which would pose a risk to tenant safety from being approved as a tenant. HUD suggests the tenant can show the facts or circumstances surrounding the criminal conduct, such as age at the time of the crime, evidence of being a good tenant before and/or after the conviction, and evidence of rehabilitation.
HUD also suggests that a landlord delay consideration of criminal history until after financial (credit and employment) requirements are met, to minimize the costs to the landlord associated with a more thorough and individualized review of the tenant’s criminal history. All this time, I thought I was doing tenants a favor by screening criminal records first, and refunding their fee for a credit report if I rejected on criminal history first.
One type of criminal conviction can absolutely still be a bar to tenancy, without violating the Fair Housing Act. Section 807(b)(4) of the Act provides that refusing rental to one convicted of “illegal manufacture or distribution of a controlled substance” does not violate the Act. Note that conviction is required, and it must be for drug trafficking, not possession.
So, what is a landlord to do? Stop automatic litmus tests for conviction records as a bar to tenancy, except for drug trafficking convictions. Consider the length of time since the conviction, the nature of the specific offense and its relation to the safety of other residents, and consider what the prospective tenant has done both before the conviction and rehabilitation since the conviction before rejecting the applicant. Maybe it is easier to run the credit report and verify employment first!
The author, Bradley S. Dornish is a licensed attorney, title insurance agent and real estate instructor in Pennsylvania. He can be reached at bdornish@dornish.net.
May 2016
