What Schooling Bought

Essay · economics · October 2026

In 2000, four Portuguese adults in five had at most nine years of school. In 2025 the figure was about one in three, and young Portuguese are now as schooled as young Europeans. It was the largest catch-up in Europe. The degree still pays well, and young adults read almost as well as their European peers, but output per hour has barely grown: the gain from schooling was cancelled by a fall in capital per hour of work. This essay looks at the boom and what it paid, and looks at the next problem: more than half the teachers are 50 or older.

Reading time ≈ 11 min The boom · the skills · the reward · the puzzle · the teachers

The question

The usual explanation for Portugal’s low wages and productivity is education: a country that reached the 1970s with most adults schooled for four years or less could not work like its neighbours. The Two-Thirds Country showed that an hour of Portuguese work produces about two-thirds of the European average, and The Stalled Hour that output per hour has barely grown since 2013. Over the same period Portugal schooled itself faster than any other country in Europe. This essay asks what that schooling bought: skills, pay, productivity, or mainly diplomas.

81% → 36%adults aged 25–64 with at most nine years of school, 2000 and 2025; EU 19%

6%of 18–24-year-olds left school early in 2025, against 44% in 2000; EU 9%

235adult literacy score in 2023, against an OECD average of 260

2.0×median hourly pay of a graduate relative to secondary school, 2022; EU 1.4×

The boom

From four years of school to twelve

In 1960, the average Portuguese adult had about two years of schooling, and 42 per cent had none; Spain was at four. By 2015 the Portuguese average was over nine years, against eleven in Spain.[1] The pace of the last quarter-century stands out. In 2000, 81 per cent of adults aged 25 to 64 had at most lower secondary education, nine years of school or less; in 2025, 36 per cent did (Figure 1). Among young adults aged 25 to 34, the share with a degree rose from 13 to 43 per cent, about level with the EU’s 45.[2]

Fig. 1 — Adults aged 25–64 with at most lower secondary education (green), and adults aged 25–34 with a tertiary degree (red), Portugal (solid) and EU27 (dashed), 2000–2025, per cent; EU series from 2002. The 2021 peak in Portuguese graduates is a break in the survey, not a real rise and fall. Data: Eurostat, Labour Force Survey (edat_lfse_03).

Early school leaving fell even faster. In 2000, 44 per cent of 18-to-24-year-olds had left school without finishing secondary education and were in no training; in 2025, 6 per cent had, below the EU’s 9 per cent and below Germany or France.[3] Since 2009, schooling has been compulsory to the age of 18 or the end of secondary school.[4]

What remains is an age gap. Among Portuguese aged 25 to 34, only 15 per cent stopped at nine years of school or less, against 14 per cent in the EU. Among those aged 55 to 64, 59 per cent did, against 25 per cent in the EU.[2] The national average is held down by people who were children before 1974, and that gap will close only as they retire.

The skills

More school, and mostly more skill

Years in school are not the same as what people can do, so the question is whether the extra years brought skills. For fifteen-year-olds, the answer is yes. In the OECD’s PISA tests, Portuguese students scored 470 in reading in 2000, well below the average of rich countries; they reached 498 in 2015, the best Portuguese result, and were at 477 in 2022, level with the OECD average of 476 after a fall that happened almost everywhere.[5]

Adults are another matter. In 2023 Portugal took part for the first time in the OECD’s Survey of Adult Skills, which tests people aged 16 to 65. Portugal scored 235 in literacy against an OECD average of 260, and 238 in numeracy against 263, near the bottom of the participating countries, level with Lithuania and Poland in literacy. Forty-two per cent of Portuguese adults read at the lowest level or below, against 26 per cent across the OECD.[6]

The age profile explains most of it (Figure 2). The youngest adults, those aged 16 to 24, are about 13 points below the OECD average of their age; adults aged 55 to 65 are 28 points below theirs. Compared level by level, Portuguese adults are close to their OECD peers: those with a degree score 272 against 282, those with secondary education about 240 against 250. The national average is low mainly because so many adults have little schooling, not because a Portuguese diploma is worth less.[6]

Fig. 2 — Mean literacy score by age group, Portugal and the OECD average, Survey of Adult Skills 2023. Values are read from a published chart and are accurate to about two points; the numbers between the lines are the gaps. Data: OECD, Survey of Adult Skills 2023, Table A.2.4, as presented by Pessoas 2030 and ANQEP (November 2025).

