by Isabel de Brigard
There is unusual consensus amongst politicians, policymakers, parents, etc., about how important schooling and education are. However, in many ways, consensus stops there. Schools, for example, are deemed to be either the best vehicle for social mobility, the one true driver of social change which can lessen social inequalities on all fronts, or – if you side with cultural reproduction theorists – perpetuators of social stratifications and amplifiers of the effects of socioeconomic background.
School children in Florida (Valle), in Colombia
This sort of double bind of the social effects of education and schools makes it critical to know, when examining concrete education policy choices, how they might trace back to existing and future inequalities. It is key to understand how to turn schools into places where everyone learns in equal measure and is equally able to take advantage of what is taught.
However, the platitude that given the benefits of schooling, societies should strive to get as many kids to school as they can, is faced in low-income countries with limited institutional capacities due to the equally obvious limits in resources available. Thus, two-shift schooling – by which the same classrooms and the same teachers accommodate two or more shifts of students per day – has been introduced in many countries to mitigate heavy budgetary constraints. In Colombia, it was implemented for the first time in the 1960s, and to this day, continues to be the only accessible option for over 70% of children attending school. Since then, great improvements have been made in the education sector on measures such as average education level or coverage of the school system. But the effects of shift-schooling on the inequities the school system produces are also substantial. And mostly underlooked, in a debate that has more often been politicised than informed.
This study assesses the effects of two-shift schooling on standardized test performance, exploring if they are the same for all children or what differences might tell us about the role schools play as vehicles for social equity. Besides updating the few available results of the impact of full-day schooling on test performance, this research contributes in at least two substantive ways. First, because data disaggregated at the student level has only recently become available, previous results only went as far as the school or district level. While these are valuable, they do not allow for explorations into the importance of individual-level factors for test performance, such as family background, socioeconomic status, or gender. Second, existing research does not include any measures of school quality that might contribute to understanding the impact of this additional time in school. Rather, studies mostly interpret the average performance of students as a measure of school quality, which can be considered a flawed approach since good schools are arguably not the ones with the best students but rather those doing the most for the ones they have.
The model
Panel data was used from the mathematics and language Pruebas Saber, a standardised test designed to test the performance of students and schools throughout the basic and secondary education cycles, as well as data compiled by the Colombian Institute for the Evaluation of Education (ICFES) about the quality of schools in Colombia and gathered in the ‘synthetic index of education quality’ (ISCE). The effects on the mathematics and language scores were studied, and rather than report pooled results in the analysis, discrepancies that appear between subjects were included. Because simple regression models were very likely to suffer from omitted variable bias and to take advantage of the within-individual comparisons that panel data allows, a fixed-effects model was used. To study effect heterogeneity and as a set of robustness tests intended to validate my results, various fixed-effects models were run on subgroups of the sample according to time-constant traits, such as the location of the school (rural vs. urban), the quality of the school, or the socioeconomic level of the student.
Isabel de Brigard presenting her work in paeradigms Dialogue series
The results
So, what were the findings? First, and perhaps somewhat unsurprisingly, across all models, more instruction time led to better test scores in mathematics and language. Although somewhat anticlimactic, it is nonetheless reassuring to see that the school system is making some difference in the ability of students to answer academically designed questions in an academic setting. However, there are substantial differences in the size of these effects between subjects, with effects in math always larger than in language. The explanation for these differences is unclear. It might come from the fact that time in school impacts math scores more than language because schools tend to play a larger role in teaching mathematics, while language skills can be acquired at home. It might also be a consequence of how students are tested for each subject and how aligned testing and teaching are in each case.
Regarding the impact of school quality on the academic gains made in full-length school days, the results are very puzzling. Contrary to what could be expected, the relationship between a school's quality and its ability to transform instruction time into good grades is not linear, because additional time in school seems to be doing the most difference at both ends of the quality spectrum and dipping for medium quality schools. These results point to the somewhat counterintuitive notion that in schools of poor quality, every extra moment in class helps.
Finally, results show that an increase in the time children spend in school has bigger effects for rural and public schools, compared to urban and private schools. It also makes a bigger difference in schools with a lower average socioeconomic status and for individualstudents of less advantaged backgrounds.The key finding here is that more time in school makes more of a difference wherever other resourcesare more limited.If the chances a child has ofdevoting time to reading or homework are limited to the time they spend in school, and iftheir living conditions make it less likely for them to engage in after school activities thatmight contribute to their learning, then how much they get out ofschooling cruciallydepends on how long they get to be there every day.
Open discussions at Paeradigms Dialogue monthly series event
The gist of it
As things stand, two-shift schooling is widening the gap between children in different contexts and ensuring that the education system perpetuates existing inequalities instead of contributing to levelling the playing field. But these results also indicate that schools are complex but valuable social institutions. If nothing else, they are an essential part of the development and flourishing of kids who are otherwise dealt a rough hand. Thus, offering every child the opportunity to have a full day in school every day is a crucial first step towards turning the education system into a tool for closing the enormous social gaps that are still present in Colombia.
Watch the full presentation: https://www.youtube.com/watch?v=QeMgqdpfNbY&ab_channel=PaeradigmsNGO
Isabel de Brigard, Education Policy Expert, Colombia
A Colombian born education policy specialist based in Berlin. She holds a master’s degree in public policy from the Hertie School of Governance in Berlin. Her master’s thesis “Is more time in school worth it? Effects of two-shift schooling on test performance from a longitudinal study in Colombia” is an example of such work and can be found on GitHub: @isabeldebrigard.