MDRC’s Findings on School Improvement Strategies

By Joshua Malbin, Rekha Balu

The Every Student Succeeds Act (ESSA) — which replaced No Child Left Behind as the main federal K-12 education law — gives states greater responsibility for choosing strategies to improve underperforming schools.[1] For over a decade, MDRC’s K-12 Education Policy Area has rigorously evaluated school improvement strategies, collecting evidence that can help states determine which strategies are likely to work. This Issue Focus describes four of MDRC’s most recent studies.

Small Schools of Choice

Beginning in 2002, New York City closed 31 large, failing high schools and opened more than 200 new, small high schools in their place. These schools were developed and approved through a competitive proposal process designed to elicit innovative ideas from a range of stakeholders and institutions. They emphasize academic rigor; strong, sustained relationships between students and faculty members; and community partnerships to offer relevant learning opportunities outside the classroom. Each school also received start-up funding as well as assistance and policy support from the district and other players to facilitate leadership development, hiring, and implementation.

Also beginning in 2002, New York moved to a district-wide high school choice process for all students entering ninth grade, which included a lottery system that kicked in when too many students picked the same school. The lottery system provided a random assignment-like experimental condition that allowed researchers to estimate the effects of attending one of these new, small schools.

For the full sample — over 21,000 children — students at small high schools had a graduation rate 9 percentage points higher than their control group counterparts. These effects were seen among nearly all subgroups, including special education students. Attending a small high school also increased the percentage of students who graduated from high school in four years and enrolled the next year in college by 8 percentage points. The small high schools achieved these gains at a lower cost per graduate — roughly 14 percent to 16 percent lower — largely because more students successfully graduated from small high schools and fewer needed to attend an expensive fifth year.

Success for All

Success for All aims to improve the reading skills of all elementary school children, especially those from low-income families. Its reading program emphasizes phonics for beginning readers and comprehension for all students, while its whole-school reform model also incorporates professional development, frequent assessments of students’ progress, tutoring, a team to solve nonacademic problems, a facilitator for teachers, and leadership development.

Based on earlier, promising evaluations, in 2010 Success for All won a U.S. Department of Education Investing in Innovation grant, which it used to expand to 447 new schools with a total enrollment of some 218,000 students. During this expansion, MDRC evaluated the model in 37 schools, 19 of which were randomly assigned to receive Success for All and 18 of which were not.

The evidence indicates that for a modest investment (about $227 per student), Success for All did effectively teach phonics: In Success for All schools, second-grade students performed better in phonics than students in control group schools. Students in Success for All schools did not outperform control group students in reading fluency or comprehension. The model seemed especially effective for students who entered school with poor preliteracy skills. Although the great majority of schools that implemented Success for All did so faithfully, resource constraints prevented some schools from putting in place some of its features, including the full-time facilitator and computerized tutoring program. Program and control group schools were different in some respects (for example, Success for All schools were unique in placing students in cross-grade ability groups for reading, and Success for All teachers made more use of cooperative learning), but similar in others.

Communities In Schools

Founded in 1977, Communities In Schools serves more than 1.3 million students in over 2,000 schools. The model aims to reduce dropout rates among low-income students in some of the nation’s poorest-performing schools by providing so-called “Level 1” and “Level 2” services. Level 1 services are broadly available to all students or to groups of students and are usually short-term, low-intensity activities (for example, making clothing or school supplies available or organizing school-wide career fairs). Communities In Schools site coordinators — those responsible for school-based operations — spend much of their time on more intensive Level 2 case management, which they provide to students who display one or more significant risk factors for dropping out.

MDRC is conducting a two-part evaluation of Communities In Schools. One study focuses on the implementation and impact of Level 2 services, randomly assigning some students to receive case management and others to receive the other forms of support available in their schools. The second study investigates both Level 1 and Level 2 services, using a quasi-experimental design to compare schools that implemented the model with similar schools that did not.

So far, MDRC has published only a year of results from the first study; those show that while case-managed students have received more services than control group students, those services have not yet had an impact on students’ attendance or course performance, or on the behaviors that lead to disciplinary action. There were some effects on students’ adult and peer relationships and on how much they value education. “Higher-risk” students did not receive more services than others. Additional results should be published in the coming year from both halves of the evaluation.

Diplomas Now

Diplomas Now is a partnership of three national organizations — Talent Development Secondary, City Year, and Communities In Schools — to transform urban secondary schools so fewer students drop out and more graduate ready for postsecondary education and work. The model combines a comprehensive school reform strategy aimed at transforming the academic experience of all students with targeted interventions for students at risk of dropping out. By using early warning indicators related to attendance, behavior, and course performance to identify at-risk students and then intervening with targeted support, Diplomas Now attempts to get failing students back on track.

MDRC and ICF International are conducting an experimental evaluation of the impact and implementation of Diplomas Now that includes 62 middle and high schools in 11 school districts. Thirty-two of these schools were randomly assigned to implement the Diplomas Now model, while the other 30 schools were assigned to a control group. So far, schools have implemented a majority of program components as the model’s designers intended, but still showed room for growth to meet ideal implementation. Implementation has produced differences in organization and services between program and control schools, suggesting that schools’ practices have changed in ways that accord with the goals of Diplomas Now. Program staff members have been most successful in implementing the model when they could align the program with schools’ own goals and garner school administrator and teacher support. The first impact results should be available this year.

E-mail Rekha Balu or William Corrin for more on these and other school improvement strategies.


[1]Daarel Burnette II, “States, Districts to Call Shots on Turnarounds Under ESSA,” Education Week (January 6, 2016).

Document Details

Publication Type
Issue Focus
April 2016
Malbin, Joshua and Rekha Balu. 2016. “MDRC’s Findings on School Improvement Strategies.” New York: MDRC.