Media ownership, bias & political influence

Four distinctions this section refuses to collapse

Most public argument about media power fails because it treats these as the same thing. Keeping them apart is what makes the rest of the section usable by someone who disagrees with you.

1. 1

   Ownership is not control

   A parent company is not always the controlling party. Companies House persons-with-significant-control records are the evidence, and this section cites them by company number.

2. 2

   Editorial position is not bias

   Newspapers are legally entitled to campaign. An endorsement is an observable, dated fact. A "bias rating" is a judgement dressed as a measurement, and this site does not publish one.

3. 3

   Newspapers are not broadcasters

   Broadcasters are bound by Ofcom’s due impartiality rules. The press is not. Applying one standard to both is the most common error in this whole argument.

4. 4

   Reach is not trust is not influence

   Three different measurements that point in three different directions. The fastest-growing platforms are the least trusted, and the most trusted brands are not the most used.

Three ways media influence is supposed to work

These are the established mechanisms in political communication research. They are not equally well evidenced, and the one people argue about most is the weakest.

• Agenda setting

  Evidence: well established

  Which issues get sustained attention?

  Outlets cannot tell people what to think, but they largely determine what people are thinking about. The clearest UK example is measurable: during the referendum campaign, coverage of immigration more than tripled and it led 99 front pages, more than any other issue including the economy.

  Basis: KCL content analysis of the 2016 campaign; a large international political-communication literature going back to the 1970s.

• Framing

  Evidence: well established

  How is the issue described?

  The same fact can be reported as a cost or a contribution, a wave or an arrival, a warning or a scare. KCL found the word "scaremongering" used 737 times in referendum coverage, overwhelmingly to describe economic warnings — a frame that converts a forecast into an accusation.

  Basis: KCL content analysis; the study also found sovereignty referenced in almost 2,000 articles but law-making powers in only 6% of them.

• Endorsement and persuasion

  Evidence: weak

  Does telling readers how to vote change how they vote?

  This is the mechanism people mean when they say the press decides elections, and it is the one with the weakest support. Readers select outlets that already match their views, so a correlation between what a paper says and how its readers vote is exactly what you would expect if the paper changed nobody’s mind.

  Basis: No UK study establishes a reliable causal effect of newspaper endorsement on vote choice at national scale. In 2024, most of Fleet Street including the Sun backed Labour and Labour won; in 2019 most backed the Conservatives and the Conservatives won. A theory that predicts both is not doing any work.

The defensible position is narrow and worth stating precisely: there is strong evidence that UK media shape which issues are salient and how those issues are described, and no reliable evidence that endorsements deliver elections. Those two sentences are compatible, and most arguments about media power fail because they treat the first as proof of the second.