Alex Cohen is the Director of Learning and Evaluation for the Richard M. Fairbanks Foundation.

High-performing school leaders have a substantial impact on student achievement. Studies that aim to isolate the impact of leaders on student test scores find that moving from a low- to a high-quality principal can mean several additional months of learning. And the impacts are larger for low-income students.

But can schools do anything to boost the quality of their school leaders? Or are the traits that lead to better leaders too difficult to predict or develop and the best leaders too hard to retain?

To help answer these questions, the Wallace Foundation funded and provided technical assistance to six districts from 2011 to 2016 as part of its Principal Pipeline Initiative. The initiative included preservice preparation for new and prospective school leaders, systems for identifying and placing new leaders, and on-the-job training and support. The districts included were large, urban districts spread throughout the U.S. (North Carolina, Colorado, Florida, Georgia, New York, Maryland).

Over the course of the initiative, the six participating districts were able to implement all components of the new principal pipeline model at scale, revamping their process for attracting, preparing and retaining high-quality leaders. But did it lead to any changes for schools or students?

An evaluation by RAND Corporation found that after the launch of the initiative in these districts, students saw larger improvements in student outcomes—by 6.22 percentile points in reading and 2.87 percentile points in math—than similar students in comparison schools using business-as-usual principal pipeline practices. And the initiative’s positive effects tended to be largest for schools who were initially the lowest performing.

The researchers also found leaders that went through the new pipeline program were nearly 8 percentage points more likely to stay at their school for three or more years, relative to leaders in comparison schools.

While the initiative involved comprehensive changes in district practice, the cost was relatively low at an average of $42 per student or 0.5% of the district budgets for the districts included in the study.

Combined with studies of other principal training programs, these findings suggest it is possible (and financially feasible) for schools to increase leader quality—and see improvements in student outcomes that come from having a high-quality school leader.