RULE FOUR

Step Back and Enjoy the View

The shortest-lived creatures on the Disc were mayflies, which barely make it through twenty-four hours. Two of the oldest zigzagged aimlessly over the waters of a trout stream, discussing history with some younger members of the evening hatching.

 “You don’t get the kind of sun now that you used to get,” said one of them.

 “You’re right there. We had proper sun in the good old hours. It were all yellow. None of this red stuff.”

 “It were higher, too.”

 “It was. You’re right.”

• Terry Pratchett, Reaper Man

The newspapers had an alarming message for Londoners in April 2018:

“London’s Murder Rate Is Higher Than New York’s for the First Time Ever!” The headlines played into a narrative of gangs gone wild. And if we ignore for a moment that the very definition of “murder” differs on either side of the Atlantic, this claim is also perfectly true. In February 2018, there were fourteen murders in New York City, but fifteen in London.1

But what should we conclude? Nothing.

We should conclude nothing because that pair of numbers alone tells us very little. If we want to understand what’s happening, we need to step back and take in a broader perspective.

Here are a few facts worth knowing about murders in London and New York. London had 184 murders in 1990, while New York suffered 2,262—more than ten times as many. It’s with that image in mind of New York as a murderous place that Londoners are alarmed at the idea that they might have become as rotten as the Big Apple. But London’s murder rate has fallen, not risen, since 1990. In 2017, there were 130 murders in London, including ten people killed in terrorist attacks. London was safe in 1990 and it’s a little bit safer today. As for New York, murders fell to 292 in 2017. That means New York is still more dangerous than London, but much, much safer than in 1990.

(We should really look at the murder rate per million people rather than the murder total, but the populations of New York City and London are similar, so let’s not worry about that.)

Now that New York is vastly safer, very occasionally it has a good month and London has a bad one, and New York’s monthly murder count dips below London’s. The thing about numbers is that over time, they do tend to go up and down a bit.*

So while the newspaper headlines are narrowly correct, they point us away from the truth rather than toward it: the news is good, not bad; London has become safer, not more dangerous; and London remains safer than the fast-improving New York. We get the real story only with context.

In 1965, two Norwegian social scientists, Johan Galtung and Mari Ruge, made a fascinating observation: What counts as “news” depends very much on the frequency with which we pay attention.2 If media outlets know most of their audience is checking in every day, or every few hours, they will naturally tell us the most attention-grabbing event that’s happened in that time.

Consider the financial news. There is a big difference between the rolling business coverage of Bloomberg TV, the daily rhythm of the newspaper the Financial Times (my employer), and the weekly take of The Economist, even if the three outlets have a similar interest in business, economics, and geopolitics. Bloomberg might pick up on sharp market moves over the past hour. The same moves won’t merit a mention in The Economist. Weekly, daily, hourly—the metronome of the news clock changes the very nature of what is news.

Now imagine a much slower rhythm of news: a twenty-five-year newspaper, say. What would the latest edition say? It would be packed with updates, some hopeful and some grim; it would describe the rise of China, the World Wide Web and smartphones, the emergence of al-Qaeda and the collapse of Lehman Brothers. There might be a small feature article on crime, noting that the murder count had fallen in London, but not nearly as much as in New York. Nobody would spare a syllable on the idea that London was experiencing a killing spree; such an observation could only make sense in a fast-twitch media outlet.

How about a fifty-year newspaper? Max Roser, a young economist who created the Our World in Data website, proposed that idea, inspired by Galtung and Ruge. Roser imagines a newspaper published in 1918, 1968, and 2018. Topics that seemed earth-shattering to the daily newspapers of the time might not be mentioned at all, while huge changes in the world would scream from the front pages.3

What would the front page of the fifty-year newspaper say in 2018? One possibility might be a story about something that didn’t happen: “Phew! World Avoids Nuclear Armageddon!” Readers of the 1968 newspaper would have read anxiously about how, over the previous three decades, the atomic bomb had been invented, developed, used on Japan with catastrophic effect, then superseded by vast arsenals of much more powerful hydrogen bombs, and how the superpowers had flirted with nuclear conflict repeatedly—in the Korean War, during the Cuban missile crisis, and more than once over Berlin. For a reader picking up a newspaper in 2018 for the first time since 1968, it would be big news that the Cold War had simply ended without a nuclear exchange of any kind—even if no daily newspaper would have been tempted in the meantime to run with the headline “No H-Bombs Dropped Today.”

