Does the Shift to a Four-Day School Week Increase Juvenile Crime?

Najam and Thompson (2026) examine whether adopting a four-day school week affects juvenile crime. They analyze incident-level crime data from the National Incident-Based Reporting System across six states, combined with longitudinal data on school schedule adoption from 2005–2019. Using a difference-in-differences design, they find that four-day school weeks increase juvenile crime by about 12%, driven by a 20% rise in property crime and a 9% rise in violent crime. They also find a decrease in drug- and alcohol-related offenses during school hours. The effects are largest in non-rural and higher-enrollment areas, with crime increases concentrated off school grounds.

Why This Article Was Selected for The Policy Scientist

The policy relevance of four-day school weeks extends well beyond education finance or scheduling efficiency, implicating broader questions about youth supervision, social structure, and public safety. As more districts adopt compressed schedules—particularly post-pandemic—understanding unintended behavioral consequences becomes increasingly salient. Najam and Thompson (2026), who have contributed extensively to this literature, provide timely evidence linking reduced school exposure to measurable changes in juvenile crime. Their use of detailed NIBRS data across multiple states strengthens external validity, though generalizability to dense urban systems remains uncertain. The difference-in-differences design represents a credible causal inference approach, though future work incorporating randomized or quasi-experimental variation would further strengthen identification.

Full Citation and Link to Article

Najam, R., & Thompson, P. N. (2026). Impacts of the four-day school week on juvenile crime. Journal of Policy Analysis and Management, 45, e70101. https://doi.org/10.1002/pam.70101

Central Research QuestionThis article examines whether the adoption of a four-day school week causally affects juvenile crime. Specifically, it asks how reducing formal school exposure—by eliminating one instructional day per week—alters the frequency, type, timing, and location of criminal activity among adolescents. The authors are particularly attentive to whether these effects vary across geographic contexts, such as rural versus non-rural areas, and across institutional environments, such as small versus large jurisdictions. A central component of the inquiry is disentangling how changes in supervision, routine structure, and daily schedules influence behavioral outcomes. The study also explores whether effects are concentrated on the nonschool weekday or spill over into remaining weekdays and weekends, thereby providing insight into underlying mechanisms.

Previous LiteratureThe study builds on a well-established literature linking education to crime, where schooling is typically understood to reduce criminal behavior through increased opportunity costs and structured time use. Foundational work by Lochner and Moretti (2004) and subsequent studies has demonstrated that increased educational attainment lowers crime rates, while other research has examined how schooling policies—such as compulsory attendance laws or school finance reforms—affect criminal activity. A related strand focuses on contemporaneous exposure to school environments, showing that time spent in school can directly reduce juvenile crime by limiting unsupervised activity.

Within this context, the four-day school week represents a relatively understudied but increasingly relevant policy shift. Prior evidence, particularly Fischer and Argyle (2018), suggests that reduced school days may increase juvenile crime, especially property offenses. However, existing studies are geographically limited and do not fully examine heterogeneity or temporal dynamics. Najam and Thompson extend this literature by providing broader spatial coverage, more granular data, and a systematic examination of mechanisms, thereby contributing to a more comprehensive understanding of how altered school schedules influence youth behavior.

DataThe authors assemble a detailed, multi-source dataset that combines crime, education policy, and demographic information. The primary outcome data come from the National Incident-Based Reporting System (NIBRS), which provides incident-level records of crime, including offender age, crime type, location, and timing. The analysis focuses on juveniles aged 14–17 and aggregates these incidents at the law enforcement agency–year level from 1999 to 2019.

These crime data are matched with a proprietary longitudinal dataset documenting four-day school week adoption across six states: Colorado, Idaho, Montana, Oklahoma, South Dakota, and Oregon. This dataset includes both the timing of adoption and the intensity of exposure, measured as the share of students enrolled in four-day schedules within a given jurisdiction. The authors further integrate school-level demographic data from the National Center for Education Statistics, including enrollment, student-teacher ratios, and socioeconomic composition.

