So are you an economic missionary, or is this just something that you love to do?
It began as something that I love to do--and I think I am now starting to get a sense of it being a mission. It has actually developed from More or Less, my Radio 4 show.
I realised that actually people can use economics and they can use statistics and numbers to get at the truth and there is a real appetite for doing so. It got me thinking that actually there is a--I know this is such a BBC thing to say--public service mission to be fulfilled in educating people about economics. But that is something that has come to me quite recently. When I wrote The Undercover Economist it was all about my pure enthusiasm for the subject; the book is full of stuff I wanted to say and that is always the thing with the books: they are always such fun to write.
Do you think that because The Undercover Economist was part of the first wave of economics books--everlasting light bulbs and so forth--that people these days are generally more economically literate?
They are, for two reasons: firstly, Freakonomics and The Undercover Economist were actually part of the second wave. There were very popular economics books in the 1990s and before that, but Freakonomics obviously made a big splash.
But of course, people are now aware of economics for various other reasons. There are the problems with the economy--there is always more interest in economics when it is all going wrong--so, yes, it has been very interesting to see those two things go hand in hand. Both the behavioural economics revolution and also this sense that people want the world explained to them.
Is the new book, Adapt, your movement away from economic rationalist to management guru? Are you going to cast your eye over bigger problems?
The two changes in Adapt are that I have tried to start with the problem, rather than saying, “I have got a hammer--I’m going to look for a nail.” I started with a nail and said, “Ok, look, I need to get this hammered in.” So I have started with the problem and then looked anywhere for solutions. And the second thing is that I have tried to be write with more of a narrative. This is not a Malcolm Gladwell book, but I really admire the way that people like Gladwell get quite complex ideas across because they get you interested in the story and that is something that I have tried to do more of here. I am not too worried about it because I know that I am never going to turn into Malcolm Gladwell--I am always going to be Tim Harford--but it doesn’t hurt to nudge in a certain direction.
Where is the border line in your new book between economics and sociology?
I don’t draw a border line, and particularly not with the new book. The Undercover Economist was basically all the cool economics I could think of and The Logic of Life was me investigating a particular part of economics. All of the references in The Logic of Life were academic economics papers that I had related--and hopefully made more fun. This new book, Adapt, is very different. I have started by asking what is wrong with the world, what needs fixing, how does it work--and if economics can tell us something about that (which it can) then I have used it. And if economics is not the tool that you need, if you need to turn to sociology or engineering or biology or psychology, I have in fact turned to all of them in this book. If that’s what you need, then that’s where I have gone. So I have written this book in a different way: I started with a problem and tried to figure out how to solve it.
What specific subjects do you tackle?
To be a bit more specific, the book is about how difficult problems get solved and I look at quick change; the banking crisis; poverty; innovation, as I think there is an innovation slow-down; and the war in Iraq. Also, I look at both problems in business and in everyday life. Those are the big problems that I look at--and my conclusion is that these sorts of problems only ever get solved by trial and error, so when they are being solved, they are being solved through experimentation, trial and error, which is often a bottom-up process. When they are not being solved it is because we are not willing to experiment, or to use trial and error. And of course, once I started doing that, I also started getting very interested in the psychological questions: “Well, why do we experiment? If it is so effective, why don’t politicians experiment? Why don’t businesses experiment more? Why don’t we experiment in our own lives? Why do we lose that?” A lot of people go to university and find that time of their life can be quite an experimental phase for them--they have got new ideas, new friends, new hobbies, they’re trying out everything as they are in this safe space and nothing can really go wrong, and then they get their first job--and in principle, it should be the same, right? It is a new place, new experiences, new skills, new friends and colleagues… but somehow, it just doesn’t feel the same at all, and it is because the feeling is that the experiment is over--and we lose that willingness to try a whole bunch of stuff and see what is working. So I think that getting interested in the psychology of it, as well as from a purely results-oriented view of the world; you should experiment because it works to say: “Well, why don’t we?”
Do you think companies will change to be much more experimental, with more decisions placed in the hands of employees?
