What really drives education quality around the world? Mostly money.

Several LinkedIn charts this month rank "best and worst education systems" using the new PISA 2025 results. Most are sloppy at best. One plots Russia, which did not take the test. Another puts "China" on top, when the OECD tested four wealthy regions, not the country. Some use a cost-of-living axis that says nothing about education.

So, to get the facts right, I checked the data, in accordance with one of my leadership principles - "Substance: accurate facts and critical thinking".

Conclusion: What explains PISA scores, first and foremost, is whether a country is rich. GDP per capita alone explains two-thirds of the variation of PISA scores across 83 countries.

The lessons, if any, are to be found in the other third: countries far above or below what their income predicts.

Above the line: Japan, Estonia and Korea score 60 to 70 points more than their wealth suggests. Vietnam, at $14,000 of GDP per capita, beats Argentina at $27,000. Singapore, as rich as Luxembourg, scores almost 100 points more.

Below the line: Qatar, Saudi Arabia, Brunei, Luxembourg, Norway and Iceland score 30 to 75 points less than their income predicts.

A few hypotheses, that may compound:

Who teaches. Singapore, Korea and Japan recruit teachers from the top of the class and treat the job as a profession with status. Many underperformers do not.

After school. In Japan and Korea, a large share of 15-year-olds spend evenings in paid tutoring. The state system gets credit for work done by the parents!

Stability. Argentina has had decades of inflation, teacher strikes and policy reversals. Schools need sustained direction to show results.

And of course a few methodological issues. For example the sample: Vietnam's PISA sample does not cover all its 15-year-olds, and those left out are more likely to have left school. Part of its gap with Argentina is coverage, not teaching. Or where GDP comes from. Oil, gas and finance can raise GDP per capita artificially. Luxembourg is a typical case with GDP/capita being inflated by workers who commute in.

One more point (mostly for my French readers). In France, one hears constantly that the American school system is far worse than the French one. Let's debunk this myth. In fact both countries sit right on the line for their income. The US is richer, and does score higher: 485 against 466. Income alone predicts an 18-point gap. The actual gap is 19. Both do exactly what their wealth predicts.

Analysis: NeuronVentures

Method: PISA 2025 mean of science, reading and mathematics for each country and economy in the analysis, plotted against 2024 GDP per capita at purchasing power parity (World Bank), with a line fitted against log GDP per capita.