Media bias

In my day job (subscription required), I say investors should not worry about the fall in the dollar. Such a view is obviously the consensus in equity and bond markets; if you were to look at just these in the last few months, you’d never guess the dollar were in trouble at all.

But it’s in sharp contrast to the language of newspaper stories. Plunge, slide, slump and even crisis are the descriptions of choice for the greenback.

This difference has set me thinking about the nature of media bias. Not obvious political bias, but slightly subtler selection biases, that can discolour our view of the world. Here are five such biases.

1) News meeting bias. A journalist who offers his editor the line “dollar falls a little but no-one cares” will get a boot up the jacksy. One who offers “dollar slump spells end of the world” is offering a story, even if it’s nonsense. In this way, there’s a selection bias towards exaggerated stories; Matthew Parris (my favourite of the old media columnists) touches on this point here. Most of us, most of the time, don’t care about most things. Newspapers don’t represent this fact.

2) Illusion of knowledge bias. Journalists are selected for their ability to exaggerate their knowledge, to create a story from a few facts, or less. This has two unfortunate consequences. First, ignorance and doubt gets under-reported. The fact that even experts know little about lots of important things is only reported, if at all, as a “crisis”, not as a normal state of affairs. Second, when journalists write comments and leaders, they overstate the case for state intervention, simply because they underweight the contrary argument – that governments cannot foresee the consequences of their actions.

3) Availability bias. Journalists report from where they are. Murders in Baghdad therefore get more headlines than murders in, say, the Ivory Coast. I don’t know many people are being killed there, but how bad do things have to be for people to seek refuge in Liberia? Equally, would you really have worked out from day-to-day reports that almost 100,000 people were killed in Algeria in the early 1990s, or millions in the Congo, where fighting might be about to increase again?

4) Anecdotes over statistics bias. Journalists prefer the vivid anecdote to statistical information that may (notwithstanding my previous post) be more representative. No journalist in history ever said of anything: "it's only a half standard deviation event - not worth reporting."  There are many ways in which this can mislead us – for example in the reporting of job losses. Here’s another example. A numerate news editor (an oxymoron) considering how much attention to give to murders in Baghdad could logically reason as follows: “The murder rate in Colombia or El Salvador is over 50 per 100,000. We don’t consider death rates of this magnitude important, because we haven’t reported them. If such a murder rate existed in Baghdad, with a population of 5.8 million, there’d be around 250 killings a month. So I’ll only report deaths in Baghdad if they rise above this rate.”

5) The party politics bias. Journalists still think party politicians are worth reporting. This gives exaggerated coverage to one narrow strand of political thinking – centrist managerialism – and insufficient coverage to more philosophically sophisticated ideas. You’d never guess from newspapers that there’s a millennia-old discipline known as political philosophy. And you’d never guess that two-fifths of the population, among them me, didn’t vote in the last general election.

I’ve made no effort at completeness here. No doubt there are other biases. My point is simply that there’s a trade-off between news and truth. And this exists not (just) because individual journalists are stupid, but because of the nature of journalism itself.

Comments

[Andrew]

I'd add a subset to point 3 - availability bias - relating to the language of an originating report. English language accounts are more likely to propagate faster and further than other language accounts. Which partly explains why France can be pretty brutal in their old African colonies, in a way America never could, and largely get away with it.