Highlights

The Problem With Facts

1.

Just before Christmas 1953, the bosses of America’s leading tobacco companies met John Hill, the founder and chief executive of one of America’s leading public relations firms, Hill & Knowlton. Despite the impressive surroundings — the Plaza Hotel, overlooking Central Park in New York — the mood was one of crisis.

Scientists were publishing solid evidence of a link between smoking and cancer. From the viewpoint of Big Tobacco, more worrying was that the world’s most read publication, The Reader’s Digest, had already reported on this evidence in a 1952 article, “Cancer by the Carton”. The journalist Alistair Cooke, writing in 1954, predicted that the publication of the next big scientific study into smoking and cancer might finish off the industry.

It did not. PR guru John Hill had a plan — and the plan, with hindsight, proved tremendously effective. Despite the fact that its product was addictive and deadly, the tobacco industry was able to fend off regulation, litigation and the idea in the minds of many smokers that its products were fatal for decades.

So successful was Big Tobacco in postponing that day of reckoning that their tactics have been widely imitated ever since. They have also inspired a thriving corner of academia exploring how the trick was achieved. In 1995, Robert Proctor, a historian at Stanford University who has studied the tobacco case closely, coined the word “agnotology”. This is the study of how ignorance is deliberately produced; the entire field was started by Proctor’s observation of the tobacco industry. The facts about smoking — indisputable facts, from unquestionable sources — did not carry the day. The indisputable facts were disputed. The unquestionable sources were questioned. Facts, it turns out, are important, but facts are not enough to win this kind of argument.

2.

Agnotology has never been more important. “We live in a golden age of ignorance,” says Proctor today. “And Trump and Brexit are part of that.”

In the UK’s EU referendum, the Leave side pushed the false claim that the UK sent £350m a week to the EU. It is hard to think of a previous example in modern western politics of a campaign leading with a transparent untruth, maintaining it when refuted by independent experts, and going on to triumph anyway. That performance was soon to be eclipsed by Donald Trump, who offered wave upon shameless wave of demonstrable falsehood, only to be rewarded with the presidency. The Oxford Dictionaries declared “post-truth” the word of 2016. Facts just didn’t seem to matter any more.

The instinctive reaction from those of us who still care about the truth — journalists, academics and many ordinary citizens — has been to double down on the facts. Fact-checking organisations, such as Full Fact in the UK and PolitiFact in the US, evaluate prominent claims by politicians and journalists. I should confess a personal bias: I have served as a fact checker myself on the BBC radio programme More or Less, and I often rely on fact-checking websites. They judge what’s true rather than faithfully reporting both sides as a traditional journalist would. Public, transparent fact checking has become such a feature of today’s political reporting that it’s easy to forget it’s barely a decade old.

Mainstream journalists, too, are starting to embrace the idea that lies or errors should be prominently identified. Consider a story on the NPR website about Donald Trump’s speech to the CIA in January: “He falsely denied that he had ever criticised the agency, falsely inflated the crowd size at his inauguration on Friday . . . —” It’s a bracing departure from the norms of American journalism, but then President Trump has been a bracing departure from the norms of American politics.

Facebook has also drafted in the fact checkers, announcing a crackdown on the “fake news” stories that had become prominent on the network after the election. Facebook now allows users to report hoaxes. The site will send questionable headlines to independent fact checkers, flag discredited stories as “disputed”, and perhaps downgrade them in the algorithm that decides what each user sees when visiting the site.

We need some agreement about facts or the situation is hopeless. And yet: will this sudden focus on facts actually lead to a more informed electorate, better decisions, a renewed respect for the truth? The history of tobacco suggests not. The link between cigarettes and cancer was supported by the world’s leading medical scientists and, in 1964, the US surgeon general himself. The story was covered by well-trained journalists committed to the values of objectivity. Yet the tobacco lobbyists ran rings round them.

In the 1950s and 1960s, journalists had an excuse for their stumbles: the tobacco industry’s tactics were clever, complex and new. First, the industry appeared to engage, promising high-quality research into the issue. The public were assured that the best people were on the case. The second stage was to complicate the question and sow doubt: lung cancer might have any number of causes, after all. And wasn’t lung cancer, not cigarettes, what really mattered? Stage three was to undermine serious research and expertise. Autopsy reports would be dismissed as anecdotal, epidemiological work as merely statistical, and animal studies as irrelevant. Finally came normalisation: the industry would point out that the tobacco-cancer story was stale news. Couldn’t journalists find something new and interesting to say?

Such tactics are now well documented — and researchers have carefully examined the psychological tendencies they exploited. So we should be able to spot their re-emergence on the political battlefield.

“It’s as if the president’s team were using the tobacco industry’s playbook,” says Jon Christensen, a journalist turned professor at the University of California, Los Angeles, who wrote a notable study in 2008 of the way the tobacco industry tugged on the strings of journalistic tradition.

One infamous internal memo from the Brown & Williamson tobacco company, typed up in the summer of 1969, sets out the thinking very clearly: “Doubt is our product.” Why? Because doubt “is the best means of competing with the ‘body of fact’ that exists in the mind of the general public. It is also the means of establishing a controversy.” Big Tobacco’s mantra: keep the controversy alive.

Doubt is usually not hard to produce, and facts alone aren’t enough to dispel it. We should have learnt this lesson already; now we’re going to have to learn it all over again.

3.

Tempting as it is to fight lies with facts, there are three problems with that strategy. The first is that a simple untruth can beat off a complicated set of facts simply by being easier to understand and remember. When doubt prevails, people will often end up believing whatever sticks in the mind. In 1994, psychologists Hollyn Johnson and Colleen Seifert conducted an experiment in which people read an account of an explosive warehouse fire. The account mentioned petrol cans and paint but later explained that petrol and paint hadn’t been present at the scene after all. The experimental subjects, tested on their comprehension, recalled that paint wasn’t actually there. But when asked to explain facts about the fire (“why so much smoke?”), they would mention the paint. Lacking an alternative explanation, they fell back on a claim they had already acknowledged was wrong. Once we’ve heard an untrue claim, we can’t simply unhear it.

This should warn us not to let lie-and-rebuttal take over the news cycle. Several studies have shown that repeating a false claim, even in the context of debunking that claim, can make it stick. The myth-busting seems to work but then our memories fade and we remember only the myth. The myth, after all, was the thing that kept being repeated. In trying to dispel the falsehood, the endless rebuttals simply make the enchantment stronger.

With this in mind, consider the Leave campaign’s infamous bus-mounted claim: “We send the EU £350m a week.” Simple. Memorable. False. But how to rebut it? A typical effort from The Guardian newspaper was headlined, “Why Vote Leave’s £350m weekly EU cost claim is wrong”, repeating the claim before devoting hundreds of words to gnarly details and the dictionary definition of the word “send”. This sort of fact-checking article is invaluable to a fellow journalist who needs the issues set out and hyperlinked. But for an ordinary voter, the likely message would be: “You can’t trust politicians but we do seem to send a lot of money to the EU.” Doubt suited the Leave campaign just fine.

This is an inbuilt vulnerability of the fact-checking trade. Fact checkers are right to be particular, to cover all the details and to show their working out. But that’s why the fact-checking job can only be a part of ensuring that the truth is heard.

Andrew Lilico, a thoughtful proponent of leaving the EU, told me during the campaign that he wished the bus had displayed a more defensible figure, such as £240m. But Lilico now acknowledges that the false claim was the more effective one. “In cynical campaigning terms, the use of the £350m figure was perfect,” he says. “It created a trap that Remain campaigners kept insisting on jumping into again and again and again.”

Quite so. But not just Remain campaigners — fact-checking journalists too, myself included. The false claim was vastly more powerful than a true one would have been, not because it was bigger, but because everybody kept talking about it.

Proctor, the tobacco industry historian turned agnotologist, warns of a similar effect in the US: “Fact checkers can become Trump’s poodle, running around like an errand boy checking someone else’s facts. If all your time is [spent] checking someone else’s facts, then what are you doing?”

4.

There’s a second reason why facts don’t seem to have the traction that one might hope. Facts can be boring. The world is full of things to pay attention to, from reality TV to your argumentative children, from a friend’s Instagram to a tax bill. Why bother with anything so tedious as facts?

Last year, three researchers — Seth Flaxman, Sharad Goel and Justin Rao — published a study of how people read news online. The study was, on the face of it, an inquiry into the polarisation of news sources. The researchers began with data from 1.2 million internet users but ended up examining only 50,000. Why? Because only 4 per cent of the sample read enough serious news to be worth including in such a study. (The hurdle was 10 articles and two opinion pieces over three months.) Many commentators worry that we’re segregating ourselves in ideological bubbles, exposed only to the views of those who think the same way we do. There’s something in that concern. But for 96 per cent of these web surfers the bubble that mattered wasn’t liberal or conservative, it was: “Don’t bother with the news.”

In the war of ideas, boredom and distraction are powerful weapons. A recent study of Chinese propaganda examined the tactics of the paid pro-government hacks (known as the “50 cent army”, after the amount contributors were alleged to be paid per post) who put comments on social media. The researchers, Gary King, Jennifer Pan and Margaret Roberts, conclude: “Almost none of the Chinese government’s 50c party posts engage in debate or argument of any kind . . . they seem to avoid controversial issues entirely . . . the strategic objective of the regime is to distract and redirect public attention.”

Trump, a reality TV star, knows the value of an entertaining distraction: simply pick a fight with Megyn Kelly, The New York Times or even Arnold Schwarzenegger. Isn’t that more eye-catching than a discussion of healthcare reform?

The tobacco industry also understood this point, although it took a more highbrow approach to generating distractions. “Do you know about Stanley Prusiner?” asks Proctor.

Prusiner is a neurologist. In 1972, he was a young researcher who’d just encountered a patient suffering from Creutzfeldt-Jakob disease. It was a dreadful degenerative condition then thought to be caused by a slow-acting virus. After many years of study, Prusiner concluded that the disease was caused instead, unprecedentedly, by a kind of rogue protein. The idea seemed absurd to most experts at the time, and Prusiner’s career began to founder. Promotions and research grants dried up. But Prusiner received a source of private-sector funding that enabled him to continue his work. He was eventually vindicated in the most spectacular way possible: with a Nobel Prize in Medicine in 1997. In his autobiographical essay on the Nobel Prize website, Prusiner thanked his private-sector benefactors for their “crucial” support: RJ Reynolds, maker of Camel cigarettes.

The tobacco industry was a generous source of research funds, and Prusiner wasn’t the only scientist to receive both tobacco funding and a Nobel Prize. Proctor reckons at least 10 Nobel laureates are in that position. To be clear, this wasn’t an attempt at bribery. In Proctor’s view, it was far more subtle. “The tobacco industry was the leading funder of research into genetics, viruses, immunology, air pollution,” says Proctor. Almost anything, in short, except tobacco. “It was a massive ‘distraction research’ project.” The funding helped position Big Tobacco as a public-spirited industry but Proctor considers its main purpose was to produce interesting new speculative science. Creutzfeldt-Jakob disease may be rare, but it was exciting news. Smoking-related diseases such as lung cancer and heart disease aren’t news at all.

The endgame of these distractions is that matters of vital importance become too boring to bother reporting. Proctor describes it as “the opposite of terrorism: trivialism”. Terrorism provokes a huge media reaction; smoking does not. Yet, according to the US Centers for Disease Control, smoking kills 480,000 Americans a year. This is more than 50 deaths an hour. Terrorists have rarely managed to kill that many Americans in an entire year. But the terrorists succeed in grabbing the headlines; the trivialists succeed in avoiding them.

Tobacco industry lobbyists became well-practised at persuading the media to withhold or downplay stories about the dangers of cigarettes. “That record is scratched,” they’d say. Hadn’t we heard such things before?

Experienced tobacco watchers now worry that Trump may achieve the same effect. In the end, will people simply start to yawn at the spectacle? Jon Christensen, at UCLA, says: “I think it’s the most frightening prospect.”

On the other hand, says Christensen, there is one saving grace. It is almost impossible for the US president not to be news. The tobacco lobby, like the Chinese government, proved highly adept at pointing the spotlight elsewhere. There are reasons to believe that will be difficult for Trump.

5.

There’s a final problem with trying to persuade people by giving them facts: the truth can feel threatening, and threatening people tends to backfire. “People respond in the opposite direction,” says Jason Reifler, a political scientist at Exeter University. This “backfire effect” is now the focus of several researchers, including Reifler and his colleague Brendan Nyhan of Dartmouth.

In one study, conducted in 2011, Nyhan, Reifler and others ran a randomised trial in which parents with young children were either shown or not shown scientific information debunking an imaginary but widely feared link between vaccines and autism. At first glance, the facts were persuasive: parents who saw the myth-busting science were less likely to believe that the vaccine could cause autism. But parents who were already wary of vaccines were actually less likely to say they’d vaccinate their children after being exposed to the facts — despite apparently believing those facts.

What’s going on? “People accept the corrective information but then resist in other ways,” says Reifler. A person who feels anxious about vaccination will subconsciously push back by summoning to mind all the other reasons why they feel vaccination is a bad idea. The fear of autism might recede, but all the other fears are stronger than before.

It’s easy to see how this might play out in a political campaign. Say you’re worried that the UK will soon be swamped by Turkish immigrants because a Brexit campaigner has told you (falsely) that Turkey will soon join the EU. A fact checker can explain that no Turkish entry is likely in the foreseeable future. Reifler’s research suggests that you’ll accept the narrow fact that Turkey is not about to join the EU. But you’ll also summon to mind all sorts of other anxieties: immigration, loss of control, the proximity of Turkey to Syria’s war and to Isis, terrorism and so on. The original lie has been disproved, yet its seductive magic lingers.

The problem here is that while we like to think of ourselves as rational beings, our rationality didn’t just evolve to solve practical problems, such as building an elephant trap, but to navigate social situations. We need to keep others on our side. Practical reasoning is often less about figuring out what’s true, and more about staying in the right tribe.

An early indicator of how tribal our logic can be was a study conducted in 1954 by Albert Hastorf, a psychologist at Dartmouth, and Hadley Cantril, his counterpart at Princeton. Hastorf and Cantril screened footage of a game of American football between the two college teams. It had been a rough game. One quarterback had suffered a broken leg. Hastorf and Cantril asked their students to tot up the fouls and assess their severity. The Dartmouth students tended to overlook Dartmouth fouls but were quick to pick up on the sins of the Princeton players. The Princeton students had the opposite inclination. They concluded that, despite being shown the same footage, the Dartmouth and Princeton students didn’t really see the same events. Each student had his own perception, closely shaped by his tribal loyalties. The title of the research paper was “They Saw a Game”.

A more recent study revisited the same idea in the context of political tribes. The researchers showed students footage of a demonstration and spun a yarn about what it was about. Some students were told it was a protest by gay-rights protesters outside an army recruitment office against the military’s (then) policy of “don’t ask, don’t tell”. Others were told that it was an anti-abortion protest in front of an abortion clinic.

Despite looking at exactly the same footage, the experimental subjects had sharply different views of what was happening — views that were shaped by their political loyalties. Liberal students were relaxed about the behaviour of people they thought were gay-rights protesters but worried about what the pro-life protesters were doing; conservative students took the opposite view. As with “They Saw a Game”, this disagreement was not about the general principles but about specifics: did the protesters scream at bystanders? Did they block access to the building? We see what we want to see — and we reject the facts that threaten our sense of who we are.

When we reach the conclusion that we want to reach, we’re engaging in “motivated reasoning”. Motivated reasoning was a powerful ally of the tobacco industry. If you’re addicted to a product, and many scientists tell you it’s deadly, but the tobacco lobby tells you that more research is needed, what would you like to believe? Christensen’s study of the tobacco public relations campaign revealed that the industry often got a sympathetic hearing in the press because many journalists were smokers. These journalists desperately wanted to believe their habit was benign, making them ideal messengers for the industry.

Even in a debate polluted by motivated reasoning, one might expect that facts will help. Not necessarily: when we hear facts that challenge us, we selectively amplify what suits us, ignore what does not, and reinterpret whatever we can. More facts mean more grist to the motivated reasoning mill. The French dramatist Molière once wrote: “A learned fool is more foolish than an ignorant one.” Modern social science agrees.

On a politically charged issue such as climate change, it feels as though providing accurate information about the science should bring people together. The opposite is true, says Dan Kahan, a law and psychology professor at Yale and one of the researchers on the study into perceptions of a political protest. Kahan writes: “Groups with opposing values often become more polarised, not less, when exposed to scientifically sound information.”

When people are seeking the truth, facts help. But when people are selectively reasoning about their political identity, the facts can backfire.

6.

All this adds up to a depressing picture for those of us who aren’t ready to live in a post-truth world. Facts, it seems, are toothless. Trying to refute a bold, memorable lie with a fiddly set of facts can often serve to reinforce the myth. Important truths are often stale and dull, and it is easy to manufacture new, more engaging claims. And giving people more facts can backfire, as those facts provoke a defensive reaction in someone who badly wants to stick to their existing world view. “This is dark stuff,” says Reifler. “We’re in a pretty scary and dark time.”

