Recent conversations about racial bias in policing have highlighted our criminal justice system’s heavy-handedness, particularly in low-income communities and communities of color. An arrest or conviction reduces the likelihood of finding a job or housing for many years to come, contributing to cycles of poverty. In this way, our criminal justice system casts a long shadow over the lives of those it touches. As we decriminalize low-level offenses (like minor drug possession) and recognize that police, judges, and prosecutors treat white and black civilians differently, many have wondered if people with past convictions should have the opportunity to wipe their slate clean and start over without the stigma of a criminal record.

There is plenty of evidence that employers discriminate against people with criminal records, either because they’re worried about legal liability or the job applicant’s reliability and productivity. While discrimination may be rational for an individual employer, it is quite destructive for society as a whole, given the importance of employment to reducing recidivism. It is also perverse and unfair to tell formerly incarcerated people that we expect them to go on the straight and narrow, only  to repeatedly slam the doors of honest work in their faces. 

With this in mind, a growing chorus has called for broader sealing (also called “expungement”) of criminal records. Sealing an official criminal record hides it from view by employers, landlords, and others doing formal background checks. However, relatively few criminal records are currently eligible for sealing (typically minor offenses committed a long time ago), and the process to have a record sealed is complicated and costly. Advocates are pushing to expand the set of records eligible for sealing—and to leverage technology to make sealing automatic. These are often referred to as ‘Clean Slate’ policies.

Advocates hope that these policies will make it easier for people who have paid their dues to find jobs and build rewarding lives without their criminal records haunting them at every turn. This is a laudable goal, and just last year, at least 31 states implemented record sealing policies. 

Unfortunately, the evidence we have suggests these policies will not work. In fact, record-sealing could increase discrimination based on race when criminal records are not visible. Even without these unintended consequences, so long as employers want information on criminal records, sealing official records is unlikely to truly hide the evidence of someone’s criminal justice involvement.

The case of Ban the Box: Hiding information about criminal records did more harm than good

This is not the first time that advocates have tried to reduce discrimination against people with criminal records by hiding those records. So-called “Ban the Box” policies prevent employers from asking about applicants’ criminal records until late in the hiring process. But because they don’t directly address employers’ concerns about safety and fit for the job, not being able to ask about criminal records upfront simply leaves employers to guess. In the United States, young men of color are more likely to have recent criminal convictions. Because of this, Ban the Box incentivizes employers to discriminate against applicants from this group. 

Research on this policy’s effects shows that employment for young men of color with limited education actually decreased in areas that passed Ban the Box. While it is possible that the policy helped some people with records get a job, the best evidence suggests that any beneficial effects were small. This is likely because employers could still check applicants’ criminal records at the end of the hiring process. If they didn’t want to hire people with records, they could turn those applicants away at that point. It appears that most did. 

The motivation behind Clean Slate is similar. Once criminal records are sealed, advocates hope that employers who do not want to hire people with records will simply not discriminate. But as the case of Ban the Box shows, this policy might broaden discrimination rather than reduce it.

The effects of Clean Slate will hinge largely on why employers discriminate against people with records in the first place. Clean Slate could reduce legal liability concerns for the subset of applicants whose records are sealed, since an employee’s sealed record cannot be held against an employer in court. However, if employers are worried about something else—such as reliability or productivity—we probably will not see such benefits. Because Clean Slate targets minor and older offenses, it could simply have no effect at all. Research shows that these types of records already result in less discrimination than more serious and recent offenses do.

To the extent that Clean Slate seals records that employers care about, it could leave them guessing, just as Ban the Box did. Employers who want to avoid hiring people with a criminal past may then discriminate against applicants from groups where criminal records are more common. Once again, this would most likely harm young men of color who never had a criminal record.

Clean Slate will not hide or correct criminal record information

Clean Slate could lead to broader, race-based discrimination if record-sealing effectively hides criminal records that employers want to see. But record sealing also has the opposite problem: it might actually do too little. This means that even if our concerns about racial discrimination are overblown, this legislation might not be effective at hiding evidence of criminal justice involvement or correcting the inaccurate information circulating on the internet. 

