It's sometimes said that Johnson is in trouble because his erstwhile supporters in the media have turned against him. I've no interest in that claim, but it raises an overlooked issue: why does the media still have so much influence despite collapsing newspaper circulation?

For example, since 1997 the Daily Mail's circulation has fallen from 2.3m to under one million, and the Guardian's from 428,000 to 109,000. And although the Sun, Telegraph and Times no longer publicise their circulation, data from 2019-20 show falls of over 50%. Of course, digital readership has risen, but it's not clear that this sustains their political influence: you can read the Mail's sidebar of shame without succumbing to the paper's politics.

Such collapses in circulation, however, have not been matched by a decline in their influence. Nobody says the media is only one-third as powerful as it was in the 90s, although it would be if its influence had matched its sales.

Although it is easy to overstate this power - people can be right-wing idiots without the help of the Sun or Mail - it still exists. Two facts tell us this. One is the age-voting gradient. Older people consume much more legacy media (including the BBC) than others, and they vote right. In the 2019 election over-65s split 62%-18% Tories to Labour, whilst 25-34 year-olds went 23% Tory-55% Labour. The other is that In Liverpool, where the Sun is not sold, the Labour (and Remain) vote has been bigger than in otherwise comparable areas. As Mic Wright says:

Despite declining revenues and readership, the British press (and the broadcast media that hews closely to the agenda it sets) does not simply report the news, it creates it.

Why is this?

One explanation is that the legacy media still has a big comparative advantage. Granted, it is mostly useless at describing longer-term social changes and emergent processes or at scrutinizing legislation: think of the comparative lack of coverage of the police bill. But it is better than other sources at breaking Westminster bubble stories, such as those about the Downing Street parties. This fact alone gives it some influence over party politics.

But there might be more to it than this.

One possibility is that a belief in the power of the press is self-fulfilling. As Duncan Robinson says:

[The tabloids] maintain influence because they shape elite opinion - among ministers and the BBC - not because they speak for the people, who no longer buy them.

The media are like the Wizard of Oz, possessing power because people believe in that power. If a politician believes he can be made or broken by Murdoch or the Mail, he'll stay in their good books and so they will indeed have power.

Secondly, political careers have become shorter. David Cameron, for example, spent only 15 years in parliament: Thatcher and Wilson were MPs for longer than that before becoming PM. And several New Labour MPs who would now have accumulated to the experience to be "big beasts" have long left national politics, such as David Miliband, Ruth Kelly, Ed Balls or James Purnell. With parliamentary careers truncated, politicians can no longer build reputations through campaigning within the party (the old "rubber chicken" circuit) or through diligent backbench work. Which means they need quicker routes to prominence, and the media offer these.

Thirdly, people don't see politics as a discrete and separate activity with its own distinct criteria of excellence and failure. In fact, many politicians don't even realize what the purpose of politics is - the solution or amelioration of collective action problems. Instead, we watch politics in the way our nans used to watch the wrestling, cheering the good guys whilst trying to hit the baddies with their handbags.

This gives influence to the dumbed-down media. Back in 2018 John Humphrys said of Brexit - on one of the BBC's "flagship" programmes - that this is "all getting a wee bit technical and I'm sure people are fed up to the back teeth of all this talk of stuff most of us don't clearly understand." This revealed journalists' disdain for policy analysis if favour of the "drama" of Westminster shenanigans. If we think of politics as being about policy, such attitudes would render the media irrelevant. But with so many regarding politics as cheap entertainment, so it retains influence.

Fourthly, party membership has fallen. Labour has less than 500,000 members, only half as many as in the mid-1950s, whilst the Tories have around 200,000 compared to a peak of almost three million. This means people are much less likely to hear political news and opinion from doorstep canvassers or (more importantly I suspect) from friends and work colleagues. The media's influence has thus increased simply because an alternate source of influence has diminished. This is all the more true because people have limited attention. Few want to actively seek out information and ideas, preferring to consume it passively.

Fifthly, there's an ageing population. Older people are not just more likely to buy a newspaper (pdf) than young ones, but are also more likely to watch TV news: 72% of viewers of BBC One's news are aged 55 or more. And they are cut off from the effects of being in work - not only from material concerns about wages, working conditions and the state of the economy but also the influence of conversations with co-workers. This means media influence is greater, because a countervailing influence has disappeared - a fact especially true for the many older people who live lonely lives.

The ageing population less rooted in material conditions whilst being big media consumers makes it easier for the media to push the agenda of fanatics and hobbyists, be it Brexit or the culture wars. It might be no accident, therefore, that the salience of culture war issues has increased with the number of retired people.

I say all this for a simple reason. It's common enough to bemoan the influence of the media - and reasonable, given that insofar as this influence exists it is largely pernicious. What we should ask, though, is: what socio-economic factors sustain this influence? The answer, I fear, is not pretty.