29 October 2016

Tim Harford: Why the world needs to be a lot more messy


There are journalists who master economics, and economists who master journalism. But few bridge the gap with quite such fluency as Tim Harford.

As the Financial Times’s “Undercover Economist”, as well as a BBC radio presenter and prolific author, Harford has a gift for explaining economic theories and findings in the most accessible way possible.

His new book “Messy” (subtitle: “How to be Creative and Resilient in a Tidy-Minded World”) is, in part, a how-to guide for getting the best out of life. But it provides also a lucid explanation of how our attempts to impose order on the world may be going off the rails.

In the week of its publication, he spoke to CapX about why the world needs a little less order and a lot more chaos.

Given what you’re writing about, I hope you’ll take it as a compliment when I saw that this is quite a messy book. You cover a fascinating range of topics, from Rommel, to Martin Luther King, to stomach bacteria, to how to organise your inbox, to how improv comedy can help dementia patients. How would you describe the linking theme between them? 

I think the theme is the overreach of tidy-mindedness. Having observed that tidy organisational schemes – scripts, targets, filing cabinets – are sometimes great ways to approach the world, we make the mistake of applying them in all kinds of situations where they’re dysfunctional.

In those cases, a looser, vaguer, or more improvised approach works much better – even though it often makes us anxious.

What are your favourite examples?

A simple, literal example is how we organise our desks. It turns out that “premature filing” is a common woe – we try to tidy up documents before we’ve really understood what they mean or how they fit together. The result is a tidy desk but a filing cabinet full of folder labels that no longer make sense to us.

In contrast, a messier desk is subtly self-organising, with relevant material staying near the surface. Several persuasive studies find that the messier approach is more practical for many office workers.

For a more conceptual example of embracing mess, there’s Brian Eno’s “Oblique Strategies” cards. Eno – a composer and producer who’s worked with everyone from U2 and Coldplay to David Bowie – uses these cards and their odd, gnomic instructions to push musicians away from their routines.

The process is discomfiting – Eno reduced Phil Collins to hurling beer cans across the studio – but the results speak for themselves: “Heroes”, Music for Airports, Achtung Baby

What I find interesting about this book is that the message is generally positive – ‘Here’s how we can make our lives, or jobs, or society better by introducing a bit of messiness.’

Yet at the same time it feels like it would have been very easy to use exactly the same material to make a case that the world is racing in the wrong direction…

That’s true enough. For me, the argument comes first: here’s what I believe, here’s why I believe it, and here’s the research that supports my view. The storytelling comes later – that’s part of the craft of writing a book.

My aim is to tell stories that are persuasive and memorable and keep readers turning the pages to find out what happens next. But the stories are the emotional colour to what I hope is a rational argument.

A nice example of this – about a decade ago there was a great book published called “A Perfect Mess”. There’s a modest amount of overlap with my book and I’m definitely in agreement with the authors. But one of their examples is the failure of tidy Apple vs the success of messy Microsoft. That’s no longer such a persuasive case study.

One fascinating thing about the book is that you don’t just talk about physical messiness, but social too.

For example, we have this tendency to cluster with people like ourselves, and to define ourselves against everyone else – even though we’d actually be better off shaking things up. There’s this fantastic example of these kids who are separated into Eagles and Rattlers…

The story of the Eagles and Rattlers is astonishing – I think it’s one of the most remarkable pieces of social science ever done, and not nearly as famous as (say) Milgram’s obedience experiments.

This story involved psychologists sending two groups of boys on a summer camp to see how they responded to the perception of a rival “tribe”. It’s remarkable piece of work – I won’t spoil the ending, except to say that it suggests tribalism is a powerful force, but it’s a force that can be overcome.

More broadly, the chapter about collaboration underlines a well-known point – that we get more done when diverse people work together – by discussing just how strong our resistance is to diversity.

We avoid strangers, we avoid people who disagree with us, and when we’re forced to work with strangers we achieve great results and then deny that the stranger helped at all.

I think that seeking out people with different experiences and opinions is one of the things we all know we should do and it’s one of the hardest things to actually embrace. I certainly find it hard.

Do you see a comparison there to immigration – with the way that it brings greater diversity and productivity, but unsettles people in the process?

For sure.

Another theme of the book is architecture and spaces – whether that be playgrounds for children or offices for adults. What would your ideal office space look like?

The most important thing is that my ideal office space is mine. It won’t look like your ideal office space.

It’s tempting to try to force other people to tidy up – whether it’s children’s bedrooms or a clear desk policy – but it’s usually a bad idea.

The psychological research I discuss in Messy shows that people really resent such policies. They’re infantilising and humiliating. We feel angry, we get less done – sometimes we feel physically unwell – if control over our own space is taken away from us.

But the truth is that most of us are not professional designers, so when everyone gets control of his or her own space the result will look like a messy – garden gnomes and mascots and coffee cups and all kinds of crap. Tough: autonomy is more important than tidiness.

That’s completely convincing. But I can see a situation where some CEO reads this book without paying proper attention and then becomes the worst boss in the world – constantly pulling teams apart or bringing in Eno’s Oblique Strategies or barrelling ahead with a dozen different initiatives because “it’s what Jeff Bezos does”…

Yes – and there’s also a tension in the book, because sometimes I’m saying that bosses need to shake people up by giving them new people to work with or difficult challenges, while at other times I’m saying that bosses need to leave their workers alone to make a mess.

