Using Data to Improve Case Screening & Selection

The Mid-Atlantic Innocence Project (MAIP) works to correct the conviction of innocent people in the District of Columbia, Maryland, and Virginia. To do this, MAIP reviews approximately 300 requests for help each year from individuals claiming innocence. The goal of the case-review process is to ensure that its screening process is as accurate and efficient as possible. This means isolating strong innocence claims without: (1) erroneously rejecting the cases of applicants with provable innocence claims; (2) focusing too many resources on meritless claims; or (3) unduly delaying the review of cases at any stage of the process.

    Over the past several years, MAIP has prioritized the first goal because it never wants to miss the case of an innocent person. This has resulted in a significant number of cases that MAIP has flagged as needing some additional investigation, whether it is by requesting public records, interviewing witnesses, or other means. Each of these tasks can take months or even years, which has meant that MAIP has more cases at this stage of review than it can address with its current staff and resources. This proposal aims to solve the problem through strategies that would reduce the existing backlog and prevent it from growing in the future.
 

    This project will address this problem both reactively and prospectively by: (1) increasing MAIP’s investigative resources so the most compelling cases are more quickly isolated and moved to litigation; (2) allowing staff attorneys to focus on the most promising cases by transferring document requests to MAIP’s new full-time Screening Manager and by hiring a grant-funded Paralegal; and (3) gathering and studying MAIP’s case data to see whether there are predictive factors that can be used to more accurately assess innocence claims at an earlier stage in the case-review process. By implementing both reactive and proactive solutions, MAIP will be able to solve this long-term problem and keep it from occurring in the future.