Education Businesses Sell a Promise That Takes Years to Test
The customer pays upfront for an outcome that arrives later and cannot be evaluated in advance. That gap explains most of the regulation and most of the failures.
The Structural Difficulty
A student pays for education before knowing whether it will deliver what was promised. The outcome, employment or earnings improvement, is observable only years later, and by then the money is spent.
This is a severe information asymmetry. The provider knows far more about the likely outcome than the buyer, and the buyer typically makes the decision once in their life with no repeat experience.
Markets work well when buyers can evaluate what they bought and buy again. Education fails both conditions, which is why it is one of the most regulated consumer sectors.
Who Actually Pays
| Model | Payer | Key sensitivity |
|---|---|---|
| Public institution | Government | Budget allocations |
| Private, tuition funded | Student, often via loans | Loan availability |
| Corporate training | Employer | Corporate budgets |
| Consumer online learning | Individual | Discretionary spending, high churn |
The middle case is the one that generates most controversy. Where students borrow government backed loans to pay private providers, the provider receives money regardless of whether the student completes or benefits, and the student carries the debt.
That structure produced serious problems in several countries, with providers expanding enrolment aggressively while completion rates and employment outcomes deteriorated. The regulatory response has generally been outcome based rules tying eligibility for funding to completion and earnings measures.
The Unit Economics
Traditional institutions carry heavy fixed costs in campuses and faculty, so enrolment relative to capacity drives profitability in the same way occupancy drives it for hotels.
Online providers have low marginal delivery cost, which appears to make scale enormously profitable. In practice the cost moves to customer acquisition, and consumer education has some of the highest marketing spend as a proportion of revenue of any sector.
Completion rates in self directed online learning are low, which affects both outcomes and reputation, and word of mouth is weak when most buyers do not finish.
Why Reputation Compounds
Because outcomes are hard to observe, buyers use proxies: institutional reputation, selectivity, rankings, employer recognition.
Those proxies are slow to change in either direction, which makes established institutions extraordinarily durable and new entrants extraordinarily difficult to establish. An institution with a century of reputation has a moat that cannot be bought.
It also means quality decline is not punished quickly. A provider can coast on reputation for a long time before demand responds.
The Credential Question
Part of what education sells is a credential that signals capability to employers, separate from the learning itself.
Where the credential carries the value, the incentive to lower standards is real, since a credential granted to more people generates more revenue while diluting the signal it depends on.
Professional accreditation exists partly to prevent this by putting the standard outside the institution collecting the fees.
What to Examine in a Provider
Completion rates rather than enrolment. Employment and earnings outcomes where disclosed. The proportion of revenue from government sources and the rules attached. Marketing spend as a share of revenue, since very high figures indicate acquisition dependence rather than demand. And the trend in cost per enrolment, which reveals whether growth is getting harder.
The Bottom Line
Education sells an outcome the buyer cannot evaluate until years afterwards, which produces heavy regulation and allows enrolment growth to diverge from quality for a long time. Where government funding flows to private providers, the incentive is to enrol rather than to complete, and outcome based rules exist to correct it. Reputation is the durable asset and it changes far more slowly than the underlying quality does.