Association for Public Policy Analysis & Management

A central focus of educational research is to identify explanations for (and interventions to reduce) this fanning out of achievement during the school-age years. To better guide the field’s explorations of this key question, we build on the existing literature to document the extent to which this fanning out occurs during the academic year in school settings or—instead—during the summer. Even if, over the summer, students tend to lose ground, summers will only contribute to widening achievement disparities if students exhibit meaningful variation around the typical summer pattern. We therefore focus on the degree of variability across students in summer learning losses, relative to school-year learning gains. The overall purpose of the current paper is to update the existing knowledge base about overall first through eighth grade school year learning gains and subsequent summer loss patterns, document the degree of variability in those patterns, and characterize the extent to which end-of-school achievement disparities arise during summers.

Using data provided by the Northwest Evaluation Association (NWEA), we estimate means and variances in summer learning loss from K to 8 from a dataset that contains over 200 million test scores for nearly 18 million students in 7,500 districts across all 50 states in a very recent time period (2008 through 2016). In the current paper, we use this powerful dataset to characterize the contribution of summer learning loss to end-of-school achievement disparities.

It is clear that the summer period is a particularly variable time for students. We find that many students can in fact maintain average school-year learning rates during the summer in the absence of formal schooling. Other students, however, will lose nearly as much as what is typically gained in the preceding school year. This remarkable variability in summer learning rates appears to be a strong contributor to widening achievement disparities as students move through school. On average, about 19 percent of the fluctuations in students’ achievement between first and fifth grade occurs during the summer. However, for some students, the summer accounts for as much as 30 percent. This is a particularly outsized influence given that the summer is about a third the length of the school year. Ultimately, we show that—even if all the inequality in school-year learning rates could be entirely eliminated, students would still end school with very different achievement levels because of what happens during the summer.