Media sameness

Why do rival news outlets chasing the same audience end up sounding almost identical?

Put the evening bulletins of three competing networks side by side, or the front pages of two rival papers, and the overlap is uncanny: the same five stories, in nearly the same order, treated at nearly the same length and in nearly the same register. This is not what competition is supposed to look like. Competing firms are meant to differentiate. So why does competition for an audience push in the opposite direction?

Line the audience up along some dimension of taste — how much politics they want, how serious a tone, how much sport. Suppose everyone consumes whichever outlet sits nearest their own position, and suppose you are one of two outlets. Where should you stand?

Work it through from any starting position. If you sit at the one-third mark and your rival at two-thirds, each of you takes half. Now shift a little toward the middle. You keep everyone on your side — they still have nobody nearer — and you take a slice of the contested centre. Your rival's best reply is to do the same. The only place neither of you can profitably move from is the middle, side by side, which is where the process stops. This is Harold Hotelling's result, and it is exactly the model behind retail clustering: the reason two shoe shops end up on the same corner and the reason two bulletins end up with the same running order are one argument, applied once to physical distance and once to distance in taste.

That is the skeleton. Two further mechanisms thicken it, and both are more mundane than the model.

The first is shared inputs. Most outlets draw on the same wire services and the same handful of agencies for raw copy and footage. If the initial pool of what happened is common, the output cannot diverge much, whatever the editorial intent — and the smaller the newsroom, the more of the bulletin comes straight off the wire.

The second is shared criteria. Editors are not choosing from all events but applying a widely shared sense of what makes something newsworthy — scale, unexpectedness, proximity, a nameable person to attach it to, a fit with something already running. Johan Galtung and Mari Ruge set out this list of news values in the sixties, and it is largely a description of how journalists everywhere were trained. Apply the same filter to the same wire feed and identical bulletins are not a conspiracy, they are arithmetic.

On top of that sits pack behaviour. Ask what it costs an editor to miss a story every rival is running, and compare it with the cost of running a story that turns out to be thin along with everyone else. The first is a visible failure attributable to you, the second is a shared miss nobody is blamed for. When errors are judged against what peers did rather than against the facts, the safest position is the middle of the pack — the same reputational herding that shows up among fund managers and forecasters.

That last clause is not a quibble; it is the condition the whole result rests on, and it is the reason the picture has changed. Hotelling convergence needs the people at the extremes to accept a distant product rather than abandon the market. That was a fair description of an era with three television channels and one metropolitan daily — if the middle-of-the-road bulletin was all there was, you watched it.

Cheap distribution broke that. With effectively unlimited channels, the person at the end of the spectrum has somewhere else to go, so a central position no longer holds them. And a niche outlet does not need a plurality to survive — it needs a viable slice, which changes the optimisation entirely. The profitable move becomes staking out a distinctive position and holding it intensely, which is what partisan cable, newsletters and podcasts are. So media fragmentation is not a refutation of the model but the same model run with a different assumption about the audience's outside option.

Two honest qualifications. The convergence and fragmentation stories now run at once, at different levels: outlets have diverged in stance while converging harder than ever on format, story selection and rhythm, since the wires and the news values are unchanged. And Hotelling's own result is fragile in its details — with three or more competitors, or with meaningful price and quality differences, the tidy middle can dissolve. Treat it as a pressure that explains a tendency, not a theorem about newsrooms.

• Whether audience fragmentation genuinely reduces shared information or mainly redistributes it.
• How public-service broadcasters, funded without competing for the same margin, sit differently in this model.

Every term the collection defines is gathered in the glossary.