I don’t think it is necessarily a trend, and the reason is that the market itself is highly experimental, so if your company isn’t experimental it may just happen to have a really great, successful idea--and that’s fine; if it doesn’t, it will go bankrupt and you know, that is fine too--from the point of view of the wider world; so I am not saying you have to experiment inside a company, because the market itself is experimental. But that said, it is very interesting to look at the range of companies who have got very into experimentation—the ones who have very systematised experimentation and are very into employee empowerment and--as you point out--they range from the key-cutting chain Timpson’s to Google; you can’t get more different than those two firms, but actually the language is very similar, the recruitment policies are similar, the way the employees get paid is similar.
It is quite weird to look at the parallels between these companies because they both basically said, “Look, the people on the ground are the people that know what is going on, and they have got to make the decision,” and then to draw that parallel with, say, the US Army in Iraq. To find that the US Army have these incredibly tragic, painful and awful experiences—both awful for the army and awful for the people in Iraq--and the US Army has gradually taught itself the lesson that it is actually the guys on the ground who have to make the decisions, not the guys in the command centre.
Do you think that those companies are overcoming their size, that their management structures are their way of saying, “Ok, we are a bigger organisation, we have to be--but our way of dealing with that is to let the guys on the ground make the decisions.”
Yeah. It doesn’t just have to be about size but it is about decentralising power and there has been a trend in that regard: you can measure a trend of decentralisation in business in general. Job descriptions seem to be being pushed further down the pyramid and it’s not just title inflation; people seem to be able to get hold of a CEO more quickly, or there are fewer levels of hierarchy, and that is probably because the world is faster-moving and contrary to what [American economist] Tom Friedman says, the world is not flat--it is very different in different places and if things are changing really fast and the world is different in different places, then that is when you need local decision-making.
If the world is the same everywhere and does not change very much, surely you should have centralised decision-making, because then you can get all the best experts and bring their expertise together. But the other interesting thing that I discovered in the process of writing the book is that technology empowers decentralisation. There has been this fantasy, that technology is this guy sitting in front of a big wall of computer screens who is taking it all in and then sending out all the information to the front line--and at one stage we had that fantasy in Iraq, and the former Soviet Union powers had the same fantasy, this idea that technology has got to empower the corporate centre. But, actually, it is usually the other way round--not always, but usually, it is empowering the people on the front line. The people in the front line have much more information about the big picture than they used to, and of course only they know the day-to-day details of the front line.
The “strap line” of the book is that “Success always starts with failure.” You are a successful author… so what was the failure that set you up for success?
Well, I mean, I was working on a book before The Undercover Economist--more the scratchings of a book--which was going to be a terrible, terrible book, and it took me about a year to figure out exactly the book I really needed to write instead. It was going to be a sort of Adrian Mole/Bridget Jones’ Diary-styled fictional comedy, in which the hero was this economist and through the hilarious things that happened to him, all these economic principles would be explained--which is a great idea—but the trouble is that I am not actually funny. So you can imagine this sort of terrible… unfunny… it was just awful. But that is fine--why would you know how to write the first time around? I think that each book has adapted as I have gone along; I have learned more about being a writer, and of course there are failures elsewhere. My first job was as a management consultant… and I was a terrible management consultant. A colleague of mine said, “Look, you only need to do two things to do this job well. You need to wear a suit and talk very slick--and you can’t do either of them.” So I crashed out after a few months. Much better that, than to stick with the job for two or three years--which a lot of people say, you have got to do, to show your commitment. Taking the job was a mistake--why would I need to show my commitment to a mistake? Better to realise you made a mistake, stop and do something else, which I did--and that has been fine.
This is an interesting topic, because people do stick with mistakes much longer than they often should. You know, we say that people don’t have sticking power--yet people have enormous sticking power when they are failing, because as long as they stick with it, they haven’t failed yet. Economists have studied Deal or No Deal, poker, the way people play the stock market. Basically, if it is going badly, people will stick with it, because the moment they stop they have failed and as long as they are still going they haven’t failed yet. People are incredibly unwilling to admit that they have failed.
