OECD Average in PISA Falls to a Historic Low. Austrian Students Read and Do Maths Worse Than Three Years Ago, Yet Sit Above It

The Organization for Economic Cooperation and Development published the results of the PISA 2025 testing on Tuesday and measured the lowest average in the history of measurement. The decline is so deep that it changes the meaning of the ranking itself: Austrian students read and calculate worse than three years ago, yet they are above the average of OECD countries. And one Spanish region dropped out of the comparison entirely, because it left almost a quarter of the selected students out of the sample.

The Ranking Fell So Deep That You Can Rise In It Without Improvement

The average score of OECD countries in mathematics fell from 485 points in 2015 to the current 463. In reading, 28 points were lost over the same period, which the OECD converts to roughly one and a half years of school curriculum. Today, one in five fifteen-year-olds on average in OECD countries does not achieve the basic level in any of the three areas, whereas in 2022 it was 16 percent. OECD ↗ The 2025 cycle was the most extensive measurement in the project's history, involving 760,000 students from 91 countries and economies. ANSA ↗

The depth of this decline created a peculiar effect: you can advance in the ranking even when your own results decline. After this cycle, Austria is for the first time since testing was introduced significantly above the OECD average in all measured disciplines simultaneously. In reading, it lost thirteen points and in mathematics ten, improving only in science. More than a quarter of Austrian youth do not achieve the minimum level in reading or mathematics. Die Presse ↗

The same effect applies to Slovakia in mathematics. Economist Róbert Chovanculiak from the INESS institute recalled that the most discussed success was aligning with Finland, but Finnish results have been declining for twenty years until they approached our level. "If I had to sum it up in one sentence, we are stagnating and it is not a great success, but since the results of other countries continue to worsen and decline, it is such bittersweet joy," he said for Aktuality.sk. Aktuality.sk ↗

Who Is Falling and Who Is Standing Still

The difference in tone with which various governments discuss the results has largely a factual basis: countries are on different trajectories.

Germany achieved 465 points in reading, 464 in mathematics, and 486 in science. Compared to the OECD average of 461, 463, and 482 respectively, these are values comparable to the average, not significantly better. However, the movement is significant: in reading, Germany fell from 480 points in 2022 and in mathematics from 475, which is the lowest result it has ever achieved in this test. Federal Minister Karin Prien speaks of a dramatic and worrying finding, in which the social scissors continue to widen, and employer associations warn of subsequent costs in the trillions of euros. Tagesspiegel ↗ The SPD faction proposed calling an education summit, and the question of banning private mobile phones in schools also returned to the debate. Die Zeit ↗

In Slovakia, both reading and mathematics were statistically stable compared to 2022. In reading, it achieved 448 points compared to the OECD average of 461, in mathematics 469 compared to 463, and in science 475 compared to 482. Compared to 2022, reading and mathematics did not change statistically significantly, and in science Slovakia improved by thirteen points. Ministry of Education of the SR ↗ Minister of Education Tomáš Drucker therefore, in evaluating the overall results, speaks of cautious optimism and links it to implemented reforms, although the test itself does not demonstrate the effect of a specific reform. Postoj ↗ Slovak Deputy Secretary General of the OECD František Ružička at the Bratislava presentation said that Slovakia has no reason to be ashamed, while also noting that Slovak students need to read more. TASR ↗

What stability masks is distribution. More than a third of Slovak students, specifically 34.4 percent, do not achieve even basic reading literacy, and in vocational programmes that do not end with a school-leaving exam, it is 78.1 percent in reading and 72.9 percent in mathematics.

There was also a sharp reaction in Flanders, where almost a third of fifteen-year-olds do not achieve the basic level in reading. Minister of Education Zuhal Demir from the nationalist N-VA at the presentation of results said: "I am ashamed that I have to present this." De Standaard ↗ In the French-speaking part of Belgium, which improved compared to its neighbors, Minister Valérie Glatigny refused to celebrate, saying they cannot be satisfied with being "the least bad." La Libre ↗ In Bulgaria, 387 points were measured in reading, the lowest score in the country's entire history of participation. Dnevnik ↗ Minister of Education Georgi Valčev summed it up by saying that the results show one thing: that there is a problem in the Bulgarian education model. 24 Chasa ↗ In France, with presidential elections approaching, candidates speak of an earthquake, an alarm signal, and the need for a cultural revolution. Le Figaro ↗

Turkey Grew Against the Trend

Turkey recorded the largest increase in scores among OECD countries, in all three areas simultaneously. Compared to 2022, it improved by eighteen points in science to 494 and by sixteen points in reading to 472, achieving 462 points in mathematics. In science and reading, it is above the OECD average; in mathematics, one point below it. In the ranking among all participants, it moved from 34th place to 19th in science, from 36th to 18th in reading, and from 39th to 28th in mathematics, with the number of participating countries varying between cycles. Turkish Minute ↗

Turkey's average score in reading is thus higher than Austria's, Germany's, and Slovakia's. The French daily Le Figaro devoted a separate analysis to the question of how Turkey achieved this. Le Figaro ↗ Within the Visegrad Four, Poland stands out, and according to a Polish research institute it remains among the top performers in the European Union. TVN24 ↗

Catalan Data Was Devalued by Excessive Student Exclusion From the Sample

Spain recorded the worst result in its history and also faced a sampling problem. Catalonia excluded 591 students from testing, that is 23.2 percent of the originally selected sample of 2,545 fifteen-year-olds, although the rules allow a maximum of five percent. The stated reasons were disability or insufficient command of the test language. The OECD therefore did not include Catalan results in the main report and notes that with such an extent of exclusion, they are probably overestimated. ARA ↗

The Catalan administration speaks of an error in sample composition. According to the test documentation, school principals received instructions on which groups of students could be exempted from the test, but they were not informed that the overall cap was five percent. Secretary General of Education Teresa Sambola was dismissed on August 31, with Education Minister Esther Niubó publicly denying a direct link between the dismissal and PISA results and speaking of a new stage for the ministry. El Nacional ↗

Catalonia was not alone. The nationwide exclusion rate in Spain reached 7.89 percent, also above the permitted cap, and Murcia excluded 12.7 percent of students. The OECD handled this differently than with Catalonia: it retained its results but marked them with a warning. ARA ↗ Spanish results as a whole are also commented on by El País, according to which it is a general defeat across autonomous communities. El País ↗

Limits of This Perspective

Belgian newspaper De Morgen also points out the limits of the tool itself. In comments on the results, it writes that the score is not a measure of education quality and that the country ranking should be taken with great reservation. De Morgen ↗ This is an objection that also applies to this text: comparing statements by ministers across countries is not proof that one government acts better than another. Political language is governed by the domestic cycle as much as by results.

Caution is also required when comparing scores between countries. A difference of a few points may not be statistically significant, and the OECD therefore usually works with ranges rather than rankings. Slovak improvement in science is statistically significant and is not an optical illusion from others' decline, even though it is still fourteen points lower than in 2006, when science was the main measurement area for the first time. The proportion of Slovak students in the two highest levels also significantly increased in mathematics.

One piece of data outside the main numbers is also interesting. According to the questionnaire that students completed alongside the test, nine percent of Slovak fifteen-year-olds skipped at least one day of school without excuse during the two weeks before testing, compared to 22 percent on average in OECD countries.

The Spanish case ultimately shows why methodology is important. When discipline in sample composition is relaxed, the ranking stops measuring what it claims to measure, regardless of whether this was due to error or intent.

What hotinfo Is Monitoring