Information on arrests and convictions is typically a matter of public record, which means anyone can access it. There is a good reason for this: locking someone up is an infringement on their liberty, and we should know how often our government is doing this and to whom. However, once criminal record data is made public, it is unregulated. Because criminal record information is so decentralized in the United States, there is big money to be made from collecting and aggregating this information across agencies and places, and selling it online.

Even official criminal record data is notoriously incomplete. An incorrect court docket or missing case disposition could undermine the automated systems proposed by Clean Slate advocates. But the unofficial criminal records available online are even less reliable, and there’s no way to correct this information. Record sealing policies do not address this problem, which may pose a greater hurdle than official background checks ever did.

There is a risk that Clean Slate could actually incentivize employers to rely more on the private market for criminal records, if they view the official record as hiding relevant information. Mugshots and other information about arrests and convictions are readily available from private vendors. Advocates promise job applicants that criminal records are sealed, but interested parties could easily find a sealed record through a simple web search. 

Let the research guide policy 

But this is all just speculation. Perhaps we’re too pessimistic! What effect does record sealing have in practice? 

We don’t know yet, because there is no research evidence on the effects of broadening access to record-sealing. The research that advocates typically point to suffers from a major shortcoming: It simply compares outcomes for the few people who successfully navigated the existing maze of sealing requirements to outcomes for those who did not. These studies find that those who have their records sealed are better off—but this could be due to differences in motivation or resources that enabled them to have their records sealed in the first place. It is not yet clear if record-sealing itself has any benefits, or if those benefits outweigh any unintended costs.

In our view, the rush to broaden record sealing is premature. Many places have already implemented Clean Slate policies, giving researchers an opportunity to study the effects of such changes. We should wait and see whether they find net benefits, costs, or no effect at all, before scaling these policies further.

More promising alternatives exist

We see at least three other policies that are likely to be more effective at helping people with criminal records.

Centralize criminal record data and improve accuracy: Record sealing policy emphasizes sealing a single version of a record. It could be far more helpful if states centralized and restructured their criminal record data practices to ensure that available records are complete and accurate. We should increase the use of state record repositories to manage a single version of a person’s criminal record, integrating data between police, courts, and corrections. States could then provide qualified employers and public access to a single, high-quality source; this would limit demand for private vendors’ inaccurate and incomplete data, making it less lucrative for such vendors to exist in the first place. People should also have free and direct access to their own criminal records so that they can check for and correct inaccuracies. In other words, we should treat criminal records the way we now treat credit reports and medical records. These are administrative records that affect people’s lives; they deserve privacy and integrity. 

Broaden the use of deferred adjudications for non-violent, first-time felony offenses: Helping people avoid a record in the first place is likely to be the best way to help them move forward. Some states allow “deferred adjudications” for certain charges, usually non-violent offenses. Defendants who receive such deferrals are essentially placed on probation, and if they complete those requirements, the original charges are dismissed. Research from Texas provides compelling evidence that these policies have considerable benefits when applied to non-violent, first-time felony defendants: they dramatically reduce the likelihood of future convictions, and increase employment. In other words, giving people a real second chance before putting a (first) felony conviction on their record is an effective way to ensure that one bad decision doesn’t derail their lives.

Invest in rehabilitation and training: For most people coming out of jail and prison, the criminal record itself may not be the primary barrier to employment. We currently do little to give individuals the skills and support they need to compete in the labor market. Investing in education, job training, cognitive behavioral therapy, and health care for those with recent convictions could all go a long way toward helping people reintegrate into society – and address employers’ concerns in the process. Policymakers should also experiment with designing systems to increase the information available to employers about individuals’ rehabilitation and job-readiness (for instance, court-issued rehabilitation certificates). 

None of these alternatives provide the magic wand that advocates seek. But they might actually work.


Jennifer Doleac is an Associate Professor of Economics and Director of the Justice Tech Lab at Texas A&M University. She hosts the Probable Causation podcast about law, economics, and crime.

Sarah Lageson is an Assistant Professor of Criminal Justice at Rutgers University. She is the author of Digital Punishment: Privacy, Stigma, and the Harms of Data-Driven Criminal Justice.