Perhaps the resolution of that apparent contradiction is in whether people have autonomy. Eno and Bezos were asking people to take initiative to solve near-impossible problems. That’s tough – and I would never, ever want to work for Jeff Bezos! But at least they weren’t removing elementary decisions from people on the grounds that “corporate policy says only one A5 picture on your desk”.

On a similar note, you talk about how much Paul Erdos contributed to mathematics – but also make it clear that he was a pretty exacting person to be around, leading this nomadic existence where he collaborated with hundreds of people essentially by couch-surfing from campus to campus and country to country. How much mess is too much mess?

Yes, Erdos was an impossible house guest – he was adorable but helpless. He’d wake you up at four in the morning because he was hungry. He wore only silk underwear because of allergies – but someone else would have to hand-wash it. And his relentless collaboration was an act of pure genius – nobody could hold him up as a practical role model, and I don’t.

So how much mess is too much mess? An impossible question because it depends on the situation. But what I can say is that in most situations, we’re biased towards not enough disorder, too much control, too much quantification, too much preparation.

So if you ask me how much mess, my answer is “a little bit more that you’d prefer will do you no harm”.

Getting topical for a second, one of the examples you give of messiness is people who create and profit from chaos – like Rommel in the Second World War and more recently Donald Trump.

Given that the forces of order (aka Hillary Clinton) seem to have crushed him, is this a case of messiness reaching its limits?

Well, Rommel also ran out of road when he met Montgomery. But nothing I’ve seen has changed my view that mess can be a very powerful weapon – especially in the hands of underdogs like Rommel and Trump.

It’s hard to remember now – but Trump was an utter laughing stock when he announced his candidacy against the likes of Bush and Rubio. All the forecasts are that he’ll finish second in this presidential campaign, but that’s an effective performance given that he started it as the host of The Apprentice.

You talk powerfully in the book about the problem with targets and the way they distort our behaviour, whether in banking or healthcare.

Is there a case that economics as a whole is suffering from the same problem? I’m thinking in particular of the dependence on GDP or the unemployment rate…

Yes, there’s that risk – although I think economists in general are quite well attuned to this problem, which is why they tend to be very suspicious of economic planning.

In my experience, it’s often the non-economists who say: “We should embrace this policy because it will be good for GDP.”

So if we suddenly appointed you Prime Minister – or head of the NHS, let’s say – how would you go about measuring and improving performance?

I think NHS regulation would be much better served by deep and somewhat unpredictable inspections rather than broad, shallow and predictable box-ticking exercises.

Show up and have a good hard look at something chosen at random, rather than try to cover everything and instead cover nothing. And I know that the country’s leading medical stats gurus feel the same way.

And how would you introduce messiness?

If I was in charge of the country? Ha! Give that job to someone else, please…

But I think the obvious thing to do would be to move civil servants around between departments, central and regional and local governments, encourage them to take secondments in the private sector, get private sector secondees in, do job swaps with other countries – the kind of thing that 3M does as it rotates researchers around the company.

The more we have to work with unfamiliar people or in unfamiliar settings, the more we build our own skills.

One of the most powerful points you make in the book is about computers, and our excessive reliance on them – there’s this harrowing description of the crash of Air France flight 447, where the pilots were so used to being molly-coddled by the autopilot that they completely froze when its sensors went offline.

I don’t want to sound like a Luddite here – of course planes are much safer than they used to be. The concern I have is that autopilots are de-skilling pilots themselves.

This is a risk wherever we hand over decisions to a computer – algorithms for making hiring decisions, for instance, or relying on Google Maps to navigate a city for us.

I think we need to think more seriously about how humans and computers work together. It’s clear that chess software has made human players better. But it’s not at all clear that autopilots have improved the skills of pilots.

One problem may be who does what: at the moment we ask human pilots to babysit the computer. We might do much better to flip that so that the computer (who does not get tired) babysits the humans (who get to practice their skills).

You also make this point that an algorithm that’s ten times less likely to make a mistake than a human, but is a thousand times faster, will generate a hundred times as many mistakes. Is our reliance on data and algorithms making the world too tidy for our own good? 

Well, algorithms fail all the time. That’s not necessarily a problem – humans fail too. But I see worrying signs that we now give the algorithms an authority they don’t deserve.

Because we don’t really understand them, we have a tendency to think they can’t fail. And that means that when there are miscarriages of justice – after all, algorithms are now being used (or misused) to send people to prison – we need to pay more attention to an appeals procedure.

How did the idea for the book come to you?


And what about your own life? You talk about how you prioritise projects – by pinning up and shuffling round cards with the things you need to do on them.

But are you a Ben Franklin figure, who’s constantly fretting about the chaos of his own life while accomplishing great things despite it? Or do you actually practise what you preach, and lean into your own messiness?

I’m a pretty tidy person, actually – except when it comes to my desk, which is chaos. But writing the book has made me realise that there’s actually a good reason why even though my kitchen is tidy, my desk isn’t. And I’m trying to cut myself a bit of slack.

One big change has been my attitude to my children’s bedrooms. After studying the psychological research on disempowering people by forcing them to tidy up, I gave up on that project. Their bedrooms aren’t any more messy than they used to be. But now they’re in control.

The one thing that I really struggle with is working with strangers – I’m a shy person, so I find it hard to take my own advice. But I do try.

‘Messy: How to be Creative and Resilient in a Tidy-Minded World’ by Tim Harford is out now from Little, Brown

Robert Colvile is Editor of CapX.