Or perhaps the editors would splash with a story on climate change. Since early research on the greenhouse effect probably wouldn’t have merited a mention in the 1968 edition, the 2018 newspaper would have to start with an explanation of the basic problem: burning fossil fuels such as gas, oil, and coal turns out to alter the composition of the atmosphere in a way that helps it trap heat. (Headline: “Gah! Burning Coal Turns Out to Be a Terrible Idea!”) That explanation would be illustrated by an alarming graph showing the increase in global temperatures.

Climate change is a difficult thing to report over a short time horizon. On an annual basis global temperatures bounce up and down; you can find almost as many years when they have fallen as when they have risen—which is raw material for the manufacturing of doubt. The fifty-year newspaper, however, conveys the grim news clearly: temperatures have risen by about 0.75°C (1.35°F) since the 1960s, depending on exactly what temperature measure you look at and between which years.4 Alas, from the right perspective, the trend is clearly that the planet is heating up.

How about a hundred-year newspaper? The perspective changes again. Thinking about readers who last consulted a newspaper in 1918, you might decide to offer a leading story about the miracle of safe childhood: “Child Mortality Falls by a Factor of Eight!” Imagine a school set up to receive a hundred five-year-olds, randomly chosen at birth from around the world. In 1918, only sixty-eight children would have turned up for the first day of school; thirty-two would have died before reaching the age of five. This wasn’t some temporary catastrophe because of the terrible 1914–18 war, or the global influenza outbreak of 1918: in 1900 the statistic would have been even worse. Now ninety-six children show up safely for their first year in school; just four die before reaching school age. Remember, they’re selected from all over the world, including the poorest, most isolated, and most strife-torn of countries. That is astonishing progress.5

For a two-hundred-year newspaper, the editorial board might take yet another angle: “Most People Aren’t Poor!” There are still a lot of poor people, of course—between 600 million and 700 million now live in what we call extreme poverty, according to the World Bank’s definition of less than around $1.90 per day. That’s not far from one in ten of the world’s population. But in the early nineteenth century, almost everyone—nineteen people out of twenty—lived in that state of destitution. That’s wonderful progress, and it becomes apparent only if we step back and change our perspective.

So far I’ve talked about perspective mainly in terms of time. We can get useful context from other kinds of comparison, too.

Let’s return to our case study of income inequality from the previous chapter, where we learned there are many plausible ways to measure it—such as the 50/10 ratio, or the income share of the top 1 percent. What if we could produce some sort of composite measure that summarizes the whole of the income distribution? These composite measures exist, and we’ve already mentioned the most famous—the Gini coefficient, named after the early-twentieth-century Italian statistician Corrado Gini.

Like any other measure of inequality, the Gini coefficient doesn’t tell us everything. On a global scale, the coefficient has been falling—that is, incomes are becoming more equal. That’s because lots of previously very poor people, many in China and India, have become a lot better off—and in the mathematical calculations that go into the Gini coefficient, that outweighs inequality rising in the upper half of the income scale, with the very rich leaving the moderately well-to-do in their wake.6 No single number could communicate all that. But the Gini coefficient does elegantly reflect the experience of everyone across the income spectrum. Moving a dollar from a billionaire to a millionaire will not change the top 1 percent share of income, since that dollar stays in the hands of someone in the top 1 percent. But moving a dollar from a richer person to a poorer person, no matter how rich or poor either of them might be, will reduce the Gini coefficient.

One big problem with the Gini coefficient, however, is getting an intuitive feel for what it actually means. It’s easy enough to picture a country with a Gini coefficient of zero—there, everyone gets exactly the same income. Likewise, we can readily imagine a country in which the Gini coefficient is 100 percent—there, the despotic president has cornered all the income and everyone else gets precisely nothing. But what would it be like to live in a country where the Gini coefficient of income is, say, 34 percent?

As it happens, if you live in the UK, you can answer that question.7 But even a specialist in income distribution would probably understand a Gini coefficient of 34 percent only in reference to the Gini coefficients of other countries. It’s 50 in China, for example, 42 in the United States, 25 in Finland. Globally, including everyone who lives in the poorest sub-Saharan nations and the richest petrostates, the Gini coefficient of income is 65 percent, higher than in any individual nation.8

But we can get an even better intuitive feel for what the Gini coefficient means by doing the same calculation on things other than income. Take life itself. Like income, life is unequally distributed. Some babies die almost immediately after being born; others live for a hundred years. But these extremes are relatively unusual: most people live for at least sixty years, and few live for more than ninety. So we would expect the global Gini coefficient of life expectancy to be fairly low, and it is—less than 20 percent.

How about the height of adults? We all have an intuitive sense of how little that varies, so it can be another useful reference point. If my back-of-the-envelope calculation is correct, the Gini coefficient is less than 5 percent.

