The Schooling Rule That Followed People Into Adulthood
Mexico’s 1993 extension of compulsory schooling is the subject of this series. The question is not simply whether more education eventually raises wages, but whether a legal change that kept children in school longer altered the shape of adult life and the prospects of the next generation. The evidence offers a way to examine that question with unusual clarity: follow the cohorts touched by the reform from educational attainment into family life, work, mobility, and their children’s enrolment.
The practical judgment this supports is equally specific. A schooling mandate should not be judged only by its average effect or by a wage premium attached to a credential. It should be judged by who it reaches, which threshold they cross, and what new options become available afterward. In Mexico, the people most affected were not evenly distributed across the country. The largest changes appeared among Indigenous and rural populations, where lower-secondary education had been least attainable. That uneven reach is not a side note. It is the central clue to understanding what the reform did.
A law that reached beyond the classroom
In 1993, Mexico changed its General Education Law and extended compulsory schooling from six to nine years. Lower secondary education—grades 7 through 9—became mandatory nationwide. The rule applied to children younger than 12 at the beginning of the 1993 school year, creating an eligibility boundary around the 1981 birth cohort.
The legal change did not guarantee that every eligible child completed three more years. Public lower-secondary education was already free, the reform did not impose sanctions for non-compliance, and schools did not expand at the same speed everywhere. Its importance lay in changing the institutional expectation and, in places where supply could respond, the practical opportunity to remain in school. Mexico also expanded telesecundaria, a distance-learning model designed to reach rural and remote communities without conventional secondary schools. The mandate was therefore not a solitary command issued above society; it operated through a particular arrangement of law, school access, infrastructure, and local constraints.
That distinction between a legal entitlement and a completed year of schooling is essential. A government can change the age or grade at which attendance becomes compulsory without removing the obstacles that make attendance difficult. A secondary school may still be distant, a household may still depend on a young person’s labour, and instruction may still vary in quality. The reform’s effect is therefore best understood as a change in exposure and opportunity. It created a pressure toward continued schooling, and the study asks what happened to people whose education actually rose because of that pressure.
That arrangement makes the reform a useful subject for studying life-course change. People do not experience schooling as an isolated number. An additional year can change when education ends, which jobs are reachable, the opportunity cost of having another child, or the resources and expectations brought into a household. None of those possibilities is automatic, and the reform study does not claim that every pathway was separately identified. But a policy that changes attainment at a formative age can leave traces well beyond the classroom.
The study follows those traces using Mexico’s 2020 Population and Housing Census. Its main sample contains more than a million people born close to the reform cutoff, observed in adulthood. That timing matters: the affected cohorts had largely completed schooling, established labour-market trajectories, and reached ages at which much of fertility and some of their children’s education could be observed. The result is not a snapshot of students immediately after a policy announcement. It is an attempt to see whether a schooling rule remained visible decades later.
The timing also changes what counts as a meaningful result. An immediate rise in enrolment would show that schools absorbed more students. An adult outcome shows that the change survived long enough to affect a person’s accumulated choices. A child’s enrolment shows that the consequence may continue through a household. These are different claims, and they should not be collapsed into one measure of “educational success.” The reform is valuable as a case because it allows the reader to ask how one institutional change appears at several points in a life, while still keeping each outcome’s evidence separate.
Why wages are too narrow a frame
Wages are an important way to think about education, but they are not the only outcome that can carry its value. The reform study could not provide a clean causal wage estimate because census earnings data had high non-response. That limitation is useful rather than fatal: it forces a broader question about what a life-course effect looks like when income is not treated as the sole scoreboard.
The relevant outcomes include family size and child survival, whether adults work and which sectors they enter, whether movement is permanent or recent, and whether their children enrol in secondary school. These outcomes are connected by a common fact: schooling can change the set of choices people face. A person who stays in school longer may delay or reconsider family formation, qualify for work outside agriculture, respond differently to opportunities in another municipality, or bring different knowledge and expectations into decisions about a child’s education. The evidence can establish some of these changes as effects of reform-induced schooling. It cannot reduce them to one universal mechanism.
This is why the reform is better understood as a change in life-course capacity than as a simple investment with one financial return. The phrase does not mean that schooling liberates people from structural limits, nor that every choice becomes easier. It means that the policy may alter several margins at once: the timing of family decisions, the survival of children, the allocation of labour, the kind of mobility available, and the educational starting point passed to the next generation.
It also keeps the interpretation honest. If education raises employment for one group but mainly changes sector for another, those are not interchangeable signs of success. If it increases recent movement while reducing the chance of living outside one’s birth state, “more mobility” is too crude a description. If children become more likely to enrol in secondary school but not yet in higher education, the result is a threshold effect, not proof of complete intergenerational mobility. Looking beyond wages adds outcomes; it does not remove the need to define them carefully.
The diagram is a map of the evidence, not a claim that every arrow is a separately measured mechanism. The study estimates a common reform-induced schooling variation and observes its relationship with several later outcomes. It does not, for example, prove that a specific change in cognitive skill caused a specific fertility response. The value of the map is that it keeps the outcomes in one frame while preserving the distinction between what is estimated and what is inferred.
The same rule did not reach everyone equally
At the national level, the reform’s attainment effect was modest: cohorts just after the eligibility cutoff completed about 0.17 additional years of schooling on average. They were also 1.79 percentage points more likely to complete at least nine years of education. Those figures are meaningful, but they do not resemble the legal extension’s nominal three years. They describe the average change in completed schooling generated by an imperfect policy operating across very different places.
The subgroup results change the interpretation. Among people who spoke an Indigenous language, the reform increased completed schooling by about half a year, compared with the much smaller national shift. Literacy rose by roughly 4.5 percentage points, and lower-secondary completion by about 3.8 points from a baseline of only 29 percent. Effects were also larger in smaller rural localities. In settings where secondary education was already common, the mandate had less margin to operate. Where it had been blocked by distance, language, infrastructure, or the opportunity cost of leaving work at home, a legal requirement could move more people across a consequential threshold.
That pattern gives the reform its wider significance. An average can make a policy look moderate precisely because it combines places where the policy had little room to change behaviour with places where it changed the educational path of people previously left behind. The evidence does not show that the reform erased inequality, or that all Indigenous or rural people experienced the same effect. It shows that uptake and gains were concentrated among populations with lower initial attainment and weaker access.
The observable measure here also matters. The study identifies Indigenous populations through language status in the census analysis. That is useful for locating unequal reach, but it is not a complete description of Indigenous identity, community experience, or every barrier to education. Likewise, “rural” is a classification of locality size, not a single social condition. These distinctions prevent a compelling pattern from becoming an overly simple story about groups. They also point toward the practical question of whether a mandate can work without the supply and support needed in the places where the margin is greatest.
This series therefore follows two questions together. What did an additional year of reform-induced schooling change in adult lives? And why did those changes appear most strongly among people who began with the fewest educational options? Answering the first without the second would turn a distributional result into a generic story about education. Answering the second without tracing later outcomes would reduce the reform to an enrolment statistic.
The next step is to examine how the study can distinguish the reform’s effect from the many other differences between people born before and after 1981. That distinction matters because the life-course outcomes are not self-interpreting. A credible account must first show how eligibility became a measurable schooling difference, who complied, and what the resulting estimates actually represent.