The gap that remains among the young is real, though. A decade after the best PISA results, those students now in their twenties still read and calculate less well than young adults across the OECD, and the decline since 2015 means the next generation will not close it on its own.

The reward

A degree still pays, but less

If degrees were being handed out faster than they were needed, their reward would have collapsed. It has fallen, but from a high level. In 2006, the median Portuguese graduate earned 2.4 times as much per hour as the median worker with secondary education; in 2022, 2.0 times, still the highest ratio among the countries compared and well above the EU’s 1.4 (Figure 3).[7] A different measure from the OECD, which counts annual earnings of all workers, shows no fall at all: Portuguese graduates earned 77 per cent more than people with secondary education in 2024, sixth highest in the OECD.[8]

Fig. 3 — Median gross hourly earnings of employees with a tertiary degree divided by those of employees with upper secondary education, firms with ten or more employees, 2006–2022. Data: Eurostat, Structure of Earnings Survey (earn_ses_pub2i); author’s ratios.

Studies by the Banco de Portugal and the government’s planning office agree on the trend. The reward for a degree rose in the 1990s, when graduates were scarce, peaked around 1996, and has declined slowly since, more for women, the young and low earners, and more for the first degree, shortened by the Bologna reform; the reward for master’s degrees and doctorates has risen. Even in 2016, a degree still returned about 5 per cent a year on what it cost in forgone pay, and the Banco de Portugal concluded that it “remains a profitable investment”.[9][10]

The fall in real pay is more striking. At 2022 prices, the median graduate earned €15.55 an hour in 2006 and €11.59 in 2022, a quarter less. Much of this is composition: the graduate pool more than doubled and became younger, so the typical graduate in 2022 was a less experienced worker than in 2006. The same effect explains why average pay rose while pay within each level of education stagnated, as The Floor showed.[7] At the other end, the extra pay for finishing secondary school has almost vanished: the median worker with secondary education earned only 8 per cent more per hour than one with nine years of school in 2022, against 33 per cent in 2006, squeezed by the rising minimum wage.

Two other measures argue against the idea that Portugal produced too many graduates. Only 16 per cent of employed graduates work in jobs that do not need a degree, against 21 per cent in the EU and 34 per cent in Spain.[11] And 88 per cent of recent graduates were employed in 2025, slightly above the EU average, although young graduates are still more often unemployed than their European peers (8.1 against 6.8 per cent among those aged 25 to 29). Recent graduates of vocational secondary courses do worse: 75 per cent employed, against 80 per cent in the EU.[12]

The puzzle

Where did the productivity go?

On standard assumptions, the rise in schooling alone should have added about half a percentage point a year to the growth of output per hour since 2013. Measured output per hour grew by 0.17 per cent a year, a third of that (Figure 4). From 2000 to 2013 the opposite was true: productivity grew faster than schooling alone would explain, partly because capital was still being added and, after 2008, the least productive jobs were being shed.[13] The calculation is a rough one: it counts only years of schooling and applies a standard weight to them, and it cannot say what offset the schooling. But the gap since 2013 is too large to be a rounding error.

The detailed European accounts, which split productivity growth into the workforce, capital and everything else, show where the schooling went. It did register: a better-educated workforce added about 0.9 points a year to the growth of output per hour in 2013–2019, more than in any other country compared. But capital per hour fell enough to subtract about 0.6 points a year, and everything else, which includes how firms are managed, added nothing. Where the Schooling Went sets out the numbers and how far each cause can be measured (correction, October 2026: an earlier version of this essay left open whether the schooling had registered at all).

