In Latin America, children now receive better education than their parents. Absolute educational mobility has progressed in a significant portion of the region, but relative mobility remains weak and unequal. According to the World Bank, which relies on the GDIM database covering 148 economies, 75% of adolescents in the region fail to reach level 2 in mathematics on PISA 2022, the minimum competency threshold for this discipline. The region has experienced increased school enrollment, but competencies remain weak, diploma returns are heterogeneous, and mismatches between education and employment are substantial.
The Essentials
- Intergenerational educational mobility is progressing in Latin America: children of poorly educated parents are gaining greater access to secondary and higher education than in the past (OECD, 50 national surveys, December 2025).
- Yet 75% of Latin American adolescents remain below level 2 in PISA 2022 mathematics, the floor of competencies deemed necessary to participate in a productive economy.
- A system can distribute more diplomas and enroll more students while stagnating on actual learning outcomes: the rise in enrollment rates masks stagnation in learning.
- This disconnect directly questions the value of the diploma as a signal in the labor market, and the region’s capacity to benefit from the digital transition that is reshaping global employment.
- Targeted experiments—structured tutoring, competency-based curricula, robust national assessments—show that reversing the trend is possible, but it requires political choices that few governments have yet made.
Measurable Progress and Missing Progress
For half a century, Latin America has accomplished something real. Millions of children born in households without diplomas have reached secondary school, then university. Enrollment rates have jumped. Countries like Brazil, Colombia, and Mexico have multiplied campuses, scholarships, and access programs for rural and indigenous populations. Measured in years of education or diplomas obtained, intergenerational mobility has progressed in documented fashion, and this is good news not to be erased.
The problem lies in what is being measured. Diane Coyle, in The Measure of Progress, poses a thesis that applies word for word to this case: counting diplomas produced by an educational system creates the same illusion as counting GDP to measure well-being. The indicator aggregates, smooths, masks. It says something is increasing without saying whether that something is worth what we think. When disaggregated by measured competencies, by ability to solve a problem, by aptitude to read a contract or understand a bank statement, the picture changes radically.
PISA 2022 provides this disaggregation. Across all Latin American and Caribbean countries participating in the assessment, three out of four adolescents fail to reach level 2 in mathematics. At level 2, students can extract relevant information from tables, graphs, or other slightly more complex representations and use it in simple problem-solving strategies. Level 2 corresponds notably to the ability to mobilize simple strategies, understand basic functional relationships, and solve problems involving simple ratios. Being below level 2 signals gaps in fundamental mathematics likely to complicate certain training courses and tasks, without allowing one to infer a general inability to perform a given job.
Educational mobility has progressed.
Actual competency, meanwhile, stagnates.
The Escalator That Goes Up but Doesn’t Move Forward
To understand how these two trends coexist, one must see how an educational system can expand without strengthening.
The pressure on Latin American governments since the 1990s has focused on access: building schools, recruiting teachers, keeping children in class. Conditional cash transfer programs, Bolsa Família in Brazil, Oportunidades in Mexico, played a decisive role in reducing early dropout. These policies worked on their primary objective. Attendance rates increased, access inequalities between socioeconomic groups narrowed, children of parents without diplomas gained access to education levels their parents had never reached.
But access is not learning. A student who remains in class until age sixteen without mastering basic competencies obtains a diploma that signals their presence in the system, not their capacity to leave it armed with knowledge. This is what education economists call the “credential effect” disconnected from the “skill effect”: the diploma as a social entry ticket, not as proof of competency.
This disconnect has multiple causes. Teacher training remains insufficient in most countries in the region: according to the OECD, a large proportion of primary and secondary teachers have never been exposed to structured teaching methods for mathematics. Curricula are often overloaded and poorly sequenced, accumulating learning objectives without ensuring that foundations are solidly laid before advancing. National assessments, when they exist, rarely serve to guide classroom teaching practices. And resources—textbooks, digital tools, effective teaching time—remain very unevenly distributed between urban and rural schools, public and private institutions.
The Signals the Market Detects Beyond the Diploma
In the labor market, the disconnect becomes visible in wage data. Measured competencies and diplomas are each associated with wages in statistical analyses, but these associations are insufficient to establish two completely independent causal returns; digital skills can sometimes compensate for a lower level of formal qualification, but the hierarchy of their premiums varies by country and type of skill. A high school graduate who can solve a complex mathematical problem, read a spreadsheet, or write a structured report earns more than another high school graduate who cannot do these things. The diploma does not always allow one to distinguish between them.
Intergenerational mobility thus risks becoming hollow. If the son of an agricultural worker obtains a high school diploma without having acquired the competencies this diploma is supposed to certify, he formally accesses a higher socioprofessional category, but without the actual productivity that should accompany it. In the short term, he does better than his parents. In the medium term, when the market detects the gap between the signal and actual competency, the premium erodes.
This phenomenon is not unique to Latin America. It is observed in Southeast Asia, where entire cohorts of graduates find themselves in overqualified-for jobs lacking competencies recognized by employers, a dynamic this journal has already documented for the region. But in Latin America, the high level of underperformance of 15-year-old students signals a major risk of future gaps in fundamental competencies, without measuring the entire workforce.
