Within- and between-school socioeconomic inequality in concurrent low performance. Evidence from PISA 2025 Colombia

Every year, millions of students around the world take a standardized test called PISA, which measures how well they can apply their knowledge in reading, mathematics, and science. For decades, researchers have known that a student's family background—specifically their access to money, books, and cultural resources—strongly predicts how well they will do on these tests. This relationship is so consistent that it serves as a global benchmark for educational equity. However, looking at a single test score or a single subject only tells part of the story. In many countries, including Colombia, a significant number of students struggle not just in one area, but in all three simultaneously. This "concurrent low performance" represents a deep accumulation of learning difficulties, where a student falls below a basic proficiency threshold in math, reading, and science all at once. Understanding this phenomenon requires more than just counting how many students are struggling; it demands looking at where these students are sitting. Are they struggling because of their own family circumstances, or because they are surrounded by peers from similar backgrounds in schools that lack resources? The question is whether the inequality happens mostly between different schools, or if it exists even among students sitting in the same classroom.

A recent study using data from the 2025 PISA cycle in Colombia set out to map this landscape with unprecedented precision. The researchers analyzed responses from nearly 7,000 students across 266 schools to understand how socioeconomic disadvantage shapes the experience of struggling in all three core subjects at the same time. They found that nearly 44 percent of the students represented in the sample were performing below the basic level in math, reading, and science simultaneously. This is a massive group, larger than the number of students struggling in just one or two subjects. The study's primary goal was to untangle the social structure of this struggle: does it happen mostly because poor students are clustered in poor schools, or does it happen because poor students struggle even when they are in schools with wealthier classmates?

The analysis revealed that both forces are at work, but they operate with very different strengths. Within a single school, there is a clear gradient: students from wealthier backgrounds are less likely to be struggling in all three subjects than their poorer classmates. However, the difference between schools is far more dramatic. The odds of a student struggling in all three subjects drop sharply as the average wealth of the school's student body increases. In fact, the gap between a school with a low average socioeconomic status and one with a high average status is much larger than the gap between a rich and a poor student within the same building. This suggests that the social composition of a school acts as a powerful context that shapes outcomes for everyone inside it, regardless of their individual family background.

The researchers also investigated whether the way students struggle in all three subjects is fundamentally different from how they struggle in just one. Some might assume that failing in math, reading, and science at the same time creates a unique type of disadvantage with its own social rules. The data, however, showed no such distinction. The pattern of inequality for students failing in all three subjects was statistically identical to the pattern seen when looking at math, reading, or science alone. The "concurrent" nature of the struggle is simply a matter of accumulation; it identifies a large group of students who are falling behind across the board, but the social forces driving that failure are the same ones that drive failure in any single subject.

Another layer of the study examined whether the relationship between a student's background and their performance changes from school to school. In some educational systems, the gap between rich and poor students is wide in some schools and narrow in others. The Colombian data supported the idea that this gap does vary across schools, but the researchers could not pinpoint exactly why. They tested whether the overall wealth of a school could explain why some schools had steeper gaps than others, but the answer was no. The variation in these gaps remained largely unexplained by the basic characteristics of the schools or the students' grade levels. This suggests that while we know the gap exists and varies, the specific institutional factors that widen or narrow it in different settings remain a mystery.

The study also addressed a common concern in educational research: how much of the data is missing or unreliable. In this dataset, a significant number of students did not provide information about their family's economic status. The researchers tested whether ignoring these missing cases or making different assumptions about them would change the results. They found that the main conclusions held firm. The strong link between school wealth and student struggle remained clear, and the pattern of inequality within schools persisted, even when the analysis was adjusted for these missing data points. This robustness gives confidence that the findings reflect a real structural feature of the Colombian education system rather than an artifact of how the data was collected.

Ultimately, the study paints a picture of an education system where the location of a student's school is a decisive factor in their academic fate. While a student's own family background matters, the collective background of the school they attend matters even more. The findings challenge the idea that we can understand educational inequality simply by looking at individual students. Instead, they show that inequality is deeply embedded in the social architecture of schools themselves. For policymakers, this means that improving outcomes for the nearly half of students struggling in all three subjects will require more than just helping individual students catch up; it will likely require addressing the profound differences in resources and environments between schools. The data does not offer a simple fix, but it provides a clear map of where the problem lies: the divide is not just between rich and poor families, but between the schools those families can access.