Comparative and international education: what actually travels

What comparison can and cannot show

International assessments and cross-country studies are unusually powerful and unusually easy to misuse. What they establish well is variation: that outcomes differ between systems by amounts too large to be noise, that the differences are patterned, and that some patterns persist across cycles. That is a real contribution, and it has repeatedly overturned complacent national assumptions.

What they cannot establish, on their own, is cause. A high-performing system differs from a low-performing one in hundreds of ways at once: curriculum, teacher selection, class size, tracking age, inequality, child poverty, language, the status of the profession, the age at which the test is taken. Picking one of those and calling it the reason is not a finding; it is a preference with a citation attached.

The most common specific error is comparing systems on an average while their intakes differ. Where one system tests a selected population and another tests nearly all children of an age, the comparison measures selection at least as much as teaching.

Policy borrowing, and why it usually disappoints

Policy borrowing (importing a practice from a high-performing system) has a poor record, and the reason is structural rather than accidental. A practice in its home system is held in place by things that did not travel with it: how teachers are recruited and paid, what parents expect, what the assessment at the end of schooling rewards, how much inequality the surrounding society tolerates.

Take the recurring interest in Finnish schooling that this domain's catalogue documented. The visible features (later formal schooling, little standardised testing, high teacher autonomy) sit inside a system where teaching is a competitive graduate profession and child poverty is comparatively low. Importing the visible features into a system without those conditions imports the parts that are cheap and leaves the parts that are load-bearing.

This is not an argument against learning from other systems. It is an argument for asking, before any transfer, what holds the practice in place at home and whether that support exists here. The same question, asked of a workplace rather than a country, is the transfer problem from the adult education brief.

Education and development

Research on education in developing systems has moved through recognisable phases: expanding access, then measuring whether access produced learning, then asking what conditions make schooling effective where resources are scarce. The uncomfortable finding from the second phase — that enrolment can rise substantially while measured learning does not — reshaped the field's assumptions and is still working through its policy.

Two cautions belong on any page that summarises this literature. First, aggregate national statistics conceal the variation that matters: within-country differences between regions, languages and income groups are often larger than the differences between countries. Second, the research is disproportionately produced outside the systems it describes, which shapes which questions get asked, a point the decolonial literature in this area makes at length and which is worth taking as methodology rather than as politics.

Global citizenship education

Global citizenship education is the part of this subject area where aims and evidence are furthest apart, and it is worth being plain about it. The aims are stated as dispositions (awareness, empathy, responsibility towards people elsewhere) which are difficult to define operationally and harder to measure durably. That does not make the goal illegitimate; it makes claims of effectiveness weak, and programmes in this area should be argued for on the grounds of what they teach rather than on measured outcomes they cannot show.

The critical strand argues something sharper: that a curriculum can address distant others in a way that leaves the relationship between the learner's own society and those others unexamined, and that the resulting lesson is more comfortable than the topic warrants. That criticism is about curriculum design rather than about measurement, and it is answerable by design.

Shadow education, the system nobody planned

Private supplementary tutoring (paid instruction in school subjects, outside school) is one of the largest education sectors in the world and one of the least represented in policy documents about education systems. It is substantial across East Asia, the Mediterranean, Eastern Europe and increasingly elsewhere, and this domain's catalogue included a full-length study of the Mediterranean case.

Three findings matter for anyone reading system-level evidence. First, shadow education confounds comparison: a system's measured performance may reflect what families purchased after school rather than what schools provided, and no international assessment separates the two. Second, it is regressive by construction: access follows income, so a sector that raises average attainment can widen the gap while doing it. Third, it feeds back into the school system: where tutoring is widespread, teachers can come to rely on it having happened, and the curriculum drifts towards what tutored pupils already know.

The policy responses on record range from prohibition to regulation to public provision of the same service, and none has a clean evidence base. What is clear is that a national reform designed as though schooling were the only instruction pupils receive is designing for a system that does not exist.

Reading a cross-country claim

• Who was tested, and were the intakes comparable? Selection differences outrank most teaching differences.
• Is the comparison over time within a system, or between systems at one point? The first can support a much stronger causal reading than the second.
• Which single factor is being credited, and what else differs? Name three other differences before accepting the attribution.
• What holds the borrowed practice in place at home? If the answer is not in the proposal, the proposal is incomplete.

Works referred to

1. Education and Development in Zimbabwe: A Social, Political and Economic Analysis — Sense Publishers, 2012.
2. Decolonizing Global Citizenship Education — Sense Publishers, 2015.
3. Private Tutoring across the Mediterranean — Sense Publishers, 2013.
4. Understanding Israel/Palestine: Race, Nation, and Human Rights in the Conflict — Sense Publishers, 2015.
5. Finnish Innovations and Technologies in Schools — Sense Publishers, 2014.

Questions readers send

Do international assessments show which country has the best schools?

They show which systems score highest on that assessment, for the population it tested. Whether that reflects teaching depends on who sat the test, at what age, after how much out-of-school tutoring, and in what language, none of which the headline number carries.

Can a country simply copy a high-performing system?

Rarely with the expected result. The visible practices are held in place by teacher recruitment, pay, parental expectations and the assessment at the end of schooling, and those do not travel. Ask what supports the practice at home before importing it.

Why did rising enrolment in developing systems not raise learning?

Because access and learning are different outcomes. Expanding places without adding teachers, materials, language support and time increases the number of children in school and not the amount taught, which is what the measurement phase of that literature established.

Is private tutoring a problem or a solution?

Both, and which one depends on the question. It raises attainment for those who buy it, so it looks like a solution to families; it follows income, so it widens gaps and confounds any comparison of school systems.

Does global citizenship education work?

Its stated aims are dispositions, which are hard to define operationally and harder to measure durably, so claims of measured effectiveness should be treated sceptically. That is an argument for defending such programmes on what they teach rather than on outcome data they do not have.

The briefs that cover this in depth

1. Learning theories The four families of learning theory compared on the same three questions: what each one can explain, what it cannot, and which teaching decisions it licenses.
2. Adult education What adult education research supports and what it only asserts: the six andragogical assumptions checked one at a time, plus the participation evidence.
3. Curriculum design Backward design, spiral sequencing and constructive alignment explained as a coverage-versus-depth trade, with the cost of each choice stated in teaching hours.