Tim Harford The Undercover Economist

Articles published in July, 2012

“The Undercover Economist” – a free chapter

The second edition of The Undercover Economist was published last year in the UK, and recently as an eBook in the US.

The biggest change from the first edition was a new chapter about the financial crisis. Lots of people have written to ask whether they can get this chapter without buying the entire book again. That seems only reasonable, and you can now download it here or as a backup, here. Enjoy.

Three PowerPoint tips you really need to know

Microsoft bought PowerPoint 25 years ago. Happy anniversary.

PowerPoint has a curious status these days – it’s ubiquitous and yet widely loathed. Both the ubiquity and the loathing are overdone.

Here are three tips I’ve found very useful as a speaker.

1) There are three things you can do with PowerPoint (or most of its rivals). You can put visual aids on a screen; you can create bullet-point speaker’s notes; and you can produce handouts for people to take home. All of these uses are perfectly legitimate, but you can’t do them all at once. Your speaker’s notes should be on small cards in your hand; your handouts can have contact details, sources, a bibliography, or dense data; your visuals should be simple and look awesome. If you feel you need to do all three, fine: you will need to create three completely different presentations.

2) If you don’t have anything useful to display for a particular section of your talk, display nothing. During slideshow mode, press B to show a black slide, or W for a white one. Or if you don’t have direct access to the computer while presenting, insert blank slides as necessary. There’s nothing wrong with giving a talk during which you only show one or two slides – but don’t leave them up as wallpaper.

3) You don’t have to use any visual aids at all. You might be surprised at how much people focus on you when you stop competing with yourself for attention.

When simplicity is a real asset

Perhaps the whole ‘don’t put all your eggs in one basket’ school of portfolio allocation is financial wisdom enough

When James Tobin won the Nobel memorial prize in 1981, a journalist asked him to summarise his research in simple language. The great macroeconomist attempted to respond to this challenge, and one wire service dutifully reported that Professor Tobin had won the prize “for his work on the principle of not putting all your eggs in one basket”.

A newspaper cartoon then appeared announcing the award of a Nobel prize for “an apple a day keeps the doctor away”.

But Tobin perhaps anticipated the awkward history of the Nobel memorial prize and financial economics. Robert Merton and Myron Scholes won in 1997 for their work on option pricing – less than a year before the dramatic bailout of Long-Term Capital Management, a hedge fund in which Merton and Scholes were closely involved.

Harry Markowitz, who shared the prize in 1990, was really the founder of the whole “don’t put all your eggs in one basket” school of portfolio allocation. Markowitz showed how investors could pick an optimal portfolio of assets, minimising risk for any given expected return, or maximising expected return for any given risk. (The basic idea is simple enough to be worthy of Tobin: if you hold shares in a sunblock manufacturer and an umbrella company, your finances will be fine in all weathers.)

In 1952, Markowitz had had the perfect opportunity to put his theory to good use. He joined the Rand corporation and had to decide how to invest his pension. Did he compute the efficient risk mitigating frontier? He did not. He split his contributions 50/50 between stocks and bonds. So there.

Here’s a question, though: are these practical tips from Markowitz and Tobin as useful as their sophisticated academic theories? Could it be that simply dividing your money equally between a bunch of different assets – known as the “1/N” strategy – a perfectly good approach to investment?

It might seem implausible: after all, the “1/N” strategy is arbitrary and ignores useful information about historical risks, returns and correlations across asset classes. We know, thanks to the research of the behavioural economists Shlomo Benartzi and Richard Thaler, that many investors do exactly what Markowitz did. Surely this is an error, or at least clear evidence of our cognitive limitations?

Perhaps. But here’s the intriguing thing about the financial theory that Markowitz developed: it’s extremely difficult to apply in practice. If you know for certain the distribution of returns for all the assets in which you are investing, you can compute an efficient frontier. But you don’t. You can only guess.

One problem is that historical correlations are poor guides to future ones. Imagine the shares of two oil companies, for instance: as the oil price rises and falls, so would the shares, which would seem highly correlated. If one company then ran into some kind of trouble – another Deepwater Horizon, for instance – then the shares might well become negatively correlated as the unaffected company picked up market share from the affected one.

A second problem is that even with lots of historical data, it is hard to estimate the likelihood of rare events. (By definition, there will be few or no historical examples.)

Portfolio theorists have produced a variety of sophisticated methods to try to update Markowitz’s ideas for an uncertain world. But in research published in 2009 in the Journal of Financial Studies, Victor DeMiguel, Lorenzo Garlappi and Raman Uppal showed that the naive 1/N approach outperforms far more complex calculations until a vast amount of historical data are available with which to calibrate them. How much data? For a 50-asset portfolio, about 500 years. Perhaps “don’t put all your eggs in one basket” is financial wisdom enough.

Also published at ft.com.

Cash in hand is worth £35bn in the tax bush

‘Getting a discount with your plumber by paying cash in hand is something that is a big cost to the Revenue and means others have to pay more in tax. I think it is morally wrong.’