Daron Acemoglu poses the question of who captures the gains from technological progress; this question takes on a precise form here. In an economy that automates its repetitive tasks and values analytical skills, the digital transition benefits those with the skills to accompany it. Companies investing in robotics and artificial intelligence need workers capable of supervising, adapting, and understanding. Weak fundamental and digital competencies can limit workers’ and companies’ capacity to benefit from the digital transition.
Countries Attempting to Loosen the Knot
Signaling the problem without documenting ongoing responses would be incomplete. Several countries in the region have launched serious experiments with results worth monitoring.
Chile has invested in establishing regular national learning assessments since the 2000s (SIMCE system), with results published by school. The objective: create pressure through transparency, force schools to confront their actual results rather than their enrollment rates. Effects have been limited as long as assessments were not accompanied by concrete pedagogical support, but the logic—measuring what matters rather than what is easy to measure—is correct.
Brazil, with the IDEB (Basic Education Development Index), aggregates learning rates and school progression rates in a composite indicator. Municipalities that improve their IDEB access additional federal funding. The incentive has produced effects in certain states, notably Ceará, often cited as a textbook case of strong recovery in learning outcomes in high poverty areas.
Structured tutoring—sessions of individual or very small group mentoring focused on unmastered basic competencies—is being piloted in Mexico, Colombia, and Ecuador, sometimes in conjunction with international NGOs. Published evaluation data on these programs are promising: two to four additional months of effective mathematics learning compared with unsupported students. These results do not close the overall gap, but they show the trajectory can change.
These initiatives share a common trait: they take learning as a measurable objective, not access as an end in itself. They fit into what education researchers call the transition from “input” policy (building schools, hiring teachers) to “output” policy (verifying what students actually know how to do).
A Weak but Real Signal: What Mobility Data Say Anyway
A more granular reading of OECD data is necessary before closing the matter. Intergenerational mobility is progressing in Latin America, and this finding deserves to be taken seriously.
Children of poorly educated parents who access secondary and higher education do not necessarily return to square one even if their diploma is undervalued. They acquire networks, lifestyles, a relationship to formal work that breaks with the informal sector in which their parents operated. They have greater access to social protections, health coverage, formal financial markets. Educational mobility produces real effects, even when measurable competencies stagnate.
This tension is precisely what Diane Coyle points to as the risk of any composite measure: by focusing on a single indicator, however more rigorous like PISA, one risks missing what the other captures. Both readings are true simultaneously. Children in Latin America are doing better than their parents on several real dimensions. And the competencies they acquire in the system are insufficient to navigate the economy awaiting them.
The contradiction is not abstract. It plays out concretely in the labor markets of the 2020s: a Brazilian nurse, a Colombian technician, a Mexican logistics employee have accessed formal jobs their parents never had, but they enter these jobs with weak analytical skills that limit their ability to progress, to retrain, to resist the automation of the tasks they perform. Mobility has opened a door. It has not provided the tools to go far beyond it.
The Challenge of the 2030s: Transforming School Enrollment into Real Productive Capital
The 2030-2035 horizon places the question under additional tension. Latin American economies will continue to transform under the effect of digital technology and automation. Certain repetitive administrative and productive tasks are exposed to automation or technological transformation, but the scope and direction of employment effects vary by sector, country, and technology. Jobs being created demand greater analytical competencies, ability to work with digital tools, to interpret data, to solve non-routine problems.
The fact that 75% of 15-year-old students in participating countries are below level 2 in mathematics reveals a major structural risk for the future competencies of the workforce, without corresponding directly to all new labor market entrants. Companies investing in robotics and digitization may need to train their workforce, recruit competencies, or limit their transformation to certain segments. None of these options is satisfactory for a region seeking to move up economically.
The response to this challenge depends on specific and barely visible political choices: reforming initial and continuing teacher training, aligning curricula with basic competencies before ambitious objectives, deploying learning assessments with real consequences for schools, funding structured tutoring at scale. These policies exist as experiments in several countries. The question is whether they scale up, and with what budgetary priority.
The expansion of access to education and public investments in it have contributed, but the progression of educational mobility results from several factors and cannot be attributed univocally to a single policy decision. The next step—making actual learning a priority on equal footing—assumes the same type of political decision, with different tools, stronger institutional resistance, and less visible effects in the short term. No government makes speeches about progress in PISA points with the same ease as about inaugurating a campus. But it is on this silent ground that the real quality of promised mobility is decided.
Sources
- OECD, Social Mobility and Inequality in Latin America and the Caribbean, December 2025
- Diane Coyle, The Measure of Progress: Counting What Really Matters, Princeton University Press
- OECD, PISA 2022 Results, Latin America and Caribbean results (available at oecd.org/pisa)
- World Bank, IDEB Program Brazil, learning assessments (data available at worldbank.org)
- Journal d’un Progressiste, Le tutorat contre les inégalités scolaires
- Journal d’un Progressiste, Les diplômés d’Asie du Sud-Est condamnés à l’émigration
