Skip to content

Unlocking the Power of Data

Across the nation, policy labs are pairing data experts with government officials to solve our most pressing social problems.

Arnold A decorative icon
Illustration of silhouetted person against data background

A common thread emerged as researcher Elysia Clemens took a closer look at Colorado’s foster programs: Many youths were dropping out of high school. Exactly how many, she couldn’t determine. At the time, educational databases and child welfare databases were housed separately and run by different state agencies.

So Clemens and her team at the Colorado Evaluation and Action Lab, one of nine policy labs funded by Arnold Ventures, partnered with the Colorado Departments of Human Services and Education to develop an integrated sharing system that could provide a 360-degree view of student behavior.

What they found was astounding: Colorado students in foster care, on average, changed public schools 3.46 times during their first four years of high school. Worse still, as the number of school changes during this time increased, the odds of earning a high school diploma decreased drastically: Less than one-third of foster youth were graduating high school on time, compared with 75 percent of Colorado’s general student population.

“When I first looked at it, I thought, ‘This can’t be right,’” Clemens told The Denver Post in May 2018. “It was very disheartening and surprising to see how much we need to do in order to change the odds for these young people.”

The Colorado Senate meets during the body's closing session in May 2017. (David Zalubowsk/The Associated Press)

The Colorado policy lab, which is housed at the University of Denver, shared these findings with lawmakers, and in 2018, Gov. John Hickenlooper signed into law a major bill to fund the transportation of foster youth to their schools of origin. The legislation drew heavily from Clemens’ research about the underlying causes of dropout rates — research that couldn’t have been done without the partnership her team formed with state officials.

That’s the power of policy labs: They bring together academics, who have the expertise to do research but not the access to data, with policymakers, who have the data but not always the resources to invest in research — all in an effort to make the government work better.

“There is untapped information sitting on computers in the state Capitol building right now that could help us understand how we get more kids through college or break the cycle of incarceration or move families out of poverty,” said Kelli Rhee, President and Chief Executive Officer of Arnold Ventures. “We want to bring together the best researchers in the country with public officials to unlock the power of that data and use it to make real progress on these problems.”

The process begins when public officials identify a policy challenge (jail overcrowding, homelessness); a program that could be improved in its efficiency or effectiveness (school suspensions, youth employment); or a new approach they want tested. They then hand over whatever relevant data they have to the policy lab, where data analysis experts and social scientists start studying the issue.

There is untapped information sitting on computers in the state Capitol building right now.
Kelli Rhee President and Chief Executive Officer of Arnold Ventures

The lab team helps define the problem or opportunity and then draws on existing research and analysis to suggest a useful approach. Programs that are shown to be effective can be replicated and scaled, while those that do not deliver the hoped-for effects can be improved or replaced.

In just two years, the multimillion-dollar policy labs — which are spread across six states and the District of Columbia — have used data and research to unlock solutions in multiple areas. In addition to helping Colorado officials uncover why foster care youth dropped out of high school at much higher rates than their peers, policy labs tested ways to reduce jail overcrowding in California. And they conducted a seminal study evaluating the impact of school suspensions on academic performance, which led Texas to ban the practice for nonviolent offenses from pre-kindergarten through second grade.

While several policy labs are already seeing the fruits of their work (see box), younger labs are taking root in growing communities. On the campus of Rice University in Houston, dignitaries gathered in September to break ground on a social sciences building that will house the Texas Policy Lab. The lab’s research agenda will be set by state agencies and Texas policymakers — one of its first projects will be a collaboration with the Department of Criminal Justice as it seeks to lower the rate of incarceration in the Texas prison system — with a shared goal of helping Texans lead healthier, happier, more productive lives, said Rice Dean of Social Sciences Antonio Merlo.

“Using cutting edge research, the Texas Policy Lab will work directly with Texas policymakers to develop measurable solutions to the state’s most pressing problems,” Merlo said. “We will provide scientific, independent, and timely analysis to our partners in order to ensure policies and programs are producing optimal outcomes for the constituents they aim to serve.”

What They’re Doing

Photo by Matt Rourke/The Associated Press

Here’s a look at some of the research happening in policy labs across the country:

Washington, D.C.: The Lab @ DC partnered with the District of Columbia police department in 2017 to study whether having officers wear body cameras would have an effect on police use of force, civilian complaints, policing activity, or judicial outcomes. What they found: The cameras had no discernible impact — on police or civilian behavior. “I think we’re surprised by the result,” said Chief of Police Peter Newsham to NPR. “There was no indication that the cameras changed behavior at all.”

California: The California Policy Lab paired with the San Francisco Public Defender’s Office to evaluate the impact of a new Pretrial Release Unit that provides attorney representation to indigent defendants earlier in the process than is customary. The study found that the pilot program doubled the likelihood of release at arraignment for those who received interventions and reduced pre-trial incarceration by 44 percent. From the time the unit began seeing defendants in October 2017 to the end of February 2018, the program saved an average of 940 jail bed days a month.

Rhode Island: Responding to Gov. Gina Raimondo’s goal to reduce the state’s three-year recidivism rate from 52 percent to 44 percent by 2020, the Rhode Island Policy Lab got to work. It discovered that individuals who enroll in the federal Supplemental Nutrition Assistance Program post-release are 7.4 percentage points less likely to reoffend within the first six months than their non-SNAP counterparts. The problem: Only 40 percent enroll in SNAP post-release even though nearly all qualify. So the policy lab, in collaboration with the Rhode Island Departments of Corrections and Homeland Security, developed a program called Connect for Success to ensure all inmates have activated benefits cards post-release.

Michigan: The Youth Policy Lab at the University of Michigan researched the topic of early childhood maltreatment and its effects on academic outcomes. It found that about 18 percent of Michigan third-graders have been investigated by Child Protective Services for possible exposure to maltreatment and that early childhood maltreatment is associated with significantly lower academic outcomes, even after controlling for key demographics. Among its list of recommendations, the lab called on state officials to design and implement databases to allow for easier and more-timely sharing of data between the education and child welfare systems.

160
Data lab projects underway in six states and the District of Columbia
$38.1m
Money committed for policy labs from Arnold Ventures
>30
Researchers working in policy labs
$150m
Money reallocated to more effective and efficient programs as a result of data lab research