David Gauke, Treasury minister, July 24

I feel bad now. I always pay my cleaner in cash.

Don’t feel too bad. I always pay my cleaner in cash, too, and for that matter I use cash to buy coffee. There are legitimate reasons for paying with cash, as just about every other government minister has squeaked over the past few days.

But if a plumber or an electrician offers a heavy discount in exchange for cash payment, surely you have to admit that something is fishy?

It might still be legitimate – prompt payment can save a business and cheques can bounce. But of course Mr Gauke has a point: if somebody wanted to evade tax they’d find it much easier with cash in hand.

And if I agree to pay cash in hand, am I helping the plumber to evade tax?

Economically speaking, the plumber is helping you to evade tax. Lawyers and accountants look at the legal responsibility for paying tax and that lies with the plumber. But economists look at something called “tax incidence”, which is who ultimately bears the cost of the tax. A tax on plumbing might well increase the post tax cost of plumbing far more than it lowers the post-tax wages of plumbers.

Why do you say that?

Let’s make a couple of assumptions. Assume that most plumbing work is not price sensitive: if the lavatory is leaking, you pay to have it fixed, whatever the cost; at the same time, nobody has cheap plumbing work done for the joy of a bargain. Assume also that the supply of plumbing services is quite elastic; if plumbing becomes more profitable, more plumbers will arrive from Poland and prices will fall; if plumbing becomes less profitable the Polish plumbers will ply their trade elsewhere.

You and your assumptions will get you into trouble one day, but go on.

If these assumptions are broadly true then any tax on plumbing would ultimately be paid by consumers, not by plumbers. Tax evasion would help consumers themselves, not the plumbers.

Interesting. But isn’t this a fuss about nothing? Isn’t the really big-deal tax evasion by zillionaires and

It’s certainly not a fuss about nothing. We could look at estimates from Revenue & Customs about how much tax they think is going missing. They call it the “tax gap”, and it represents the difference between how much tax they think they should be receiving in accordance with both the letter and the spirit of the law, and how much is being paid. That gap is partly due to error, partly due to tax evasion (violating the letter of the law), partly due to tax avoidance (violating the spirit of the law) and partly due to outright fraud.

How big is the tax gap?

We’re not sure. After all, not many people make explicit declarations of how much tax they are evading. But we have estimates and in 2009-10, the most recent year available, the tax gap was about £35bn.

Is that a lot?

It is big enough to worry about: 7 per cent of total revenue, give or take, and more than £500 for every man, woman and child in the country.

How much of that is down to cash-in-hand payments?

Potentially quite a bit. About half of the tax gap can be blamed on small and medium-sized companies, a category that would include most cleaners and plumbers, as well as companies with 100 employees or more. A quarter is attributed to large companies and their battalions of tax lawyers; 17 per cent to outright criminal scams; 11 per cent to individuals. The problem goes beyond any one particular category of scapegoats. And, alas, I am unable to find any estimates that break down culpability by income level.

So tax evasion is really a blue-collar crime, then?

No, I think that is wildly over-interpreting. But it is certainly not the exclusive preserve of the Monaco set. One important thing is missing from the HMRC definition of “tax gap”, however. It’s an estimate of how much tax is demanded yet not paid. But remember that how much tax is demanded is itself a function of opportunities to evade tax. If multinational companies are footloose, for instance, governments may offer low rates and generous tax breaks. The tax gap would then be small on paper but large relative to the taxes the Treasury might fantasise about levying.

Also published at ft.com.

Just for clicks: the Google ad model

Traditionally, advertising is sold by salespeople who quote prices for ads. The tech company decided to auction advertising space instead

When Hal Varian arrived at Google as a part-time economic adviser in 2002, he asked the then chief executive, Eric Schmidt, how he might make himself useful. Schmidt suggested that Varian might “take a look” at the way Google sold its advertising because “it might make us a little money”. That was an understatement: in 2011, Google’s ad revenues were more than $36bn.

Nice as it would be to give Varian all the credit for this – his textbook was my microeconomics bible – the foundation stones of Google’s advertising success had been laid before he arrived. Traditionally, advertising is sold by salespeople who quote prices for advertisements. Google decided to auction advertising space instead. And when Varian, who is now Google’s chief economist, took “a look” at the auctions that Google’s computer scientists had designed, he found that they were pretty much perfect.

If you search on Google, one set of Google computers tries to deliver the best possible search results; a second set is running an auction with the aim of delivering the most effective ads to be displayed alongside them, in 11 different positions of varying prominence. An auction is run every time a user searches – billions a day.

Beyond the sheer computational demands, there are two reasons why these auctions are tricky to run well. First, these advertising spaces are substitutes for each other. If I sell flights to Reykjavik and you Google “flights to Iceland”, I want one of those ad spaces. I probably don’t want all of them, and that might also irritate users, which is in the long-run interests of nobody (except possibly Yahoo and Microsoft, Google’s rivals in this business).