For a newspaper column, I once calculated the Gini coefficient of recent sexual activity in the UK among thirty-five- to forty-four-year-olds. I know you’re curious: it’s 58 percent, much higher than the UK income Gini of 34 percent.9 Should we be surprised that it is higher than the income Gini? I’m not sure. But it is. It seems that a tenfold gap in sexual activity—with one person having sex once a month, and another having sex ten times a month—is far more common than a tenfold gap in income. A tenfold gap in longevity—a centenarian and a child who dies at the age of ten—is thankfully rarer still. A tenfold gap in adult height? Unheard of, even in the record books.

Another way to step back and enjoy the view is to give yourself a sense of scale. Faced with a statistic, simply ask yourself, “Is that a big number?” The creators of More or Less, Michael Blastland and Andrew Dilnot, made a habit of asking this unassuming but powerful question.10

Take, for example, the claim that Donald Trump’s border wall between the United States and Mexico would cost $25 billion to build. Is that a big number? It certainly sounds biggish, but to really understand the number you need something to compare it with. For example, the US defense budget is a little under $700 billion, or $2 billion a day. The wall would fund about two weeks of US military operations. Or, alternatively, the wall would cost about $75 a person: there are about 325 million people in the United States, and $25 billion divided by 325 million is about $75.* Big number? Small number? You can be the judge of that, but I’m guessing your judgment will be better informed having made these comparisons.

Andrew Elliott—an entrepreneur who likes the question so much he published a book with the title Is That a Big Number?—suggests that we should all carry a few “landmark numbers” in our heads to allow easy comparison.11 A few examples:

Personally, I like to carry a few of these numbers around in my head. I’m a geek that way. And I find that the more landmarks I have, the more sense all the other landmarks make. But the truth is that we don’t have to remember any of these numbers. We can look any of them up, from any number of reputable sources, using any reference book or internet connection—and in many cases it will be worth double-checking anyway.

Once we have some landmark numbers to hand, they’re easy to use. You can compare one thing with another (this 10,000-word report seems long, but an ordinary novel is ten times longer), or you can divide one thing by another (the US defense budget is more than $2,000 per American, per year). Memorize, or look up, some handy numbers, and then do some simple arithmetic—with a calculator if you want. It isn’t hard. But it is remarkably illuminating.*

It would be nice if we didn’t have to do this—if we could rely on the media outlets that present us with statistics to also helpfully provide all the context and perspective we need to make sense of them. The better ones will indeed try to do this. But context and perspective are never going to be on the front page above the fold.

We’ve seen one reason for this: our frequency of engagement. The splash of a daily newspaper, the lead story on a TV bulletin, and the top item on a website will all focus on the most dramatic, engaging, and significant events, since the typical news consumer will last have checked in a few hours previously. Some media critics believe there is another reason media outlets don’t emphasize context and perspective: people are drawn to bad news. Hans Rosling, coauthor of Factfulness and a wonderful campaigner for more realistic views of the world based on good data, calls this “the negativity instinct.” And it’s generally easier to make news seem bad if you omit the context.

I’m cautious about the idea that we’re biased toward bad news, because in general we tend to be rather optimistic; psychologist Tali Sharot reckons that 80 percent of us suffer from an “optimism bias,” systematically overestimating our longevity, our career prospects, and our talents while being blind to the risk of illness, incompetence, or divorce.12 Daniel Kahneman, Nobel laureate and one of the fathers of behavioral economics, calls overconfidence “the most significant of the cognitive biases.”13 In many ways we humans are actually pretty positive creatures—perhaps a little too positive, sometimes.

A more plausible explanation is that we are drawn to surprising news, and surprising news is more often bad than good.14 If media outlets had a bias merely toward the negative, one might expect them to report regularly on, say, smoking-related deaths. Ten times as many US residents died from smoking-related diseases as from terrorism in September 2001, the month that saw the most deadly terrorist attack in the country’s history.15 Even a weekly magazine could honestly have noted at the end of that terrible week that cigarettes had killed more people than al-Qaeda. The newspapers ignored the deaths from cigarettes because they had a bias toward the shocking.

It’s possible, of course, for shocking news to be positive. But the psychologist Steven Pinker has argued that good news tends to unfold slowly, while bad news is often more sudden.16 That sounds right—it is, after all, quicker to knock something down than to build it. Following a thought experiment the great psychologist Amos Tversky once shared with a young Pinker,17 imagine the best possible thing that could happen to you today. You could win the lottery, I suppose. (Would that really be good news?) There are certain other moments where something wonderful could happen: you could have been hoping for a baby after many months of fruitless trying, and finally the pregnancy test comes back positive; you might have applied for a promotion or a place at university, and you get it. But for most people, on most days, the possibility of some dramatic and surprising life improvement is fairly limited. Life is already good for many people; when life isn’t good, it is likely to improve slowly rather than thanks to some sudden miracle.