Is there an answer? Perhaps there is.

We know that scientific literacy can actually widen the gap between different political tribes on issues such as climate change — that is, well-informed liberals and well-informed conservatives are further apart in their views than liberals and conservatives who know little about the science. But a new research paper from Dan Kahan, Asheley Landrum, Katie Carpenter, Laura Helft and Kathleen Hall Jamieson explores the role not of scientific literacy but of scientific curiosity.

The researchers measured scientific curiosity by asking their experimental subjects a variety of questions about their hobbies and interests. The subjects were offered a choice of websites to read for a comprehension test. Some went for ESPN, some for Yahoo Finance, but those who chose Science were demonstrating scientific curiosity. Scientifically curious people were also happier to watch science documentaries than celebrity gossip TV shows. As one might expect, there’s a correlation between scientific knowledge and scientific curiosity, but the two measures are distinct.

What Kahan and his colleagues found, to their surprise, was that while politically motivated reasoning trumps scientific knowledge, “politically motivated reasoning . . . appears to be negated by science curiosity”. Scientifically literate people, remember, were more likely to be polarised in their answers to politically charged scientific questions. But scientifically curious people were not. Curiosity brought people together in a way that mere facts did not. The researchers muse that curious people have an extra reason to seek out the facts: “To experience the pleasure of contemplating surprising insights into how the world works.”

So how can we encourage curiosity? It’s hard to make banking reform or the reversibility of Article 50 more engaging than football, Game of Thrones or baking cakes. But it does seem to be what’s called for. “We need to bring people into the story, into the human narratives of science, to show people how science works,” says Christensen.

We journalists and policy wonks can’t force anyone to pay attention to the facts. We have to find a way to make people want to seek them out. Curiosity is the seed from which sensible democratic decisions can grow. It seems to be one of the only cures for politically motivated reasoning but it’s also, into the bargain, the cure for a society where most people just don’t pay attention to the news because they find it boring or confusing.

What we need is a Carl Sagan or David Attenborough of social science — somebody who can create a sense of wonder and fascination not just at the structure of the solar system or struggles of life in a tropical rainforest, but at the workings of our own civilisation: health, migration, finance, education and diplomacy.

One candidate would have been Swedish doctor and statistician Hans Rosling, who died in February. He reached an astonishingly wide audience with what were, at their heart, simply presentations of official data from the likes of the World Bank.

He characterised his task as telling people the facts — “to describe the world”. But the facts need a champion. Facts rarely stand up for themselves — they need someone to make us care about them, to make us curious. That’s what Rosling did. And faced with the apocalyptic possibility of a world where the facts don’t matter, that is the example we must follow.

Written for and first published in the Financial Times.

My book “Messy” is available online in the US and UK or in good bookshops everywhere.

What makes the perfect office?

In 1923, the father of modern architecture, Le Corbusier, was commissioned by a French industrialist to design some homes for workers in his factory near Bordeaux. Le Corbusier duly delivered brightly-hued concrete blocks of pure modernism. The humble factory workers did not take to Le Corbusier’s visionary geometry. They added rustic shutters, pitched roofs, and picket-fenced gardens. And they decorated the gardens in the least modernist way imaginable: with gnomes.
Companies no longer hire star architects to design housing for an industrial workforce. The architects are instead put to work producing the most magazine-shoot worthy office spaces. A pioneer was the uber-cool advertising agency, Chiat-Day, which in 1993 hired the playful Italian architect Gaetano Pesce to create a New York space for them (hot-lips mural, luminous floor, spring-loaded chairs). Their Los Angeles office (four-storey binoculars, brainstorming pods commandeered from fairground rides) was designed by Frank Gehry, whom Chiat-Day’s boss, Jay Chiat, had spotted before Gehry created the Guggenheim Bilbao and became the most famous architect on the planet.
Jay Chiat believed that design was for the professionals. Give workers control over their own space and they would simply clutter up Frank Gehry’s vision, so Jay Chiat decreed that his employees be given tiny lockers for “their dog pictures, or whatever”.
Now everyone is hiring the high priests of architecture. Google has asked Thomas Heatherwick, creator of the 2012 Olympic torch, to create a new Googleplex. Apple’s new headquarters will be a gigantic glass donut over a mile around, designed by Norman Foster and partners.
The most famous corporate architect was not an architect at all: the late Steve Jobs, the boss of Apple, owned much of the film studio Pixar and stamped his taste all over Pixar’s headquarters. Jobs pored over the finest details, choosing an Arkansas steel mill that produced steels of the perfect hue (bolted, not welded).
Jobs believed that a building could shape the way people interacted with each other, and hit upon the notion that Pixar would have just a single pair of washrooms, just off the main lobby. Every time nature called, there was only one place for the entire company to go, and serendipitous new connections would be made.
But what if all these efforts are basically repeating Le Corbusier’s error? What if the ideal office isn’t the coolest or the most aesthetically visionary? What if the ideal office is the one, dog pictures and gnomes and all, that workers make their own?
In 2010, two psychologists conducted an experiment to test that idea. Alex Haslam and Craig Knight set up simple office spaces where they asked experimental subjects to spend an hour doing simple administrative tasks. Haslam and Knight wanted to understand what sort of office space made people productive and happy, and they tested four different layouts.
Two of the layouts were familiar. One was stripped down – bare desk, swivel chair, pencil, paper, nothing else. Most participants found it rather oppressive. “You couldn’t relax in it,” said one. The other layout was softened with pot plants and tasteful close-up photographs of flowers, faintly reminiscent of Georgia O’Keefe paintings. Workers got more and better work done there, and enjoyed themselves more.
The next two layouts produced dramatically different outcomes – and yet, photographs of the spaces would offer few clues as to why. They used the same basic elements and the same botanical decorations. But the appearance wasn’t what mattered; what mattered was who got to decide.
In the third and fourth layouts, workers were given the plants and pictures and invited to use them to decorate the space – or not – before they started work. But in the fourth, the experimenter came in after the subject had finished setting everything out to her satisfaction, and then rearranged it all. The office space itself was not much different, but difference in productivity and job satisfaction was dramatic.
When workers were empowered to design their own space, they had fun and worked hard and accurately, producing 30 per cent more work than in the minimalist office and 15 per cent more than in the decorated office. When workers were deliberately disempowered, their work suffered and of course, they hated it. “I wanted to hit you,” one participant later admitted to an experimenter.
Haslam and Knight have confirmed what other researchers have long suspected – that lack of control over one’s physical environment is stressful and distracting. But this perspective is in stark contrast to those who see office design as too important to be left to the people who work in offices.
At least Le Corbusier had a vision, but many office spaces are ruled instead by an aesthetic that is mean and petty. The Wall Street Journal reported on Kyocera’s clipboard-wielding “inspectors” not only enforcing a clear-desk policy, but pulling open drawers and cabinets, photographing messy contents and demanding improvements. The Australian Financial Review published an 11-page clean-desk manual leaked from the mining giant BHP Billiton; apparently copper and coal cannot be mined if office workers do not respect the limit of one A5 picture frame on each desk. (The frame may display a family photo, or an award certificate, but it was forbidden to display both). Haslam and Knight told of a Sydney-based bank that changed the layout of its IT department 36 times in four years at the whim of senior management.
It is unclear why any of this top-down design is thought desirable. Official explanations are often empty or circular: that clean desks are more professional, or look tidier. In some cases, streamlined practices from the production line have been copied mindlessly into general office spaces, where they serve no purpose. Whatever the reason, it is folly. It can be satisfying to straighten up all the pens on your desk; but to order an underling to straighten their own pens is sociopathic.
When the likes of Steve Jobs or Frank Gehry are in charge, we can at least expect a workplace that will look beautiful. But that does not make it functional. Truly creative spaces aren’t constantly being made over for photoshoots in glossy business magazines. Just ask veterans of M.I.T., many of whom will identify as their favourite and most creative space a building that didn’t even have a proper name, which was designed in an afternoon and built to last just a couple of years. Building 20 was 200,000 square feet of plywood, cinderblock and asbestos, a squat, dusty firetrap originally designed to accommodate the wartime radar research effort, but which eked out an existence as M.I.T.’s junk-filled attic until 1998.
Building 20 was an unbelievably fertile mess. The successes started with the wartime RadLab, which produced nine Nobel prizes and the radar systems that won the second world war. But the outpouring continued for more than half a century. The first commercial atomic clock; one of the earliest particle accelerators; Harold Edgerton’s iconic high-speed photographs of a bullet passing through an apple – all sprang from Building 20. So did computer hacking and the first arcade video game, Spacewar. So did the pioneering technology companies DEC, BBN, and Bose. Cognitive science was revolutionised in Building 20 by the researcher Jerry Lettvin, while Noam Chomsky did the same for linguistics.
All this happened in the cheapest, nastiest space that M.I.T. could offer. But that was no coincidence. Building 20 was where the university put odd projects, student hobbyists and anything else that didn’t seem to matter, producing new collaborations.
And Building 20’s ugliness was functional. The water and cabling was exposed, running across the ceilings in brackets. Researchers thought nothing of tapping in to them for their experimental needs – or for that matter for knocking down a wall. When the atomic clock was being developed, the team removed two floors to accommodate it. This was the result not of design but of neglect. In the words of Stewart Brand, author of How Buildings Learn, “nobody cares what you do in there.”
And that was all Building 20’s residents wanted: to be left alone to create, to make whatever mess they wanted to make. When, inevitably, M.I.T. finally replaced Building 20 with a $300m structure designed by Frank Gehry himself, its former residents held a memorial wake. The new building might have been cutting-edge architecture, but one unhappy resident summed up the problem perfectly: “I didn’t ask for it.”
Of course nobody cares what the people who actually do the work might want or need. Chief executives exult in bold architectural statements, and universities find it easier to raise money for new buildings than for research. And so the grand buildings continue to be built, especially by the most profitable companies and the most prestigious seats of learning.
But we’re often guilty of confusing causation here, believing that great architecture underpins the success of great universities, or that Google flourishes because of the vibrancy of the helter skelters and ping pong tables in the Googleplex. A moment’s reflection reminds us that the innovation comes first, and the stunt architecture comes later.
Remember that for the first two years of Google’s history, there were no headquarters at all. The company’s founders, Sergey Brin and Larry Page, made the breakthroughs at Stanford University. Then came the cliché of a garage in Menlo Park, with desks made from doors set horizontally across sawhorses. The company grew and grew, into one crude space after another – and with engineers always free to hack things about. One knocked down the wall of his office, decided he didn’t like the results, and rebuilt the wall. That made for an ugly space – but a space that worked for the people who worked in it. The spirit of Building 20 lived on at Google.
So how should the ideal office look? In the most prestigious offices at the most prestigious companies, the ones which are being photographed by Wired, the answer to that question is: this place should look the way the boss’s pet architect wants it to look.
But Building 20, and Google’s early offices, and some of the great creative spaces all over the world, suggest a very different answer to the same question: how this place looks doesn’t matter.
Back in 1977, the editor of Psychology Today, T George Harris, put his finger on the problem:
“The office is a highly personal tool shop, often the home of the soul… this fact may sound simple, but it eludes most architects… They have a mania for uniformity, in space as in furniture, and a horror over how the messy side of human nature clutters up an office landscape that would otherwise be as tidy as a national cemetery.”
Harris scoured the academic literature for any evidence that good design helped people to get things done, or to be happier in the office. He couldn’t find it. “People suddenly put into “good design” did not seem to wake up and love it,” he wrote. What people love, instead, is the ability to control the space in which they work – even if they end up filling the space with kitsch, or dog photos, or even – shudder – garden gnomes.
Strangely enough, it was Steve Jobs himself – notorious as a dictatorial arbiter of good taste – who came to appreciate this at Pixar. When he unveiled his plan for the single pair of serendipity-inducing uber-bathrooms, he faced a rebellion from pregnant women at Pixar who didn’t want to have to make the long walk ten times a day. Jobs was aghast that people didn’t appreciate the importance of his vision. But then he did something unexpected: he backed down and agreed to install extra bathrooms.
Steve Jobs found other ways to encourage serendipitous interactions. More importantly, he showed that even on a question that mattered deeply to him, junior staff were able to defy him. Milled Arkansas steels be damned: it is the autonomy that really matters.
“The animators who work here are free to – no, encouraged to – decorate their work spaces in whatever style they wish,” explains Pixar’s boss Ed Catmull in his book Creativity, Inc. “They spend their days inside pink dollhouses whose ceilings are hung with miniature chandeliers, tiki huts made of real bamboo, and castles whose meticulously painted, fifteen-foot-high Styrofoam turrets appear to be carved from stone.”
I suspect that there may be a garden gnome in there, too.

The ideas in this article are adapted from my book “Messy“, which is available online in the US and UK or in good bookshops everywhere.

Newsnight film featuring Brian Eno

I made a short film for BBC’s Newsnight.  Lots of fun:

(If the embed function doesn’t work you can try this.)

There’s magic in mess: Why you should embrace a disorderly desk

In 1726, during a long voyage from London to Philadelphia, a young printer hatched the idea of using a notebook to systematically chart his efforts to become a better man. He set out 13 virtues — including industry, justice, tranquillity and temperance — and his plan was to focus on each in turn in an endless quest for self-improvement, recording failures with a black spot in his journal. The virtue journal worked, and the black marks became scarcer and scarcer.

Benjamin Franklin kept up this practice for his entire life. What a life it was: Franklin invented bifocals and a clean-burning stove; he proved that lightning was a form of electricity and then tamed it with the lightning conductor; he charted the Gulf Stream. He organised a lending library, a fire brigade and a college. He was America’s first postmaster-general, its ambassador to France, even the president of Pennsylvania.

And yet the great man had a weakness — or so he thought. His third virtue was Order. “Let all your things have their places; let each part of your business have its time,” he wrote. While all the other virtues were mastered, one by one, Franklin never quite managed to get his desk or his diary tidy.

“My scheme of Order gave me the most trouble,” he reflected six decades later. “My faults in it vexed me so much, and I made so little progress in amendment, and had such frequent relapses, that I was almost ready to give up the attempt.” Observers agreed. One described how callers on Franklin “were amazed to behold papers of the greatest importance scattered in the most careless way over the table and floor”.

Franklin was a messy fellow his entire life, despite 60 years of trying to reform himself, and remained convinced that if only he could learn to tidy up, he would become a more successful and productive person. But any outsider can see that it is absurd to think such a rich life could have been yet further enriched by assiduous use of a filing cabinet. Franklin was deluding himself. But his error is commonplace; we’re all tidy-minded people, admiring ourselves when we keep a clean desk and uneasy when we do not. Tidiness can be useful but it’s not always a virtue. Even though Franklin never let himself admit it, there can be a kind of magic in mess.

…

Why is it so difficult to keep things tidy? A clue comes in Franklin’s motto, “Let all your things have their places … ” That seems to make sense. Humans tend to have an excellent spatial memory. The trouble is that modern office life presents us with a continuous stream of disparate documents arriving not only by post but via email and social media. What are the “places”, both physical and digital, for this torrent of miscellanea?

Categorising documents of any kind is harder than it seems. The writer and philosopher Jorge Luis Borges once told of a fabled Chinese encyclopaedia, the “Celestial Emporium of Benevolent Knowledge”, which organised animals into categories such as: a) belonging to the emperor, c) tame, d) sucking pigs, f) fabulous, h) included in the present classification, and m) having just broken the water pitcher.

Borges’s joke has a point: categories are difficult. Distinctions that seem practically useful — who owns what, who did what, what might make a tasty supper — are utterly unusable when taken as a whole. The problem is harder still when we must file many incoming emails an hour, building folder structures that need to make sense months or years down the line. Borgesian email folders might include: a) coming from the boss, b) tedious, c) containing appointments, d) sent to the entire company, e) urgent, f) sexually explicit, g) complaints, h) personal, i) pertaining to the year-end review, and j) about to exceed the memory allocation on the server.

Regrettably, many of these emails fit into more than one category and while each grouping itself is perfectly meaningful, they do not fit together. Some emails clearly fit into a pattern, but many do not. One may be the start of a major project or the start of nothing at all, and it will rarely be clear which is which at the moment that email arrives in your inbox. Giving documents — whether physical or digital — a proper place, as Franklin’s motto recommends, requires clairvoyance. Failing that, we muddle through the miscellany, hurriedly imposing some kind of practical organising principle on what is a rapid and fundamentally messy flow of information.

When it comes to actual paper, there’s always the following beautiful approach. Invented in the early 1990s by Yukio Noguchi, an emeritus professor at Hitotsubashi University in Tokyo and author of books such as Super Organised Method, Noguchi doesn’t try to categorise anything. Instead, he places each incoming document in a large envelope. He writes the envelope’s contents neatly on its edge, and lines them up on a bookshelf, their contents visible like the spines of books. Now the moment of genius: each time he uses an envelope, Noguchi places it back on the left of the shelf. Over time, recently used documents will shuffle themselves towards the left, and never-used documents will accumulate on the right. Archiving is easy: every now and again, Noguchi removes the documents on the right. To find any document in this system, he simply asks himself how recently he has seen it. It is a filing system that all but organises itself.

