Demographics and AI: the two forces that will redraw the map of educational institutions

There are two ways to look at the boom in education sector transactions. The first: capital has discovered a counter-cyclical sector with recurring revenue. The second, less comfortable one: the sector is consolidating because two structural forces are redrawing the map, and whoever understands them first will decide from a better position.

Neither force is a hypothesis. Both are already here. They just run on different calendars.

The slow, certain force: demographics

Births in Spain are at historic lows: 318,005 births in 2024 and fertility at 1.10 children per woman, according to the National Statistics Institute (INE) — nearly half the replacement level. And it is not a Spanish anomaly: Italy and Portugal follow similar curves, and Latin America — for decades the demographic relay of Spanish-language education — is falling faster than Europe did. Chile is already among the lowest fertility rates in the world; Colombia has strung together double-digit annual declines in births; Brazil has been below replacement for years.

What makes demographics unique as a market force is its certainty. The students who will enrol in university in 2044 have already been born — or already not been born. No commercial strategy changes that number. Only how it is distributed.

And the contraction moves through the system with the precision of a clock:

• Already visible: nurseries and primary schools closing classrooms year after year. In Spain, birth cohorts a third smaller than fifteen years ago.
• In progress: secondary education absorbs the contraction through this decade.
• From ~2030: higher education leaves its years of large cohorts behind. Every campus will compete for a structurally smaller domestic market.

Against this there are three buffers — post-compulsory participation rates, international students and lifelong learning — and all three have something in common: they do not benefit all institutions equally. They benefit those with scale, brand, online capability and a differentiated proposition. In other words: demographics does not shrink the sector proportionally; it polarises it. And polarisation is the classic engine of consolidation.

The fast, uncertain force: AI

If demographics shrinks the market, AI questions what is sold in that market.

For centuries, the educational business model rested on a scarcity: access to knowledge and to those who could explain and certify it. AI has made an enormous part of that scarcity abundant. Any student today carries an infinite, instant, almost free tutor in their pocket.

This does not kill educational institutions — I have written it many times: it makes the ones that develop what AI does not replace more valuable (judgment, real experiences, community, evidence of capabilities) and exposes the ones that only transmitted content. AI works like a selective acid: it does not attack the sector, it attacks a specific model within the sector.

At the same time, it lowers the entry cost for new competitors: challenger models without legacy structures, platforms that scale without campuses, corporate academies, alternative credentials. The competitor of a €15,000 on-campus master’s is no longer just another master’s: it is a whole range of new ways to achieve the same outcome.

The four-position map: where each institution stands

Each force alone would be manageable. Together they form a scissor: fewer students choosing among more alternatives. From that crossing emerges the sector’s new map. I call it the four-position map, and every educational institution stands today in one of these positions:

1. Institutions that will buy. Scale, brand, financial muscle and — the best ones — an educational model with a future. For them, demographics is a market-share opportunity and AI an investment they can afford.
2. Institutions that will be bought. Solid, beloved projects that are sub-critical: without scale to invest in technology, without generational succession, without capacity to internationalise. Their best future may be to integrate well — into the right group, at the right time, protecting their identity.
3. Institutions that will transform on their own. Those with the strategic clarity and institutional energy to renew their learning architecture in time. They are fewer than those who believe themselves capable.
4. Institutions that will close. Those that wait until both curves are visible in their income statement. By then, neither the sale will be good nor the transformation possible.

The most honest question a leadership team or an owner can ask today is not “how will next year go?”, but “which of the four positions are we in — and which one do we want to be in?”

What does this mean for each actor?

For those who lead or own an institution: structural decisions — selling, merging, transforming, finding a partner — are worth more the earlier they are taken. An institution with stable enrolment, a solid faculty and a living project negotiates; an institution with a visible downward trend pleads. The difference between those two scenes is usually three to five years.

For those who invest or buy: the demographic analysis of the catchment area and the educational model’s resilience to AI should weigh as much as normalised EBITDA. A reasonable multiple on a structurally exposed asset is more expensive than a demanding multiple on a learning architecture with a future. The key due diligence question is no longer “how much does it generate?” but “why will it keep generating it?”.

For the system: consolidation is not intrinsically bad. Done well, it can save educational projects that would not survive alone, and finance transformations no sub-critical institution could afford. Done badly — as pure financial engineering on an asset that is not understood — it destroys in three years what took fifty to build.

Both forces are already in motion. The map will be redrawn either way. The only open variable is who redraws it with educational judgment — and who with nothing but a spreadsheet.

Frequently asked questions

When will the demographic decline reach universities? In Spain, from around 2030: the cohorts born after 2008 — each smaller than the last — reach university age then. Early childhood and primary education are already living the contraction today; secondary education absorbs it through this decade.

What is the four-position map? A framework to place any educational institution facing the double pressure of demographics and AI: it will buy, be bought, transform on its own, or close. The position is not fixed — but it changes price with every year that passes without a decision.

How does this affect a specific transaction? The demographic analysis of the catchment area and the model’s resilience to AI should weigh in the valuation as much as EBITDA. It is the basis of the academic due diligence I describe in A new educational reality demands a new M&A paradigm.

I support owners, funds and education groups in transactions through the analysis of the real value of the educational project. More on this line of work at Education M&A.