Skip to content

← Thinking Clearly

The news

Chapter 27 · The phone

In this chapter 5 sections
  1. How a report gets made
  2. News, opinion and the headline
  3. Two sides, when there aren’t
  4. What never happens on a Tuesday
  5. Reading the news well

Here is a headline that would have been true every year for the last thirty-four: “Fewer young children died in Nepal this year than last.” Spread over the days, the fall came to about seven fewer deaths a day, on average, across that whole stretch. It’s about as good as news gets. Each of those children was someone’s son or daughter who lived past five. And it’s not the kind of thing anyone expects to see on a front page.

The German economist Max Roser, who founded a website called Our World in Data, once made the same point about the whole world. A newspaper, he wrote, could have run “The number of people in extreme poverty fell by 130,000 since yesterday” every single day since 1990, and been right on average every time. Good news that comes slowly, he added, never makes the headlines. The reason isn’t that reporters are lazy or gloomy. It has to do with what news is. Understanding that is the last piece of reading the news well, and it comes after three simpler ones: how a report gets made, what a headline does to it, and what “both sides” can hide.

How a report gets made

A news report is a strange and rather wonderful thing. Someone whose job it is goes to find out what happened. They go to the place, or ring the hospital, or read the court papers. They talk to the people who saw it and the people it happened to, and they ask the person accused what they have to say. Then an editor, who wasn’t there, reads the report and asks how they know each thing in it. Only then is it published, under the name of a paper that will look foolish, and may be sued, if it’s wrong.

Two American journalists, Bill Kovach and Tom Rosenstiel, wrote a book about what journalism is for, and they put the answer in one line: “the discipline of verification is what separates journalism from entertainment, propaganda, fiction, or art.” Checking is the whole trade. Nepal’s Journalist Code of Conduct, issued by the Press Council in 2016, says much the same. Its rules for news websites say that the authenticity of any news “should be verified before publishing,” and when a report turns out to be wrong, the paper should “rectify such error as soon as possible.”

None of this makes the news always right. A magazine bought Hitler’s fake diaries, and plenty of papers have printed things that weren’t so. But there is a world of difference between a report and a forward. The forward came from nobody in particular and answers to nobody. The report came from a named person at a named paper, who had to tell an editor how they knew, and whose paper will print a correction if it’s caught out. That is exactly why it was reporters at the Nepali Times, in December 2000, who found the film star’s interview and played the whole thing through.

So a news report is usually the most reliable thing on your phone about something that happened yesterday. The trouble lies in three other places.

News, opinion and the headline

A newspaper holds two quite different kinds of writing, and they are easy to mix up. A report tells you what happened: who, what, where, when, and how the paper knows. An opinion piece tells you what someone thinks about it, and argues for it. Both are honest work. An editorial, written by the paper itself, and a column by a professor or a politician are meant to persuade you, and they say so by where they are printed. In Kantipur they go under विचार (opinion), in The Kathmandu Post under Opinion. Nepal’s code asks papers to publish “news, articles and opinions with a clear distinction among these items,” and forbids presenting “an advertisement as news.”

The distinction matters because the questions you ask are different. Of a report you ask whether it’s accurate. Of an opinion you ask whether its reasons hold up, what it leaves out, and whether it treats the other side fairly. An opinion piece can be full of true facts and still argue for something wrong, and a reader who takes it for a report takes its conclusion as one of the facts.

The online version of the same mix-up is the advertisement dressed as an article. Researchers at Stanford showed American students the website of a news magazine with two articles about climate change, one a news story and one marked “sponsored content”, paid for by an oil company and carrying its logo. Nearly 70% of the more than 200 high school students judged the oil company’s piece the more reliable. It had a pie chart. When younger students were given a similar task, 82% failed to recognize an item clearly marked “sponsored content” as an advertisement. The label was there, and they didn’t know what it meant: somebody paid for this to look like news.

