America's Paper Prisons: The Second Chance Gap

95 Pages Posted: 4 Nov 2018 Last revised: 9 Mar 2021

Date Written: December 16, 2020


Over the last decade, all fifty states and the federal government have enacted “second chance” reforms that increase the eligibility of individuals arrested, charged, or convicted of crimes to shorten their sentences, clear their criminal records, and/or regain the right to vote. While much fanfare has accompa- nied the increasing availability of “second chances,” little attention has been paid to their delivery. This study introduces the concept of the “second chance gap,” which it defines as the difference between eligibility and delivery of second chance relief; explores its causes; and approximates the size of the gap in connection with several second chance laws and initiatives. Using administrative and other data, it finds that among a host of petition-based second chance opportunities, to shorten sentences, restore one’s vote, and clear one’s criminal convictions, only a small fraction (less than 10 percent) of those eligible for relief actually received it. Extrapolating based on a novel analysis of around sixty thousand criminal histories of persons primarily seeking gig- economy work and of the expungement laws governing nonconvictions of all fifty states, this study estimates that at least twenty to thirty million American adults, or 30–40 percent of those with criminal records, fall into the “second chance expungement gap,” living burdened with criminal records that persist despite appearing to be partially or fully clearable under existing law.
These findings suggest that tens of millions of American are stuck in a paper prison, held back by deficiencies in the administration of second chances that have left them incarcerated, disenfranchised, or burdened by convictions beyond what the law requires. Some of the barriers to relief are structural and related to debt, overburdened bureaucracies, and the contested nature of second chance rules that unwind past judgments and policies. But others are harder to see and stem from administrative failures like unworkable standards, missing and incomplete criminal justice information (“dirty data”), a lack of awareness of second chance opportunities, and costly and complex processes. Addressing them—by moving administrative burdens from the defendant and onto the state and algorithms through automation, standardization, and ruthless iteration—can narrow the second chance relief gap. Leveraging them, “Clean Slate” initiatives to automatically clear eligible criminal records can have the potential to help the millions of Americans in the second chance expungement gap. However, the ability of such second chance initiatives to improve outcomes depends on how they are implement- ed. Debt-related barriers and dirty data can contribute to incomplete automation, leading to “second second chance gaps.” In the realm of expungement, application of the expungement criteria to minor but not major offenses can also have the effect of exacerbating, not narrowing, existing racial disparities within the population of people with records, while improving them within the general population. Further research is needed to under- stand the impact of automated clearance under different scenarios, such as when the defendant is not notified of the relief received or there is a risk of statistical discrimination making things worse, not better. Overall, however, though other hurdles may remain, automation can remove the unfair collateral punishments, not steel bars, holding back tens of millions of Americans.

Keywords: criminal records, cost-benefit analysis, uptake analysis, computational policy

JEL Classification: K14, K4

Suggested Citation

Chien, Colleen V., America's Paper Prisons: The Second Chance Gap (December 16, 2020). 119 MICH. L. REV. 519 (2020), Available at SSRN: or

Colleen V. Chien (Contact Author)

UC Berkeley School of Law ( email )

302 JSP
2240 Piedmont Ave
Berkeley, CA 94720
United States
510-664-5254 (Phone)

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
PlumX Metrics