That idea that “failure breeds success” is central to most entrepreneurs. Do you think we need more of it in the UK? Does the UK as a country--as distinct from the US or from Europe--need more of a culture of failure?
I think that the real problem is not failure rates in business; the problem is failure rates in politics. We need to have much higher failure rates in politics. What actually happens is politicians--and this is true of all political parties--have got some initiative and they’ll say, “Right, we are going to do this thing,” and it is quite likely that idea is a bad idea--because most ideas fail; the world is complicated and while I don’t have the numbers for this, most ideas are, as it turns out, not good ideas.
But they never collect the data, or whatever it is they need to measure, to find out where their idea is failing. So they have this bad idea, roll this bad idea out and the bad idea sticks, costs the country hundreds, or millions, or billions of pounds, and then the bad idea is finally reversed by the next party on purely ideological grounds and you never find out whether it really worked or not. So we have this very, very low willingness to collect the data that would be necessary to demonstrate failure, which is the bit we actually need.
To give a couple of brief examples: Ken Livingstone, as Mayor of London, came along and introduced these big long bendy buses. Boris Johnson came along and said, “If you elect me, I am going to get rid of those big bendy buses and replace them with double decker buses.” He was elected and he did it, so… which one of them is right? I don’t know. I mean, isn’t that crazy? I know democracy is a wonderful thing and we voted for Ken Livingstone and we voted for Boris Johnson, but it would be nice to actually have the data on passenger injury rates, how quickly people can get on and off these buses, whether disabled people are using these buses… the sort of basic evidence you would want to collect.
As a second example, I spoke to an Education Evaluation expert at the University of York. She had been commissioned by our last Government to figure out the best strategy for teaching kids to read. So she comes back and she says, “I have collected all the data, all the evidence, and it is pretty weak. We think there is this one way of doing it--synthetic phonics--that is probably the best way, but we really just don’t have the data to be sure. But, what we could do is trial this one method in half the schools and the other method in the other half of the schools and while we don’t know which method is best, clearly they are both acceptable and both of them are in widespread use. If we systematically push the two ways of teaching out and measure what happens, then at the end of one year--in fact, it would take six months, because kids learn to read really fast--in six months’ time I will be able to tell you which of these two systems you should use,” and the Department for Education said, “You said ‘synthetic phonics’, ok, fine.” They just weren’t interested in the experiments.
And what drives that? Is that political pressure? Or is it that the media drive the political culture?
I don’t think it is just something in the UK, although there is slightly more willingness to experiment in the US, but not a lot more. It is partly driven in the US by the fact that you have got lots of different states with lots of different registers and things. There is more of an element of somebody doing something ‘crazy’, and everyone having a look to see if it works. But I think this country has no appetite for that. We don’t vote for a politician who says, “I don’t know what will work. I am going to try half a dozen things, let’s see.” People don’t vote for that. They should, but they don’t. It is partly the media. The media is not interested in the scientific method and not interested in evidence. There are some exceptions--people like Ben Goldacre--but most are not interested. I think politicians instinctively know that the failure rate is quite some distance from the popular perception. So if you collect the data, you will demonstrate that lots of your ideas are failing--which is no disgrace, because you should be shutting them down and saving money—but there is a possibility of political damage. There is also short-termism, because many politicians are not in their jobs for more than a couple of years and usually it takes longer--not always, but usually--to find out the truth.
Based on that theory then, are you a supporter of David Cameron’s “Big Society”, which in some senses favours local experimentation over central government planning?
Well, I have some sympathy for the idea of local experimentation, but what worries me is that we have to have some mechanism that is going to tell you what is working and what is not--and there is no proposal for that. Cameron’s Tories seem to have the view that ‘if it is local then it will work.’ In my book, I have all kinds of interesting case studies of situations where localism really would have worked incredibly well, as in, say, the US Army in Iraq. But I have also got examples of where localism did not work well at all--such as a corruption-fighting drive in Indonesia. There was a big project to build roads in Indonesia and they came up with these incredibly clever experiments to find out the right way to fight corruption, but it turns out it was centralised through an audit; so they had this village-level empowerment to report corruption and it didn’t work at all. I have sympathy with localism, but you can’t just assume that that is always going to be the way the world works. The world is too complicated for any one formula to work.
