Seven Takeaways

Why This Case Matters for Freedom of Expression

Takeaway 1

A people silenced to protect a relationship with an autocratic president

Voices of the People, DEMOS’s study of public opinion in Guinea-Bissau, was made to give the country’s citizens a public voice: the first comprehensive, scientific account of their own political views and social values. The European Union (EU) helped fund it, praised it, and then stopped its publication. On 26 November 2021, Rita Laranjinha, Managing Director for Africa of the European External Action Service, the EU’s diplomatic service, met President Umaro Sissoco Embaló in private. Three days later, the EU Delegation in Bissau wrote: “Do NOT go to print.” The European Commission’s explanation came almost three years later: some of the study’s conclusions “could undermine the EU’s relations with Guinea-Bissau.”

To protect a relationship with an autocratic president, the people’s voices were silenced, and the academic freedom of the research team was infringed. Such conduct contradicts the values the EU professes: freedom of expression, academic independence, and support for democracy and civil society. The European Ombudsman, having opened an inquiry into the alleged human rights violations, did not even address censorship or the restrictions on academic freedom in its Decision of July 2025.

Source: The Price of a Relationship, Executive Summary (Context, Points 1–2), Point 2.2 and Conclusion; European Ombudsman, Decision in case OI/4/2024/MIK, 8 July 2025, para. 3; vozesdopovo.org, pages 2 and 5.

Takeaway 2

Censorship that lowered the cost of political repression

After 2020, Guinea-Bissau was a democracy in free fall. President Embaló brought the Supreme Court under his control, restricted civil liberties and, in 2023, dissolved the National Assembly; already in 2022, Varieties of Democracy (V-Dem) reclassified the country as an electoral autocracy. In that setting, the EU’s censorship of the Voices of the People study sent an unambiguous signal, read in civil society and political circles alike: critical independent research, even when contracted and subsidized by the EU itself, was not safe if it inconvenienced the president. Public presentations were canceled, and a climate of caution set in.

Robert Dahl, among the most influential democratic theorists of the twentieth century, set out the axiom that explains why this matters: the likelihood that a government will tolerate an opposition increases as the expected costs of repression increase. External actors can raise or lower that cost. When an institution formally committed to democracy signals accommodation to an autocratizing regime, it lowers the cost of political repression. That is what the EU did. Because of its restrictions, an initiative meant to give the people a voice and strengthen democracy produced the opposite outcome.

Source: The Price of a Relationship, Introduction and Point 7 (7.1–7.3); V-Dem Institute, University of Gothenburg, V-Dem dataset v16 (2026), Regimes of the World; Robert A. Dahl, Polyarchy: Participation and Opposition (Yale University Press, 1971), pp. 15–16.

Takeaway 3

The case exposes the limits of EU accountability

Accountability failed at every level where it was meant to operate. The EU Delegation halted publication without explanation, and left DEMOS’s proposals for a way out unanswered. The European Commission maintained the restrictions for years through shifting justifications, and closed every avenue short of litigation, refusing even the mediation provided for in the contract. The European Ombudsman, the institution charged with holding EU institutions to account, opened an exceptional inquiry on human rights grounds.

Yet it did not examine the evidence before it in an independent, probing way. It accepted the Commission’s account as fact, without testing it against the contrary evidence or seeking its own legal analysis of the contracts, and left key evidence unexamined. Even the simplest question was never put: who took the decision, and why. It found one fault — late notice of the free-distribution condition — yet offered no remedy. The inquiry lacked diligence. The next three takeaways show it also lacked critical discernment.

Source: The Price of a Relationship, Part II (Executive Summary, Points 3–4; 3.1, 3.2 and 4.1).

Takeaway 4

Censorship can be explicit or veiled: oversight bodies must recognize it in both forms

Overt censorship is easy to see, and to condemn. The veiled forms — oblique, limited, by-product — present themselves as something else: a delay, a procedure, a condition of contract, a hidden consequence. They are no less damaging, only harder to name. This case moved from one to the other. An overt ban — “Do NOT go to print” (2021), any publication “strictly prohibited” (June 2022) — gave way to permission to publish in October 2022, and then, in March 2023, to a condition: free distribution only.

