Timeline — what the law has done
Click a node (or use arrow keys) to read source and statute. The same law can both restrict certain speech and protect groups' freedom — that duality is marked with a scale.
The Freedom of the Press Act
The world's first freedom-of-the-press act is adopted in Sweden. Censorship of printed material is abolished and the principle of public access is introduced.
The measurement — how large is the space today?
Sweden · SOM 2021–2022
53%
agree with the statement:
”In today's political climate I cannot openly express my views because others might find them offensive.”
USA · Cato/YouGov 2020
62%
agree with a similar — but not identical — question:
”The political climate these days prevents me from saying things I believe because others might find them offensive.”
The wording is similar but not identical — direct percentage comparisons should be read as benchmarks, not exact differences. The Swedish measurement is conducted within the research project "The Open Society" at Uppsala University, funded by the Swedish Psychological Defence Agency and the Swedish Research Council — publicly funded research, not an opinion poll.
Who self-censors — and why
Share agreeing with the SOM question, by party sympathy 2021–2022 (ESO 2024:1). The differences between groups are large — part of why this chapter cannot be reduced to a single political side.
Source: ESO 2024:1 — Talande tystnad? (Widmalm m.fl.), bearbetning av SOM-data.
Researchers point to social sanction as the driver — the risk of losing a job, reputation or relationships. This is the measurable version of the often-used phrase "colder debate climate". The measurement does not support more than that.
The silence comes from many directions
Threats against elected officials and journalists
According to Brå's Politicians' Safety Survey (PTU), around one third of Sweden's elected officials were subjected to threats, hate or violence in a single year. The Journalist Survey (JMG) shows similar levels for editorial staff.
Brå PTU · JMG/Göteborgs universitet
Artistic self-censorship after threats
Lars Vilks lived for ten years under police protection after an image-related conflict with Islamist circles. Elisabeth Ohlson Wallin has faced threats and censorship after works perceived as blasphemous in a Christian context. Two entirely different conflicts — the same mechanism.
Konstnärsnämnden · Beckman
Silence in the public sector
Statskontoret and the Swedish National Audit Office have repeatedly reported that a substantial share of state and municipal employees hesitate to flag malpractice — out of fear of reprisals or career consequences. That too is freedom of expression — in the workplace.
Statskontoret · Riksrevisionen
Point: the space is shrinking for several groups, for several reasons — threats, legislation and social sanction. Not a one-sided story, and not a party issue.
The other side
Where the line should be drawn is an ongoing, legitimate debate — not settled. Two positions are presented here without being weighed against each other.
Self-censorship as protection
Holding back an opinion can be politeness, consideration or simply wise. Hate-crime laws — including hate speech against groups — are justified as protecting groups' freedom to live without threats. Unlike the US, Sweden does not have absolute freedom of expression.
Self-censorship as a democratic problem
Widmalm et al. argue that a substantial share of citizens routinely choosing silence out of fear is a democratic problem in itself — regardless of what they would otherwise have said. Public debate loses information about what voters actually think.
What the data does NOT say
- — Self-censorship is measured via survey (self-reporting). It captures perception, not necessarily actually silenced statements.
- — The SOM time series for this particular question is relatively short. Long-run trends should be interpreted with caution.
- — That the space has shrunk is documented as perception. The causes (law, threats, culture, social media) are partly interpretation and should be kept separate from the measurement.
- — International comparisons are based on similar, not identical, questions. Percentages should be read as benchmarks.