But wait a moment. Eric Abrahamson and David Freedman, authors of A Perfect Mess, offer the following suggestion: “Turn the row of envelopes so that the envelopes are stacked vertically instead of horizontally, place the stack on your desktop, and get rid of the envelopes.” Those instructions transform the shelf described in Super Organised Method into an old-fashioned pile of papers on a messy desk. Every time a document arrives or is consulted, it goes back on the top of the pile. Unused documents gradually settle at the bottom. Less elegant, perhaps, but basically the same system.

Computer scientists may recognise something rather familiar about this arrangement: it mirrors the way that computers handle their memory systems. Computers use memory “caches”, which are small but swift to access. A critical issue is which data should be prioritised and put in the fastest cache. This cache management problem is analogous to asking which paper you should keep on your desk, which should be in your desk drawer, and which should be in offsite storage in New Jersey. Getting the decision right makes computers a lot faster — and it can make you faster too.

Fifty years ago, computer scientist Laszlo Belady proved that one of the fastest and most effective simple algorithms is to wait until the cache is full, then start ejecting the data that haven’t been used recently. This rule is called “Least Recently Used” or LRU — and it works because in computing, as in life, the fact that you’ve recently needed to use something is a good indication that you will need it again soon.

As Brian Christian and Tom Griffiths observe in their recent book Algorithms to Live By, while a computer might use LRU to manage a memory cache, Noguchi’s Super Organised Method uses the same rule to manage paper: recently used stuff on the left, stuff that you haven’t looked at for ages on the right. A pile of documents also implements LRU: recently touched stuff on the top, everything else sinks to the bottom.

This isn’t to say that a pile of paper is always the very best organisational system. That depends on what is being filed, and whether several people have to make sense of the same filing system or not. But the pile of papers is not random. It has its own pragmatic structure based simply on the fact that whatever you’re using tends to stay visible and accessible. Obsolete stuff sinks out of sight. Your desk may look messy to other people but you know that, thanks to the LRU rule, it’s really an efficient self-organising rapid-access cache.

…

If all this sounds to you like self-justifying blather from untidy colleagues, you might just be a “filer” rather than a “piler”. The distinction between the two was first made in the 1980s by Thomas Malone, a management professor at the Massachusetts Institute of Technology. Filers like to establish a formal organisational structure for their paper documents. Pilers, by contrast, let pieces of paper build up around their desks or, as we have now learnt to say, implement an LRU-cache.

To most of us, it may seem obvious that piling is dysfunctional while filing is the act of a serious professional. Yet when researchers from the office design company Herman Miller looked at high-performing office workers, they found that they tended to be pilers. They let documents accumulate on their desks, used their physical presence as a reminder to do work, and relied on subtle cues — physical alignment, dog-ears, or a stray Post-it note — to orient themselves.

In 2001, Steve Whittaker and Julia Hirschberg, researchers at AT&T Labs, studied filers and pilers in a real office environment, and discovered why the messy approach works better than it seemingly has any right to. They tracked the behaviour of the filers and pilers over time. Who accumulated the biggest volume of documents? Whose archives worked best? And who struggled most when an office relocation forced everyone to throw paperwork away?

One might expect that disciplined filers would have produced small, useful filing systems. But Whittaker and Hirschberg found, instead, that they were sagging under the weight of bloated, useless archives. The problem was a bad case of premature filing. Paperwork would arrive, and then the filer would have to decide what to do with it. It couldn’t be left on the desk — that would be untidy. So it had to be filed somewhere. But most documents have no long-term value, so in an effort to keep their desks clear, the filers were using filing cabinets as highly structured waste-paper baskets. Useful material was hemmed in by well-organised dross.

“You can’t know your own information future,” says Whittaker, who is now a professor of psychology at University of California Santa Cruz, and co-author of The Science of Managing Our Digital Stuff. People would create folder structures that made sense at the time but that would simply be baffling to their own creators months or years later. Organisational categories multiplied. One person told Whittaker and Hirschberg: “I had so much stuff filed. I didn’t know where everything was, and I’d found that I had created second files for something in what seemed like a logical place, but not the only logical place … I ended up having the same thing in two places or I had the same business unit stuff in five different places.”

As for the office move, it was torture for the filers. They had too much material and had invested too much time in organising it. One commented that it was “gruesome … you’re casting off your first-born”. Whittaker reminds me that these people were not discarding their children. They weren’t even discarding family photographs and keepsakes. They were throwing away office memos and dog-eared corporate reports. “It’s very visceral,” Whittaker says. “People’s identity is wrapped up in their jobs, and in the information professions your identity is wrapped up with your information.” And yet the happy-go-lucky pilers, in their messy way, coped far better. They used their desks as temporary caches for documents. The good stuff would remain close at hand, easy to use and to throw away when finished. Occasionally, the pilers would grab a pile, riffle through it and throw most of it away. And when they did file material, they did so in archives that were small, practical and actively used.

Whittaker points out that the filers struggled because the categories they created turned out not to work well as times changed. This suggests that tidiness can work, but only when documents or emails arrive with an obvious structure. My own desk is messy but my financial records are neat — not because they’re more important but because the record-keeping required for accountancy is predictable.

One might object that whatever researchers have concluded about paper documents is obsolete, as most documents are now digital. Surely the obvious point of stress is now the email inbox? But Whittaker’s interest in premature filing actually started in 1996 with an early study of email overload. “The thing we observed was failed folders,” he says. “Tiny email folders with one or two items.”

It turns out that the fundamental problem with email is the same as the problem with paper on the desk: people try to clear their inbox by sorting the email into folders but end up prematurely filing in folder structures that turn out not to work well. In 2011, Whittaker and colleagues published a research paper with the title “Am I Wasting My Time Organizing Email?”. The answer is: yes, you are. People who use the search function find their email more quickly than those who click through carefully constructed systems of folders. The folder system feels better organised but, unless the information arrives with a predictable structure, creating folders is laborious and worse than useless.

…

So we know that carefully filing paper documents is often counterproductive. Email should be dumped in a few broad folders — or one big archive — rather than a careful folder hierarchy. What then should we do with our calendars? There are two broad approaches. One — analogous to the “filer” approach — is to organise one’s time tightly, scheduling each task in advance and using the calendar as a to-do list. As Benjamin Franklin expressed it: “Let each part of your business have its time.” The alternative avoids the calendar as much as possible, noting only fixed appointments. Intuitively, both approaches have something going for them, so which works best?

Fortunately we don’t need to guess, because three psychologists, Daniel Kirschenbaum, Laura Humphrey and Sheldon Malett, have already run the experiment. Thirty-five years ago, Kirschenbaum and his colleagues recruited a group of undergraduates for a short course designed to improve their study skills. The students were randomly assigned one of three possible pieces of coaching. There was a control group, which was given simple time-management advice such as, “Take breaks of five to 10 minutes after every ½-1½ hour study session.” The other two groups got those tips but they were also given much more specific advice as to how to use their calendars. The “monthly plan” group were instructed to set goals and organise study activities across the space of a month; in contrast, the “daily plan” group were told to micromanage their time, planning activities and setting goals within the span of a single day.

The researchers assumed that the planners who set quantifiable daily goals would do better than those with vaguer monthly plans. In fact, the daily planners started brightly but quickly became hopelessly demotivated, with their study effort collapsing to eight hours a week — even worse than the 10 hours for those with no plan at all. But the students on the monthly plans maintained a consistent study habit of 25 hours a week throughout the course. The students’ grades, unsurprisingly, reflected their work effort.

The problem is that the daily plans get derailed. Life is unpredictable. A missed alarm, a broken washing machine, a dental appointment, a friend calling by for a coffee — or even the simple everyday fact that everything takes longer than you expect — all these obstacles proved crushing for people who had used their calendar as a to-do list.

Like the document pilers, the monthly planners adopted a loose, imperfect and changeable system that happens to work just fine in a loose, imperfect and changeable world. The daily planners, like the filers, imposed a tight, tidy-minded system that shattered on contact with a messy world.

Some people manage to take this lesson to extremes. Marc Andreessen — billionaire entrepreneur and venture capitalist — decided a decade ago to stop writing anything in his calendar. If something was worth doing, he figured, it was worth doing immediately. “I’ve been trying this tactic as an experiment,” he wrote in 2007. “And I am so much happier, I can’t even tell you.”

Arnold Schwarzenegger has adopted much the same approach. He insisted on keeping his diary clear when he was a film star. He even tried the same policy when governor of California. “Appointments are always a no-no. Planning ahead is a no-no,” he told The New York Times. Politicians, lobbyists and activists had to treat him like a popular walk-up restaurant: they showed up and hoped to get a slot. Of course, this was in part a pure status play. But it was more than that. Schwarzenegger knew that an overstuffed diary allows no room to adapt to circumstances.

Naturally, Schwarzenegger and Andreessen can make the world wait to meet them. You and I can’t. But we probably could take a few steps in the same direction, making fewer firm commitments to others and to ourselves, leaving us the flexibility to respond to what life throws at us. A plan that is too finely woven will soon lie in tatters. Daily plans are tidy but life is messy.

…

The truth is that getting organised is often a matter of soothing our anxieties — or the anxieties of tidy-minded colleagues. It can simply be an artful way of feeling busy while doing nothing terribly useful. Productivity guru Merlin Mann, host of a podcast called Back To Work, has a telling metaphor. Imagine making sandwiches in a deli, says Mann. In comes the first sandwich order. You’re about to reach for the mayonnaise and a couple of slices of sourdough. But then more orders start coming in.

Mann knows all too well how we tend to react. Instead of making the first sandwich, we start to ponder organisational systems. Separate the vegetarian and the meat? Should toasted sandwiches take priority?

There are two problems here. First, there is no perfect way to organise a fast-moving sandwich queue. Second, the time we spend trying to get organised is time we don’t spend getting things done. Just make the first sandwich. If we just got more things done, decisively, we might find we had less need to get organised.

Of course, sometimes we need a careful checklist (if, say, we’re building a house) or a sophisticated reference system (if we’re maintaining a library, for example). But most office workers are neither construction managers nor librarians. Yet we share Benjamin Franklin’s mistaken belief that if only we were more neatly organised, then we would live more productive and more admirable lives. Franklin was too busy inventing bifocals and catching lightning to get around to tidying up his life. If he had been working in a deli, you can bet he wouldn’t have been organising sandwich orders. He would have been making sandwiches.
Image by Benjamin Swanson. This article was first published in the Financial Times magazine and is inspired by ideas from my new book, “Messy“. (US) (UK)

How Politicians Poisoned Statistics

We have more data — and the tools to analyse and share them — than ever before. So why is the truth so hard to pin down?

In January 2015, a few months before the British general election, a proud newspaper resigned itself to the view that little good could come from the use of statistics by politicians. An editorial in the Guardian argued that in a campaign that would be “the most fact-blitzed in history”, numerical claims would settle no arguments and persuade no voters. Not only were numbers useless for winning power, it added, they were useless for wielding it, too. Numbers could tell us little. “The project of replacing a clash of ideas with a policy calculus was always dubious,” concluded the newspaper. “Anyone still hankering for it should admit their number’s up.”

This statistical capitulation was a dismaying read for anyone still wedded to the idea — apparently a quaint one — that gathering statistical information might help us understand and improve our world. But the Guardian’s cynicism can hardly be a surprise. It is a natural response to the rise of “statistical bullshit” — the casual slinging around of numbers not because they are true, or false, but to sell a message.

Politicians weren’t always so ready to use numbers as part of the sales pitch. Recall Ronald Reagan’s famous suggestion to voters on the eve of his landslide defeat of President Carter: “Ask yourself, ‘Are you better off now than you were four years ago?’” Reagan didn’t add any statistical garnish. He knew that voters would reach their own conclusions.

The British election campaign of spring last year, by contrast, was characterised by a relentless statistical crossfire. The shadow chancellor of the day, Ed Balls, declared that a couple with children (he didn’t say which couple) had lost £1,800 thanks to the government’s increase in value added tax. David Cameron, the prime minister, countered that 94 per cent of working households were better off thanks to recent tax changes, while the then deputy prime minister Nick Clegg was proud to say that 27 million people were £825 better off in terms of the income tax they paid.

Could any of this be true? Yes — all three claims were. But Ed Balls had reached his figure by summing up extra VAT payments over several years, a strange method. If you offer to hire someone for £100,000, and then later admit you meant £25,000 a year for a four-year contract, you haven’t really lied — but neither have you really told the truth. And Balls had looked only at one tax. Why not also consider income tax, which the government had cut? Clegg boasted about income-tax cuts but ignored the larger rise in VAT. And Cameron asked to be evaluated only on his pre-election giveaway budget rather than the tax rises he had introduced earlier in the parliament — the equivalent of punching someone on the nose, then giving them a bunch of flowers and pointing out that, in floral terms, they were ahead on the deal.

Each claim was narrowly true but broadly misleading. Not only did the clashing numbers confuse but none of them helped answer the crucial question of whether Cameron and Clegg had made good decisions in office.

To ask whether the claims were true is to fall into a trap. None of these politicians had any interest in playing that game. They were engaged in another pastime entirely.

…

Thirty years ago, the Princeton philosopher Harry Frankfurt published an essay in an obscure academic journal, Raritan. The essay’s title was “On Bullshit”. (Much later, it was republished as a slim volume that became a bestseller.) Frankfurt was on a quest to understand the meaning of bullshit — what was it, how did it differ from lies, and why was there so much of it about?

Frankfurt concluded that the difference between the liar and the bullshitter was that the liar cared about the truth — cared so much that he wanted to obscure it — while the bullshitter did not. The bullshitter, said Frankfurt, was indifferent to whether the statements he uttered were true or not. “He just picks them out, or makes them up, to suit his purpose.”

Statistical bullshit is a special case of bullshit in general, and it appears to be on the rise. This is partly because social media — a natural vector for statements made purely for effect — are also on the rise. On Instagram and Twitter we like to share attention-grabbing graphics, surprising headlines and figures that resonate with how we already see the world. Unfortunately, very few claims are eye-catching, surprising or emotionally resonant because they are true and fair. Statistical bullshit spreads easily these days; all it takes is a click.

Consider a widely shared list of homicide “statistics” attributed to the “Crime Statistics Bureau — San Francisco”, asserting that 81 per cent of white homicide victims were killed by “blacks”. It takes little effort to establish that the Crime Statistics Bureau of San Francisco does not exist, and not much more digging to discover that the data are utterly false. Most murder victims in the United States are killed by people of their own race; the FBI’s crime statistics from 2014 suggest that more than 80 per cent of white murder victims were killed by other white people.

Somebody, somewhere, invented the image in the hope that it would spread, and spread it did, helped by a tweet from Donald Trump, the current frontrunner for the Republican presidential nomination, that was retweeted more than 8,000 times. One can only speculate as to why Trump lent his megaphone to bogus statistics, but when challenged on Fox News by the political commentator Bill O’Reilly, he replied, “Hey, Bill, Bill, am I gonna check every statistic?”

Harry Frankfurt’s description of the bullshitter would seem to fit Trump perfectly: “He does not care whether the things he says describe reality correctly.”

While we can’t rule out the possibility that Trump knew the truth and was actively trying to deceive his followers, a simpler explanation is that he wanted to win attention and to say something that would resonate with them. One might also guess that he did not check whether the numbers were true because he did not much care one way or the other. This is not a game of true and false. This is a game of politics.

…

While much statistical bullshit is careless, it can also be finely crafted. “The notion of carefully wrought bullshit involves … a certain inner strain,” wrote Harry Frankfurt but, nevertheless, the bullshit produced by spin-doctors can be meticulous. More conventional politicians than Trump may not much care about the truth but they do care about being caught lying.

Carefully wrought bullshit was much in evidence during last year’s British general election campaign. I needed to stick my nose in and take a good sniff on a regular basis because I was fact-checking on behalf of the BBC’s More or Less programme. Again and again I would find myself being asked on air, “Is that claim true?” and finding that the only reasonable answer began with “It’s complicated”.

Take Ed Miliband’s claim before the last election that “people are £1,600 a year worse off” than they were when the coalition government came to power. Was that claim true? Arguably, yes.

But we need to be clear that by “people”, the then Labour leader was excluding half the adult population. He was not referring to pensioners, benefit recipients, part-time workers or the self-employed. He meant only full-time employees, and, more specifically, only their earnings before taxes and benefits.

Even this narrower question of what was happening to full-time earnings is a surprisingly slippery one. We need to take an average, of course. But what kind of average? Labour looked at the change in median wages, which were stagnating in nominal terms and falling after inflation was taken into account.

That seems reasonable — but the median is a problematic measure in this case. Imagine nine people, the lowest-paid with a wage of £1, the next with a wage of £2, up to the highest-paid person with a wage of £9. The median wage is the wage of the person in the middle: it’s £5.

Now imagine that everyone receives a promotion and a pay rise of £1. The lowly worker with a wage of £1 sees his pay packet double to £2. The next worker up was earning £2 and now she gets £3. And so on. But there’s also a change in the composition of the workforce: the best-paid worker retires and a new apprentice is hired at a wage of £1. What’s happened to people’s pay? In a sense, it has stagnated. The pattern of wages hasn’t changed at all and the median is still £5.

