Resentencing is a promising strategy to reduce lengthy prison sentences, especially those imposed during the rise of punitive sentencing. With the possibility of scaling existing resentencing mechanisms hindered by the time needed to review cases, For The People in California is piloting new technology that could dramatically shorten the time needed to identify and assess eligible cases and could reduce the time spent incarcerated for thousands of people.
Resentencing is the review and adjustment of extreme prison sentences in the criminal legal system. At the time of this publication, 37 states have introduced resentencing legislation that would allow the opportunity to review and potentially reduce overly long sentences if they meet certain criteria. Nine states–California, Colorado, Connecticut, Illinois, Louisiana, Minnesota, New Mexico, Oregon, Washington–and the District of Columbia have passed some form of “Second Look” measure specifically; Second Look is one type of resentencing mechanism.
Prosecutor-Initiated Resentencing (PIR), launched by For The People and first passed in California in 2018, is another such emerging resentencing mechanism. Through PIR, district attorneys can revisit sentences, consider whether circumstances have changed since the original sentencing, and determine whether further confinement is in the interest of justice. If it is decided that an incarcerated person can safely return to the community, the district attorney can recommend to the court that their case be reissued a lesser sentence; this can result in release if the lesser sentence has been served already.
PIR has resentenced over 850 people nationally, and it could fuel a further reduction of the US prison population and significant diversion of taxpayer dollars from prisons into communities.
PIR also has the potential to alleviate racial disparities in the prison system. Black people are overrepresented in the category commonly used to determine initial eligibility for resentencing (i.e., crimes that were nonserious, nonviolent, and nonsexual in nature). In addition, the proportion of Black and Brown people who have already served lengthy sentences is greater than the proportion of white people who meet the same eligibility criteria (PDF). Resentencing presents opportunities to reduce the disproportionate representation of Black people in America’s prisons overall.
Passing resentencing legislation is only the first step. Once a law passes, implementation presents a new challenge. Logistical barriers can slow the identification of eligible cases and their subsequent review. For example, as part of the resentencing process in California, a district attorney must identify cases eligible for resentencing by reading through the prison records, called a central file (or C-file), of a person who has been convicted and is incarcerated in the California Department of Corrections and Rehabilitation. As these files contain substantial information about their cases and people’s time postconviction, the individual’s consent is needed to obtain their C-file.
Reading the C-files, which can hold more than 20 years of information on people’s incarceration, is only the initial step in determining eligibility. Just this step can take up to eight hours per case. Using this estimate, reviewing the C-files of all cases currently deemed eligible for resentencing in California alone would take decades.
As it stands, district attorney staff, legal interns, or even law school clinics have to review cases manually, which includes reading and summarizing key data. But as many district attorney’s offices are understaffed, they are in need of support and additional resources in order to implement this work. The resentencing law is not mandatory, so district attorneys may choose not to take on this increase in workload. This ultimately results in eligible people remaining behind bars.
A Catalyst Grant Program grantee, For The People, is working with the Possibility Lab to develop and pilot a case summary data automation tool to reduce the burden on district attorneys, streamline the initial review process, and increase the efficiency of PIR in practice.
When developing the C-file automation tool, For The People and the Possibility Lab connected with prosecutors in multiple California counties to learn what challenges the PIR process presents and how data automation could be most helpful. Their data collection resulted in the following key takeaways:
- Prosecutors need to identify who has been convicted of eligible offenses and determine how people have been spending their time in prison.
- Prosecutors want to quickly and easily refer to the original C-file and are looking for certain key data as a starting point.
- Prosecutors need to engage with developers during the design and early implementation of the C-file automation tool in order to trust the tool’s authenticity and efficacy and buy in to wide-scale adoption.
- Prosecutors prefer to personally review handwritten C-file sections.
Accounting for these considerations, the automation tool first converts the C-file to a more readable format, combs through and systematically extracts the information needed to review the case, and stores this information in a dataset to prepare a case summary. Because the tool focuses on systematically locating and extracting exact information rather than using algorithm-based summarization of text, prosecutors can be confident that the tool is not making any preliminary decisions without their oversight and expertise. Though C-files can be hundreds of pages long, the tool extracts text to generate a summary file, which is typically several pages on average, creates a condensed version of the C-file, and identifies handwritten forms that may be of interest. With the tool, prosecutors are more readily able to take a holistic look at a person’s record, including eligibility determinants and mitigating and aggravating postconviction factors. The tool also streamlines the review process by pulling keywords from the handwritten documents that meet criteria for PIR consideration and flagging page numbers for quick reference.
With their tool, not only do For The People and the Possibility Lab estimate that the initial case review time would be reduced from eight hours per case to one, but the reach of the PIR legislation would likely expand as the burden of undertaking this review would be vastly reduced. And prosecutors would have more time to engage with victims, victims’ advocates, reentry service providers, and public defenders.
The tool will be free and allow prosecutors to look at past sentences more readily, reduce excessive sentences, and safely bring people home from prison.
The next step in this work is to document lessons learned from the pilot phase to share with the field. For The People and the Possibility Lab are also planning to scale up to a desktop application available for all prosecutors.
Findings from this project highlight the potential of resentencing legislation, identify areas of existing approaches that can be improved, and can inform future strategies on reducing extreme prison sentences. To this end, those working to implement and streamline resentencing policies and other prosecutorial reforms should seek to do the following:
- Identify the barriers and limitations of existing approaches.
- Expand data collection and automation in resentencing and other prosecutorial endeavors.
- Pursue intentional community engagement from actors such as prosecutors, judges, and defense counsel throughout the legal process, especially when developing innovative solutions.
- Encourage broad updates in original sentencing policies in addition to these individual resentencing efforts.
This continuation and promotion of For The People and the Possibility Lab’s thoughtfully developed data automation tool has great potential to be helpful in expediting the process of identifying eligible cases, a key step in the resentencing process. With this tool, resentencing could reach more broadly, reduce excessive sentences, and increase equity in the criminal legal system. Moreover, the legislative progress being seen with Prosecutor-Initiated Resentencing could indicate an evolving social and judicial consciousness that, with the continued success of “looking back” through resentencing, could lead to “looking forward” and ending excessive sentences imposed on the front end.
The Catalyst Grant Program is a collaboration between the Urban Institute and the Microsoft Justice Reform Initiative to help nonprofit organizations use data and technology to advance racial equity and reform in the criminal legal system. Visit the Catalyst Grant Program Insights page for more resources and stories about the grantees.