In essence, that meant no publication with an academic press: no publisher could recover its costs through sales, and no funds were provided for open access. The Ombudsman’s inquiry team probed this veiled restriction; the Decision found the condition reasonable. If an oversight body cannot see censorship when it changes form, can it see democratic erosion when it happens?

Source: Restrictions on Academic Freedom (definitions and “Using the matrix”); The Price of a Relationship, Points 2.4–2.5; vozesdopovo.org, page 2.

Takeaway 5

Public funding is not public access

The Commission argued that the study, produced with EU funds, is a “public good,” and may therefore only be given away, never sold. The Ombudsman accepted the claim without scrutiny. It is a category error. A good is public when no one can be excluded from it and one person’s use does not diminish what is available to others. An English edition kept from academic publishers, and from its readers, fails the first of these tests. A good is not made public by being publicly funded. It is made public by being publicly accessible. A commitment to free access does not require prohibiting every sale. The EU’s own open-science framework says as much.

In arguing otherwise, the Commission hollowed out the concept of the public interest. The genuine public interest lies in the widest possible circulation of knowledge — in its availability for the common deliberation on which democratic life depends. From different perspectives, Norberto Bobbio, Hannah Arendt and Jürgen Habermas laid the foundations for understanding that link: the visibility of power, a common world that gathers people together, and public debate. Yet in December 2025, the Commission made the sales ban an established rule for managing its external-action contracts. The rule that silenced this study is now taught as procedure.

Source: The Price of a Relationship, Executive Summary (Point 5) and Points 5.1–5.3; Norberto Bobbio, The Future of Democracy (1987); Hannah Arendt, The Human Condition (1958); Jürgen Habermas, The Structural Transformation of the Public Sphere (1989).

Takeaway 6

Concerns over fundamental freedoms can be dissolved

The Ombudsman’s Decision did not refute the case for academic freedom. It dissolved it. Meetings were held with one side only: the inquiry team met the Commission, but held no equivalent meeting with DEMOS, and no official who administered the contract or ordered the halt was questioned. The Commission’s answers to the Ombudsman’s four further questions informed the Decision; DEMOS never saw them.

The reasoning advanced through small shifts, always in the Commission’s favor. DEMOS’s textual analysis found that, of the 21 positions the Commission set out before the inquiry team, the Decision adopts 15 in the Ombudsman’s own voice, and rejects none. The words on which the case was opened — censorship, academic freedom, human rights — do not appear at all.

Source: The Devil Is in the Details, §1, Illustration I and §4, “The Pattern”; The Price of a Relationship, Points 3–4.

Takeaway 7

What harms democracy abroad harms democracy at home

Democratic erosion proceeds through what scholars have called “death by a thousand cuts” — the gradual accommodation of violations by the institutions designed to prevent them, through signals sent, inferences not drawn, questions not asked. The EU acquiesced in Guinea-Bissau’s autocratization. What is tolerated abroad does not stay abroad. In Strasbourg, the Ombudsman’s Decision normalized a class of conduct — the use of contractual power to suppress inconvenient knowledge — and sent a signal to EU officials. The Commission then wrote its rationale into guidance for its staff.

The case raises a question the current moment makes urgent: can EU accountability institutions be trusted to name democratic erosion when they encounter it — including when it originates in the conduct of EU officials themselves?

Source: The Price of a Relationship, Executive Summary (Point 8), Points 8.3–8.4 and Conclusion; Susan C. Stokes, The Backsliders (Princeton University Press, 2025).

What We Ask

The remedy DEMOS asks for is modest, and squarely within what the European Union already agreed to. First, bring the book to the people of Guinea-Bissau, including its diaspora. Second, let the English edition exist, at the Union’s cost. Third, safeguard academic freedom in EU-funded research.

Public funds should open knowledge to the public, never lock it away.

Source: The Price of a Relationship, Coda: A Path Forward; vozesdopovo.org, page 7.