"I never understand why 'economist makes forecast' is ever news": Tim Harford. ©PopTech
What are the limits of big data? Should we take economic forecasts seriously? And why should economists be more like dentists? I spoke to writer and broadcaster Tim Harford, ahead of his talk with Prospect this Thursday in association with Little, Brown, about these and other quandries. Harford is the author of a weekly Financial Times column, five books including the bestselling The Undercover Economist, and is the presenter of Radio 4 series More or Less. By peering beneath the bonnet of everyday situations and problems, he enlivens and explains dry economic principles in a manner which is always engaging.
Update: Listen to a recording of Tim Harford's talk on Irving Fisher: "The lost economist"
You write about economics for a wide audience, is that because you believe there’s some benefit in wider economic literacy?
I do, but I also just find it fascinating. I don’t think Brian Cox does TheWonders of the Solar System because he believes the world would be a better place if people understood about the rings of Saturn, I just think he finds physics extremely interesting. It brings him joy and he wants to spread the love. I feel the same about economics. Our society is intertwined with the economy that we’ve built, which is a fantastically complex system. I hope that my writing about it might do some good, but that’s not why I do it.
What do you think the wider public perception is of economics?
There’s a sort of split personality here. People widely feel that economics did not perform very well in the crisis, and yet they keep coming to economics for forecasts. Somehow people can’t quite kick the habit of asking economists for advice.
Economists are seen as prophets, then?
I think the association of economics with forecasting is unfortunate, and is down to the fact that one great way to get an investment bank’s name on business television is to hire a guy called a Chief Economist who will go and prognosticate. John Maynard Keynes famously said it would be splendid if economists were a competent humble profession like dentists. You would never ask your dentist to predict how many teeth you will have in 20 years’ time, or whether you will need to have your wisdom teeth out. You ask your dentist for advice on keeping your teeth healthy, and for help if you have a problem. Equally, economists should primarily aspire to give good advice about keeping the economy healthy.
Do you find it frustrating when economic bodies’ words get co-opted by journalists and politicians?
I never understand why “economist makes forecast” is ever a headline. Whether the economist in question is from the International Monetary Fund, a City forecasting group or the Treasury—a forecast is still not news.
That leads me to ask about “big data”—a subject that is attracting a lot of hype. You wrote a piece recently where you expressed skepticism about some of its practical applications. You’re concerned that these focus too much on correlation rather than causation, is that right?
Big data offers a big opportunity but it doesn’t change the laws of statistics. Some of the hype surrounding big data suggests that the old economic lessons that we spent several hundred years learning don’t matter anymore, when in fact they do. An example is Boston’s Streetbump app [a project run by the city’s Mayor’s Office of New Urban Mechanics] which uses smartphone accelerometers to detect potholes. When you run over a pothole your smartphone relays that information to City Hall, who will now aim to fix the roads in response to these smartphone pings. But there’s a sampling error there; it’s systematically biased towards affluent areas because they’re the people who are most likely to have smartphones. So this is an old lesson, getting lost in the excitement of new technology.
Do you think we’re seeing a bias towards solving the problems of geeky white men, and not looking at wider issues?
It’s not just a white male geek bias, though there is something in that. There’s a general bias in our innovation system towards problems that are quick and easy to solve. When someone says “I work in technology,” you now assume they are talking about a website, or an app. You don’t think of, say, nuclear fusion. Innovation on the big problems is getting increasingly challenging. It involves interdisciplinary work, a long-term perspective and serious amounts of money. I’m not anti-tech industry and I am a big consumer of technological products—they’re just not the only thing [we should focus on].
How important do you think the idea of open data is? For example getting Governments to free up some of their data?
There are two elements here; governments releasing their data, which I think is generally a good thing, and corporations who are revealing data. There is some data that it’s perfectly reasonable to keep commercially confidential. But there is clearly a role for governments mandating the release of certain kinds of corporate data. For example data on pricing would enable you to download a file from your mobile phone or electricity provider, which gives details on your usage. You could then go to a comparison website, upload the file, and the website would assess your usage pattern and suggest the cheapest provider. That’s an initiative that Downing Street have been looking at. I don’t know how far they’ve got with it, but that’s something that’s helpful in an increasingly confusing commercial space.
As for governments releasing their own data? I think they should release as much as possible. My slight hesitation is that our tools for visualising data are now so powerful that it’s possible to get hold of data, format it, make it look beautiful and produce some conclusions, without ever once thinking. That’s a problem.
It seems that, as data visualisation becomes more popular, information can be easily distorted and yet people will trust it just because it’s data. Do you think that’s an issue?
Yes, I do. If you look at advertising from 50 to 100 years ago, you think: “These claims are so ridiculous how could anybody have believed this?” People weren’t used to advertising then—their brains hadn’t got to grips with the fact that advertisers were lying to them. But today’s consumers are smarter and more sophisticated. It’s a similar case with data visualisation. Currently, it’s still very new and there are some grotesque misrepresentations, or incredibly clumsy errors, going viral on Twitter, or being printed in the newspapers.
I’d like to touch on what you’re going to be talking about at your event with us this Thursday, which is the relationship between Keynes and his fellow economist Irving Fisher, is that right?
I’m going to talk about their experience as investors. They were both caught out by the Wall Street Crash of 1929. They both lost a lot of money, and reacted in different ways. While Fisher was discredited and his career went into a death spiral, Keynes recovered and triumphed, both as an investor and an economist. I want to draw that comparison and explore the psychology behind it: why did Keynes succeed and Fisher fail when there was, apparently, initially little to separate them?
So we’re back to the economist as forecaster issue?
Absolutely, and Fisher was a very enthusiastic forecaster, and Keynes… didn’t go in for public forecasts very much.
Lastly, early in The Undercover Economist Strikes Back, you include a quote from PJ O’Rourke: “Microeconomics concerns things economists are specifically wrong about, while macroeconomics concerns things economists are wrong about generally.” Why did you put that in the book? Is that something that keeps you up at night?
No. My two-year-old son keeps me up at night. I included that because it makes me laugh. You can’t be a good economist without recognising that the world is a complicated place, and that economics is fallible. Understanding the economy is tremendously difficult—it involves psychology, history, mathematics and diplomacy—and economists are still working on finding the best approach. It’s a wonderful subject to study, and a wonderful time to be an economist, because today’s problems are deeply complex and are not going to be solved any time soon.
