Do two weeks of instruction time matter?

The question of how the school year length affects academic outcomes is important for many stakeholders, including policymakers and school principals. Instruction time is the primary input for the formation of human capital and has been a central element in several education reforms. OECD emphasizes the essential role of instruction time in improving the quality and quantity of education outcomes.

The term length for primary and secondary school students differs markedly across countries. The 2021 OECD indicators show that children go to school, on average, 184 days per year. School lasts 175 days in Spain, 162 in France and 190 in the UK.

An unexpected reform affecting the term length of ninth-grade students in the Madrid region (the Comunidad de Madrid, as it is known locally) during the 2017/2018 school year is an interesting case do delve into. This reform changed the resit exams to an earlier date in June, reducing the total instruction time for students who had already passed these exams by around two weeks. Students with passing grades did not receive formal teaching but were involved in cultural activities. This reform also intended to allow families to have more accessible summer holiday breaks by eliminating resit exams in September.

The same measure was adopted in other Spanish regions such as the Basque Country, Navarra, La Rioja, and Cantabria and was subsequently extended to the entire country. But reducing two weeks of instruction negatively affected student performance, especially in Spanish and Math.  The results of PISA 2018 also blamed this reform for the unexpected negative scores obtained by these communities. The report explicitly noted that this negative “impact is larger on the results of the five subnational entities with early high-stakes exams”. Most notably, the top-performing students in public schools felt the brunt of this decision, causing the gap between them and their lagging peers to shrink.

Two particularly relevant characteristics of this reform led to interesting results. First, it affected public and charter schools (treatment group) but not private schools (control group), enabling a counterfactual analysis. A second important feature is that the region of Madrid conducted standardized external exams that measured the performance of tenth-grade students before and after the reform. Based on the previous arguments, we employed a difference-in-differences approach, comparing results between the treatment group and the control group over time, to estimate the differential effect of this reform. The difference-in-differences technique, also known as DID, is a statistical method used to estimate causal effects by gradually comparing changes in outcomes between a treatment group and a control group. This technique helps to isolate the impact of a specific intervention by accounting for trends that would have occurred even without the intervention. We rely on the parallel trend assumption which states that, in the absence of treatment, the difference in students' performance in participating and non-participating schools would be constant.

The difference between the treatment and control group is a two-week period: while both groups used to sit their final exam on the third week of June, the control group (private and charter schools) had that final exam on the first week of June instead. Both groups then sat the same external and standardized test.

This is a visual representation of the timeline of the experiment and what both groups experienced. The main difference is the two-week period between one and the other.

The relevance of this analysis is twofold. First, it evaluates a proposed solution to a persistent problem in the Spanish education system, which is excessive grade retention. Grade retention is the process of a student having to repeat a grade after having failed the previous year, and it affects more than 30% of all compulsory education students in Spain. Second, and more importantly, this analysis has wider implications.

Despite some limitations and concerns — the reform was unexpected and enforced in all free schools at the end of the year; the response variable was a performance evolution happening the next March; we used a student-level sample with detailed information on students' socioeconomic characteristics — we have clear evidence that even a small change in instruction time (just two weeks) can have substantial implications for students' academic outcomes: around 0.13 times of a standard deviation. The effect was especially significant in Spanish and Mathematics, and high-performing students seemed to be most affected by the change. Comparatively, the impact of this reduction in learning time is similar to the effect of increasing a classroom size by about five students.

In conclusion, and especially in the wake of the COVID-19 pandemic, we believe further research is needed to better understand the long-term implications and the role of online versus face-to-face instruction.

This article is based on a paper written in 2021 by Ismael Sanz and J.D. Tena and published in the journal Kyklos two years later entitled “Do 2 weeks of instruction time matter? Using a natural experiment to estimate the effect of a calendar change on students' performance”. The article is available in full here.