A notable strength of the data lies in its granularity. The NIBRS data allow for disaggregation by crime type (property, violent, drug-related), location (on- versus off-school grounds), and timing (school hours, weekdays, weekends). This level of detail enables the authors to move beyond aggregate effects and identify specific behavioral responses. At the same time, the geographic scope—while broader than prior work—is still concentrated in states where four-day school weeks are prevalent, which has implications for generalizability.

MethodsThe empirical strategy relies primarily on a difference-in-differences framework that exploits variation in the timing and intensity of four-day school week adoption across jurisdictions. By comparing changes in juvenile crime in areas that adopt the policy to those that do not, while controlling for agency and year fixed effects, the authors aim to isolate the causal impact of the policy. The inclusion of time-varying controls—such as demographic and school characteristics—further reduces potential confounding.

To address concerns about staggered treatment adoption and heterogeneous treatment effects, the authors employ modern difference-in-differences estimators, including the Sun and Abraham (2021) approach, as well as alternative specifications proposed by Wooldridge (2021) and Borusyak et al. (2024). Event study analyses are used to assess pre-treatment trends, providing evidence in support of the parallel trends assumption. Additional robustness checks include Poisson models for count data and triple-difference specifications comparing juvenile and adult crime trends.

From a methodological standpoint, the study reflects a strong commitment to causal inference within an observational framework. The use of staggered DiD estimators and extensive robustness checks strengthens identification relative to standard multivariate regression approaches. However, the absence of randomized assignment or a clearly exogenous policy shock means that residual concerns about unobserved confounding cannot be entirely eliminated. Future research leveraging quasi-experimental designs with stronger exogeneity—or, where feasible, randomized interventions—would further enhance causal credibility.

Findings/Size EffectsThe central finding is that the adoption of a four-day school week leads to a statistically and substantively significant increase in juvenile crime. The preferred specification indicates that total juvenile crime rises by approximately 12 percent relative to the mean. This effect is primarily driven by increases in property crime (approximately 20 percent) and, to a lesser extent, violent crime (approximately 9 percent). These effects are concentrated in off-school locations, suggesting that reduced supervision outside of school plays a key role.

In contrast, the authors find a decrease in drug- and alcohol-related offenses during school hours, on the order of roughly 9 percent. This pattern is consistent with a shift in the timing and location of such behaviors rather than a reduction in overall activity. The temporal analysis reveals that crime increases are not confined to the nonschool weekday but extend to other weekdays and weekends, indicating spillover effects. This finding challenges a simple “one-day” mechanism and suggests broader adjustments in routines and supervision.

Heterogeneity analyses show that effects are larger in non-rural and higher-enrollment areas, where increases in total juvenile crime exceed 10 incidents per 1,000 students in some specifications. In rural areas, effects are smaller but still present, particularly for property crime. Similarly, large jurisdictions experience substantial increases in property crime, while smaller jurisdictions show more limited effects, primarily in violent crime.

Overall, the magnitude of the estimated effects is consistent with prior research but extends it by demonstrating that these impacts are robust across multiple contexts and specifications. The results also highlight that the policy’s behavioral consequences are multifaceted, affecting not only the level of crime but also its composition, timing, and spatial distribution.

ConclusionThis study provides evidence that reducing school exposure through four-day school weeks is associated with measurable increases in juvenile crime, particularly property offenses occurring off school grounds. The findings contribute to a broader understanding of how institutional structures—such as school schedules—shape youth behavior through mechanisms related to supervision and time use. By integrating detailed crime data with policy variation across multiple states, the authors offer one of the most comprehensive analyses of this issue to date.

The study’s contribution lies not only in its empirical findings but also in its methodological approach, which applies contemporary causal inference techniques to a policy-relevant question. While the difference-in-differences design is well-executed and supported by robustness checks, future research could strengthen causal claims by leveraging more exogenous sources of variation. In terms of external validity, the results are most directly applicable to regions with similar demographic and institutional characteristics, though the underlying mechanisms likely extend more broadly.