There are troubling trends in education: In 2024, reading scores declined across both fourth and eighth grades, with no state seeing gains in either grade compared with 2022. A third of eighth-graders are reading below the most basic standard—more than ever before. Schools and state educational leaders need evidence-based interventions that can help students succeed. Among the options available, high-dosage tutoring (where students work with a trained tutor in small groups at least three times per week) stands out for its strong track record of effectiveness. But recent research has demonstrated that it has smaller effects when implemented on a large scale.

The Personalized Learning Initiative (PLI), a collaboration between the University of Chicago Education Lab and MDRC, is partnering with school districts around the country to implement and evaluate a range of high-dosage tutoring models. The models vary in a number of ways, and the goal of the study is to assess which models work best for which students in which contexts.

What the PLI work has found so far is that biggest factor contributing to the success of such programs is implementing tutoring programs well enough that students actually receive enough tutoring sessions. Making sure students receive consistent, high-quality instruction probably matters much more than other variations (for example, the size of the tutoring group or whether tutoring is delivered in person or virtually).

Consider New Mexico, which faced a pressing gap in math achievement, with only 13 percent of eighth-graders scoring proficient in math in 2022. The New Mexico Public Education Department pilot tested virtual tutoring to try to address this gap, offering free, high-quality algebra instruction on evenings and weekends. Just 1.5 percent of eligible students enrolled. Several thousand students were eligible for the virtual tutoring program, and state officials received fewer than 500 applications. Rural students faced particular barriers, citing concerns about work and family commitments, internet connectivity, and competing priorities.

Rather than abandon the virtual model, New Mexico made one pivotal change for the 2023-2024 school year: It moved tutoring from evenings and weekends into the regular school day. Attendance rates dramatically improved. Small, rural schools found they could provide tutoring to entire grade levels simultaneously. While some schools repurposed electives, others carved out time for tutoring without sacrificing electives by creating specific intervention periods.

Lessons for State Leaders Interested in Tutoring Programs

1. Encourage tutoring during the school day and create models for consistent student participation. Doing so may mean giving districts more flexibility to restructure schedules, using intervention periods, electives, or other instructional time.
2. Plan to pilot test your program and leave ample preparation time to bring the program from its pilot phase to full scale. State funding for tutoring programs should include sufficient time to plan an effective program. It may be helpful for schools or school districts to launch a small pilot test in the middle of a school year, with a clear plan to evaluate success and make adjustments for wider implementation the following school year.
3. Require reporting on whom tutoring is serving, attendance in tutoring sessions, and outcomes. Such reporting requirements (a feature of outcome-based contracts) can help ensure that tutoring programs actually use data-collection systems in ways that allow leaders to identify and address challenges with student attendance and other implementation issues. State leaders should not only require reporting but also direct funding to give schools and districts the tools they need to collect data efficiently.

To fulfill its promise, tutoring must be more than well designed, it must be consistently offered and attended. The story of New Mexico’s postpandemic tutoring offers a clear lesson: Offering tutoring during the school day and backing it with strong implementation systems can lead to higher participation and better outcomes. State leaders can transform promising programs into powerful engines of learning—if they focus on getting the delivery right.