Skip to content
Nations Inquiry — home

AnalysisDemocracy & Civil LibertiesEducation & Knowledge

Academic freedom in 2025: where it fell, where it rose, and how far since 1990

Academic freedom fell in 93 of 176 countries in 2025 and rose in 57. The middle country read 0.619, its lowest since 1990. Czechia is highest, at 0.978.

Published
Share

The figures in this piece are as they stood on . The pages it is based on are kept current; this piece is not, so a number here may since have moved on those pages.

Based on: Academic freedom · Academic Freedom Index, 2025: full table · Academic Freedom Index, 1990: full table · Democracy & Civil Liberties

The V-Dem Institute's Academic Freedom Index asks how free scholars and universities are in practice, on a scale from 0 to 1, where a higher figure means more freedom. This piece reads V-Dem's 2026 release, which covers 176 countries in 2025 and re-estimates every year back to 1990 on one method, so a difference between two years is a change in the estimates rather than a revision of them. The figures are as they stood on 14 November 2026; the statistic page is the one kept current.

More countries fell than rose in 2025

Of the 176 countries measured in both 2024 and 2025, 93 read lower in 2025, 57 read higher and 26 did not move. The falls were also the larger moves: 39 countries fell by 0.05 or more on the 0–1 scale, against 8 that rose by as much. The middle country of the field read 0.679 in 2024 and 0.619 in 2025.

The falls were not confined to one part of the world. On every continent more countries fell than rose: 22 of Europe's 40 fell and 12 rose; in Asia 26 fell and 14 rose; in Africa 29 fell and 20 rose; in the Americas 13 fell and 9 rose; and in Oceania, with six countries measured, 3 fell and 2 rose.

Where it is highest and lowest

Czechia has the highest reading in 2025, alone, at 0.978. Estonia follows at 0.974, then Belgium at 0.946, Jamaica at 0.939 and Slovenia at 0.937; Honduras and Seychelles share 0.931 just below. The list is not a list of the richest countries, and it was never meant to be: the index measures how freely universities work, not how well funded they are.

At the other end, Nicaragua has the lowest reading, alone, at 0.019, followed by Myanmar at 0.023, Eritrea at 0.034, North Korea at 0.036 and South Sudan at 0.054. In all, 57 countries read 0.8 or more and 32 read below 0.2.

Both ends have moved over the last ten years, and in the same direction. In 2015, 36 countries read 0.9 or more; in 2025, 19 do. The group at 0.8 or more has shrunk from 77 to 57, and the group below 0.2 has grown from 24 to 32. The extremes themselves barely moved — the highest reading was 0.974 in 2015 and is 0.978 now, the lowest 0.012 and 0.019 — so the change is in the large group of countries between them.

ContinentCountries measuredMiddle country, 2015Middle country, 2025
Oceania60.8840.857
Europe400.9220.820
Americas270.8550.761
Africa550.6440.506
Asia480.3800.278

Each figure is the reading of the continent's middle country, unweighted by population. Every continent's middle is lower in 2025 than in 2015.

The largest moves of 2025

The five largest falls between 2024 and 2025 were Ecuador, from 0.731 to 0.433; the United States, from 0.677 to 0.397; Burkina Faso, from 0.840 to 0.562; Kosovo, from 0.821 to 0.621; and Uganda, from 0.375 to 0.196.

The five largest rises were Mauritius, from 0.346 to 0.592; Bangladesh, from 0.286 to 0.519; Iraq, from 0.412 to 0.604; Somaliland, from 0.567 to 0.709; and Portugal, from 0.715 to 0.826. Two of those are partial returns rather than new highs: Mauritius read 0.653 in 2015 and Portugal 0.960, and both are still below those readings.

The file records these moves; it does not record their causes, and this piece does not supply them. V-Dem's own reports discuss individual countries, and a reader who wants an account of any one of them should take it from there.

The longer view

In 1990 the middle country of the 161 then measured read 0.552. The 1990s brought the largest gains in the series: in 1991 alone 65 countries rose and 12 fell, and by 2000 the middle country read 0.730. For more than two decades after that it barely moved, staying between 0.725 and 0.745 every year until 2022. Then it fell three years running, to 0.694 in 2023, 0.679 in 2024 and 0.619 in 2025 — its lowest reading since 1990, the first year of the series.

Falls have outnumbered rises in every year since 2019, and in every year but one since 2011. Over the whole span the picture is more mixed: of the 161 countries measured in both 1990 and 2025, 83 read higher and 76 lower. The largest rises over the span are Albania, from 0.125 to 0.760; Malawi, from 0.161 to 0.795; Kenya, from 0.266 to 0.864; and South Africa, from 0.340 to 0.831. The largest falls are Venezuela, from 0.934 to 0.129; Nicaragua, from 0.793 to 0.019; Hungary, from 0.946 to 0.299; Hong Kong, from 0.855 to 0.237; and India, from 0.682 to 0.136.

The last ten years

Between 2015 and 2025, 118 of the 176 countries fell and 56 rose. The largest falls over the decade are El Salvador, from 0.808 to 0.165; Mali, from 0.855 to 0.281; the United States, from 0.921 to 0.397; and Chad and Afghanistan, which fell by the same amount, 0.427 — Chad from 0.559 to 0.132, Afghanistan from 0.513 to 0.086. The largest rises are Gambia, from 0.525 to 0.821; Bangladesh, from 0.235 to 0.519; Fiji, from 0.371 to 0.636; Thailand, from 0.118 to 0.382; and Uzbekistan, from 0.047 to 0.248.

Pakistan is one case among the fallers. It read 0.636 in 1990 and 0.563 in 2015, and 0.344 in 2024 and 0.268 in 2025, below the middle of its continent.

How the index is built

V-Dem builds the index from five indicators: the freedom to research and teach, the freedom of academic exchange and dissemination, the institutional autonomy of universities, campus integrity — freedom from surveillance and intimidation on campus — and the freedom of academic and cultural expression. Each indicator is rated year by year by country experts, several for every country, and V-Dem's Bayesian measurement model combines their ratings, allowing for the fact that different experts read the same scale differently. The index is the model's estimate of what the five indicators share.

Three things follow for a reader. First, the figures are estimates, which V-Dem publishes with an uncertainty range; this site holds only the central figure, so two countries a few hundredths apart may not differ at all, and a small move in one year is weak evidence on its own. Second, each release re-estimates the whole history, so a year's figure here can differ from the one printed in an earlier V-Dem report. Third, V-Dem publishes values and orders nobody: the position shown on the statistic page is this site's position · derived, worked out by ordering the published values, and is labelled as ours.