When his phone rang one afternoon in January 2015, Jeffrey Liebman, Wiener Professor of Public Policy and director of the Taubman Center for State and Local Government, was grading exams. “I decided to answer anyway,” Liebman says, “which is good, because it turned out to be the governor of Rhode Island asking for help.”

Governor Gina Raimondo had just taken office, inheriting a troubled child welfare agency $16 million over budget and with a higher percentage of children in group care settings than almost any state in the country. An audit would reveal that the agency was signing away tens of millions of dollars without financial controls or oversight and was producing poor outcomes for the children and families in its care.

“When I took office, the Department of Children, Youth and Families faced a number of significant operations and budget challenges that had to be addressed quickly so that we could provide better outcomes for our children,” Raimondo says. “I knew that Jeff had started an organization at the Kennedy School that could actually send staff to come work full-time, pro bono on the ground supporting transformational initiatives. I asked him to send as many people as he could.”

In response to Governor Raimondo’s request, Liebman agreed to embed a team from the Government Performance Lab (GPL) with Rhode Island’s Department of Children, Youth, and Families (DCYF) to work with agency staff to figure out solutions to the agency’s most pressing problems.

Liebman had founded the Government Performance Lab in 2011 after serving in leadership roles at President Obama’s Office of Management and Budget. The GPL’s mission is to help state and local leaders make progress on social problems by improving the way governments contract for and manage social services. The GPL developed a hands-on model of research, embedding staff in government agencies to provide technical assistance. These team members, including many HKS alums, typically work with an agency for one to two years to help develop, test, and implement innovative management strategies.

Liebman had heard about Pay for Success (PFS) projects in the United Kingdom and was interested in testing their use in the United States. Under PFS contracts, also called social impact bonds, the government agrees to pay for services only if they are shown to be successful, allowing it to test promising new interventions and pull together disparate stakeholders in long-term collaborations. The GPL’s inaugural projects assisted Massachusetts and New York in setting up the first two state-level PFS project in the country, one aimed at reducing recidivism and the other at helping young people exiting the juvenile justice system.

Jeffrey Liebman headshot.

These hands-on engagements are where the best new ideas for public sector management are going to come from.

Jeffrey Liebman


“I thought I was going to do a few small projects and then get back to regular teaching and research,” Liebman says. But he discovered there was enormous appetite for the type of data-driven, boots-on-the-ground technical assistance his fledgling team could provide. Before the first PFS project had even launched, 25 more governments had applied for assistance. Liebman began building a staff with backgrounds in data analysis, public policy, and implementation, and placing them in the field with governments selected through national competitions. Over the next few years, the GPL helped launch 12 of the country’s first 18 PFS projects, including initiatives in maternal and infant health, high-quality pre-K, and homelessness, expanding over $135 million in preventative services to more than 20,000 individuals.

It quickly became clear that the GPL could have a greater impact if it moved beyond PFS and instead helped agencies improve their overall operations. “Our government partners began to ask whether the same models we developed in PFS could be applied to help government social service agencies run their core operations better,” says Ryan Gillette MPP 2012, a program director who launched PFS projects in Massachusetts, Chicago, Connecticut, Michigan, and Alameda County, California. “We’d be working on a small PFS project with an agency that was spending $200 million a year on core services with no strategy and no use of data. Agency heads began to say, ‘We want you to take what you did in PFS and apply it to all of our work.’ PFS became the admissions ticket to broader systems change.”

In 2016 and 2017, the GPL invited applications from governments seeking assistance with high-priority initiatives and was overwhelmed by responses. Raising over $22 million in grant funding, the GPL expanded at a breathtaking pace—today it is a 45-person organization and has worked with 65 governments in 31 states.