But the possibility of some dramatic turn for the worse? That’s easy to imagine. You, or a loved one, could be diagnosed with cancer, hit by a truck, or violently assaulted. Your house could be burgled, or it could burn down. You could be sacked from your job. You could be accused of a crime you didn’t commit. You could discover that your partner is having an affair, or wants a divorce. I didn’t have to think hard to reel those ideas off, and I’m sure you could add more without breaking sweat—or perhaps the cold sweat would break all too swiftly. The list of catastrophes could go on indefinitely.

So when media outlets want to grab our attention, they look for stories that are novel and unexpected over a short time horizon—and these stories are more likely to be bad than good.

The need to grab attention also skews the tactics of politicians, charities, and other campaigners. They know that if they want to get into the headlines, they need to make surprising claims. For example, in May 2015 the British media published the alarming news that strokes were on the increase in middle-aged people; this conclusion was based on official statistics highlighted by the Stroke Association, whose chief executive commented, “There is an alarming increase in the numbers of people having a stroke in working age.”18 Fortunately, this is incorrect. Strokes are becoming rarer, thanks to improved diet, better treatment, and public awareness campaigns; but those same public awareness campaigns encouraged people to present themselves at hospital at the first sign of a minor stroke. As a result, hospital admissions for strokes in younger people increased—or “rocketed,” as the Stroke Association put it—and the Stroke Association was on the story. The good news is that the incidence of stroke in the UK has for a long time been falling steadily and substantially across most age groups. But how could the Stroke Association be noticed with a story like that? And if it isn’t noticed, it can’t raise money.

Or consider Oxfam’s lament, late in 2016, that “the highly successful fight against global poverty is being lost badly in one critical area—people’s minds. A new global survey . . . reveals that 87% of people around the world believe that global poverty has either stayed the same or gotten worse over the past 20 years, when the exact opposite is true—it has more than halved.”19 This press release didn’t win nearly as much attention as the one we discussed in the previous chapter, which said that eighty-five people (or was it eight?) owned as much wealth as half the world (or was it everyone else?). When the alarmist press releases get the headlines, no wonder people think the plight of the world is getting worse.

In the UK, people are not hugely worried about issues such as immigration, teenage pregnancy, crime, and unemployment in their own areas—but they are profoundly anxious about these issues in the country as a whole. Similar results emerge if you ask people about their personal job situation versus their view of their country’s economy: most people think that all is well for them personally but are worried about the society they live in.20 Presumably this is because we personally experience our own localities, but we rely on the news for information about the wider world. The “negativity instinct” may not be a driver of negativity in news coverage, but it certainly seems to be a result.

In 1993, Martyn Lewis, then the most popular news anchor in the UK, argued that the media should spend more time covering good news stories.21 He was sneered at by fellow journalists, who caricatured his argument as simply a request for more cheery “And finally . . .” stories of skateboarding dogs, slotted in at the end of a news program to sprinkle a little sugar over the evening bulletin’s bitter offerings. This was unfair;* Lewis explicitly called for substantive good news stories rather than the precursors of today’s videos of cats surfing on Roombas.

“Good stories are there,” he wrote, “made all the more memorable by their rarity.” Happily, this is precisely wrong. Since Lewis wrote this in 1993, 154,000 people have escaped from extreme poverty every day.22 In 1980, the vaccines for illnesses such as measles, diphtheria, and polio were given to about 20 percent of one-year-olds. Eighty percent missed out. Recently at least 85 percent of one-year-olds received these vaccines, and the challenge is to maintain this in the coronavirus age.23 Child mortality, as we’ve seen, has fallen dramatically. The good stories are everywhere. They are not made memorable by their rarity; they are made forgettable by their ubiquity. Good things happen so often that they cannot seriously be considered for inclusion in a newspaper. “An Estimated 154,000 People Escaped from Poverty Yesterday!” True, but not news.