Fig. 4 — Portugal: contribution of rising schooling to the growth of output per hour on standard assumptions (growth of the Penn World Table human-capital index times a labour share of 0.64; to 2023), against measured growth of real output per hour worked, per cent a year. A back-of-envelope calculation, not an estimate. Data: Penn World Table 11.0; Eurostat (nama_10_lp_ulc); Banco de Portugal (2019) for the labour share; author’s calculation.

Three explanations were on the table, and the numbers favour the first. The first is that education needs other things to pay off. Since 2013 investment has barely replaced worn-out capital, and new graduates went to work in the same small firms, with the same equipment, as before (The Stalled Hour covers this). The second is that part of the extra schooling is a signal rather than a skill, which would mean the accounts overstate its contribution: when everyone has a degree, a degree tells employers less. The shorter post-Bologna degree losing value while master’s degrees gained is consistent with this, though direct tests elsewhere find little evidence of pure signalling.[9][14] The third is that some of the new skills left: by 2010 almost one Portuguese-born graduate in seven lived in another OECD country; a rough calculation puts the cost at 2 to 3 per cent of output per hour, small next to the fall in capital. Emigration is the subject of The Leavers.[15]

The first explanation has the strongest Portuguese evidence, and it points at the people who run firms. In 2010, 58 per cent of employers in Portugal had at most nine years of school; in 2024, 38 per cent did, while 30 per cent had a degree; on the EU survey’s measure, 28 per cent of Portuguese employers had a degree, against 41 per cent across the EU.[16] A study of every Portuguese firm from 1995 to 2015 found that firms founded by more educated entrepreneurs start larger, grow faster and adopt more technology, and that the effect comes from the entrepreneur’s schooling, not from that of the workers; one more year of the founder’s schooling is associated with about 5 per cent higher productivity.[17] The Banco de Portugal concluded in 2019 that “CEOs in Portugal have still fairly low levels of education”.[18] The educated young are arriving in firms faster than they are arriving at the head of them, a subject for The Business Side.

The teachers

The next shortage

Portugal paid for the boom. Public spending on education rose to 6.7 per cent of GDP in 2010, above the EU’s 5.0, and has since fallen to 4.3 per cent, below the EU’s 4.7; the fall reflects pay cuts and freezes, fewer pupils and a larger economy, in proportions not measured here.[19]

The teachers who taught the boom are now old (Figure 5). In 2024, 56 per cent of the 137,000 primary and secondary teachers were 50 or older, against 40 per cent in the EU; 19 per cent were 60 or older; only 3 per cent were under 30, against 8 per cent in the EU. Only Italy has a comparably old teaching force. And only 4 per cent of new Portuguese graduates studied education, against 9 per cent in the EU.[20][21] A wave of retirements is coming with a thin supply behind it.

Fig. 5 — Primary and secondary school teachers aged 50 and over (solid) and under 30 (dashed and lower red), Portugal and the EU27, 2013–2024, per cent of all teachers. Data: Eurostat, teachers by age (educ_uoe_perp01); author’s shares.

The balance

The schooling arrived; the rest has not

Portugal did the hard part. It went in one generation from the least schooled population in Western Europe to young adults as schooled as their European peers, without a collapse in the value of their degrees and without the over-qualification seen in Spain. The skills followed the diplomas, imperfectly: young Portuguese read and calculate close to, but still below, their peers abroad. The low national average in adult skills is mainly the past, and it will fade as the older generation retires; the Banco de Portugal projects that better education will add about 10 percentage points to GDP per person by 2050, against about 20 points lost to ageing.[22]

What schooling has not done is raise output per hour by itself. The return to education depends on what educated people find when they start work: capital to work with, firms large enough to use them, managers able to organise them, and a reason to stay. Those are the questions of the next essays. The risk on the education side is no longer quantity but who will teach the next generation.

On method and tools

This piece was written collaboratively with Claude Opus 5.5 (Anthropic): human specification, editorial direction and critical review; machine data research, analysis and drafting. The figures are computed by scripts/education.py. Attainment, early leaving, over-qualification and unemployment come from the EU Labour Force Survey, which has breaks in 2011 and 2021. Earnings ratios are medians of gross hourly pay in firms with ten or more employees. The adult-skills values by age are read from a published chart. The productivity comparison multiplies the growth of the Penn World Table human-capital index, which counts years of schooling only, by the labour share of 0.64 used by the Banco de Portugal; it is arithmetic, not an estimate. Results are in docs/education-results.json; the downloaded sources, with page or table for each number, are kept with the script’s data.

