As the Biden administration begins to implement its criminal justice reform goals, experts say that access to timely and accurate data will be essential for the success or failure of the administration’s agenda. Key to that effort is reforming the Bureau of Justice Statistics, the nation’s main source of data on the criminal legal system.
“Your policies have to rest on something, and the outcome of your policies, the goals that you want to achieve have to be measurable,” said Peter Wood, chairperson of the Crime & Justice Research Alliance. “And if you have no data, how are you going to know whether what you’re doing is having an impact?”
But decades of poor funding have left the agency struggling to fulfill its mission as the nation’s official provider of data and statistics on police, prisons, jails, courts, and other components of the criminal legal system. And politically biased leadership during Donald Trump’s presidency contributed to delays in report releases and an atmosphere of censorship, former BJS officials told The Appeal.
During its more than 40 years as the Justice Department’s statistical agency, the BJS has become the primary source for national data on the United States’s highly fragmented criminal legal system, which comprises roughly 18,000 law enforcement agencies, 3,000 jails, 55 state and territorial prison systems, and hundreds of state and municipal criminal court systems.
The agency’s largest and most expensive data collection, the National Crime Victimization Survey, gathers information from households each year about the number and types of crimes they experience. Because the majority of crimes are never reported to police, the survey provides an important counterpoint to the FBI’s Uniform Crime Reporting program, which collects crime data from law enforcement agencies.
The BJS also produces dozens of surveys and reports on subjects including deaths in prison and jails, contacts between police and the public, and probation and parole systems, which include the majority of the more than 7 million people under correctional control in the United States.
Wendy Sawyer, research director at the Prison Policy Initiative, says the BJS is one of the few sources of transparency for some of the criminal legal system’s most opaque institutions, such as local jails.
“Nobody else is putting that all together. That’s just a huge undertaking. So the BJS is really the only agency I can think of that’s positioned to give us big picture data about what’s going on in local jails, which is where we see some of the worst features of the criminal justice system,” Sawyer said.
According to Sawyer, BJS surveys also provide one of the few platforms that allow researchers and the public to learn from incarcerated people about their experiences with the criminal legal system.
“We get to learn about their family history. Do they have children? What are their medical issues?” Sawyer said. “Nobody can really collect that kind of data that’s nationally representative except for the government. They have so much access.”
Politicization under Trump
Under BJS director Jeffrey H. Anderson, appointed by the Trump administration in 2017 despite concerns that he lacked relevant experience, the agency’s publication pipeline ground to a halt.
New reports and data output plummeted, with much of it released years behind schedule. According to a former BJS statistician who worked under both Barack Obama and Trump, statisticians continued to write reports and collect data at roughly the same rate as before, but Anderson refused to release reports until he had personally reviewed them.
“[Anderson] would work on one paper at a time, and I mean one paper, and he would not pick up another one until that one paper was published,” said the statistician, who requested anonymity because they are considering working with the Department of Justice in the future. “[The publications unit has] probably got at least 50 or 60 reports, many and most of which are probably publication ready. Somebody could easily look at them and publish them. They’ve been edited multiple times.”
When he reviewed reports, Anderson edited them in ways that seemed to reflect political bias, the statistician said.
“There was, you know, an official list of banned words … certain things we couldn’t say, certain things that we couldn’t emphasize, certain headlines that they wanted to bury,” said the statistician.
Banned words and “politically roadblocked” topics included words and phrases like “re-entry” and research on access to legal representation, according to the statistician. In 2017, the Washington Post reported on a similar list of “language to avoid” at the Justice Department’s Office of Juvenile Justice and Delinquency Prevention, including “substance use disorder,” “underserved youth,” and “overrepresentation of minorities (in the juvenile justice system).”
Before his appointment, Anderson was a fellow at the Hudson Institute, a conservative think tank. He also briefly ran the 2017 Project, a political organization that sought to influence Trump’s agenda during his first year in office, before joining the Department of Health and Human Services in 2017. And he was a regular contributor to the National Review, where he penned “Why Trump Should Oppose ‘Criminal-Justice Reform,’” a May 2016 article that argued that criminal justice reform, along with immigration reform, was only supported by “D.C.-N.Y. elites.” The article also falsely claimed that crime rose during the Obama administration and incorrectly predicted that federal legislation that reduced sentences for some drug offenses would lead to a spike in drug crime.
Under Anderson, the BJS released reports and press releases that independent analysts and advocates said presented a misleading picture of crimes committed by non-U.S. citizens. “They painted this very rosy picture, unless it had to do with undocumented [immigrants] committing crimes,” the statistician said.
The agency also released reports that “fuel[ed] myths” and “overstate[d]” recidivism rates, according to Sawyer and other researchers.
Wood, the Crime & Justice Research Alliance chairperson, heard from multiple academics who complained about politicization from the BJS and its sister agency, the National Institute of Justice.
“They were being asked to change language in their reports, to remove language that they thought was appropriately descriptive, and then replace it with what arguably was language that might have a political bent to it,” Wood said.
Anderson did not respond to requests for comment sent to his personal email address and social media profile before publication.