“We began to see common challenges across governments,” says GPL Program Director Danielle Cerny MPP 2014. “They have talented staff but are so consumed with putting out fires and keeping things running that they don’t have the capacity to strategically connect their daily activities with their long-term goals. Are there high-needs individuals that they ought to be reaching who are not connected to services or have fallen through the cracks? Are the services the agency inherited actually the ones that their clients need? How can they use real-time data to improve the outcomes they get from their providers? Working across jurisdictions has allowed us to develop a set of approaches that responds to the recurring issues we see governments grappling with.”

Liebman and his growing team have become experts at using analytics, outcomes-focused contracts, and cross-sector collaborations to help improve social service programs. The GPL helps governments address thorny questions, like how to use data to find individuals who had fallen through the cracks, or how to restructure contracts to incentivize providers to serve the most difficult clients. “From these initial projects, it became clear that the best way to figure out how to help governments administer and manage programs better is to try it out. These hands-on engagements are where the best new ideas for public sector management are going to come from,” Liebman says. “And once our government partners see the value of having a change agent on their team, they often create positions to hire our fellows directly after our formal technical assistance comes to an end. Our work lets us train a new generation of passionate public servants and connect them to governments that are hungry for talent.”



One of the key insights Liebman’s team had uncovered during its PFS work was that procurement and contracting were too often seen as back-office compliance functions instead of used as powerful levers for systems change. GPL developed a framework for Results-Driven Contracting to help governments achieve high-priority strategic goals and drive improvements in service delivery. Its success in helping governments make significant changes in day-to-day operations caught the eye of the Bloomberg Philanthropies’ What Works Cities Initiative, which awarded the GPL $5 million to coach more than 26 U.S. cities on implementing Results-Driven Contracting practices.

Through the initiative, Seattle applied for help on its homelessness contracts. Over the course of a decade, the city’s budget for homelessness services had almost doubled to $50 million, but homelessness was still increasing at 13 percent every year. “We had 168 contracts with homelessness service providers, but we couldn’t tell if they were adding up to actual progress,” says Tyler Running Deer, the city’s former organizational performance director. “We needed more detailed data and more frequent data review points. It was clear that it wasn’t just a question of spending more money, but spending it smarter.”

The GPL helped the city review its homelessness contracts and identified 26, worth $8.5 million, to rework. The GPL found that rather than measuring meaningful metrics of client progress, provider staff were asked to report on measures of activity such as the number of beds filled or the numbers of showers administered—not only were such metrics meaningless, but they took up a significant amount of staff time to compile.

The GPL helped the city simplify and streamline the contracts around a core set of metrics related to the outcomes the city cared most about—whether individuals were progressing to stable housing situations rather than cycling back into homelessness. Those new metrics are allowing the city to measure progress in real time and implement data-focused collaborations between service providers and agency staff that are improving performance. “In homelessness, the challenge is often overcoming a perpetual crisis state in order to focus on preventative and long-term solutions,” says Tyler Jaeckel MPP 2013, the program director who leads the GPL’s homelessness and behavioral health practice. “Our homelessness projects in Seattle, Denver, and Massachusetts take innovative approaches to addressing the crisis and develop new methods for shifting resources to solutions that are likely to generate lasting change.”



Following Governor Raimondo’s call to Liebman in 2015, a GPL team arrived in Rhode Island’s Department of Children Youth and Families (DCYF) soon after the appointment of an interim director tasked with leading a full agency turnaround. They began their work by painting a basic picture of what the agency was buying. “Our fellow literally went desk to desk finding contracts in drawers and built a database of the money the agency owed for services,” says Scott Kleiman, the program director overseeing the GPL’s work with DCYF. The resulting dataset offered the agency’s new leadership team its first glimpse at historical service utilization trends and enabled them to project future funding needs. It also allowed them to see that the department’s array of services had not kept up with the needs of children and families in care. Automatically extended year after year, many contracts had remained frozen in time, even as needs had shifted. In recent years, for example, a growing number of young women had been ensnared in sex trafficking, yet DCYF lacked any specialized services to address their unique needs.