What is your position on George Osborne’s economic vision--should we be cutting faster, should we be going slower, or does it not matter?
It matters--but unfortunately we don’t know, and this is one of those examples of a situation where you won’t know what the right thing to do was until it is too late. My personal instincts were that Osborne should have looked a bit more at things that would have fundamentally put the government’s finances on an even keel on a very long-term basis, like 17 years for instance pensions.
They have done a bit of that, and some very dramatic short term cuts. But the truth is that the last Labour government were planning very big cuts: roughly £50 billion a year of tax rises and spending cuts, and George Osborne added £30 million a year to that £50 billion. So Alistair Darling had already gone more than halfway towards where George Osborne wanted to go. That is just the situation we were in. Government had spent a bit too much before the crisis; the crisis suddenly revealed that we weren’t as rich as we thought we were and so we had to cut back, so if Alistair Darling was Chancellor right now we would still be facing big, big cuts. That’s what people have to recognise.
On that note, what should we do about UK banks? Should they be broken up? Better regulated? Should they be left alone, or should we continue with the status quo, which incorporates a little regulation while largely leaving the banks to act as they see fit?
I think the main push of regulation at the moment is to stop banks from failing in the future; I think that is the wrong approach. People forget, by the way, that banks were very heavily regulated before the crashes; it wasn’t that they could do what they liked. They were regulated—but they were regulated incompetently, so we always need to bear that in mind. Just because we have some regulation it doesn’t mean it is the right regulation.
So the push now is to stop banks from failing and that’s fine as far as it goes, but they will fail. Everything fails sooner or later--and what we haven’t got enough of is effectively decoupling the financial system. Think of a domino falling. When people set up these domino-toppling exhibitions, they place safety gates between the dominos so that if something happens they won’t all fall over at once before they are ready. We don’t have those safety gates in banking. Everyone wants banking regulators to try and ensure that Lehman Brothers never collapses; what I want them to ensure is that when in the future Lehman Brothers does collapse, we are fine, and they are sufficiently isolated. That will require simplifying the structure of the banks, altering the kinds of assets they can hold… there is a whole bunch of different things it is going to involve. It is not going to be easy. But the ideal is to make failure survivable rather than make failure impossible, because it is always possible.
Actually, one of the big themes of the chapter on banking in the book is to make connections between industrial accidents-- Kunii Island, Piper Alpha, Deepwater Horizon--and banking, and what safety engineers and psychologists have learnt from looking at these accidents about why complex systems go wrong and how one tiny thing can blow everything up. It turns out to be really useful when thinking about banking and one of the things that the safety engineers have learnt is you can’t just bang on more and more sophisticated safety systems in an attempt to make sure that nothing ever goes wrong, because the more baroque the safety systems get, the more unreliable they become. The banking crisis basically took place in a context of an incredible number of financial safety systems, the credit default swaps that you hear about, all these fancy three-letter acronyms that [BBC economist] Robert Peston talks about. They were all safety systems and they all failed--they actually made things worse--so you learn from talking to the safety engineers, and a lot of it is about simplification and decoupling and making stuff robust enough to stand on its own two feet, even if some other part of the system goes wrong. That’s what financial regulators need to think about.
Do you think that the reaction to the banking crisis was wrong, or was it the right reaction given the situation?
I think it was the right reaction given the situation that we were in. If Lehman Brothers and AIG were both allowed to go down, we would probably have lost other banks and worsened the situation we are in now. I mean, I don’t like bailing out banks either, but the situation we would have been in is that you would have gone to a cash machine and you wouldn’t have gotten your cash out--and it wouldn’t have been because there was no money in your account; it would have been because the bank was out of money. Your salary would not have been paid and the reason for that would not be that your employer was out of money, but because your employer’s bank was out of money. You would have gone to the supermarket, and the point-of-sale card machines would have failed for the same reason.
Those are three separate systems by the way: the cash machine, point of sale and direct payments. They are all separate systems so they can’t go down in the event of a computer failure, but if the banks had gone down that’s the situation we would have been in. The Government would have had to step in somehow, because all the stuff would still be there: the cards would still be out there, the information about who owes what would all still be out there.