This wouldn’t be the first time political concerns have influenced BJS research. In 2005, George W. Bush’s administration faced criticism after it fired then-BJS director Lawrence Greenfeld. According to the New York Times, “political supervisors” at the department asked Greenfeld to delete references to racial disparities in a press release announcing the results of a study on police and traffic stops. Greenfeld refused. In May 2017, he and three other former BJS directors wrote a letter to then-Attorney General Jeff Sessions urging that he appoint a director who would support “evidence-based, data-driven policy making.”
Despite its importance, the BJS has been effectively flat-funded for years, meaning that its budget has not kept pace with the rate of inflation.
As a result, “the agency has for some years walked a tight line of small cuts of sample or measurement, short delays of publications, and temporary hiring freezes—each of these tolerable in itself, but cumulating over the years such that core functions have broken down,” according to a 2009 review of the agency by the National Research Council.
In the 11 years since that report’s release, the BJS’s budget fell from $60 million in fiscal year 2010 to $45 million in fiscal year 2021, a more than 30 percent reduction after adjusting for inflation.
Partially as a result of these shortfalls, many BJS data collections have been canceled or postponed over the last two decades, including the Survey of Inmates in Local Jails (last released in 2002), the National Survey of Prosecutors (last released in 2007), the State Court Processing Statistics data collection (last released in 2009), and the Arrest-Related Deaths Program (last released in 2012).
“Other government research agencies do get properly funded and do have regular data releases,” Sawyer said. “I think that really reflects what a low priority criminal justice reform or criminal justice policy in general has been.”
Missing data has hurt researchers and policymakers
Politicization, mismanagement, and funding issues have left researchers and policymakers waiting, sometimes years, for basic information about the criminal legal system.
Policymakers would often call the agency asking questions that the agency wasn’t able to answer, the former BJS statistician said.
“Do you know how many times I was asked, how much did it cost to incarcerate somebody in this state for this? How much does this facility spend for mental health care? What’s the average cost for such and such?” the statistician said. “And I would just be like, ‘I don’t know. Nobody in the nation collected it.’”
Citations of BJS products in congressional records and testimony fell by more than 60 percent between fiscal year 2016 and fiscal year 2019 because of “a decrease in the number of reports released during that time period,” according to a DOJ report last year.
According to John Pfaff, a law professor at Fordham University and author of “Locked In: The True Causes of Mass Incarceration—and How to Achieve Real Reform,” a lack of data has hindered reform efforts, particularly around what he calls the “middle” of the criminal legal system: prosecution and sentencing.
“In this time of a major push for ‘progressive prosecution,’ we’ve lost what little ability we had to see what prosecutors are doing, or to see how (or whether!) their behavior is changing, or what the impact of any such changes are,” he wrote in an email to The Appeal. “All these omissions make it really hard to figure out what is working and what is not, no matter where on the reform continuum one falls.”
Addressing these problems, of course, requires more money. In 2009, the National Research Council noted that “to fill any of these gaps in BJS’s portfolio would require increased and sustained support from Congress and the administration in terms of staff and fiscal resources.”
Although Congress voted to increase the agency’s budget by $2 million for 2021, it specified that the agency must use $3 million to fund a new study on suicide in law enforcement, effectively cutting funds for the agency’s other programs by $1 million.
The BJS budget accounts for less than 1 percent of the Department of Justice’s total spending and pales in comparison to what is spent on incarcerating people in the U.S. For perspective, doubling the agency’s now $45 million budget costs roughly the same as incarcerating 1,200 people in federal prison for one year. There are roughly 152,000 people in federal prison today.
Despite budget cuts, the next director still has ample opportunities to expand the BJS’s role as a resource for researchers and policymakers, said former BJS director William Sabol, who served under Obama.
During Sabol’s tenure, the BJS collaborated with other federal agencies and used open-source resources, such as reports of police shootings collected by media outlets like the Washington Post, to gather more data on arrest-related deaths. It found more than 100 deaths that were never reported in the press during a three-month period in 2015.
The BJS can take a similar approach in the future, according to Sabol. For instance, he suggested that the agency could use records of arrest and prosecution (“rap sheets”) collected by state governments to answer questions about prosecution and sentencing, an approach some researchers have used to estimate the effect of guilty pleas on sentence length.
“My guess is that’s not a one-year project, it’s a multi-year project,” Sabol said. “It’s like planting trees, and the metrics are the fruit.”
Sabol also suggested that the BJS could shed light on police funding by combining data on police expenditures from the Census Bureau with its own information and that of the Office of Community-Oriented Policing Services (COPS) and the FBI. The data could help researchers and the public better understand why police budgets have remained high even as crime has approached historic lows in many jurisdictions: Are agencies hiring more officers? Buying more equipment? Expanding their civilian workforces?
“There’s a whole slew of efforts that could rely on expanding this data-collecting community,” Sabol said. “The same thing could happen in corrections, jail, all that.”
Regardless of how the BJS expands its data portfolio, Wood says that simply getting more data would be a huge boon for policymakers.
“You need the best data that you can get in order to have effective policy and find out whether it’s working or not, or whether it’s counterproductive,” Wood said. “To the degree to which the information you’re getting is seven or eight years old, or has some political taint associated with it, that’s a problem.”