But if you asked the individual workers about their experiences, they would all tell you that they had received a generous pay rise. (The exceptions are the newly hired apprentice and the recent retiree.) While this example is hypothetical, at the time Miliband made his comments something similar was happening in the real labour market. The median wage was stagnating — but among people who had worked for the same employer for at least a year, the median worker was receiving a pay rise, albeit a modest one.

Another source of confusion: if wages for the low-paid and the high-paid are rising but wages in the middle are sagging, then the median wage can fall, even though the median wage increase is healthy. The UK labour market has long been prone to this kind of “job polarisation”, where demand for jobs is strongest for the highest and lowest-paid in the economy. Job polarisation means that the median pay rise can be sizeable even if median pay has not risen.

Confused? Good. The world is a complicated place; it defies description by sound bite statistics. No single number could ever answer Ronald Reagan’s question — “Are you better off now than you were four years ago?” — for everyone in a country.

So, to produce Labour’s figure of “£1,600 worse off”, the party’s press office had to ignore the self-employed, the part-timers, the non-workers, compositional effects and job polarisation. They even changed the basis of their calculation over time, switching between different measures of wages and different measures of inflation, yet miraculously managing to produce a consistent answer of £1,600. Sometimes it’s easier to make the calculation produce the number you want than it is to reprint all your election flyers.

Very few claims are eye-catching, surprising or emotionally resonant because they are true and fair

Such careful statistical spin-doctoring might seem a world away from Trump’s reckless retweeting of racially charged lies. But in one sense they were very similar: a political use of statistics conducted with little interest in understanding or describing reality. Miliband’s project was not “What is the truth?” but “What can I say without being shown up as a liar?”

Unlike the state of the UK job market, his incentives were easy to understand. Miliband needed to hammer home a talking point that made the government look bad. As Harry Frankfurt wrote back in the 1980s, the bullshitter “is neither on the side of the true nor on the side of the false. His eye is not on the facts at all … except insofar as they may be pertinent to his interest in getting away with what he says.”

Such complexities put fact-checkers in an awkward position. Should they say that Ed Miliband had lied? No: he had not. Should they say, instead, that he had been deceptive or misleading? Again, no: it was reasonable to say that living standards had indeed been disappointing under the coalition government.

Nevertheless, there was a lot going on in the British economy that the figure omitted — much of it rather more flattering to the government. Full Fact, an independent fact-checking organisation, carefully worked through the paper trail and linked to all the relevant claims. But it was powerless to produce a fair and representative snapshot of the British labour market that had as much power as Ed Miliband’s seven-word sound bite. No such snapshot exists. Truth is usually a lot more complicated than statistical bullshit.

…

On July 16 2015, the UK health phentermine secretary Jeremy Hunt declared: “Around 6,000 people lose their lives every year because we do not have a proper seven-day service in hospitals. You are 15 per cent more likely to die if you are admitted on a Sunday compared to being admitted on a Wednesday.”

This was a statistic with a purpose. Hunt wanted to change doctors’ contracts with the aim of getting more weekend work out of them, and bluntly declared that the doctors’ union, the British Medical Association, was out of touch and that he would not let it block his plans: “I can give them 6,000 reasons why.”

Despite bitter opposition and strike action from doctors, Hunt’s policy remained firm over the following months. Yet the numbers he cited to support it did not. In parliament in October, Hunt was sticking to the 15 per cent figure, but the 6,000 deaths had almost doubled: “According to an independent study conducted by the BMJ, there are 11,000 excess deaths because we do not staff our hospitals properly at weekends.”

Arithmetically, this was puzzling: how could the elevated risk of death stay the same but the number of deaths double? To add to the suspicions about Hunt’s mathematics, the editor-in-chief of the British Medical Journal, Fiona Godlee, promptly responded that the health phentermine secretary had publicly misrepresented the BMJ research.

Undaunted, the health phentermine secretary bounced back in January with the same policy and some fresh facts: “At the moment we have an NHS where if you have a stroke at the weekends, you’re 20 per cent more likely to die. That can’t be acceptable.”

All this is finely wrought bullshit — a series of ever-shifting claims that can be easily repeated but are difficult to unpick. As Hunt jumped from one form of words to another, he skipped lightly ahead of fact-checkers as they tried to pin him down. Full Fact concluded that Hunt’s statement about 11,000 excess deaths had been untrue, and asked him to correct the parliamentary record. His office responded with a spectacular piece of bullshit, saying (I paraphrase) that whether or not the claim about 11,000 excess deaths was true, similar claims could be made that were.

So, is it true? Do 6,000 people — or 11,000 — die needlessly in NHS hospitals because of poor weekend care? Nobody knows for sure; Jeremy Hunt certainly does not. It’s not enough to show that people admitted to hospital at the weekend are at an increased risk of dying there. We need to understand why — a question that is essential for good policy but inconvenient for politicians.

One possible explanation for the elevated death rate for weekend admissions is that the NHS provides patchy care and people die as a result. That is the interpretation presented as bald fact by Jeremy Hunt. But a more straightforward explanation is that people are only admitted to hospital at the weekend if they are seriously ill. Less urgent cases wait until weekdays. If weekend patients are sicker, it is hardly a surprise that they are more likely to die. Allowing non-urgent cases into NHS hospitals at weekends wouldn’t save any lives, but it would certainly make the statistics look more flattering. Of course, epidemiologists try to correct for the fact that weekend patients tend to be more seriously ill, but few experts have any confidence that they have succeeded.

A more subtle explanation is that shortfalls in the palliative care system may create the illusion that hospitals are dangerous. Sometimes a patient is certain to die, but the question is where — in a hospital or a palliative hospice? If hospice care is patchy at weekends then a patient may instead be admitted to hospital and die there. That would certainly reflect poor weekend care. It would also add to the tally of excess weekend hospital deaths, because during the week that patient would have been admitted to, and died in, a palliative hospice. But it is not true that the death was avoidable.

Does it seem like we’re getting stuck in the details? Well, yes, perhaps we are. But improving NHS care requires an interest in the details. If there is a problem in palliative care hospices, it will not be fixed by improving staffing in hospitals.

“Even if you accept that there’s a difference in death rates,” says John Appleby, the chief economist of the King’s Fund health think-tank, “nobody is able to say why it is. Is it lack of diagnostic services? Lack of consultants? We’re jumping too quickly from a statistic to a solution.”

When one claim is discredited, Jeremy Hunt’s office simply asserts that another one can be found to take its place

This matters — the NHS has a limited budget. There are many things we might want to spend money on, which is why we have the National Institute for Health and Care Excellence (Nice) to weigh up the likely benefits of new treatments and decide which offer the best value for money.

Would Jeremy Hunt’s push towards a seven-day NHS pass the Nice cost-benefit threshold? Probably not. Our best guess comes from a 2015 study by health economists Rachel Meacock, Tim Doran and Matt Sutton, which estimates that the NHS has many cheaper ways to save lives. A more comprehensive assessment might reach a different conclusion but we don’t have one because the Department for Health, oddly, hasn’t carried out a formal health impact assessment of the policy it is trying to implement.

This is a depressing situation. The government has devoted considerable effort to producing a killer number: Jeremy Hunt’s “6,000 reasons” why he won’t let the British Medical Association stand in his way. It continues to produce statistical claims that spring up like hydra heads: when one claim is discredited, Hunt’s office simply asserts that another one can be found to take its place. Yet the government doesn’t seem to have bothered to gather the statistics that would actually answer the question of how the NHS could work better.

This is the real tragedy. It’s not that politicians spin things their way — of course they do. That is politics. It’s that politicians have grown so used to misusing numbers as weapons that they have forgotten that used properly, they are tools.

…

You complain that your report would be dry. The dryer the better. Statistics should be the dryest of all reading,” wrote the great medical statistician William Farr in a letter in 1861. Farr sounds like a caricature of a statistician, and his prescription — convey the greatest possible volume of information with the smallest possible amount of editorial colour — seems absurdly ill-suited to the modern world.

But there is a middle ground between the statistical bullshitter, who pays no attention to the truth, and William Farr, for whom the truth must be presented without adornment. That middle ground is embodied by the recipient of William Farr’s letter advising dryness. She was the first woman to be elected to the Royal Statistical Society: Florence Nightingale.

Nightingale is the most celebrated nurse in British history, famous for her lamplit patrols of the Barrack Hospital in Scutari, now a district of Istanbul. The hospital was a death trap, with thousands of soldiers from the Crimean front succumbing to typhus, cholera and dysentery as they tried to recover from their wounds in cramped conditions next to the sewers. Nightingale, who did her best, initially believed that the death toll was due to lack of food and supplies. Then, in the spring of 1855, a sanitary commission sent from London cleaned up the hospital, whitewashing the walls, carting away filth and dead animals and flushing out the sewers. The death rate fell sharply.

Nightingale returned to Britain and reviewed the statistics, concluding that she had paid too little attention to sanitation and that most military and medical professions were making the same mistake, leading to hundreds of thousands of deaths. She began to campaign for better public health measures, tighter laws on hygiene in rented properties, and improvements to sanitation in barracks and hospitals across the country. In doing so, a mere nurse had to convince the country’s medical and military establishments, led by England’s chief medical officer, John Simon, that they had been doing things wrong all their lives.

A key weapon in this lopsided battle was statistical evidence. But Nightingale disagreed with Farr on how that evidence should be presented. “The dryer the better” would not serve her purposes. Instead, in 1857, she crafted what has become known as the Rose Diagram, a beautiful array of coloured wedges showing the deaths from infectious diseases before and after the sanitary improvements at Scutari.

When challenged by Bill O’Reilly on Fox News, Trump replied, ‘Hey Bill, Bill, am I gonna check every statistic?’

The Rose Diagram isn’t a dry presentation of statistical truth. It tells a story. Its structure divides the death toll into two periods — before the sanitary improvements, and after. In doing so, it highlights a sharp break that is less than clear in the raw data. And the Rose Diagram also gently obscures other possible interpretations of the numbers — that, for example, the death toll dropped not because of improved hygiene but because winter was over. The Rose Diagram is a marketing pitch for an idea. The idea was true and vital, and Nightingale’s campaign was successful. One of her biographers, Hugh Small, argues that the Rose Diagram ushered in health improvements that raised life expectancy in the UK by 20 years and saved millions of lives.

What makes Nightingale’s story so striking is that she was able to see that statistics could be tools and weapons at the same time. She educated herself using the data, before giving it the makeover it required to convince others. Though the Rose Diagram is a long way from “the dryest of all reading”, it is also a long way from bullshit. Florence Nightingale realised that the truth about public health was so vital that it could not simply be recited in a monotone. It needed to sing.

…

The idea that a graph could change the world seems hard to imagine today. Cynicism has set in about statistics. Many journalists draw no distinction between a systematic review of peer-reviewed evidence and a survey whipped up in an afternoon to sell biscuits or package holidays: it’s all described as “new research”. Politicians treat statistics not as the foundation of their argument but as decoration — “spray-on evidence” is the phrase used by jaded civil servants. But a freshly painted policy without foundations will not last long before the cracks show through.

“Politicians need to remember: there is a real world and you want to try to change it,” says Will Moy, the director of Full Fact. “At some stage you need to engage with the real world — and that is where the statistics come in handy.”

That should be no problem, because it has never been easier to gather and analyse informative statistics. Nightingale and Farr could not have imagined the data that modern medical researchers have at their fingertips. The gold standard of statistical evidence is the randomised controlled trial, because using a randomly chosen control group protects against biased or optimistic interpretations of the evidence. Hundreds of thousands of such trials have been published, most of them within the past 25 years. In non-medical areas such as education, development aid and prison reform, randomised trials are rapidly catching on: thousands have been conducted. The British government, too, has been supporting policy trials — for example, the Education Endowment Foundation, set up with £125m of government funds just five years ago, has already backed more than 100 evaluations of educational approaches in English schools. It favours randomised trials wherever possible.

The frustrating thing is that politicians seem quite happy to ignore evidence — even when they have helped to support the researchers who produced it. For example, when the chancellor George Osborne announced in his budget last month that all English schools were to become academies, making them independent of the local government, he did so on the basis of faith alone. The Sutton Trust, an educational charity which funds numerous research projects, warned that on the question of whether academies had fulfilled their original mission of improving failing schools in poorer areas, “our evidence suggests a mixed picture”. Researchers at the LSE’s Centre for Economic Performance had a blunter description of Osborne’s new policy: “a non-evidence based shot in the dark”.

This should be no surprise. Politicians typically use statistics like a stage magician uses smoke and mirrors. Over time, they can come to view numbers with contempt. Voters and journalists will do likewise. No wonder the Guardian gave up on the idea that political arguments might be settled by anything so mundane as evidence. The spin-doctors have poisoned the statistical well.

But despite all this despair, the facts still matter. There isn’t a policy question in the world that can be settled by statistics alone but, in almost every case, understanding the statistical background is a tremendous help. Hetan Shah, the executive director of the Royal Statistical Society, has lost count of the number of times someone has teased him with the old saying about “lies, damned lies and statistics”. He points out that while it’s easy to lie with statistics, it’s even easier to lie without them.

Perhaps the lies aren’t the real enemy here. Lies can be refuted; liars can be exposed. But bullshit? Bullshit is a stickier problem. Bullshit corrodes the very idea that the truth is out there, waiting to be discovered by a careful mind. It undermines the notion that the truth matters. As Harry Frankfurt himself wrote, the bullshitter “does not reject the authority of the truth, as the liar does, and oppose himself to it. He pays no attention to it at all. By virtue of this, bullshit is a greater enemy of the truth than lies are.”

Written for and first published in the FT Magazine. 

My TED talk on how distractions make us more creative

Multi-tasking: how to survive in the 21st century

Modern life now forces us to do a multitude of things at once — but can we? Should we?

Forget invisibility or flight: the superpower we all want is the ability to do several things at once. Unlike other superpowers, however, being able to multitask is now widely regarded as a basic requirement for employability. Some of us sport computers with multiple screens, to allow tweeting while trading pork bellies and frozen orange juice. Others make do with reading a Kindle while poking at a smartphone and glancing at a television in the corner with its two rows of scrolling subtitles. We think nothing of sending an email to a colleague to suggest a quick coffee break, because we can feel confident that the email will be read within minutes.

All this is simply the way the modern world works. Multitasking is like being able to read or add up, so fundamental that it is taken for granted. Doing one thing at a time is for losers — recall Lyndon Johnson’s often bowdlerised dismissal of Gerald Ford: “He can’t fart and chew gum at the same time.”

The rise of multitasking is fuelled by technology, of course, and by social change as well. Husbands and wives no longer specialise as breadwinners and homemakers; each must now do both. Work and play blur. Your friends can reach you on your work email account at 10 o’clock in the morning, while your boss can reach you on your mobile phone at 10 o’clock at night. You can do your weekly shop sitting at your desk and you can handle a work query in the queue at the supermarket.

This is good news in many ways — how wonderful to be able to get things done in what would once have been wasted time! How delightful the variety of it all is! No longer must we live in a monotonous, Taylorist world where we must painstakingly focus on repetitive tasks until we lose our minds.

And yet we are starting to realise that the blessings of a multitasking life are mixed. We feel overwhelmed by the sheer number of things we might plausibly be doing at any one time, and by the feeling that we are on call at any moment.

And we fret about the unearthly appetite of our children to do everything at once, flipping through homework while chatting on WhatsApp, listening to music and watching Game of Thrones. (According to a recent study by Sabrina Pabilonia of the US Bureau of Labor Statistics, for over half the time that high-school students spend doing homework, they are also listening to music, watching TV or otherwise multitasking. That trend is on the increase.) Can they really handle all these inputs at once? They seem to think so, despite various studies suggesting otherwise.

And so a backlash against multitasking has begun — a kind of Luddite self-help campaign. The poster child for uni-tasking was launched on the crowdfunding website Kickstarter in December 2014. For $499 — substantially more than a multifunctional laptop — “The Hemingwrite” computer promised a nice keyboard, a small e-ink screen and an automatic cloud back-up. You couldn’t email on the Hemingwrite. You couldn’t fool around on YouTube, and you couldn’t read the news. All you could do was type. The Hemingwrite campaign raised over a third of a million dollars.

The Hemingwrite (now rebranded the Freewrite) represents an increasingly popular response to the multitasking problem: abstinence. Programs such as Freedom and Self-Control are now available to disable your browser for a preset period of time. The popular blogging platform WordPress offers “distraction-free writing”. The Villa Stéphanie, a hotel in Baden-Baden, offers what has been branded the “ultimate luxury”: a small silver switch beside the hotel bed that will activate a wireless blocker and keep the internet and all its temptations away.

The battle lines have been drawn. On one side: the culture of the modern workplace, which demands that most of us should be open to interruption at any time. On the other, the uni-tasking refuseniks who insist that multitaskers are deluding themselves, and that focus is essential. Who is right?