The cover photograph is Biblioteca Joanina by Trishhhh; CC BY 2.0, via Wikimedia Commons, cropped.

Authored by: Luis Matos Ferreira — Physicist, Developer, Writer

Related essays on this blog

1. The Two-Thirds Country — Portuguese productivity at two-thirds of Europe’s.
2. The Stalled Hour — why output per hour stopped growing.
3. The Floor — the minimum wage and the flattened pay scale.
4. The Leavers — who emigrates, and why.
5. Where the Schooling Went — how much each cause of the stall is worth.
6. Work, Time and Money — the reading guide to the whole series.

Sources

1. R. Barro and J.-W. Lee, educational attainment dataset, version 3, adults aged 25–64.
2. Eurostat, population by educational attainment level, sex and age (edat_lfse_03), updated September 2026.
3. Eurostat, early leavers from education and training (edat_lfse_14), 1992–2025.
4. Lei 85/2009, of 27 August, extending compulsory schooling to the age of 18.
5. OECD PISA mean scores 2000–2018 via World Bank EdStats; IAVE, PISA 2022: Resultados (reading and science notes), p. 1; OECD PISA 2022 country note for Portugal as reproduced by FNE, December 2023, p. 3.
6. OECD, Survey of Adult Skills 2023, Tables A.2.1–A.2.5, as presented by Pessoas 2030 and ANQEP, 25 November 2025, pp. 10–20.
7. Eurostat, Structure of Earnings Survey, median hourly earnings by education (earn_ses_pub2i), 2006–2022; HICP; author’s ratios and real values.
8. OECD, Education at a Glance, relative earnings of adults by educational attainment (upper secondary = 100), 2024.
9. M. M. Campos and H. Reis, “Is it still worth investing in education?”, in Banco de Portugal, Portuguese Economic Growth: A View on Structural Features, Blockages and Reforms, 2019, pp. 53–58.
10. PLANAPP, Nota de Análise 11, November 2024.
11. Eurostat, over-qualification rate (lfsa_eoqgan), 2025; A. C. Pimenta and M. C. Pereira, in Banco de Portugal (2019), pp. 61–66.
12. Eurostat, unemployment by education (lfsa_urgaed) and employment of recent graduates (edat_lfse_24), 2025.
13. Penn World Table 11.0, human capital index; Eurostat, real labour productivity per hour (nama_10_lp_ulc); J. Garcia, H. Reis and J. Amador, in Banco de Portugal (2019), pp. 45–46; author’s calculation.
14. D. Clark and P. Martorell, “The Signaling Value of a High School Diploma”, Journal of Political Economy 122(2), 2014.
15. OECD, Connecting with Emigrants: A Global Profile of Diasporas, 2015, p. 239 (tertiary emigration rate, 2010/11).
16. GEP/MTSSS, Quadros de Pessoal 2010–2024, employers by level of schooling (Quadro 37 in 2024); Eurostat, Labour Force Survey (lfsa_esgaed), 2024.
17. F. Queiró, “Entrepreneurial Human Capital and Firm Dynamics”, Review of Economic Studies 89(4), 2022; Conselho para a Produtividade, 2019 report, citing the working-paper version.
18. S. Sazedj, “What has the crisis shown about Portuguese top managers?”, in Banco de Portugal (2019), pp. 117–122.
19. Eurostat, general government expenditure by function, education (gov_10a_exp), 1995–2024.
20. Eurostat, teachers by age (educ_uoe_perp01), ISCED 1–3, 2013–2024; author’s shares.
21. Eurostat, graduates by field of education (educ_uoe_grad02), 2023; author’s shares.
22. J. Garcia, H. Reis and J. Amador, “What is the role of demography and education on future Portuguese output growth?”, in Banco de Portugal (2019), p. 47.