We would somehow have had to get it all up and running again from the wreckage of these bankrupt banks and it would have been terribly, terribly difficult; I don’t blame them for selling up. We are simply going to have to give these guys a load of money because that is the only way we are going to stop this happening.
But in future what we want is more than that: Northern Rock goes bankrupt and the next day all Northern Rock’s customers have bank accounts with Phoenix Bank, which is a new bank and it’s fine… and the shareholders and bond holders of Northern Rock are wiped out, the Managing Directors of Northern Rock are all sacked but the staff--the basic staff who don’t know what is going on--are still there and the customers are still there, they just keep going to the bank as usual. Which is what happened with the people who have their water from Wessex Water… their water didn’t stop coming out of the taps. They probably didn’t even know that Wessex Water was owned by Enron and had gone bankrupt--because nothing changed. We could get to that situation in the banking system, but we haven’t tried hard enough to do that yet.
Looking forward—say, ten years into the future. Are there any economic trends you can see starting now that might be big and significant in ten years’ time?
I hesitate to make those predictions, partly because one of the first studies in the book is the amazing psychologist Philip Tetlock, who did exactly that: he asked exactly those questions of several hundred experts. Between them, he got 80,000 forecasts and he just sat back waited twenty years to see whose forecasts come true--and he talked to journalists, diplomats, think-tankers, government officials and academics. He talked to the communists, sociologists, historians, psychologists, political scientists… all sorts of people, and asked them to make forecasts, and they were all terrible, terrible, terrible, terrible forecasts, all of them, every single one. So it is very interesting that you will ask experts to forecast and to me, because they were all bad, I think that doesn’t tell you that economists or sociologists are rubbish at forecasts. What it tells you is that forecasting is not possible, and the reason it is not possible is because the world is too complicated, which is a central theme of the book. It is because the world is complicated that you need to experiment, because you can’t forecast. So that’s why I am not going to answer the question.
Back to Ben Goldacre: you have talked about it previously, but do you think the UK and the media in general needs a more fact-based mode of reasoning, instead of ideological reasoning?
It would be great to understand what the theory--or the evidence—is, behind what someone is saying, and it would certainly be nice if people bothered to ask. Obviously it depends who you are talking about; the Financial Times, Radio 4, Daily Mail and BBC1 Breakfast have different audiences, so you would ask different questions, but--and this is a totally trivial thing—‘Blue Monday’ came up recently: this idea that there is the most depressing day of the week. So ‘Blue Monday’ was an equation cooked up by a former psychology tutor at Cardiff University. He had written to us and said, “Well, you know, I have developed this equation and I was paid by a PR company to do it”--he was quite happy to acknowledge that and there is no empirical evidence behind it: “I just had some informal conversations with people.” So he is quite straightforward that basically, you know, there is just this equation.
And it doesn’t mean very much, or make any sense--and yet this equation was published. He is allowed to do that; it wouldn’t be my choice on how to make money, but fine, and then the newspapers publish this equation back in 2005: “This is the most depressing day of the year, say psychologists.” There’s no interest in the fact that there is no evidence, no interest in that fact that only one psychologist has devised this and it is not backed up by peer review, no interest in the fact, even, that the equation doesn’t make any sense. If you have a high enough income, the most depressing day of 2005 is actually in 2004--that’s how the equation works out. Worse than that, in 2006 they published the story again, and in 2007 they published again… and they published it in 2011 and at that point I can’t blame the original author of the equation, because now it is has just reached this point of collective insanity. Sure, it is no big deal--it doesn’t really matter--but it doesn’t really make you very confident. When somebody is actually trying to produce evidence to support a proposition and nobody cares about evidence, that’s a bit depressing.
Every successive government claims that they are going to be different; clearly it is one of these things that the Lib Dems were always very in favour of and they always got their economic plan signed off by the Audit Commission. Do you see it any more or is it just one of these things that people say when they’re in government?
