Everything you need to know about PISA 2025 on equity

What this report does

This report assesses the state of equity, or equality of opportunity, in education in Spain and its autonomous communities, based on an original and detailed analysis of the PISA 2025 microdata. Describing and understanding educational equity in depth is highly relevant for a country’s social and economic future, for young people’s real opportunities, and for present and future human capital and social cohesion. This is especially true in Spain, where the early school leaving rate has historically been high and remains persistent.

PISA 2025 is marked by a widespread decline in results, particularly severe in Spain and especially in reading: 451 points, 23 fewer than in 2022 —more than a school year’s worth of learning—, compared with the 14 points lost by the OECD-35 average. Following this decline, Spain falls below the OECD in all three competencies assessed.

Against this backdrop, equity in Spain broadly remains where it was. By socioeconomic status, measured with indicators that are comparable across editions, there are no major changes: the decline affects all students in a similar way. In grade repetition, inequality remains very high and has even increased slightly. As for the disadvantage of students with a migrant background, it is still smaller than in the OECD and is explained above all by their socioeconomic status, although it has grown over the last three years. Finally, new inequalities are emerging in the use of new technologies, though for now they are less pronounced.

Main findings

By socioeconomic status, we find that:

• All social groups fall by a similar amount. Regardless of socioeconomic quartile, the learning loss in reading is between 19 and 23 points; the gap between the most and the least advantaged barely narrows; and the weight of social background in scores does not change (from 8.7% to 8.5%).
• Particularly striking in the Spanish case is the decline among disadvantaged students. The lowest socioeconomic quartile loses 19.3 points in reading, twice the OECD-35 average (9.4).
• In fifteen of the sixteen autonomous communities analysed, all four quartiles lose ground, with the exception of Castilla-La Mancha. In five —the Valencian Community, Navarre, Castile and León, Cantabria and La Rioja— disadvantaged students fall further than advantaged ones, and comparing top quartiles, the Basque Country has the lowest performance in the country.

As for the inequalities affecting students with a migrant background, the results show that:

• Students with a migrant background score 32.9 points lower in reading, a smaller gap than in the OECD-35 (42.1) and the EU (49.3). Once socioeconomic status is accounted for, the gap is much smaller: 13.0 points. This places Spain as a country where the migrant disadvantage is above all a socioeconomic disadvantage.
• This net gap has increased slightly in three years (from 9.6 to 13.0 points), although the difference is not statistically significant, and it is widening in half of the autonomous communities, sharply so in Andalusia (from −5.1 to 18.8 points) and La Rioja (from 11.2 to 32.9). The decline is concentrated among first-generation migrant students (26.2 points, compared with 19.8 for the second generation).
• Changes in the composition of the student population, in terms of both migrant background and socioeconomic status, do not explain the decline in results in Spain: if 2025 students had the same profile as those in 2022, the drop in reading would have been the same.

In grade repetition, inequalities are much greater in the Spanish case:

• The downward trend in grade repetition has come to a halt: 22% of 15-year-old students have repeated a grade, the same as in 2022, now that the exceptional repetition measures applied in 2020 and 2021 have been lifted. Even so, Spain remains the OECD country with the third-highest repetition rate, more than double the OECD-35 (9.8%).
• Almost all of the excess is concentrated in the lower half of the social ladder: around 7% of the top quartile repeat, compared with 5% of their OECD counterparts, while 40% of the bottom socioeconomic quartile repeat, compared with 15% internationally.
• With equal competencies, a student from the bottom quartile is 5.2 times more likely to have repeated a grade than one from the top quartile (4.9 in 2022; 2.4 in the OECD-35). This indicator remains the second highest in the OECD-35, behind only Slovakia.

Finally, regarding digital use and artificial intelligence, we observe that:

• Spanish students use digital devices slightly less than the OECD average, both for learning and for leisure. Moreover, this use has declined slightly in recent years.
• However, they are among those who rely most on AI for reading and writing: 46% use chatbots at least weekly to summarise texts or write assignments, nine points above the OECD-35. The higher the socioeconomic status, the greater the use of AI for tasks that complement students’ own work —using AI to learn or do research— but not for those that can replace it: summarising texts barely changes, from 38% to 42%.
• The classroom evens out AI use; free time does not. In class, the gap between quartiles in daily AI use is seven percentage points (from 29% for the most disadvantaged quartile to 36% for the most advantaged); outside school, where use is considerably higher, it widens to more than twelve (from 38% to 51%, respectively).

Proposals to keep improving equality of opportunity and educational equity

1. Strengthen, from the early stages and more decisively than to date, individualised or small-group support and catch-up tutoring, within or outside school hours. In particular, set goals and provide intensive, personalised support in literacy and mathematics for all highly vulnerable, low-performing students.
2. Encourage the participation of vulnerable students in safe non-formal education spaces that offer them academic support as well as alternative leisure opportunities.
3. Improve access to quality early childhood education for ages 0 to 3 for children from low socioeconomic and migrant backgrounds, by increasing investment and improving the working conditions of its staff.
4. Promote teacher training aimed at improving effectiveness and practical skills in classroom management in heterogeneous classes, as well as training to reflect on and change beliefs and teaching practices related to assessment and grade repetition.
5. Commit to longer school days, guaranteeing free school materials, meals and transport for children living in poverty, as well as their access to sufficient grants that help them stay in school and continue into post-compulsory education.
6. Provide additional resources for non-teaching staff in high-complexity schools (support for diversity, guidance counselling, social workers, reception classrooms) and consider targeted reductions in student–teacher ratios where the concentration of vulnerable students or students with specific needs justifies it.
7. Support the development of school projects in highly vulnerable contexts that are committed to the academic success of all their students, by fostering teaching staff stability, support services and additional teaching resources, data collection, and support for school leadership teams.
8. Guarantee digital training for students and teachers focused on developing competencies for a safe, critical, responsible, healthy and learning-oriented use of artificial intelligence and the digital environment; effectively implement restrictions on mobile phones in the classroom to uphold the right to disconnect and protect students’ attention; and, more generally, carry out ex ante evaluations of the digital tools used by schools.