Association for Public Policy Analysis & Management

“Long summer holidays are bad for children, especially the poor” (The Economist, 2018)--this article was released just five days ago. The impact of summer break on learning continues to be of interest to parents, researchers, policymakers, and teachers, even 40 years after the foundational study of summer learning loss (Heyns, 1978). One reason it continues to be of interest is the often-cited finding that, on average, children tend to lose academic skills over the summer (Cooper et al., 1996). Perhaps most troubling, early summer learning studies found that the gap between low- and higher-income peers was especially likely to grow over the summer (e.g., Alexander et al., 2001, 2007).

Although the phenomenon of summer learning loss appears to be well-established, more recent work highlights that the literature is actually quite mixed. Von Hippel et al. (2016) and Quinn (2015) both note that findings from early studies of summer learning loss may need to be viewed more skeptically due to limits in psychometrics and differences in modeling strategy. Importantly, the work on summer learning has been very concerned with the “average” student losing ground over the summer. The current study’s findings suggest that the “average” summer learning estimate does not reflect the actual experience of any student in the dataset. Although this in itself is not surprising, it highlights that important nuances may be missed by solely focusing on the average summer learning estimates. Thus, although the literature on summer learning appears well-established, it may be in fact still very tenuous.

Using a subsample of the Early Childhood Longitudinal Study, Kindergarten Class of 2010-11 (ECLS-K:2011; n = 4880), the current study examined summer learning in reading and math at two timepoints: between the end of kindergarten and the beginning of first-grade and between the end of first grade and the beginning of second-grade. Children who were homeschooled, and thus did not have summer vacation, were excluded from the analyses as well as those who were repeating kindergarten after summer vacation. Descriptive statistics suggest that, contrary to the often-reported “summer slide”, the majority of children gain over the summer in both reading and math and at both time points (see Table 1). In fact, only 13% of children actually lose reading or math skills between kindergarten and 1^st grade. Although the percentage of students who lose in reading or math between the spring of 1^st grade and the fall of 2^nd grade grows (22% and 24%), it is still the minority of children that are losing over the summer. Findings from this study can help change the rhetoric around summer learning—some are learning, some are not—but most are. Follow up analyses describe summer learning, focusing on who loses over the summer and why. Preliminary analysis suggests that an especially important factor for promoting summer learning is child behavior; teacher-rated approaches to learning and externalizing behaviors before summer break are highly predictive of learning over summer break.