I think we have to see that there is some evidence that some of what the government is doing is a good idea, even the non-obvious stuff. For example, free schools: there is serious evidence from the US that in some circumstances they work. So then the questions are always, “Does that evidence actually carry over? Are they going to do it the way it was done?” because lots of free schools have failed as well. So some of what the government is doing has some evidence base, but the insight that politicians don’t seem to understand is this: they can produce their own really high quality evidence through pilot studies, while they figure out how to do what they want to do. But they hardly ever do and there is no excuse for that, because the evidence you could produce is so much more high quality. It is like doing a randomised trial for a new drug: you learn so much more and it is just not that hard to do in most cases, but they don’t want to do it. They want to wait and they want to prove themselves wrong.
To be fair, it is all about us, the people. If we wanted to see proper evidence… if we cared about proper evidence… If we were happy to celebrate a minister saying “I have this new pilot, I spent £10 million, it did not work so I have shut it down. I was going to spend £1,000 million--so I have saved £990 million by doing a proper pilot” we should all be standing and cheering, but you can imagine how that is going to go: “You’ve just wasted £10 million!” So as an electorate, we have to get more grown-up. We can’t expect our politicians to be grown-up; we have be more grown-up. That’s the deal--that’s the order in which it is going to happen.
I think about that when I read the recent announcement about the police crime website, which had people in the Daily Mirror saying “They spent £300,000 building this website!”--and yet, if that is all they have spent, it is a fairly low-cost way to sort of get some benefit.
I think that is a very interesting example of a possible experiment. It is low-cost as you say, so let’s see what happens. But what I hope they are doing--and what I wonder if they really are doing—is collecting the kind of data they need to evaluate it. That comes totally naturally to a business: if Amazon changed the design of your website, I am sure you would collect data to figure out if the change is an improvement or not.
In some cases it very easy to do that, moreso for an internet-based business, but the Government are not even trying. One of the most beneficial social science experiments in the last ten years was done by a celebrity chef: Jamie Oliver pitches up in Greenwich, mobilises resources and changes all the menus of all these schools, and sure, he has camera crews all over one school, but the rest of the schools in Greenwich have all changed their menus… and nothing else has happened. They have just changed the menus and everywhere else in London, nothing has changed. And this pair of economists came along afterwards and scooped up all the data from the exam results, and they said, “Well, did the exam results improve in Greenwich, in a way that isn’t congruent with all the other values we are tracking?” and the exam results did improve. The better food did seem to improve exam results. It is an uncertain finding, certainly, but what I find incredible is that David Cameron and Tony Blair were falling over themselves to say what a brilliant idea it was that Jamie Oliver had done this. Tony Blair had been Prime Minister for eight years--he could have done it, and he could have done it properly and rigorously and found this out. It wasn’t a great mystery that better nutrition might help. We know it was possible. He could have done that a long, long time ago and got really high-quality evidence in 1998 and so we could have had 13 years of good school meals, but, no, it took a celebrity chef...
On Amazon, we recommend new book ideas to people: “If you like Tim Harford you may like…”, but what does Tim Harford also like?
I read a lot of books, mostly non-fiction and in two categories: people who I think write a lot better than I do, and people who think about economics more deeply than I do. In the first category I am reading people like Michael Lewis, Kathryn Schulz (I loved her first book, Being Wrong), Malcolm Gladwell and Alain De Botton. In the second category, I read lots of technical economics books, but I enjoy Steven Landsburg, Edward Glaeser (who has a book out now which looks good), Bill Easterly... I don’t agree with all of these people.
When I am not reading non-fiction, I am reading comic books or 1980s fantasy authors like Jack Vance.
What comic books do you like?
I have a real soft spot for Judge Dredd and Missile Mouse. Alan Moore [author of Watchmen]… all of the obvious ones. I recently read Axe Cop. That’s phenomenal--absolutely incredible. It’s this comic written by a 5-year-old and illustrated by his step-brother who is 30, so they are peering into the incredibly scattered logic and imagination of a 5-year-old: an Axe Cop, who has an axe, who just kills people. It’s fantastic!
Tim, we’re out of time, thanks for coming in to talk to us at Amazon.co.uk today.