There have been several good blog articles I have read recently on the subject of the media. Specifically on the issue of bias, how this manifests itself, what the effect is the population and how we should react to the current media landscape.
This has been a topic of some research and teaching for me, so I thought I could add some detail to some of the arguments made, and as per the name of the blog, focus in particular on the incentive structure for those involved.
Scott Alexander argues that the media rarely lie directly, but do distort the picture by selectively omitting or including certain facts, by the narrative they place around the issues they are presenting, and by the issues they choose to present. Zvi Mowshowitz writes in response about some rules to follow when reading media articles, making an interesting distinction between headlines which often do explicitly lie, and the body of an article where a direct lie might not be allowed, but very many other things are, in pursuit of a specific narrative.
What can some rigorous thinking about incentive structures add to this already excellent commentary?
From an economics perspective the media exists as a solution to acquire information. In a complex world people want information about the economy, investments and government policy, as well as titillation from gossip about celebrities and other unimportant but sensational news.
There is therefore a demand for media outlets, but what are the incentives of a media outlet? If they are a private firm like any other then the goal (to some extent at least) is to make money. How does a media organisation make money?
This can depend on the payment model. Do users directly pay, or are you funded by advertisers, or some mixture?
If users are paying directly then how do you get people to pay for your product? In an idealised world people want neutral information that allows them to make better decisions, but reality is very different. People want some facts, but they prefer facts that support their world view, make the people they support look good, and the people they don’t like look bad. We therefore get a media where most organisations have some degree of bias, if only for the reason that this is a profitable move. Try too hard to be neutral and you please few people and don’t make money.
Going back to the articles cited, the idea of a specific narrative that facts are fitted around in support of makes perfect sense. Media organisations have a direct incentive to sell this narrative to their readers who already believe the narrative and want confirmation this narrative is true.
On top of this people are drawn to sensational and surprising stories rather than the mundane. So the media will emphasise surprising but rare events and make these things seem more common than they really are (as per the bias of base rate neglect)
What is different if your business model is more reliant on advertising revenue?
Well in this case your incentive is to please your advertisers. In part you do this by appealing to readers in the same way as discussed already, more readers equals more value for advertisers. For online media your advertising revenue will be proportional to ‘clicks’ and this leads to the scourge of clickbait headlines, a jet-fuelled version of Mowshowitz’s rule that headlines can lie where article headlines are deliberately crafted for clicks and often contain almost no connection to the article content.
(I wrote about a particular example in my first article on this site:
An advertiser funded model also gives you an incentive to avoid news that will upset your advertisers. Evidence from Beattie et al (2017) shows that media downplay news on product safety recalls when they receive money from the company1.
In either payment model there will also be incentives to choose news that is more relevant to larger groups.
This can all be very different for media organisations that are funded fully, or subsidised by the state, like the BBC in the UK. Here the incentives are much harder to define. The BBC has a charter that is something vaguely to do with the public interest, but there is no way for the public to hold the BBC to account, whether you watch/listen to it or not, you pay for it all the same. The BBC will claim they try to remain neutral but this is really impossible and they are dogged by accusations of bias from both the left and right.
We may also question whether media organisations genuinely have a profit motive, or is something else going? This issue is normally raised in the context of state media as above, but also in the case of very wealthy media owners, who might be thought of subsiding their media reach in order to push a particular narrative.
All of this paints a fairly gloomy picture, but does it matter in the end? I would argue, no, not very much.
Why? Firstly there is a lot of media choice in countries like the US and the UK, and people generally select the media that fits their pre-existing bias. This begs the question of where this bias comes from in the first place, and there a lot of interesting ideas about how this relates to underlying moral frameworks, culture and genetics, but the media is not often changing people’s minds in very fundamental ways (at least not in the short-run). People have an incentive to choose the media that tells the story they want to hear and will select accordingly.
We might worry that if people have inaccurate information they do not vote wisely, but as well as the previous point we have the fact that voting just doesn’t matter very much anyway. I know some will disagree, and I would have a few years ago as well, but my current perspective is that voting is largely pointless, so who cares if some people are conned into voting the ‘wrong’ way, we end up with the same scumbags in charge either way!
Where I do worry is how media perceptions can poison your view of everyday life. We saw this most strongly during the COVID-era where people’s everyday experience of the world seemed to be trumped in so many cases by their perception from the media. If you had not followed the news how many people would truly have believed that we were living through a terrible crisis in which everyone was in mortal danger as was portrayed for two years or more?
In the end I therefore side with Bryan Caplan:
Better to ignore the media altogether if you can. Your mental health will likely be better and your view of the world will likely be more realistic based on the direct experience and information you get from everyday life than from trying to understand through the prism of media organisations.
Damn, I missed this when it came out. Excellent essay, as usual!
I am not so sure about the media changing people's minds in fundamental ways. I think that changing people's perceptions of what is real counts, as well as the function of telling people who agree with them generally what their opinions should be on matters they do not currently have an opinion about.
Your COVID example nicely covers both outcomes, I think. As you say, who would have thought there was a massive crisis if the media hadn't been there to tell us there was? Additionally, note how all "right minded people" changed on a dime with regards to masks, lockdowns, vaccines, etc to whatever their media outlet told them was the right answer. I doubt many people had preconceived opinions or biases on the matter, but they got new ones right quick.
As Caplan points out, the media lies all the time while telling you what to believe. That might not affect what you already believe, but most people are sane and don't spend a lot of time forming beliefs about things they don't have to ahead of time.
Also, editing point: Well in this case you need to please your incentive is to please your advertisers.
I think you need to remove the "is" there, so you can please your incentive to please your advertisers. Or, maybe you don't use "incentive" as a euphemism, so... maybe "you need to please" has to go. Either way.