New Research Demonstrates How Automated Sealing Laws are Living up to Their Promise
New research, “Clean Slate, Dirty Data: An Audit of Algorithmic Automated Criminal Expungement Laws,” by legal scholar Colleen V. Chien at the University of California, Berkeley — and funded by The Clean Slate Initiative — was recently published in the North Carolina Law Review.
This paper offers the first major audit of automated sealing in its application to eligible non-conviction records in several states. The findings are encouraging and demonstrate that the automated sealing of non-conviction records is effective, with a significantly greater impact than petition-based processes. The study also identifies opportunities to make automated processes even more efficient and provides recommendations to ensure automated sealing laws are implemented effectively.
When Non-Conviction Records are Sealed, They Don’t Block Access to Opportunities
The study analyzed thousands of non-conviction records for cases that were dismissed, acquitted, or not prosecuted, alongside commercial background checks, detailing 5 states (Pennsylvania, Nebraska, South Carolina, Connecticut, and New York) where the law dictates that certain non-conviction records are supposed to be sealed.1 Depending on the language of the law, the terms “make nonpublic” or “suppressed” are also used.
The audit found that 98% of non-conviction records suppressed by law did not appear in commercial background checks for certain employment, housing, and educational opportunities — effectively removing barriers to a better life for people who were never convicted of the crime their record was associated with.
Some Eligible Records Still Haven’t Been Sealed
Data limitations prevented a definitive analysis of all eligible non-conviction records in each state, but a comparison of samples before and after rule changes in two states (Pennsylvania and Nebraska) with pre- and post-automated sealing law data suggested that potentially up to 1 in 3 charges eligible for sealing across the two states were not initially sealed at the time of the study.
This means that in some cases, records designated by statute for sealing may still be visible to employers, landlords, and others, not because of the law itself, but because of potential implementation challenges. These challenges can stem from outdated or incomplete data, unclear eligibility rules, states’ lack of IT infrastructure, and insufficient oversight of how the laws are applied.
While these challenges can delay sealing, the study shows that automated processes drastically reduce the uptake gap compared to petition-based systems: research has shown that fewer than 10% of those eligible under petition based laws ever complete the process, whereas this new data suggests that about two-thirds of those eligible in this study had their nonconviction records sealed without having to navigate costly, complex, and burdensome court processes. This provides promising evidence that automation is making big steps toward closing the second-chance gap.
It is important to note that eligibility rules in different states are complex and often require data not included in court case files. Therefore, some records marked as “eligible” for suppression in the analysis might not qualify under a given state’s statute. These data modeling constraints may result in the study potentially overestimating the number of eligible records that remained unsealed.
States Can Ensure that Automated Sealing Laws Are Implemented Effectively
While the study doesn’t examine Clean Slate laws specifically, but rather a range of automatic record processes for non-convictions across multiple states, it still offers lessons for jurisdictions moving toward Clean Slate automation.
As this research shows, passing a law is just the first step. Implementation is what makes second chances a reality, so states can take several steps to ensure that automated record sealing laws deliver on their promise:
- Transparency: States should publish eligibility rules, making it easy for people to check their record status and understand if their record is eligible for automated sealing.
- Notification: People should be informed when their record is sealed, or if they’re eligible but have not yet been sealed.
- Audits: Governments should track how well automated systems are performing and whether implementation gaps persist, and there should be ways for individuals to seek review and redress of apparent discrepancies.
- Accountability for Background Check Companies: Under the Fair Credit Reporting Act (FCRA), firms are required to update their data and notify individuals about their rights. This federal law should be enforced, and firms should report discrepancies they see, for example, in government data.
Clean Slate laws work and are already making a difference in the lives of millions of people. This research shows that most people with eligible non-conviction records are receiving relief, closing the uptake gap. Fortunately, there are steps that states can take to ensure that automated record sealing laws deliver on their promises, including prioritizing good data, transparency, and accountability..
To truly remove the barriers holding millions back, we must build the systems behind the laws, ensure people know when their records are sealed, and ensure no one is left behind because of bad data.