We don’t have daily updates on how many people escape poverty, and perhaps we never will. When I worked for the World Bank in 2004–05, we were still updating our estimates of extreme poverty only once every three years. If a newspaper decided to pick up on the story, fine, but that would be just a single story once every thousand days. No self-respecting newspaper would republish the story regularly to remind its readers, “Not news, but still true!” So the fall in the most extreme form of poverty—and dozens of other true stories we could tell about improved literacy, democracy, votes for women, education for girls, access to clean water, immunizations, agricultural yields, infant mortality, the price of solar power, the number of deaths in plane crashes, or the prevalence of hunger—goes unreported.24

It’s not just because it’s a happy story; it’s because the news comes at the wrong frequency. Gloomy stories that come at the wrong frequency are often ignored, too, as we’ve seen with smoking, the world’s most persistent, and thus most boring, cause of mass fatalities. Climate change is not ignored, but it is rarely reported directly; instead, the news covers deliberate attempts to get attention for it, such as protests, summits, and the occasional scientific or government report. We see it mentioned, infuriatingly, alongside reports on the weather—but we rarely see reports on slow-moving indicators such as the world’s rising temperature.

A third example is in finance. In 2004 and 2005, my Financial Times colleague Gillian Tett highlighted the development of huge financial markets in debt and derivatives, a kind of side bet on the movements of interest rates, exchange rates, and other financial indicators. The world financial system was like an iceberg: above the surface glistened the stock markets, easy to see and discuss; beneath the waves lurked the debt and derivatives markets, vast and hidden. Stock markets publish numbers continually, including a daily close-of-market update for the evening news. But one of the most important measures of the size of the derivatives market is produced by the Bank for International Settlements once every three years. The pace of information didn’t fit the frequency of the financial newspapers, and so it was systematically underreported. Of course, this was bad news worth being aware of: problems in these markets were at the center of the catastrophic financial crisis of 2007–08, and Gillian Tett was one of the few people who could honestly say she’d been paying attention beforehand.25

Some commentators argue that the cure for all this is simply to stop reading the newspapers. The author Rolf Dobelli—amusingly, writing in the Guardian newspaper—gives us ten reasons to stop reading the news.26 Nassim Taleb, author of The Black Swan, puts it succinctly: “To be completely cured of newspapers, spend a year reading the previous week’s newspapers.”27

Because I work for a newspaper, you might expect me to protest. I have a lot of sympathy, though. I often find that my Saturday Financial Times column is unmoored from the news of the week. I’m just not very interested in producing a hot take on recent news; I find my interest far more engaged by topics that have occurred to me after reading books or academic papers—or just musing about life. And while I enjoy the way that fans of More or Less often compare it favorably to rolling radio and TV news, I sometimes feel that we’re getting credit for something that comes naturally: we operate at a different rhythm from that of the rolling news. As a weekly program we usually have a couple of days to chew over something that has been said—or missed—in the blur of a live interview. Often, we find ourselves pondering a topic for weeks or months. Why cover a story quickly when you can explore it properly? And we don’t usually have to worry about being scooped because we’re far too nerdy for anyone else to care about our stories.

Professionally, I can’t ignore the news, but I pay less attention to it than many of my colleagues—occasionally to their frustration. Daily news always seems more informative than rolling news; weekly news is typically more informative than daily news. A book is often better still. Even within a daily or a weekly newspaper, I find myself preferring the slower-paced explanation and analysis rather than the breaking news.

If you’re a news junkie I suggest that you go deeper and broader, rather than faster and faster. It is harder to do this when the news itself seems to be alarming, but it’s a good habit. Very little news requires the immediate attention that you might devote to a traffic update or a severe weather warning. If you come back in an hour—or a week—you will learn just as much. Indeed, you’ll probably learn more. You might even ask yourself: What would a weekly magazine or a weekly podcast be paying attention to that might otherwise be drowned out in the noise of rolling news?

In the crazy early days when the COVID-19 coronavirus went global, Scientific American admonished journalists, “Facts about this epidemic that have lasted a few days are far more reliable than the latest ‘facts’ that have just come out, which may be erroneous or unrepresentative and thus misleading . . . a question that today can be answered [by] only informed belief may perhaps be answered with a fact tomorrow.”28 Sound advice, and not just for journalists but for citizens, too. So however much news you choose to read, make sure you spend time looking for longer-term, slower-paced information. You will notice things—good and bad—that others ignore.

What have we learned so far about how to evaluate a statistical claim? In the first chapter, I advised trying to notice your feelings about the claim; in the second chapter, constructively sense-checking the claim against your personal experience; in the third chapter, asking yourself if you really understand what the claim means. These are all simple, commonsense suggestions, and in this chapter I’ve added a fourth: Step back and look for information that can put the claim into context. Try to get a sense of the trend. “Another terrible crime has occurred!” is perfectly consistent with “Overall, crime is way down.”

Look for something that will give you a sense of scale, such as comparing the situation in one country with the situation in other countries, or figuring out the cost per person of some proposed government expenditure.

None of these methods is technical; anyone can use them. Together they can go a long way toward providing statistical illumination. But sometimes we need to dig a little deeper into how a statistic was produced. Let’s do that now.