Then there is the headline, which sits on top of every kind of writing and usually isn’t written by the person who wrote the rest. As a columnist at The Irish Times once explained to her readers, “a reporter never writes the main headline.” An editor writes it later, to fit a space and to pull you in, in a handful of words. To fit, it has to pick one thing from the story. The thing most likely to pull you in is rarely the calmest.

Ullrich Ecker and his colleagues at the University of Western Australia tested what that does. They gave students news articles to read, the same article each time but with different headlines. One article said that burglaries had risen by 0.2% in the past year, and explained that this was a blip in a ten-year fall of about 10%. Some students saw it under “Downward trend in burglary rate” and others under “Number of burglaries going up.” Every student read the same words underneath. Yet afterward the students remembered better whichever trend their headline had named. With that article, at least, they still worked out correctly that burglaries would probably keep falling. With an opinion piece the headline worked differently. It didn’t change what readers remembered, but a headline saying genetically modified food “may pose long-term health risks” left them expecting bigger costs to health than the article argued for. The headline doesn’t only introduce a story. It tells you how to read it, and the effect survives reading the whole thing.

Many people never read the whole thing. In one month in 2015, a team at Columbia University and the French research institute Inria tracked links to articles from the BBC, CNN, Fox News, The New York Times and the Huffington Post as they were posted on Twitter. They estimated that about six in ten of those links were never clicked by anyone at all. For a lot of readers, the headline is the story.

Two sides, when there aren’t

Reporters are trained to be fair, and one way to be fair is to hear both sides. Say, to take a made-up case, a minister announces that a new bridge will open next month, and the engineer building it says it can’t be done in time. A good report gives both, and lets you weigh them. The code asks for the same: room for a reply, and “balanced opinion.”

But balance can go wrong in a way that is hard to see. Suppose ninety-nine experts who have studied a question agree, and one doesn’t. A report that gives each side one quotation and equal space hasn’t been fair. It has turned a hundred to one into one to one. This is called false balance, and the newspaper that falls into it can print nothing untrue and still leave you with a picture that is wrong.

The brothers Maxwell and Jules Boykoff measured it. They took a random sample of articles on global warming from 1988 to 2002 in four of America’s most respected papers, The New York Times, The Washington Post, the Los Angeles Times and The Wall Street Journal. In a little over half of the 340 articles that discussed whether people were causing the warming, the papers gave “roughly equal attention” to that view and to the view that natural changes alone could explain it. Among the scientists who studied climate, it wasn’t anything like half and half. The papers weren’t trying to mislead anyone. They were doing what fairness usually asks of them, in a place where it didn’t fit. The tobacco companies, earlier, had counted on exactly this habit.

The BBC, which takes impartiality as seriously as any news organization in the world, got it wrong too. In August 2017 one of its main radio programs interviewed a former British finance minister who leads a group doubting climate science, and he made false claims about climate change that the interviewer didn’t sufficiently challenge. In April 2018 the broadcasting regulator, Ofcom, ruled that the program had broken its rules on accuracy. In September 2018 the BBC’s head of news sent her staff a briefing that said, plainly, “Climate change has been a difficult subject for the BBC, and we get coverage of it wrong too often,” and: “As climate change is accepted as happening, you do not need a ‘denier’ to balance the debate.”

So the fair thing is weight according to the evidence, not equal time. When a report sets two people against each other, it’s worth asking whether each speaks for half of those who know, or one for nearly all of them and the other for nearly none.

What never happens on a Tuesday

Now the strangest part. Suppose every report is accurate, every opinion is labeled, every headline fair, and every side weighed rightly. The news can still leave you with a false picture of the world, simply because of what news is.

News is what happened today. A bus going off the road is news. A flood, a strike or a fire is news. A plane that crashed is news, and the thousands that landed are not. But a great deal of what matters most in a country happens a little each day, and so never on any particular day. A small child who doesn’t die this year, when ten years ago she might have, is not an event. Nobody can photograph it. Multiply her by thousands, over thirty years, and you have one of the biggest changes in Nepal’s history, and on no particular day was it news.