The ‘cognitive cost’

There is ample evidence in favour of the proposition that we should focus on one thing at a time. Consider a study led by David Strayer, a psychologist at the University of Utah. In 2006, Strayer and his colleagues used a high-fidelity driving simulator to compare the performance of drivers who were chatting on a mobile phone to drivers who had drunk enough alcohol to be at the legal blood-alcohol limit in the US. Chatting drivers didn’t adopt the aggressive, risk-taking style of drunk drivers but they were unsafe in other ways. They took much longer to respond to events outside the car, and they failed to notice a lot of the visual cues around them. Strayer’s infamous conclusion: driving while using a mobile phone is as dangerous as driving while drunk.

Less famous was Strayer’s finding that it made no difference whether the driver was using a handheld or hands-free phone. The problem with talking while driving is not a shortage of hands. It is a shortage of mental bandwidth.

Yet this discovery has made little impression either on public opinion or on the law. In the United Kingdom, for example, it is an offence to use a hand-held phone while driving but perfectly legal if the phone is used hands-free. We’re happy to acknowledge that we only have two hands but refuse to admit that we only have one brain.

Another study by Strayer, David Sanbonmatsu and others, suggested that we are also poor judges of our ability to multitask. The subjects who reported doing a lot of multitasking were also the ones who performed poorly on tests of multitasking ability. They systematically overrated their ability to multitask and they displayed poor impulse control. In other words, wanting to multitask is a good sign that you should not be multitasking.

We may not immediately realise how multitasking is hampering us. The first time I took to Twitter to comment on a public event was during a televised prime-ministerial debate in 2010. The sense of buzz was fun; I could watch the candidates argue and the twitterati respond, compose my own 140-character profundities and see them being shared. I felt fully engaged with everything that was happening. Yet at the end of the debate I realised, to my surprise, that I couldn’t remember anything that Brown, Cameron and Clegg had said.

A study conducted at UCLA in 2006 suggests that my experience is not unusual. Three psychologists, Karin Foerde, Barbara Knowlton and Russell Poldrack, recruited students to look at a series of flashcards with symbols on them, and then to make predictions based on patterns they had recognised. Some of these prediction tasks were done in a multitasking environment, where the students also had to listen to low- and high-pitched tones and count the high-pitched ones. You might think that making predictions while also counting beeps was too much for the students to handle. It wasn’t. They were equally competent at spotting patterns with or without the note-counting task.

But here’s the catch: when the researchers then followed up by asking more abstract questions about the patterns, the cognitive cost of the multitasking became clear. The students struggled to answer questions about the predictions they’d made in the multitasking environment. They had successfully juggled both tasks in the moment — but they hadn’t learnt anything that they could apply in a different context.

That’s an unnerving discovery. When we are sending email in the middle of a tedious meeting, we may nevertheless feel that we’re taking in what is being said. A student may be confident that neither Snapchat nor the live football is preventing them taking in their revision notes. But the UCLA findings suggest that this feeling of understanding may be an illusion and that, later, we’ll find ourselves unable to remember much, or to apply our knowledge flexibly. So, multitasking can make us forgetful — one more way in which multitaskers are a little bit like drunks.

Early multitaskers

All this is unnerving, given that the modern world makes multitasking almost inescapable. But perhaps we shouldn’t worry too much. Long before multitasking became ubiquitous, it had a long and distinguished history.

In 1958, a young psychologist named Bernice Eiduson embarked on an long-term research project — so long-term, in fact, that Eiduson died before it was completed. Eiduson studied the working methods of 40 scientists, all men. She interviewed them periodically over two decades and put them through various psychological tests. Some of these scientists found their careers fizzling out, while others went on to great success. Four won Nobel Prizes and two others were widely regarded as serious Nobel contenders. Several more were invited to join the National Academy of Sciences.

After Eiduson died, some of her colleagues published an analysis of her work. These colleagues, Robert Root-Bernstein, Maurine Bernstein and Helen Garnier, wanted to understand what determined whether a scientist would have a long productive career, a combination of genius and longevity.

There was no clue in the interviews or the psychological tests. But looking at the early publication record of these scientists — their first 100 published research papers — researchers discovered a pattern: the top scientists were constantly changing the focus of their research.

Over the course of these first 100 papers, the most productive scientists covered five different research areas and moved from one of these topics to another an average of 43 times. They would publish, and change the subject, publish again, and change the subject again. Since most scientific research takes an extended period of time, the subjects must have overlapped. The secret to a long and highly productive scientific career? It’s multitasking.

Charles Darwin thrived on spinning multiple plates. He began his first notebook on “transmutation of species” two decades before The Origin of Species was published. His A Biographical Sketch of an Infant was based on notes made after his son William was born; William was 37 when he published. Darwin spent nearly 20 years working on climbing and insectivorous plants. And Darwin published a learned book on earthworms in 1881, just before his death. He had been working on it for 44 years. When two psychologists, Howard Gruber and Sara Davis, studied Darwin and other celebrated artists and scientists they concluded that such overlapping interests were common.

Another team of psychologists, led by Mihaly Csikszentmihalyi, interviewed almost 100 exceptionally creative people from jazz pianist Oscar Peterson to science writer Stephen Jay Gould to double Nobel laureate, the physicist John Bardeen. Csikszentmihalyi is famous for developing the idea of “flow”, the blissful state of being so absorbed in a challenge that one loses track of time and sets all distractions to one side. Yet every one of Csikszentmihalyi’s interviewees made a practice of keeping several projects bubbling away simultaneously.

Just internet addiction?

If the word “multitasking” can apply to both Darwin and a teenager with a serious Instagram habit, there is probably some benefit in defining our terms. There are at least four different things we might mean when we talk about multitasking. One is genuine multitasking: patting your head while rubbing your stomach; playing the piano and singing; farting while chewing gum. Genuine multitasking is possible, but at least one of the tasks needs to be so practised as to be done without thinking.

Then there’s the challenge of creating a presentation for your boss while also fielding phone calls for your boss and keeping an eye on email in case your boss wants you. This isn’t multitasking in the same sense. A better term is task switching, as our attention flits between the presentation, the telephone and the inbox. A great deal of what we call multitasking is in fact rapid task switching.

Task switching is often confused with a third, quite different activity — the guilty pleasure of disappearing down an unending click-hole of celebrity gossip and social media updates. There is a difference between the person who reads half a page of a journal article, then stops to write some notes about a possible future project, then goes back to the article — and someone who reads half a page of a journal article before clicking on bikini pictures for the rest of the morning. “What we’re often calling multitasking is in fact internet addiction,” says Shelley Carson, a psychologist and author of Your Creative Brain. “It’s a compulsive act, not an act of multitasking.”

A final kind of multitasking isn’t a way of getting things done but simply the condition of having a lot of things to do. The car needs to be taken in for a service. Your tooth is hurting. The nanny can’t pick up the kids from school today. There’s a big sales meeting to prepare for tomorrow, and your tax return is due next week. There are so many things that have to be done, so many responsibilities to attend to. Having a lot of things to do is not the same as doing them all at once. It’s just life. And it is not necessarily a stumbling block to getting things done — as Bernice Eiduson discovered as she tracked scientists on their way to their Nobel Prizes.

The fight for focus

These four practices — multitasking, task switching, getting distracted and managing multiple projects — all fit under the label “multitasking”. This is not just because of a simple linguistic confusion. The versatile networked devices we use tend to blur the distinction, serving us as we move from task to task while also offering an unlimited buffet of distractions. But the different kinds of multitasking are linked in other ways too. In particular, the highly productive practice of having multiple projects invites the less-than-productive habit of rapid task switching.

To see why, consider a story that psychologists like to tell about a restaurant near Berlin University in the 1920s. (It is retold in Willpower, a book by Roy Baumeister and John Tierney.) The story has it that when a large group of academics descended upon the restaurant, the waiter stood and calmly nodded as each new item was added to their complicated order. He wrote nothing down, but when he returned with the food his memory had been flawless. The academics left, still talking about the prodigious feat; but when one of them hurried back to retrieve something he’d left behind, the waiter had no recollection of him. How could the waiter have suddenly become so absent-minded? “Very simple,” he said. “When the order has been completed, I forget it.”

One member of the Berlin school was a young experimental psychologist named Bluma Zeigarnik. Intrigued, she demonstrated that people have a better recollection of uncompleted tasks. This is called the “Zeigarnik effect”: when we leave things unfinished, we can’t quite let go of them mentally. Our subconscious keeps reminding us that the task needs attention.

The Zeigarnik effect may explain the connection between facing multiple responsibilities and indulging in rapid task switching. We flit from task to task to task because we can’t forget about all of the things that we haven’t yet finished. We flit from task to task to task because we’re trying to get the nagging voices in our head to shut up.

Of course, there is much to be said for “focus”. But there is much to be said for copperplate handwriting, too, and for having a butler. The world has moved on. There’s something appealing about the Hemingwrite and the hotel room that will make the internet go away, but also something futile.

It is probably not true that Facebook is all that stands between you and literary greatness. And in most office environments, the Hemingwrite is not the tool that will win you promotion. You are not Ernest Hemingway, and you do not get to simply ignore emails from your colleagues.

If focus is going to have a chance, it’s going to have to fight an asymmetric war. Focus can only survive if it can reach an accommodation with the demands of a multitasking world.

Loops and lists

The word “multitasking” wasn’t applied to humans until the 1990s, but it has been used to describe computers for half a century. According to the Oxford English Dictionary, it was first used in print in 1966, when the magazine Datamation described a computer capable of appearing to perform several operations at the same time.

Just as with humans, computers typically create the illusion of multitasking by switching tasks rapidly. Computers perform the switching more quickly, of course, and they don’t take 20 minutes to get back on track after an interruption.

Nor does a computer fret about what is not being done. While rotating a polygon and sending text to the printer, it feels no guilt that the mouse has been left unchecked for the past 16 milliseconds. The mouse’s time will come. Being a computer means never having to worry about the Zeigarnik effect.

Is there a lesson in this for distractible sacks of flesh like you and me? How can we keep a sense of control despite the incessant guilt of all the things we haven’t finished?

“Whenever you say to someone, ‘I’ll get back to you about that’, you just opened a loop in your brain,” says David Allen. Allen is the author of a cult productivity book called Getting Things Done. “That loop will keep spinning until you put a placeholder in a system you can trust.”

Modern life is always inviting us to open more of those loops. It isn’t necessarily that we have more work to do, but that we have more kinds of work that we ought to be doing at any given moment. Tasks now bleed into each other unforgivingly. Whatever we’re doing, we can’t escape the sense that perhaps we should be doing something else. It’s these overlapping possibilities that take the mental toll.

The principle behind Getting Things Done is simple: close the open loops. The details can become rather involved but the method is straightforward. For every single commitment you’ve made to yourself or to someone else, write down the very next thing you plan to do. Review your lists of next actions frequently enough to give you confidence that you won’t miss anything.

This method has a cult following, and practical experience suggests that many people find it enormously helpful — including me (see below). Only recently, however, did the psychologists E J Masicampo and Roy Baumeister find some academic evidence to explain why people find relief by using David Allen’s system. Masicampo and Baumeister found that you don’t need to complete a task to banish the Zeigarnik effect. Making a specific plan will do just as well. Write down your next action and you quiet that nagging voice at the back of your head. You are outsourcing your anxiety to a piece of paper.

A creative edge?

It is probably a wise idea to leave rapid task switching to the computers. Yet even frenetic flipping between Facebook, email and a document can have some benefits alongside the costs.

The psychologist Shelley Carson and her student Justin Moore recently recruited experimental subjects for a test of rapid task switching. Each subject was given a pair of tasks to do: crack a set of anagrams and read an article from an academic journal. These tasks were presented on a computer screen, and for half of the subjects they were presented sequentially — first solve the anagrams, then read the article. For the other half of the experimental group, the computer switched every two-and-a-half minutes between the anagrams and the journal article, forcing the subjects to change mental gears many times.

Unsurprisingly, task switching slowed the subjects down and scrambled their thinking. They solved fewer anagrams and performed poorly on a test of reading comprehension when forced to refocus every 150 seconds.

But the multitasking treatment did have a benefit. Subjects who had been task switching became more creative. To be specific, their scores on tests of “divergent” thinking improved. Such tests ask subjects to pour out multiple answers to odd questions. They might be asked to think of as many uses as possible for a rolling pin or to list all the consequences they could summon to mind of a world where everyone has three arms. Involuntary multitaskers produced a greater volume and variety of answers, and their answers were more original too.

“It seems that switching back and forth between tasks primed people for creativity,” says Carson, who is an adjunct professor at Harvard. The results of her work with Moore have not yet been published, and one might reasonably object that such tasks are trivial measures of creativity. Carson responds that scores on these laboratory tests of divergent thinking are correlated with substantial creative achievements such as publishing a novel, producing a professional stage show or creating an award-winning piece of visual art. For those who insist that great work can only be achieved through superhuman focus, think long and hard on this discovery.

Carson and colleagues have found an association between significant creative achievement and a trait psychologists term “low latent inhibition”. Latent inhibition is the filter that all mammals have that allows them to tune out apparently irrelevant stimuli. It would be crippling to listen to every conversation in the open-plan office and the hum of the air conditioning, while counting the number of people who walk past the office window. Latent inhibition is what saves us from having to do so. These subconscious filters let us walk through the world without being overwhelmed by all the different stimuli it hurls at us.

And yet people whose filters are a little bit porous have a big creative edge. Think on that, uni-taskers: while you busily try to focus on one thing at a time, the people who struggle to filter out the buzz of the world are being reviewed in The New Yorker.

“You’re letting more information into your cognitive workspace, and that information can be consciously or unconsciously combined,” says Carson. Two other psychologists, Holly White and Priti Shah, found a similar pattern for people suffering from attention deficit hyperactivity disorder (ADHD).

It would be wrong to romanticise potentially disabling conditions such as ADHD. All these studies were conducted on university students, people who had already demonstrated an ability to function well. But their conditions weren’t necessarily trivial — to participate in the White/Shah experiment, students had to have a clinical diagnosis of ADHD, meaning that their condition was troubling enough to prompt them to seek professional help.

It’s surprising to discover that being forced to switch tasks can make us more creative. It may be still more surprising to realise that in an age where we live under the threat of constant distraction, people who are particularly prone to being distracted are flourishing creatively.

Perhaps we shouldn’t be entirely surprised. It’s easier to think outside the box if the box is full of holes. And it’s also easier to think outside the box if you spend a lot of time clambering between different boxes. “The act of switching back and forth can grease the wheels of thought,” says John Kounios, a professor of psychology at Drexel University.

Kounios, who is co-author of The Eureka Factor, suggests that there are at least two other potentially creative mechanisms at play when we switch between tasks. One is that the new task can help us forget bad ideas. When solving a creative problem, it’s easy to become stuck because we think of an incorrect solution but simply can’t stop returning to it. Doing something totally new induces “fixation forgetting”, leaving us free to find the right answer.

Another is “opportunistic assimilation”. This is when the new task prompts us to think of a solution to the old one. The original Eureka moment is an example.

As the story has it, Archimedes was struggling with the task of determining whether a golden wreath truly was made of pure gold without damaging the ornate treasure. The solution was to determine whether the wreath had the same volume as a pure gold ingot with the same mass; this, in turn, could be done by submerging both the wreath and the ingot to see whether they displaced the same volume of water.

This insight, we are told, occurred to Archimedes while he was having a bath and watching the water level rise and fall as he lifted himself in and out. And if solving such a problem while having a bath isn’t multitasking, then what is?

Tim Harford is an FT columnist. His latest book is ‘The Undercover Economist Strikes Back’. Twitter: @TimHarford

Six ways to be a master of multitasking

1. Be mindful

“The ideal situation is to be able to multitask when multitasking is appropriate, and focus when focusing is important,” says psychologist Shelley Carson. Tom Chatfield, author of Live This Book, suggests making two lists, one for activities best done with internet access and one for activities best done offline. Connecting and disconnecting from the internet should be deliberate acts.

2. Write it down

The essence of David Allen’s Getting Things Done is to turn every vague guilty thought into a specific action, to write down all of the actions and to review them regularly. The point, says Allen, is to feel relaxed about what you’re doing — and about what you’ve decided not to do right now — confident that nothing will fall through the cracks.

3. Tame your smartphone

The smartphone is a great servant and a harsh master. Disable needless notifications — most people don’t need to know about incoming tweets and emails. Set up a filing system within your email so that when a message arrives that requires a proper keyboard to answer — ie 50 words or more — you can move that email out of your inbox and place it in a folder where it will be waiting for you when you fire up your computer.

4. Focus in short sprints

The “Pomodoro Technique” — named after a kitchen timer — alternates focusing for 25 minutes and breaking for five minutes, across two-hour sessions. Productivity guru Merlin Mann suggests an “email dash”, where you scan email and deal with urgent matters for a few minutes each hour. Such ideas let you focus intensely while also switching between projects several times a day.

5. Procrastinate to win

If you have several interesting projects on the go, you can procrastinate over one by working on another. (It worked for Charles Darwin.) A change is as good as a rest, they say — and as psychologist John Kounios explains, such task switching can also unlock new ideas.

