Public service broadcasting and political representation — what does the research show?
Journalists as a profession lean clearly toward the Greens and the Left, especially inside SVT/SR. That this automatically produces left-biased content, however, is not proven.
The claim
"SVT and Swedish Radio are left-biased — their journalists vote red and it shows on air."
What is supported
Kent Asp (JMG, University of Gothenburg) has measured journalists' party sympathies repeatedly since 1989. In Journalist 2011/2012 roughly 41 % of SVT journalists reported sympathy for the Greens — against about 7 % in the electorate. The Left and the Greens dominate; the Moderates and the Sweden Democrats are heavily under-represented compared to the population.
The pattern repeats in later surveys (Journalisten 2020, Novus/TU): journalists — especially in public service — vote Left/Green far more than the general population.
Janne Josefsson (SVT Uppdrag granskning) has repeatedly described an internal reluctance in the profession to scrutinise its own side — for example in the memoir Granskningen (2014) and in Swedish Radio interviews.
What is not supported
That party sympathy automatically translates into biased on-air content.
Kent Asp has also done content analyses (Swedish election campaigns 1979–2018). Some favouritism appears — but the effect is less clear-cut than the sympathy surveys, and varies by election, newsroom and topic. The Broadcasting Commission (Granskningsnämnden) upholds complaints against individual items every year, but systematic partiality is not statistically established.
That public service deliberately hides certain issues on party-political grounds.
News judgement is shaped by the profession — but also by events, competition and editorial routines. "Deliberate cover-up" requires evidence that the academic literature does not provide.
The honest conclusion
The Left/Green over-representation in surveys of the journalist profession — especially among public service respondents — is well documented. Three populations must be kept apart: journalists as a profession, employees at SVT/SR, and published programme content. The surveys above cover the first two; content analysis measures the third and yields less clear-cut results. No causal link between party sympathy and biased content is statistically established.
What this teaches us
Party sympathies in a profession and bias in published content are measured with different methods, populations and time periods. Survey data about the profession therefore cannot on its own evidence content bias; that requires content analysis. Measuring political bias is methodologically hard: results depend on the period chosen, the comparison norm and what counts as favouritism.
Sources
Explicitly excluded as sources: Flashback, Nordfront and anonymous blogs.
Updated 2026-07-08