The contract overview laid the groundwork for mapping services against needs and pinpointing areas where interventions were missing or where an approach needed to be reimagined. When the agency asked for assistance with the reprocurement of nearly $100 million in core services (about 40 percent of the agency’s budget), it became an opportunity to go back to the drawing board and find new service delivery solutions. GPL staff fanned out across the state, talking with dozens of nonprofit service providers and agency social workers, elevating often overlooked perspectives from the service frontlines. Eschewing traditional “cut and paste” contract renewals, Kleiman’s team then helped DCYF structure a multi-part “problem-based” procurement that engaged a broad set of stakeholders in proposing solutions to the challenges the agency was trying to solve.

At the time, Rhode Island was close to worst in the nation in the percentage of children in its care in group homes rather than with foster families or relatives.  The agency noted that children with clinical needs were often being sent to placement in group homes. At the agency’s first-provider summit, the GPL asked nonprofit service providers to brainstorm alternative interventions. Providers suggested an evidence-based model from Oregon that provided high-intensity support for foster families taking in children with behavioral challenges.

But the GPL recognized that simply procuring a new service array was not enough. DCYF needed better practices for effectively managing its contracts while services were being delivered. Kleiman and his team implemented an approach to engaging service providers that upended typical government practice. Like many agencies, DCYF had separate teams overseeing contract invoicing, quality assurance, and program evaluation. Each one functioned separately, and no team was tasked with using real-time data to improve service delivery. The GPL helped DCYF establish a system that engaged service providers and agency staff in high-frequency, data-driven collaborations to troubleshoot problems and identify opportunities for performance improvement. In these “Active Contract Management” meetings, GPL staff used data and case reviews to focus provider and agency participants on generating new solutions to key problems.

So when the agency was still seeing children with clinical needs linger in group care, Active Contract Management provided the structure to bring the agency and service providers together to try to tease out what was happening, says GPL fellow Alice Heath MPP 2016. “Digging into the data, we learned that many kids would have their clinical needs met but then would hit administrative barriers that created lengthy delays to placement with a foster family,” Heath says. “That was a new insight for the agency that helped change the way they targeted their efforts to make sure as many kids as possible were in homes with a family.” With the GPL’s help, the department launched a pilot to meet with families early in a child’s group home placement to determine strategies to better prepare them for a move to a foster family.

Caption goes here.

“Our GPL colleagues bring new perspectives to our work,” says DCYF Director Trista Piccola. “Whether implementing an Active Contract Management structure with our community providers or helping our team align our contracting processes, their help has been invaluable. Change is one of the most difficult issues faced by large systems like ours, and GPL has helped us create positive, lasting change for our state’s children and families.”

The results of the GPL’s engagements thus far are impressive. During the GPL’s time in Rhode Island, the share of foster children in group homes declined 29 percent, the volume of contracted foster homes rose 63 percent, and the number of children in out-of-state institutional care dropped 44 percent. At the governor’s request, the GPL has embedded fellows with eight other state agencies and led over 100 state agency staff in trainings. In 2017, Seattle announced that it would utilize the results-driven contracting framework it had piloted with the GPL in a procurement for $30 million in homelessness services. Jurisdictions that initially reached out to the GPL for assistance with a PFS contract are now applying the data-driven principles they learned to core services. For example, Connecticut’s Department of Children and Families, which worked with the GPL on a $12 million project improving services for families dealing with substance abuse, is now working with GPL fellow Matthew Tyler MPP 2017 to transform the way it connects families and children to services.

As it nears its seventh anniversary, the GPL is poised for even greater impact. “We have far more governments asking us for help than we have the resources to serve,” says Liebman. “Our focus going forward is on continuing to do innovative projects that generate new solutions while also finding cost-effective ways to spread the solutions we have already demonstrated.”

Photo by Martha Stewart

Get smart & reliable public policy insights right in your inbox.