6. Cross-fertilise

“Creative ideas come to people who are interdisciplinary, working across different organisational units or across many projects,” says author and research psychologist Keith Sawyer. (Appropriately, Sawyer is also a jazz pianist, a former management consultant and a sometime game designer for Atari.) Good ideas often come when your mind makes unexpected connections between different fields.

Tim Harford’s To-Do Lists

David Allen’s Getting Things Done system — or GTD — has reached the status of a religion among some productivity geeks. At its heart, it’s just a fancy to-do list, but it’s more powerful than a regular list because it’s comprehensive, specific and designed to prompt you when you need prompting. Here’s how I make the idea work for me.

Write everything down. I use Google Calendar for appointments and an electronic to-do list called Remember the Milk, plus an ad hoc daily list on paper. The details don’t matter. The principle is never to carry a mental commitment around in your head.

Make the list comprehensive. Mine currently has 151 items on it. (No, I don’t memorise the number. I just counted.)

Keep the list fresh. The system works its anxiety-reducing magic best if you trust your calendar and to-do list to remind you when you need reminding. I spend about 20 minutes once a week reviewing the list to note incoming deadlines and make sure the list is neither missing important commitments nor cluttered with stale projects. Review is vital — the more you trust your list, the more you use it. The more you use it, the more you trust it.

List by context as well as topic. It’s natural to list tasks by topic or project — everything associated with renovating the spare room, for instance, or next year’s annual away-day. I also list them by context (this is easy on an electronic list). Things I can do when on a plane; things I can only do when at the shops; things I need to talk about when I next see my boss.

Be specific about the next action. If you’re just writing down vague reminders, the to-do list will continue to provoke anxiety. Before you write down an ill-formed task, take the 15 seconds required to think about exactly what that task is.

Written for and first published at ft.com.

The economists’ manifesto

If Britain’s top economists were in charge, what policies would they implement? Tim Harford sets the challenge

It’s often said that economists have too much influence on policy. A critic might say that politicians are dazzled by data-driven arguments and infatuated with the free-market-fetishising practitioners of the dismal science. As a card-carrying economist, I have never been convinced that politicians are the puppets of economists. Still, the idea seemed worth exploring, so I called up some of the country’s most respected economists and presented them with this scenario: after the election, the new prime minister promises to throw his weight behind any policy you choose. What would you suggest?

My selection of economists was mainstream — no Marxists or libertarians — but arbitrary. There is no pretence of a representative survey here. But there were common threads, some of which may surprise.

Let’s start with the deficit which, if we are to judge by column inches alone, is the single most important economic issue facing the country. Yet with the chance to push any policy they wished, none of my economic advisers expressed any concern about it. Indeed several wanted some form of increased spending and were happy to see that financed through borrowing or even printing money.

Economists have a reputation for being low-tax, free-market champions. Yet none of my panel fretted about red tape, proposed any tax cuts or mentioned free trade. Other untouched issues included the National Health Service, immigration and membership of the EU. Nobody suggested any changes to the way banks are regulated or taxed.

Less surprising is that several economists suggested structures that would put decision making at arm’s length from politicians, delegating it to technocrats with the expertise and incentives to do what is right for Britain. The technocracy already has several citadels: the Bank of England’s Monetary Policy Committee, the National Institute for Health and Care Excellence, the Competition Commission and the regulators of privatised utilities. My advisers wanted more of this. That is economics for you: when a political genie offers you whatever policy wish you desire, why not simply wish to have more wishes?

Nick Stern
Former chief economist of the World Bank, professor at the London School of Economics

Nick Stern will forever be associated with the Stern Review, a report into the economics of climate change published in 2006. He hasn’t stopped banging this drum but these days he is reframing the problem as an opportunity.

“I would launch a strategy for UK cities to be the most attractive, productive and cleanest in the world,” he says. Cities hold out the hope of being productive and desirable places to live as well as environmentally efficient ones. Consider Manhattan: it is rich, iconic and, with small apartments and a subway, it boasts a much smaller environmental footprint than most of sprawling, car-loving America.

That is the aim. But what is the policy? Lord Stern offers what he calls a “collection of policies”, including an expanded green infrastructure bank and more funding for green technology. His broadest stroke is to change the governance of British cities, devolving the power to raise taxes and borrow money but imposing strong national standards on energy efficiency.

Stern would introduce a platform for congestion charging to enable cities to develop areas connected by public transport and walking/cycling routes. He’d also raise the price of emitting carbon via a direct tax or an emissions trading system. Stern suggests £25/ton of CO2, and rising. That should add a penny to the price of 100g of airfreighted vegetables and £100-£200 to a household energy bill. It would raise £10bn, less than income tax or VAT but enough to narrow the deficit or allow other tax cuts.

But Stern doesn’t dwell on taxation. His policies are “long on UK strengths such as entrepreneurship, architecture and planning”, he says. While warning of the “deep deep dangers” of climate change, he claims his package “is attractive in its own right”.

. . .

Tim’s verdict Developing these new green city centres is a challenge. Are our urban planners up to it?

• Political feasibility 3 out of 5
• Economic radicalism 3 out of 5

Jonathan Haskel
Professor of economics at Imperial College, London

“Some people think that scientists have their heads in the sky, and if you gave them more government money they would simply do weirder research,” says Jonathan Haskel. Science enthusiasts, however, would say that weird research can help: Sir Andre Geim of the University of Manchester won the Nobel Prize for his discovery of the revolutionary material graphene — but not before receiving the Ig Nobel Prize for levitating a live frog.

Supporting scientific innovation has long been an easy sell for politicians. Who could be against technological progress, after all? The more difficult question is how to encourage this innovation. For Haskel, the answer is straightforward: the government should simply spend more money directly funding scientific research. At the moment the government gives about £3bn to research councils and more than £2bn to the Higher Education Funding Council. For the sake of being specific, Haskel was happy to accept my suggestion of simply doubling this funding over the course of a five-year parliament.

Haskel’s research finds that government funding of science is the perfect complement to private, practically minded research funding. “This is an example of crowding in,” he says, meaning that if the government spends more on scientific research it is likely to draw in private funding too. There is a high correlation between the research scientists who receive government grants and those collaborating with or being funded by private sector companies. Haskel has found that sectors that attract government funding are also sectors with high productivity growth.

According to Haskel’s estimates, the rate of return on basic scientific research is around 20 per cent at current funding levels — a level that would not displease Warren Buffett himself. This would probably be less if science funding dramatically increased but, even at 15 or 10 per cent return, the case for spending more would be persuasive.

An extra £5bn is not trivial. Increasing all income tax rates by one percentage point, or raising VAT to 21 per cent, would cover the cost. But given current ultra-low interest rates, Haskel says he is happy for the government to borrow to fund this spending instead. “I would regard borrowing to fund the science base as a form of infrastructure investment,” he says. It may not be the traditional Keynesian infrastructure of roads and runways but it is investment for the future nonetheless.

. . .

Tim’s verdict It’s hard to object to scientific progress and Haskel’s evidence is persuasive. Leave a bit of cash for the social scientists, please.

• Political feasibility 4 out of 5
• Economic radicalism 2 out of 5

Gemma Tetlow
Top pensions expert at the Institute for Fiscal Studies

A confession: I may have led Gemma Tetlow astray. We begin by discussing how employers can avoid national insurance contributions by diverting some of their workers’ salaries into a pension. This, she says, is an unwarranted subsidy for the comfortably off, and abolishing the rule is “not a bad way to raise £11bn”.

As we talk, a bolder thought forms in my mind: why not just abolish national insurance entirely and replace it with higher rates of income tax? That would close Tetlow’s pension loophole and many other inconsistencies besides. I wonder if I have missed something obvious. Apparently not. Tetlow is perfectly happy to endorse the idea of a merger.

In some ways this would be a huge change: national insurance raises more than half as much as income tax does, so merging the two would mean huge increases in headline income tax rates. But while the policy would make things simpler and more transparent, it would not greatly alter the tax that most people pay.

The idea is tempting to an economist because successive governments have discovered a feat of political arbitrage. They reduce the basic rate of income tax, which gets a lot of attention, while increasing national insurance rates, which do not. Since 1979 the basic rate of income tax has fallen from 30 to 20 per cent but national insurance rates have risen and so the marginal tax rate on much employment income is still above 45 per cent, much the same as ever. A bit more honesty about this would be welcome.

As Tetlow explains, national insurance was once a contributory system, designed to cope with a male workforce in conditions of near full employment. Now national insurance is more like an income tax, where people pay if they can afford to and receive the benefits in any case.

So Tetlow and I agree that the system could comfortably be scrapped — even if we might be looking to replace it with a basic rate of income tax at 40 per cent or so. The benefits? Transparency, administrative simplicity and the end of a few unwelcome loopholes. The risks are chiefly administrative, although a decision would need to be made about whether to have a special income tax rate for people above the state pension age, who currently pay no national insurance.

Could it happen? Tetlow says she would be “astonished” if it did. Perhaps governments are too fond of pretending that the true basic rate of income tax is just 20 per cent.

. . .

Tim’s verdict I bounced Tetlow into this so can hardly object. But for our politicians, the confusion over national insurance isn’t a problem, it’s an opportunity.

• Political feasibility 2 out of 5
• Economic radicalism 1 out of 5

Simon Wren-Lewis
Macroeconomist at Merton College, Oxford

Simon Wren-Lewis has a growing audience as a trenchant critic of George Osborne’s fiscal contraction. I had expected him to make the case that the incoming government should spend more but he has something more radical in mind.

“We’re passing the period when the damage was done,” he says. For Wren-Lewis, the policy error was to tighten the fiscal screws in 2010 and 2011 — he estimates that with lower taxes and higher spending the economy today would be about 4 per cent larger, while deficit reduction could wait until the economy was stronger.

Cutting spending in a severe but temporary downturn is macroeconomically perverse but makes good sense to voters, so Wren-Lewis feels that a future government would make a similar mistake in similar circumstances. What to do?

Economists have faced a related problem before. When politicians controlled interest rates they were always tempted to cut rates before elections, overheating the economy and leading to inflation once the election was safely gone. The solution was to delegate control of monetary policy to the technocrats, as when Gordon Brown gave the Bank of England this power in 1997.

“That was a good idea,” says Wren-Lewis. But, he adds, “it was always incomplete.” The missing piece of the puzzle was what the bank should do when interest rates are nearly zero — as now — and cannot be cut further to stimulate the economy. The usual solution is a fiscal expansion — cutting taxes and increasing spending, just what George Osborne has shied away from. Wren-Lewis’s response: in future, the Bank of England should print the money and hand it to the government on condition it be used for a fiscal expansion.

This is radical — but not without precedent. Economists from Adair Turner to Ben Bernanke (in 2003) and Milton Friedman (1948) have argued that deficits could be financed by printing money rather than issuing government debt. Funding real spending from paper money might seem like nonsense: if the economy is working well, creating too much money will produce inflation. But when the economy is slack, judicious money printing can turn the waste of a depressed economy into useful output, without dangerous inflation. This is a rare free lunch.

The radicalism of Wren-Lewis’s proposal lies less in the economics than the politics: the idea that the Bank of England would decide a fiscal expansion was needed, and shove a reluctant, democratically elected government into it.

Wren-Lewis calls his idea “democratic helicopter money”. He feels the government should decide whether the stimulus takes the form of tax cuts, increased benefits or new infrastructure. But the actual decision to cut taxes and raise spending to stimulate the economy? Not something one should leave to the politicians.

. . .

Tim’s verdict I sympathise with Wren-Lewis’s wish for more stimulus spending in the last parliament but outsourcing such a basic democratic responsibility feels too bold to me.

• Political feasibility 1 out of 5
• Economic radicalism 5 out of 5

Diane Coyle
Professor of economics at the University of Manchester

“My starting point is that the extent of income inequality has got too big,” says Coyle. She points to median annual full-time earnings of just over £27,000, while the average pay of FTSE 100 chief executives is — according to Manifest, a proxy voting service — about £4.7m. “If you were to ask me whether the productivity of chief executives is really that much higher, my answer is no. Something has gone wrong with the way the market is operating here.”

That market failure is easy to diagnose: it is hard for dispersed shareholders to monitor what is going on and to insist on a more rigorous approach. So Coyle would give them a little help.

The most eye-catching suggestion is that companies should publish the ratio between what the chief executive is paid and what the median worker in the company is paid. A review body would give a strong steer as to how high that ratio could reasonably go (“I don’t know what the right number is,” says Coyle) and companies who did not comply would face unwanted scrutiny from shareholders, employees, unions and politicians.

“Just talking about this much more would start to shift the social norm,” says Coyle, who in her term on the BBC Trust (soon to end) has seen the BBC start to publish these pay ratios, which have been falling. Coyle wants a binding rule, too, that companies should not be able to pay performance bonuses linked to share price. That is too easily manipulated. Instead, these must be linked to indicators such as customer satisfaction, sales or profits.

. . .

Tim’s verdict Shareholders and citizens alike should welcome pay that is tied more closely to good management decisions. But can any of this be legislated effectively?

• Political feasibility 4 out of 5
• Economic radicalism 4 out of 5

John van Reenen
Professor at the London School of Economics

“Low productivity is the number one problem Britain faces,” says Van Reenen. Even before the crisis, it lagged behind other rich countries. The latest data suggest UK output per hour worked is 30 per cent below US levels, and 17 per cent below the G7 average (at purchasing power parity).

Such a problem has no single solution but Van Reenen wants to focus on a lack of investment in the UK’s core infrastructure — housing, energy and transport. As the FT reported recently, government capital investment has fallen by a third between 2009-10 and 2013-14, despite repeated statements by the chancellor that infrastructure is at the heart of plans for growth.

Milton Keynes, the last of the “new towns”, harks back to 1967 and has 100,000 dwellings. That gives some perspective on recent proposals to build a “garden city” at Ebbsfleet of a mere 15,000 homes. If the Barker Review’s headline number of 245,000 new homes a year is to be achieved, we need an Ebbsfleet every three weeks and have done for the past 12 years.

The HS2 high-speed rail line was first examined in 1999 and is still unlikely to be finished 30 years after that date. An observer might feel the project should either have been cancelled or completed by now. And let’s not dwell on London’s airports: in 1971 the Roskill Commission proposed a major new airport north of Aylesbury after rejecting the idea of building one in the Thames estuary. We are still weighing up the issues.

“I would propose to radically change the whole way we deliver infrastructure projects,” says Van Reenen, “with a new institutional architecture for making decisions.” There are three elements to this. First, an infrastructure strategy board to recommend long-term priorities, which would be voted up or down by parliament. Second, an infrastructure planning commission to meet those priorities and arrange compensation to those affected by the march of progress. Third, an infrastructure bank to help finance projects by borrowing from capital markets and investing alongside private-sector banks.

If this seems anti-democratic, Van Reenen’s defence is that his approach “puts politics in the right place”. MPs are concerned with the short-term and the local, which causes problems with long-term investments of national significance. Like Simon Wren-Lewis, John van Reenen has more faith in technocrats than politicians.

. . .

Tim’s verdict An approach that seems justified when facing such a chronic and serious problem. Sign me up.

• Political feasibility 3 out of 5
• Economic radicalism 1 out of 5

Kate Barker
Author of the 2004 Barker Review of Housing Supply

I am expecting Dame Kate Barker to propose something controversial but straightforward: that we should build more houses. It is, after all, her report that policy wonks have been citing for the past decade whenever they want a number for how many houses England needs. Instead, some of her solutions “are so unpopular I can hardly bring myself to suggest them to you”. This is music to my ears.

In 2002/2003, the private sector completed 125,000 houses in England; the Barker Review argued that number needed to almost double to reduce the growth in real house prices to the EU’s long-term average. But the number of private-sector housing completions in England has fallen to below 100,000 a year from 2009 through 2014. The trickle of new houses is manifestly failing to accommodate population growth.

So: more houses? Not necessarily. Barker lays out three options. The first is the status quo. It is not attractive. There will be an increasing divide between the housing haves (who enjoy capital appreciation) and the housing have-nots (who find it ever harder to buy a home).

Option two is a dramatic programme of house building, which seems logical. “We’ve built much less than the top-line number associated with my name,” says Barker. “I haven’t changed my view that we need to do more.” But she is sceptical about how feasible it is to expect house building on the scale needed, given the strength of opposition to development. She has had the ear of prime ministers before, after all, and not much changed.

And so to option three: resign ourselves to not building enough houses to meet demand, and use the tax system to soften the blow. Meaning what, exactly?

Consider someone with the finance and good fortune to buy a home in London in 1992. That person has enjoyed an enormous increase in the real value of her house. But she has paid surprisingly little tax on the windfall. Council tax is proportionately lower on expensive homes. Capital gains tax does not apply to people living in their own homes. If you become a millionaire through skill, effort or entrepreneurial spirit, you will be taxed. If you do it by buying a house in Islington at the right moment, your bounty is yours to keep in its entirety. That’s inequitable and the inequity is likely to last from one generation to the next.

Barker suggests two thrusts to the tax reform, and “ideally we would do both”. The first is to replace council tax with a land value tax. This would tax expensive homes more heavily, in line with their value, and encourage valuable land to be used intensively. But it would also weigh heavily on elderly widows living alone in large houses. The second is to charge capital gains tax on people’s principal residency. If you live in your own home and its price starts to soar, you will be taxed.