Hans Rosling spent his life on this. He was a Swedish doctor who had worked in Mozambique in the late 1970s and early 1980s, and he became famous for lectures in which the world’s health and wealth moved across the screen as animated charts. He once ended a talk by swallowing a sword. He and his family founded the Gapminder Foundation, and they asked people simple questions about how the world had changed. One was: in the last twenty years, has the share of the world’s people living in extreme poverty almost doubled, stayed about the same, or almost halved? It had almost halved. In one of their surveys, in rich countries, only 9% on average chose that answer. Rosling liked to point out that chimpanzees picking at random would have done better than most of the experts he quizzed, since they’d get a third right.

A smiling man with short reddish-brown hair and dark-rimmed glasses, in a black dinner jacket and bow tie, in front of a red wall printed with the word TIME.
Figure 27.1 Hans Rosling in 2012. He died in 2017, and his book Factfulness, written with his son and daughter-in-law, came out the next year. Photo: David Shankbone, CC BY 3.0

In that book, Factfulness, he named ten habits of mind that get in the way, which he called instincts. A few of them are this chapter’s subject. The negativity instinct: news of bad events is much more likely to reach us than news of slow improvement, so the world seems worse than it is. His remedy was to keep two things apart, how good things are and which way they’re going. Things can be bad and better at the same time. The fear instinct: frightening things grab our attention, but they are “not necessarily the most risky.” A plane crash is terrifying and rare. What makes something dangerous to you is how harmful it is multiplied by how often you meet it. And the gap instinct: a story about two groups with a gulf between them sticks in the mind, while most people usually sit somewhere in between.

He noticed one more thing that’s easy to miss. More bad news can mean the world is getting better at counting its troubles, not that it has more of them.

The real test of all this is your own head, so here are a few of Nepal’s numbers. For each one you can see where it stood at the start. Guess where it stood in 2024.

Try it

Guess Nepal’s Numbers

Three of Nepal’s numbers from the World Bank. Set your guess for 2024, then tap Show me to see the real value and the path it took.

Children who die before age 5, of every 1,000 born0501001501990200020102024guess2024: 251990: 138
Your guess
138
Real, 2024
?
Off by
–

Move the slider to where you think it stood in 2024, then tap Show me.

If your guesses were too gloomy, Rosling’s audiences would have kept you company. In 1990, about 138 of every 1,000 children born in Nepal died before their fifth birthday, roughly one in seven. By 2024 it was about 25, one in forty. In 1990 that meant about 99,000 small children a year, some 270 every day. In 2024 it meant about 14,000, some 39 a day. The difference is that headline that never gets written, added up year after year: on average about 2,500 fewer deaths each year than the year before, faster in the 1990s, slower lately as the numbers got small.

The other numbers tell the same kind of story. A baby born in Nepal in 1990 could expect to live to about 55, and one born in 2024 to about 71. Women had about five children each in 1990 and about two in 2024. Look at the shape of each line as well as where it ends. There is no year when it happened. It happened a little nearly every year.

And it is still bad as well as better. About 14,000 young children a year is still a great many children. Rosling’s point was never that everything is fine. It was that you can’t fix what is wrong if you don’t know what has worked.

Reading the news well

None of this is a reason to stop reading the news. The news is how you know what happened yesterday, and a world without reporters would be a world of forwards. It is a reason to read it as what it is.

When you open a piece, notice what kind it is: a report of what happened, an opinion arguing about it, or something somebody paid for. Read past the headline before deciding what the story says, because the headline was written by someone else to make you click. When two people are set against each other, ask how many who know stand behind each. And for the big questions, whether the country is getting healthier, richer, safer or better educated, don’t ask the news at all. The news can only tell you what happened today. Ask the numbers that are counted every year, by the census, the health surveys and the World Bank, and look at the line.

Then the morning paper and the slow story can sit side by side. The bus that went off the road was real, and so were the children who lived.

Found a mistake? Please tell us.