Last week, the United States Sentencing Commission released a second analysis of federal compassionate release motions filed during the pandemic. These reports mark an important departure from the federal government’s typical practice of shielding and hiding this type of data, and this step forward is almost certainly a response to the pressure from those who have advocated for increased data transparency during the pandemic.

The Commission’s report confirms what we have found in our own investigations over the past year: although thousands of incarcerated people are elderly and infirm, and at least 2,700 have died of COVID-19, compassionate release has been granted at paltry rates and often in an arbitrary manner during the ongoing global health crisis.

The Commission’s report shows that just 20 percent of compassionate release motions filed since January 2020 were granted. Moreover, as the pandemic spread through prisons overseen by the Federal Bureau of Prisons, the share of compassionate release petitions granted fell each month. The percentage of cases granted peaked in March 2020, at 38.8 percent. By October 2020, when active cases were soaring in federal prisons, only 13 percent of petitions were granted.

The report breaks down the rate of cases granted by individual-level characteristics including age, race, sentencing date, sentence length, and charging crime. Of the nearly 13,000 motions counted, over three quarters were filed by individuals sentenced within the past eight years, and the vast majority of individuals granted relief were male (90 percent). Roughly one-third of individuals granted relief were white, 45 percent granted relief were Black, and 17 percent were identified as Hispanic.

Because the U.S. Sentencing Commission has failed to publish the raw data that supports their findings, data questions remain about federal compassionate release during the pandemic. With access to such data, we could investigate information about the judges, counsel (pro se, private counsel, etc), and prosecutors involved in these motions. For example, how many times did an individual petition for compassionate release before they were granted? For those who were never granted relief, how many times were they denied? Did access to private counsel improve an applicant’s ability to win release? Without public raw data, we may never have authoritative answers to such questions.

Because of the lack of timely and transparent data availability in this domain, our team of researchers and data scientists has pursued a data scraping operation over the past year to obtain case-level data from the federal courts’ electronic records system, PACER (Public Access to Court Electronic Information). We were surprised to learn that from January through September 2020, the US Sentencing Commission similarly gained access to information about compassionate release denials through PACER searches. From October through December 2020, information about denials was submitted directly to the commission. However, this means that prior to October of 2020, the Commission was gathering data about denials from public sites and may have had no system in place for data collection and reporting beyond what we at the COVID-19 Behind Bars Data Project could do as an external monitoring body.

The lag time between the filing of tens of thousands of motions for compassionate release by federal prisoners in the federal courts and the publication of this data deprived advocates of timely information that might have been used to strengthen petitions and improve chances for release. In this way, the failure of BOP and the federal courts to release their data in a timely fashion could well have cost lives.

As advocates and policymakers across the country develop new strategies to address the growing crisis inside of America’s prisons, they require clear and accessible data to inform their urgent efforts. This need was especially true as the pandemic unfolded in the prisons over 2020. If empirical evidence was available to show that compassionate release motions were being rejected in such high numbers, for example, public pressure might have been brought to bear on lawmakers and other public officials to create alternative mechanisms for mass decarceration.

Good public health policy must be evidence-based and rooted in a clear understanding of the current conditions in jails and prisons. The culture of secrecy and resistance to oversight in many criminal justice agencies remains a significant threat to public health and safety. The governance of jails and prisons should not be hidden at the edge of democracy, but instead must be transparent and accountable to the public. This week’s report was a good first step, but also serves to illustrate how far we have to go.