But both these reforms are complicated. A land tax would require frequent revaluations. The capital gains tax reform would require some sort of system for postponing the bill until death or entry into a retirement home. That is fiddly but the alternative might make it impossible to move house without a punitive tax charge.

As Baker admits, this is dramatic and unpopular stuff. The people who lose out are clearly identifiable and politically influential. But the same is true of the straightforward proposal to build many more houses. The UK’s housing problem seems to be the toughest of political tough nuts.

Tim’s verdict This makes sense despite the difficulties. But Barker identified the cure for unaffordable housing more than 10 years ago — build more houses. It’s depressing that she now has to advocate palliative measures instead.

• Political feasibility 1 out of 5
• Economic radicalism 2 out of 5

. . .

Last word

So what would the UK look like with my board of economists in charge? We’d have more borrowing and considerably more investment — in housing, in big infrastructure, in science and in green cities. Taxes seem unlikely to fall but they would be rationalised, with a focus on energy efficiency and a transparent taxation of income and housing wealth. Inequality would be in the spotlight.

The economists seem happy to leave the politicians to their usual arguments about the EU, immigration, the price of beer and the problem of tax-dodging. Noting that every party makes similar promises about funding the National Health Service, the economists have let it be.

Perhaps that is for the best because if the economists have their way, one big thing will change after the election: politicians will be kept at a safe distance from the decisions that matter.

Written for and first published at ft.com.

How to see into the future

Billions of dollars are spent on experts who claim they can forecast what’s around the corner, in business, finance and economics. Most of them get it wrong. Now a groundbreaking study has unlocked the secret: it IS possible to predict the future – and a new breed of ‘superforecasters’ knows how to do it

Irving Fisher was once the most famous economist in the world. Some would say he was the greatest economist who ever lived. “Anywhere from a decade to two generations ahead of his time,” opined the first Nobel laureate economist Ragnar Frisch, in the late 1940s, more than half a century after Fisher’s genius first lit up his subject. But while Fisher’s approach to economics is firmly embedded in the modern discipline, many of those who remember him now know just one thing about him: that two weeks before the great Wall Street crash of 1929, Fisher announced, “Stocks have reached what looks like a permanently high plateau.”

In the 1920s, Fisher had two great rivals. One was a British academic: John Maynard Keynes, a rising star and Fisher’s equal as an economic theorist and policy adviser. The other was a commercial competitor, an American like Fisher. Roger Babson was a serial entrepreneur with no serious academic credentials, inspired to sell economic forecasts by the banking crisis of 1907. As Babson and Fisher locked horns over the following quarter-century, they laid the foundations of the modern economic forecasting industry.

Fisher’s rivals fared better than he did. Babson foretold the crash and made a fortune, enough to endow the well-respected Babson College. Keynes was caught out by the crisis but recovered and became rich anyway. Fisher died in poverty, ruined by the failure of his forecasts.

If Fisher and Babson could see the modern forecasting industry, it would have astonished them in its scale, range and hyperactivity. In his acerbic book The Fortune Sellers, former consultant William Sherden reckoned in 1998 that forecasting was a $200bn industry – $300bn in today’s terms – and the bulk of the money was being made in business, economic and financial forecasting.

It is true that forecasting now seems ubiquitous. Data analysts forecast demand for new products, or the impact of a discount or special offer; scenario planners (I used to be one) produce broad-based narratives with the aim of provoking fresh thinking; nowcasters look at Twitter or Google to track epidemics, actual or metaphorical, in real time; intelligence agencies look for clues about where the next geopolitical crisis will emerge; and banks, finance ministries, consultants and international agencies release regular prophecies covering dozens, even hundreds, of macroeconomic variables.

Real breakthroughs have been achieved in certain areas, especially where rich datasets have become available – for example, weather forecasting, online retailing and supply-chain management. Yet when it comes to the headline-grabbing business of geopolitical or macroeconomic forecasting, it is not clear that we are any better at the fundamental task that the industry claims to fulfil – seeing into the future.

So why is forecasting so difficult – and is there hope for improvement? And why did Babson and Keynes prosper while Fisher suffered? What did they understand that Fisher, for all his prodigious talents, did not?

In 1987, a young Canadian-born psychologist, Philip Tetlock, planted a time bomb under the forecasting industry that would not explode for 18 years. Tetlock had been trying to figure out what, if anything, the social sciences could contribute to the fundamental problem of the day, which was preventing a nuclear apocalypse. He soon found himself frustrated: frustrated by the fact that the leading political scientists, Sovietologists, historians and policy wonks took such contradictory positions about the state of the cold war; frustrated by their refusal to change their minds in the face of contradictory evidence; and frustrated by the many ways in which even failed forecasts could be justified. “I was nearly right but fortunately it was Gorbachev rather than some neo-Stalinist who took over the reins.” “I made the right mistake: far more dangerous to underestimate the Soviet threat than overestimate it.” Or, of course, the get-out for all failed stock market forecasts, “Only my timing was wrong.”

Tetlock’s response was patient, painstaking and quietly brilliant. He began to collect forecasts from almost 300 experts, eventually accumulating 27,500. The main focus was on politics and geopolitics, with a selection of questions from other areas such as economics thrown in. Tetlock sought clearly defined questions, enabling him with the benefit of hindsight to pronounce each forecast right or wrong. Then Tetlock simply waited while the results rolled in – for 18 years.

Tetlock published his conclusions in 2005, in a subtle and scholarly book, Expert Political Judgment. He found that his experts were terrible forecasters. This was true in both the simple sense that the forecasts failed to materialise and in the deeper sense that the experts had little idea of how confident they should be in making forecasts in different contexts. It is easier to make forecasts about the territorial integrity of Canada than about the territorial integrity of Syria but, beyond the most obvious cases, the experts Tetlock consulted failed to distinguish the Canadas from the Syrias.

Adding to the appeal of this tale of expert hubris, Tetlock found that the most famous experts fared somewhat worse than those outside the media spotlight. Other than that, the humiliation was evenly distributed. Regardless of political ideology, profession and academic training, experts failed to see into the future.

Most people, hearing about Tetlock’s research, simply conclude that either the world is too complex to forecast, or that experts are too stupid to forecast it, or both. Tetlock himself refused to embrace cynicism so easily. He wanted to leave open the possibility that even for these intractable human questions of macroeconomics and geopolitics, a forecasting approach might exist that would bear fruit.

. . .

In 2013, on the auspicious date of April 1, I received an email from Tetlock inviting me to join what he described as “a major new research programme funded in part by Intelligence Advanced Research Projects Activity, an agency within the US intelligence community.”

The core of the programme, which had been running since 2011, was a collection of quantifiable forecasts much like Tetlock’s long-running study. The forecasts would be of economic and geopolitical events, “real and pressing matters of the sort that concern the intelligence community – whether Greece will default, whether there will be a military strike on Iran, etc”. These forecasts took the form of a tournament with thousands of contestants; it is now at the start of its fourth and final annual season.

“You would simply log on to a website,” Tetlock’s email continued, “give your best judgment about matters you may be following anyway, and update that judgment if and when you feel it should be. When time passes and forecasts are judged, you could compare your results with those of others.”

I elected not to participate but 20,000 others have embraced the idea. Some could reasonably be described as having some professional standing, with experience in intelligence analysis, think-tanks or academia. Others are pure amateurs. Tetlock and two other psychologists, Don Moore and Barbara Mellers, have been running experiments with the co-operation of this army of volunteers. (Mellers and Tetlock are married.) Some were given training in how to turn knowledge about the world into a probabilistic forecast; some were assembled into teams; some were given information about other forecasts while others operated in isolation. The entire exercise was given the name of the Good Judgment Project, and the aim was to find better ways to see into the future.

The early years of the forecasting tournament have, wrote Tetlock, “already yielded exciting results”.

A first insight is that even brief training works: a 20-minute course about how to put a probability on a forecast, correcting for well-known biases, provides lasting improvements to performance. This might seem extraordinary – and the benefits were surprisingly large – but even experienced geopolitical seers tend to have expertise in a subject, such as Europe’s economies or Chinese foreign policy, rather than training in the task of forecasting itself.

“For people with the right talents or the right tactics, it is possible to see into the future after all”

A second insight is that teamwork helps. When the project assembled the most successful forecasters into teams who were able to discuss and argue, they produced better predictions.

But ultimately one might expect the same basic finding as always: that forecasting events is basically impossible. Wrong. To connoisseurs of the frailties of futurology, the results of the Good Judgment Project are quite astonishing. Forecasting is possible, and some people – call them “superforecasters”– can predict geopolitical events with an accuracy far outstripping chance. The superforecasters have been able to sustain and even improve their performance.

The cynics were too hasty: for people with the right talents or the right tactics, it is possible to see into the future after all.

Roger Babson, Irving Fisher’s competitor, would always have claimed as much. A serial entrepreneur, Babson made his fortune selling economic forecasts alongside information about business conditions. In 1920, the Babson Statistical Organization had 12,000 subscribers and revenue of $1.35m – almost $16m in today’s money.

“After Babson, the forecaster was an instantly recognisable figure in American business,” writes Walter Friedman, the author of Fortune Tellers, a history of Babson, Fisher and other early economic forecasters. Babson certainly understood how to sell himself and his services. He advertised heavily and wrote prolifically. He gave a complimentary subscription to Thomas Edison, hoping for a celebrity endorsement. After contracting tuberculosis, Babson turned his management of the disease into an inspirational business story. He even employed stonecutters to carve inspirational slogans into large rocks in Massachusetts (the “Babson Boulders” are still there).

On September 5 1929, Babson made a speech at a business conference in Wellesley, Massachusetts. He predicted trouble: “Sooner or later a crash is coming which will take in the leading stocks and cause a decline of from 60 to 80 points in the Dow-Jones barometer.” This would have been a fall of around 20 per cent.

So famous had Babson become that his warning was briefly a self-fulfilling prophecy. When the news tickers of New York reported Babson’s comments at around 2pm, the markets erupted into what The New York Times described as “a storm of selling”. Shares lurched down by 3 per cent. This became known as the “Babson break”.

The next day, shares bounced back and Babson, for a few weeks, appeared ridiculous. On October 29, the great crash began, and within a fortnight the market had fallen almost 50 per cent. By then, Babson had an advertisement in the New York Times pointing out, reasonably, that “Babson clients were prepared”. Subway cars were decorated with the slogan, “Be Right with Babson”. For Babson, his forecasting triumph was a great opportunity to sell more subscriptions.

But his true skill was marketing, not forecasting. His key product, the “Babson chart”, looked scientific and was inspired by the discoveries of Isaac Newton, his idol. The Babson chart operated on the Newtonian assumption that any economic expansion would be matched by an equal and opposite contraction. But for all its apparent sophistication, the Babson chart offered a simple and usually contrarian message.

“Babson offered an up-arrow or a down-arrow. People loved that,” says Walter Friedman. Whether or not Babson’s forecasts were accurate was not a matter that seemed to concern many people. When he was right, he advertised the fact heavily. When he was wrong, few noticed. And Babson had indeed been wrong for many years during the long boom of the 1920s. People taking his advice would have missed out on lucrative opportunities to invest. That simply didn’t matter: his services were popular, and his most spectacularly successful prophecy was also his most famous.

Babson’s triumph suggests an important lesson: commercial success as a forecaster has little to do with whether you are any good at seeing into the future. No doubt it helped his case when his forecasts were correct but nobody gathered systematic information about how accurate he was. The Babson Statistical Organization compiled business and economic indicators that were, in all probability, of substantial value in their own right. Babson’s prognostications were the peacock’s plumage; their effect was simply to attract attention to the services his company provided.

. . .

When Barbara Mellers, Don Moore and Philip Tetlock established the Good Judgment Project, the basic principle was to collect specific predictions about the future and then check to see if they came true. That is not the world Roger Babson inhabited and neither does it describe the task of modern pundits.

When we talk about the future, we often aren’t talking about the future at all but about the problems of today. A newspaper columnist who offers a view on the future of North Korea, or the European Union, is trying to catch the eye, support an argument, or convey in a couple of sentences a worldview that would otherwise be impossibly unwieldy to explain. A talking head in a TV studio offers predictions by way of making conversation. A government analyst or corporate planner may be trying to justify earlier decisions, engaging in bureaucratic defensiveness. And many election forecasts are simple acts of cheerleading for one side or the other.

“Some people – call them ‘superforecasters’– can predict geopolitical events with an accuracy far outstripping chance”

Unlike the predictions collected by the Good Judgment Project, many forecasts are vague enough in their details to allow the mistaken seer off the hook. Even if it was possible to pronounce that a forecast had come true or not, only in a few hotly disputed cases would anybody bother to check.

All this suggests that among the various strategies employed by the superforecasters of the Good Judgment Project, the most basic explanation of their success is that they have the single uncompromised objective of seeing into the future – and this is rare. They receive continual feedback about the success and failure of every forecast, and there are no points for radicalism, originality, boldness, conventional pieties, contrarianism or wit. The project manager of the Good Judgment Project, Terry Murray, says simply, “The only thing that matters is the right answer.”

I asked Murray for her tips on how to be a good forecaster. Her reply was, “Keep score.”

. . .

An intriguing footnote to Philip Tetlock’s original humbling of the experts was that the forecasters who did best were what Tetlock calls “foxes” rather than “hedgehogs”. He used the term to refer to a particular style of thinking: broad rather than deep, intuitive rather than logical, self-critical rather than assured, and ad hoc rather than systematic. The “foxy” thinking style is now much in vogue. Nate Silver, the data journalist most famous for his successful forecasts of US elections, adopted the fox as the mascot of his website as a symbol of “a pluralistic approach”.

The trouble is that Tetlock’s original foxes weren’t actually very good at forecasting. They were merely less awful than the hedgehogs, who deployed a methodical, logical train of thought that proved useless for predicting world affairs. That world, apparently, is too complex for any single logical framework to encompass.

More recent research by the Good Judgment Project investigators leaves foxes and hedgehogs behind but develops this idea that personality matters. Barbara Mellers told me that the thinking style most associated with making better forecasts was something psychologists call “actively open-minded thinking”. A questionnaire to diagnose this trait invites people to rate their agreement or disagreement with statements such as, “Changing your mind is a sign of weakness.” The project found that successful forecasters aren’t afraid to change their minds, are happy to seek out conflicting views and are comfortable with the notion that fresh evidence might force them to abandon an old view of the world and embrace something new.

Which brings us to the strange, sad story of Irving Fisher and John Maynard Keynes. The two men had much in common: both giants in the field of economics; both best-selling authors; both, alas, enthusiastic and prominent eugenicists. Both had immense charisma as public speakers.

Fisher and Keynes also shared a fascination with financial markets, and a conviction that their expertise in macroeconomics and in economic statistics should lead to success as an investor. Both of them, ultimately, were wrong about this. The stock market crashes of 1929 – in September in the UK and late October in the US – caught each of them by surprise, and both lost heavily.

Yet Keynes is remembered today as a successful investor. This is not unreasonable. A study by David Chambers and Elroy Dimson, two financial economists, concluded that Keynes’s track record over a quarter century running the discretionary portfolio of King’s College Cambridge was excellent, outperforming market benchmarks by an average of six percentage points a year, an impressive margin.

This wasn’t because Keynes was a great economic forecaster. His original approach had been predicated on timing the business cycle, moving into and out of different investment classes depending on which way the economy itself was moving. This investment strategy was not a success, and after several years Keynes’s portfolio was almost 20 per cent behind the market as a whole.

The secret to Keynes’s eventual profits is that he changed his approach. He abandoned macroeconomic forecasting entirely. Instead, he sought out well-managed companies with strong dividend yields, and held on to them for the long term. This approach is now associated with Warren Buffett, who quotes Keynes’s investment maxims with approval. But the key insight is that the strategy does not require macroeconomic predictions. Keynes, the most influential macroeconomist in history, realised not only that such forecasts were beyond his skill but that they were unnecessary.

Irving Fisher’s mistake was not that his forecasts were any worse than Keynes’s but that he depended on them to be right, and they weren’t. Fisher’s investments were leveraged by the use of borrowed money. This magnified his gains during the boom, his confidence, and then his losses in the crash.

But there is more to Fisher’s undoing than leverage. His pre-crash gains were large enough that he could easily have cut his losses and lived comfortably. Instead, he was convinced the market would turn again. He made several comments about how the crash was “largely psychological”, or “panic”, and how recovery was imminent. It was not.

One of Fisher’s major investments was in Remington Rand – he was on the stationery company’s board after selling them his “Index Visible” invention, a type of Rolodex. The share price tells the story: $58 before the crash, $28 by 1930. Fisher topped up his investments – and the price soon dropped to $1.

Fisher became deeper and deeper in debt to the taxman and to his brokers. Towards the end of his life, he was a marginalised figure living alone in modest circumstances, an easy target for scam artists. Sylvia Nasar writes in Grand Pursuit, a history of economic thought, “His optimism, overconfidence and stubbornness betrayed him.”

