A blanket criminal record exclusion is the screening policy most likely to produce a complaint

"No criminal record" is the most common criminal history policy in rental housing and the least defensible one.

HUD's 2016 guidance on the use of criminal records under the Fair Housing Act laid out the reasoning: because arrest and conviction rates differ substantially by race, a blanket exclusion can produce a disparate impact even with no discriminatory intent, and the burden then falls on the housing provider to show the policy is necessary to achieve a substantial, legitimate, nondiscriminatory interest.

A blanket ban is difficult to carry that burden on, because it is not tailored to anything.

Arrests are not convictions and should not be screened on

The clearest single line in the guidance, and the easiest thing to fix.

An arrest without a conviction is not evidence that the person did anything. Charges get dropped, cases get dismissed, and people get acquitted. Excluding on arrest records means excluding people for having been accused.

HUD's position is that a policy excluding based on arrests alone cannot satisfy the burden. So the first correction to most criminal history policies is to screen on convictions only, and to ignore arrests, dismissals, acquittals, and sealed or expunged records entirely.

That change alone removes a large share of the exposure, and it costs nothing in actual risk screening, because an arrest with no conviction was never predictive.

Three components that make a policy defensible

A lookback period. Recency matters. A conviction from nineteen years ago has close to no predictive value about tenancy, and a policy with no time limit is screening on someone's entire life. Lookbacks commonly run in the range of five to seven years for most offenses, with longer periods sometimes applied to a narrow set of the most serious ones. Whatever you set, set it and write it down.

A severity and relevance test. The question is whether the conviction bears on the interest you are protecting: safety of other residents and staff, and protection of the property. A conviction for violent assault relates. A twelve-year-old possession conviction does not, and screening on it is difficult to justify as necessary.

Individualized assessment. This is the component that turns a policy into a defensible one. Before a final decline on criminal history, the applicant gets an opportunity to present mitigating information: what happened, how long ago, what has changed, evidence of rehabilitation, and their tenancy history since.

That third one is where most policies stop short, because it costs staff time. It is also the part that makes the first two survivable, since a tailored policy with an exception path is meaningfully different from a rule applied mechanically.

The order of operations is now regulated in some places

A growing number of jurisdictions have adopted rules on when a criminal history inquiry can occur, modeled on fair chance employment laws.

The common structure requires a conditional approval based on income, credit, and rental history first, with the criminal history inquiry only after that, followed by an individualized assessment and a written explanation if the conditional approval is withdrawn.

Whether that applies to you is a jurisdiction question and it changes as ordinances pass. Worth confirming per market with counsel rather than assuming a national process is compliant everywhere.

This is also a place where a screening platform cannot help you. Any vendor claiming to keep current with criminal history screening law across every jurisdiction you operate in should be asked for the list and its last update date.

Consistency is the practical defense

Fair housing exposure in criminal screening comes from two sources, and the second one is more common than the first.

The written policy having a disparate impact.

And the written policy being applied inconsistently by the people executing it, so two applicants with similar records get different outcomes and the difference correlates with something protected.

The second is why a documented, structured process matters more than a well-drafted policy document. Every criminal history decision needs a record of what was found, which policy criterion applied, whether an individualized assessment happened, what the applicant presented, who decided, and when.

A file with that record is defensible. A file with a decline and no reason is not, regardless of what the policy manual says.

Where ProofUp sits, and what it does not do

Worth being unambiguous, because this post is about a category we largely do not operate in.

ProofUp verifies income, identity, and document authenticity. We do not perform criminal background checks and we do not perform credit checks. Native credit and criminal capability is on the roadmap and is not shipped.

So nothing in this post describes a ProofUp feature. It is written because the same principle governs the parts we do handle: a screening decision has to be consistent, documented, and explainable, or it is a liability whatever the underlying data quality.

What we contribute to that is the audit trail on our own side. Every verification attempt recorded, every flag with the specific signal behind it, and a Pass, Flag, or Fail verdict rather than a discretionary read. That is what makes the income and document portion of a decision defensible, and the criminal portion needs its own equivalent from whoever provides it.

We also do not give legal advice on your criminal history policy. Lookback periods, severity tests, and the individualized assessment process should be reviewed by counsel who knows your markets.

Read your own policy for one sentence

Open your written screening criteria and find the criminal history section.

Look for three things: does it distinguish arrests from convictions, does it have a lookback period, and does it describe an individualized assessment process.

A policy missing all three is the most common version there is, and it is the one a complaint is easiest to build against.

How many of the three did yours have?