How agencies can clean up criminal records to automate expungement
Automatic record clearance is a growing priority for states as criminal records delay individuals’ ability to reenter society and fill open job slots.
As organizations start investing in innovative technologies, they must consider where they will find the workers to fill the new jobs they create, said Chike Aguh, the Chief Innovation Officer at the U.S. Department of Labor.
Speaking Tuesday at the Code for America (CfA) Summit in Washington, D.C., Aguh said one potential labor source could be the millions of individuals who have expungable criminal records. Those with a criminal record are often barred from voting, getting certain jobs or accessing affordable housing or educational opportunities, so clearing their criminal records and allowing them to contribute to the economy is critical.
While individuals can request government to clear their records, petition-based processes can be complicated and often entail manual, time-consuming data-entry work for agency staff, said Alia Toran-Burrell, director of CfA’s Clear My Record program, which offers agencies technology to implement automatic record clearance policies.
Technical solutions that enable automatic record clearance “take as much manual work out of it as possible for government,” she told GCN.
For instance, in 2022 the Utah Administrative Office of the Courts partnered with CfA to expunge the conviction records of 500,000 individuals in the state. The partners relied on entity resolution software, which identifies a single subject across multiple datasets or sources, CfA wrote in a March blog post.
The first step for Utah was to determine which records were qualified for clearance, Toran-Burrell said, but users found erroneous or duplicative data stood in the way. For example, a person’s name could be misspelled in one case record but spelled correctly in a different court case five years later.
Even though those cases referred to the same person, they may not be linked to one another, which is an important step in the determination process, she added.
However, imperfect data doesn’t have to halt the progress toward automatic expungement. “If we were waiting for data systems to be perfect in order for us to embark on this work, it would take a long time,” Toran-Burrell said.
Off-the-shelf entity resolution software that used machine learning helped the agency correct and unify that disparate data. What worked in Utah shows how agencies can leverage existing systems to streamline eligibility determinations and other internal processes, Toran-Burrell said.
In recent years, a slew of states have adopted legislation that enables automatic record clearance, also known as Clean Slate laws. According to the Clean Slate Initiative, 10 states—Pennsylvania, Utah, New Jersey, Michigan, Connecticut, Delaware, Virginia, Oklahoma, Colorado and California—have passed Clean Slate policies so far.
“We're really at a stage now where government leaders are coming to the same conclusion, which is that record clearance needs to happen automatically, [and] it needs to happen at scale,” Toran-Burrell said during the Tuesday presentation.
Automatic record clearance is “not just the right thing to do, but it’s the smart thing to do for our economy,” Aguh said. “We don’t have the luxury to not utilize every person we have.”