. . .

So what is the secret of looking into the future? Initial results from the Good Judgment Project suggest the following approaches. First, some basic training in probabilistic reasoning helps to produce better forecasts. Second, teams of good forecasters produce better results than good forecasters working alone. Third, actively open-minded people prosper as forecasters.

But the Good Judgment Project also hints at why so many experts are such terrible forecasters. It’s not so much that they lack training, teamwork and open-mindedness – although some of these qualities are in shorter supply than others. It’s that most forecasters aren’t actually seriously and single-mindedly trying to see into the future. If they were, they’d keep score and try to improve their predictions based on past errors. They don’t.

“Successful forecasters aren’t afraid to change their minds and are comfortable with the notion that fresh evidence might mean abandoning an old view”

This is because our predictions are about the future only in the most superficial way. They are really advertisements, conversation pieces, declarations of tribal loyalty – or, as with Irving Fisher, statements of profound conviction about the logical structure of the world. As Roger Babson explained, not without sympathy, Fisher had failed because “he thinks the world is ruled by figures instead of feelings, or by theories instead of styles”.

Poor Fisher was trapped by his own logic, his unrelenting optimism and his repeated public declarations that stocks would recover. And he was bankrupted by an investment strategy in which he could not afford to be wrong.

Babson was perhaps wrong as often as he was right – nobody was keeping track closely enough to be sure either way – but that did not stop him making a fortune. And Keynes prospered when he moved to an investment strategy in which forecasts simply did not matter much.

Fisher once declared that “the sagacious businessman is constantly forecasting”. But Keynes famously wrote of long-term forecasts, “About these matters there is no scientific basis on which to form any calculable probability whatever. We simply do not know.”

Perhaps even more famous is a remark often attributed to Keynes. “When my information changes, I alter my conclusions. What do you do, sir?”

If only he had taught that lesson to Irving Fisher.

Also published at ft.com.

Big Data: Are we making a big mistake?

Five years ago, a team of researchers from Google announced a remarkable achievement in one of the world’s top scientific journals, Nature. Without needing the results of a single medical check-up, they were nevertheless able to track the spread of influenza across the US. What’s more, they could do it more quickly than the Centers for Disease Control and Prevention (CDC). Google’s tracking had only a day’s delay, compared with the week or more it took for the CDC to assemble a picture based on reports from doctors’ surgeries. Google was faster because it was tracking the outbreak by finding a correlation between what people searched for online and whether they had flu symptoms.

Not only was “Google Flu Trends” quick, accurate and cheap, it was theory-free. Google’s engineers didn’t bother to develop a hypothesis about what search terms – “flu symptoms” or “pharmacies near me” – might be correlated with the spread of the disease itself. The Google team just took their top 50 million search terms and let the algorithms do the work.
The success of Google Flu Trends became emblematic of the hot new trend in business, technology and science: “Big Data”. What, excited journalists asked, can science learn from Google?
As with so many buzzwords, “big data” is a vague term, often thrown around by people with something to sell. Some emphasise the sheer scale of the data sets that now exist – the Large Hadron Collider’s computers, for example, store 15 petabytes a year of data, equivalent to about 15,000 years’ worth of your favourite music.
But the “big data” that interests many companies is what we might call “found data”, the digital exhaust of web searches, credit card payments and mobiles pinging the nearest phone mast. Google Flu Trends was built on found data and it’s this sort of data that ­interests me here. Such data sets can be even bigger than the LHC data – Facebook’s is – but just as noteworthy is the fact that they are cheap to collect relative to their size, they are a messy collage of datapoints collected for disparate purposes and they can be updated in real time. As our communication, leisure and commerce have moved to the internet and the internet has moved into our phones, our cars and even our glasses, life can be recorded and quantified in a way that would have been hard to imagine just a decade ago.
Cheerleaders for big data have made four exciting claims, each one reflected in the success of Google Flu Trends: that data analysis produces uncannily accurate results; that every single data point can be captured, making old statistical sampling techniques obsolete; that it is passé to fret about what causes what, because statistical correlation tells us what we need to know; and that scientific or statistical models aren’t needed because, to quote “The End of Theory”, a provocative essay published in Wired in 2008, “with enough data, the numbers speak for themselves”.
Unfortunately, these four articles of faith are at best optimistic oversimplifications. At worst, according to David Spiegelhalter, Winton Professor of the Public Understanding of Risk at Cambridge university, they can be “complete bollocks. Absolute nonsense.”
Found data underpin the new internet economy as companies such as Google, Facebook and Amazon seek new ways to understand our lives through our data exhaust. Since Edward Snowden’s leaks about the scale and scope of US electronic surveillance it has become apparent that security services are just as fascinated with what they might learn from our data exhaust, too.
Consultants urge the data-naive to wise up to the potential of big data. A recent report from the McKinsey Global Institute reckoned that the US healthcare system could save $300bn a year – $1,000 per American – through better integration and analysis of the data produced by everything from clinical trials to health insurance transactions to smart running shoes.
But while big data promise much to scientists, entrepreneurs and governments, they are doomed to disappoint us if we ignore some very familiar statistical lessons.
“There are a lot of small data problems that occur in big data,” says Spiegelhalter. “They don’t disappear because you’ve got lots of the stuff. They get worse.”
. . .
Four years after the original Nature paper was published, Nature News had sad tidings to convey: the latest flu outbreak had claimed an unexpected victim: Google Flu Trends. After reliably providing a swift and accurate account of flu outbreaks for several winters, the theory-free, data-rich model had lost its nose for where flu was going. Google’s model pointed to a severe outbreak but when the slow-and-steady data from the CDC arrived, they showed that Google’s estimates of the spread of flu-like illnesses were overstated by almost a factor of two.
The problem was that Google did not know – could not begin to know – what linked the search terms with the spread of flu. Google’s engineers weren’t trying to figure out what caused what. They were merely finding statistical patterns in the data. They cared about ­correlation rather than causation. This is common in big data analysis. Figuring out what causes what is hard (impossible, some say). Figuring out what is correlated with what is much cheaper and easier. That is why, according to Viktor Mayer-Schönberger and Kenneth Cukier’s book, Big Data, “causality won’t be discarded, but it is being knocked off its pedestal as the primary fountain of meaning”.
But a theory-free analysis of mere correlations is inevitably fragile. If you have no idea what is behind a correlation, you have no idea what might cause that correlation to break down. One explanation of the Flu Trends failure is that the news was full of scary stories about flu in December 2012 and that these stories provoked internet searches by people who were healthy. Another possible explanation is that Google’s own search algorithm moved the goalposts when it began automatically suggesting diagnoses when people entered medical symptoms.
Google Flu Trends will bounce back, recalibrated with fresh data – and rightly so. There are many reasons to be excited about the broader opportunities offered to us by the ease with which we can gather and analyse vast data sets. But unless we learn the lessons of this episode, we will find ourselves repeating it.
Statisticians have spent the past 200 years figuring out what traps lie in wait when we try to understand the world through data. The data are bigger, faster and cheaper these days – but we must not pretend that the traps have all been made safe. They have not.
. . .
In 1936, the Republican Alfred Landon stood for election against President Franklin Delano Roosevelt. The respected magazine, The Literary Digest, shouldered the responsibility of forecasting the result. It conducted a postal opinion poll of astonishing ambition, with the aim of reaching 10 million people, a quarter of the electorate. The deluge of mailed-in replies can hardly be imagined but the Digest seemed to be relishing the scale of the task. In late August it reported, “Next week, the first answers from these ten million will begin the incoming tide of marked ballots, to be triple-checked, verified, five-times cross-classified and totalled.”
After tabulating an astonishing 2.4 million returns as they flowed in over two months, The Literary Digest announced its conclusions: Landon would win by a convincing 55 per cent to 41 per cent, with a few voters favouring a third candidate.
The election delivered a very different result: Roosevelt crushed Landon by 61 per cent to 37 per cent. To add to The Literary Digest’s agony, a far smaller survey conducted by the opinion poll pioneer George Gallup came much closer to the final vote, forecasting a comfortable victory for Roosevelt. Mr Gallup understood something that The Literary Digest did not. When it comes to data, size isn’t everything.
Opinion polls are based on samples of the voting population at large. This means that opinion pollsters need to deal with two issues: sample error and sample bias.
Sample error reflects the risk that, purely by chance, a randomly chosen sample of opinions does not reflect the true views of the population. The “margin of error” reported in opinion polls reflects this risk and the larger the sample, the smaller the margin of error. A thousand interviews is a large enough sample for many purposes and Mr Gallup is reported to have conducted 3,000 interviews.
But if 3,000 interviews were good, why weren’t 2.4 million far better? The answer is that sampling error has a far more dangerous friend: sampling bias. Sampling error is when a randomly chosen sample doesn’t reflect the underlying population purely by chance; sampling bias is when the sample isn’t randomly chosen at all. George Gallup took pains to find an unbiased sample because he knew that was far more important than finding a big one.
The Literary Digest, in its quest for a bigger data set, fumbled the question of a biased sample. It mailed out forms to people on a list it had compiled from automobile registrations and telephone directories – a sample that, at least in 1936, was disproportionately prosperous. To compound the problem, Landon supporters turned out to be more likely to mail back their answers. The combination of those two biases was enough to doom The Literary Digest’s poll. For each person George Gallup’s pollsters interviewed, The Literary Digest received 800 responses. All that gave them for their pains was a very precise estimate of the wrong answer.
The big data craze threatens to be The Literary Digest all over again. Because found data sets are so messy, it can be hard to figure out what biases lurk inside them – and because they are so large, some analysts seem to have decided the sampling problem isn’t worth worrying about. It is.
Professor Viktor Mayer-Schönberger of Oxford’s Internet Institute, co-author of Big Data, told me that his favoured definition of a big data set is one where “N = All” – where we no longer have to sample, but we have the entire background population. Returning officers do not estimate an election result with a representative tally: they count the votes – all the votes. And when “N = All” there is indeed no issue of sampling bias because the sample includes everyone.
But is “N = All” really a good description of most of the found data sets we are considering? Probably not. “I would challenge the notion that one could ever have all the data,” says Patrick Wolfe, a computer scientist and professor of statistics at University College London.
An example is Twitter. It is in principle possible to record and analyse every message on Twitter and use it to draw conclusions about the public mood. (In practice, most researchers use a subset of that vast “fire hose” of data.) But while we can look at all the tweets, Twitter users are not representative of the population as a whole. (According to the Pew Research Internet Project, in 2013, US-based Twitter users were disproportionately young, urban or suburban, and black.)
There must always be a question about who and what is missing, especially with a messy pile of found data. Kaiser Fung, a data analyst and author of Numbersense, warns against simply assuming we have everything that matters. “N = All is often an assumption rather than a fact about the data,” he says.
Consider Boston’s Street Bump smartphone app, which uses a phone’s accelerometer to detect potholes without the need for city workers to patrol the streets. As citizens of Boston download the app and drive around, their phones automatically notify City Hall of the need to repair the road surface. Solving the technical challenges involved has produced, rather beautifully, an informative data exhaust that addresses a problem in a way that would have been inconceivable a few years ago. The City of Boston proudly proclaims that the “data provides the City with real-time in­formation it uses to fix problems and plan long term investments.”
Yet what Street Bump really produces, left to its own devices, is a map of potholes that systematically favours young, affluent areas where more people own smartphones. Street Bump offers us “N = All” in the sense that every bump from every enabled phone can be recorded. That is not the same thing as recording every pothole. As Microsoft researcher Kate Crawford points out, found data contain systematic biases and it takes careful thought to spot and correct for those biases. Big data sets can seem comprehensive but the “N = All” is often a seductive illusion.
. . .
Who cares about causation or sampling bias, though, when there is money to be made? Corporations around the world must be salivating as they contemplate the uncanny success of the US discount department store Target, as famously reported by Charles Duhigg in The New York Times in 2012. Duhigg explained that Target has collected so much data on its customers, and is so skilled at analysing that data, that its insight into consumers can seem like magic.
Duhigg’s killer anecdote was of the man who stormed into a Target near Minneapolis and complained to the manager that the company was sending coupons for baby clothes and maternity wear to his teenage daughter. The manager apologised profusely and later called to apologise again – only to be told that the teenager was indeed pregnant. Her father hadn’t realised. Target, after analysing her purchases of unscented wipes and magnesium supplements, had.
Statistical sorcery? There is a more mundane explanation.
“There’s a huge false positive issue,” says Kaiser Fung, who has spent years developing similar approaches for retailers and advertisers. What Fung means is that we didn’t get to hear the countless stories about all the women who received coupons for babywear but who weren’t pregnant.
Hearing the anecdote, it’s easy to assume that Target’s algorithms are infallible – that everybody receiving coupons for onesies and wet wipes is pregnant. This is vanishingly unlikely. Indeed, it could be that pregnant women receive such offers merely because everybody on Target’s mailing list receives such offers. We should not buy the idea that Target employs mind-readers before considering how many misses attend each hit.
In Charles Duhigg’s account, Target mixes in random offers, such as coupons for wine glasses, because pregnant customers would feel spooked if they realised how intimately the company’s computers understood them.
Fung has another explanation: Target mixes up its offers not because it would be weird to send an all-baby coupon-book to a woman who was pregnant but because the company knows that many of those coupon books will be sent to women who aren’t pregnant after all.
None of this suggests that such data analysis is worthless: it may be highly profitable. Even a modest increase in the accuracy of targeted special offers would be a prize worth winning. But profitability should not be conflated with omniscience.
. . .
In 2005, John Ioannidis, an epidemiologist, published a research paper with the self-explanatory title, “Why Most Published Research Findings Are False”. The paper became famous as a provocative diagnosis of a serious issue. One of the key ideas behind Ioannidis’s work is what statisticians call the “multiple-comparisons problem”.
It is routine, when examining a pattern in data, to ask whether such a pattern might have emerged by chance. If it is unlikely that the observed pattern could have emerged at random, we call that pattern “statistically significant”.
The multiple-comparisons problem arises when a researcher looks at many possible patterns. Consider a randomised trial in which vitamins are given to some primary schoolchildren and placebos are given to others. Do the vitamins work? That all depends on what we mean by “work”. The researchers could look at the children’s height, weight, prevalence of tooth decay, classroom behaviour, test scores, even (after waiting) prison record or earnings at the age of 25. Then there are combinations to check: do the vitamins have an effect on the poorer kids, the richer kids, the boys, the girls? Test enough different correlations and fluke results will drown out the real discoveries.
There are various ways to deal with this but the problem is more serious in large data sets, because there are vastly more possible comparisons than there are data points to compare. Without careful analysis, the ratio of genuine patterns to spurious patterns – of signal to noise – quickly tends to zero.
Worse still, one of the antidotes to the ­multiple-comparisons problem is transparency, allowing other researchers to figure out how many hypotheses were tested and how many contrary results are languishing in desk drawers because they just didn’t seem interesting enough to publish. Yet found data sets are rarely transparent. Amazon and Google, Facebook and Twitter, Target and Tesco – these companies aren’t about to share their data with you or anyone else.
New, large, cheap data sets and powerful ­analytical tools will pay dividends – nobody doubts that. And there are a few cases in which analysis of very large data sets has worked miracles. David Spiegelhalter of Cambridge points to Google Translate, which operates by statistically analysing hundreds of millions of documents that have been translated by humans and looking for patterns it can copy. This is an example of what computer scientists call “machine learning”, and it can deliver astonishing results with no preprogrammed grammatical rules. Google Translate is as close to theory-free, data-driven algorithmic black box as we have – and it is, says Spiegelhalter, “an amazing achievement”. That achievement is built on the clever processing of enormous data sets.
But big data do not solve the problem that has obsessed statisticians and scientists for centuries: the problem of insight, of inferring what is going on, and figuring out how we might intervene to change a system for the better.
“We have a new resource here,” says Professor David Hand of Imperial College London. “But nobody wants ‘data’. What they want are the answers.”
To use big data to produce such answers will require large strides in statistical methods.
“It’s the wild west right now,” says Patrick Wolfe of UCL. “People who are clever and driven will twist and turn and use every tool to get sense out of these data sets, and that’s cool. But we’re flying a little bit blind at the moment.”
Statisticians are scrambling to develop new methods to seize the opportunity of big data. Such new methods are essential but they will work by building on the old statistical lessons, not by ignoring them.
Recall big data’s four articles of faith. Uncanny accuracy is easy to overrate if we simply ignore false positives, as with Target’s pregnancy predictor. The claim that causation has been “knocked off its pedestal” is fine if we are making predictions in a stable environment but not if the world is changing (as with Flu Trends) or if we ourselves hope to change it. The promise that “N = All”, and therefore that sampling bias does not matter, is simply not true in most cases that count. As for the idea that “with enough data, the numbers speak for themselves” – that seems hopelessly naive in data sets where spurious patterns vastly outnumber genuine discoveries.
“Big data” has arrived, but big insights have not. The challenge now is to solve new problems and gain new answers – without making the same old statistical mistakes on a grander scale than ever.

This article was first published in the FT Magazine, 29/30 March 2014. Read it in its original setting here.