The Evidence Base

Informing Policy in Health, Economics & Well-Being
A collaboration with
USC Dornsife Center for economic and social research

Two Percent of U.S. Children Receive High Quality Tutoring, Despite Billions Funneled into School Systems

In the aftermath of Covid-related school closures and continued school disruptions through multiple academic years, there has been intense urgency to find ways to help students with unfinished learning or who continue to struggle to keep up. NAEP’s first large-scale, nationwide, analysis of student achievement post-Covid documented historically low performance and unprecedented declines, with widening (i.e., worsening) of achievement gaps. As EducationWeek called out, “Two decades of progress, nearly gone.”

The Promise of high quality/high-impact tutoring

One area of promise for addressing students’ academic needs is tutoring. There is consensus that tutoring is effective. A recent meta-analysis has helped to summarize the evidence. Reviewing 96 randomized-controlled trials (the “gold-standard” in efficacy research) published since 1980, the authors conclude that “tutoring programs yield consistent and substantial positive impacts on learning outcomes.” A separate meta-analysis examining different types of educational interventions found that tutoring programs had the largest estimated impacts across all types of interventions examined. As one expert summarized, “Tutoring is a proven and agreed upon strategy that education leaders can be sure works.”

Further, research has coalesced around the features of highly effective (or high impact) tutoring programs that lead to positive outcomes. Backed by rigorous research, and recommended widely by the Institute for Education Sciences, “high quality” tutoring:

  • Is delivered at school
  • Is delivered during school hours
  • Groups are composed of three or fewer students to one tutor (four or fewer students can be effective with older students)
  • Is provided to a student three or more times per week
  • Lasts at least 30 minutes per session.
  • Is delivered by teachers or well-trained professional tutors rather than volunteers, peers, or parents.

Barriers to delivery

Yet delivering high-quality tutoring remains a challenge. Finding qualified tutors, navigating students’ schedules, reaching the volume of students who need extra time, communicating with families, even just motivating students to attend –these barriers continue to interfere with the delivery of services that we know will help students! Many of these challenges existed before COVID began. What is different is that in the wake of the pandemic, huge sums of COVID-relief funds were allotted to K-12 schools through the federal American Rescue Plan, with large portions earmarked for implementing and/or scaling-up tutoring programs ($27 billion by one estimate). With access to unprecedented resources like this, shouldn’t high-quality tutoring by now be reaching students?

Despite funding, high-quality tutoring is not reaching students

Our nationally representative Understanding America Study (UAS), the University of Southern California Center for Economic and Social Research’s survey of American households, suggests that the barriers are proving substantial. We found little evidence that tutoring is broadly reaching students, and even less that the tutoring delivered meets agreed-upon criteria for high quality, despite the infusion of funds.

When asking adults in December 2022 about their students’ fall semester, surprisingly few – about 15% of over 1,600 households – reported that their student was receiving any tutoring at all. This percent was appropriately higher (24%) among students whose parents reported their students typically earned grades of C or lower. There is some, albeit small, unmet need – another 12% would have their child participate if they could but reported that their school did not even offer the service to their child.

With relatively low reports of tutoring participation, and large sums of money available, one might expect that at least the tutoring being delivered meets the provision of “high quality” services – smaller student to staff ratios and more time spent per child for example. But that’s not what we found. Only 11% of those receiving tutoring described their students’ programs in a way that met the criteria for high quality1 using a relatively liberal definition. Our definition of high-quality tutoring only required that the student’s tutoring be 1) in person, 2) in groups of four or fewer students to one tutor, 3) for 30 or more minutes each session, and 4) for three or more sessions per week. We did not limit only to tutoring received during the regular school day, nor did we limit to only “well-trained” tutors. We even categorized groups of four students to one tutor as sufficiently “small” for students in any age group.

Another way to look at these same numbers – across all 1,600 households surveyed, less than two percent of students are receiving tutoring that even meets a fairly moderate definition of “high-quality.” And among those who likely need it most – students who receive grades C or lower – less than four percent are receiving high-quality tutoring. While surely not all students need tutoring, there is no question that the actual percentage of students in need of high-quality academic tutoring intervention is greater than this.2

Though disappointing, these numbers square with recent reports from the federal School Pulse Panel which posed similar questions, but to schools rather than to individual households. Approximately 37% of schools reported offering high-dosage tutoring (similar to our definition of high-quality), and staff in those schools estimated approximately 30% of students receiving high-dosage tutoring. Applied to the nationwide sample, this means only 10% of American students are receiving high-dosage tutoring!

Where to focus improvements

A closer look at our data helps to identify areas for some specific improvements. About 88% of  our tutored families reported that tutoring is provided in a face-to-face setting and 79% of those receiving tutoring are doing so in appropriately small groups (including one-on-one). But only 25% are receiving tutoring at least three times per week – a critical element of effective tutoring. And, a bit less than two-thirds (63%) are meeting for 30 minutes or longer per session. So even among those who have initial overcome barriers to initiating services, the dosage schools are providing is likely not sufficient.

The takeaway message is clear – we are far from the goal of providing high-quality tutoring to all who need it. To be sure, implementation challenges abound. But this is the rare instance where we have a good sense of what we need to do, and we have the funding to help reach those goals. Our results suggest two areas of particular focus: 1) reaching more students (which likely involves better communication with families and finding and hiring more tutors) and 2) and increasing the dosage of tutoring that is already being provided. These changes have to happen now, in advance of the next school year – we simply can’t afford to provide services that may not move the needle for millions of students!

We gratefully acknowledge financial support from the National Science Foundation Grants No.2037179, 2120194, and 2214168, and the Hewlett Foundation. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the funding organizations. We are also grateful to Marshall Garland, Michael Fienberg, Morgan Polikoff, Anna Saavedra and the UAS administration team for their contributions to this work.

1. Some parents did not know details about one or another feature of their child’s tutoring, for example 3% did not know session frequency, 11% did not know group size, 13% did not know the session length. If we remove those respondents from the calculation, the percent of students receiving high quality tutoring rises to 13%.  

2. Our survey uses probability-based sampling methods. Results are representative of the population at large.