Google doesn’t want to sell slots in parallel because advertisers fear winning multiple redundant slots. The solution is something called a “generalised second price auction”: the winning bidder gets the top slot and pays whatever the second bidder offered; the second bidder gets slot two and pays the third-highest bid. (This is a slight oversimplification, as we shall see.) Google’s willingness to accept less than each bidder actually offers might seem odd, but it encourages higher bids and may well raise more money overall.

The second problem is what the metric of bidding should actually be. Google could charge per “impression” – that is, for each time an advertisement is displayed. Or it could charge per “click” – each time a user clicks on the ad and travels to the advertiser’s website. The difficulty here is that Google’s costs – such as the forgone opportunity to sell space to someone else – are based on impressions, whereas the advertiser chiefly cares about clicks. I typed “Picasso prints” into Google and was offered the chance to buy some posters, but also to bid at an auction at Christie’s. I’m sure Christie’s gets far fewer clicks but is willing to pay much more for each of them.

Google’s solution is to create a “quality” metric, largely based on expected clicks, that serves as an exchange rate between impressions and clicks. If Christie’s is willing to pay $1,000 a click, and Google expects one such click, Art.co.uk will beat them with a bid of 10 cents a click, as long as Google expects more than 10,000 clicks – rightly so, since Google’s expected revenue from Art.co.uk is higher. Art.co.uk will pay a sum related to Christie’s bid and to the “quality” of both adverts.

Despite the wrinkles it is a simple idea, executed well. The biggest surprise for me is that many Google searches are “undersold”, with a few advertisers getting a bargain-basement rate – or no advertisers at all. Type “Hal Varian Google ad auctions” into Google and you’ll see no ads. Type “flowers” into Google and, I assure you, all 11 advertising spaces will be filled. It is on such searches that Google earns that $36bn.

Also published at ft.com.

Census is proof the Jedi force is with us

‘The population of England and Wales grew over the past decade at its fastest rate in 100 years as the result of a surge in immigration, increased longevity and a mini baby boom.’

Financial Times, July 16

This was discovered by the census, right?

Of course.

What about the population of Scotland?

The Scottish census won’t publish any figures until the 700th anniversary of the Battle of Bannockburn. Or something like that – I am hazy on the details. The Northern Irish census has been published and the population there is 1.81m. The population of Northern Ireland is less than half of the increase in population in England and Wales between 2001 and 2011. The English and Welsh population increased by 3.7m to 56.1m.

Everyone keeps talking about that – is it really such a big deal?

It was quite a surge. The English and Welsh population grew at a real clip during the last decades of Victoria’s reign, but 7.1 per cent is the fastest rate of population growth since the 1911 census. It was also a surprise: official estimates gave the population as being almost 480,000 lower.

How do you get that kind of thing wrong?

Mark Easton, the BBC’s Home Affairs editor, put it rather well, saying we are good at counting babies and corpses so the only other source of error is net migration. We must have counted too few immigrants or assumed people had left the country when they hadn’t.

How is it possible to miscount such a thing?

Perhaps I shouldn’t have said “counted”. Many statistical exercises in this country don’t count, they sample. For instance, the International Passenger Survey buttonholes people on cross-channel ferries or just after they’ve left passport control and asks questions about who they are and what they are doing. That’s just a sample of the people arriving in the country, however, and the IPS surveyors can’t force people to answer their questions, nor can they force them to tell the truth.

And the census is better because it samples a higher percentage of the population?

Actually, what determines the precision of a sample is not that it covers a high percentage of the population, but that it has a lot of people in it. A sample of a million people is plenty for most purposes, whether that’s a million people out of a population of 10m, or a million people out of a population of a billion.

So why do we have a census at all, rather than a survey of a really big sample of people?

The trouble with surveys is that the sample of people they include may not be representative of the population as a whole. There’s a long-running problem, for instance, with contacting people at home by telephone: once upon a time such polling was biased towards the opinions of the prosperous; later, it became biased towards stay-at-home types. These days, it is biased against the young, who often do not use land lines at all. Sensible surveys will try to correct such biases but one can never be entirely sure.

But not everyone fills in the census either.

Indeed. The Office for National Statistics followed up the census itself with a survey to estimate how many people actually completed it in different parts of the country, so that they could make appropriate connections.

Do they also have a follow-up to the follow-up survey?

No. Behave.

How much did this cost?

£480m, apparently.

That’s a lot.

Perhaps. It’s less than one-10th of 1 per cent of annual government spending and, of course, we only do it once a decade. I suppose the question is, is it worth spending £1 in every £10,000 to help work out how to spend the other £9,999?

If you think it will help, that is. Speaking of which, have we found out how many Jedi Knights there are in
England and Wales?

We don’t yet know. Almost 400,000 people stated that their religion was “Jedi” in the 2001 census, making belief in The Force the fourth most popular religion. I have no idea whether Jedism is an evangelical religion or not. If it is, that number might well have risen.

It all seems rather confusing to me.

That’s just an old Jedi mind trick.

Also published at ft.com.

PopTech talk: Preventing Financial Meltdowns

On poverty, WiFi and The Wealth of Nations

“Family of four ‘need £36K for decent standard of living’. Joseph Rowntree Foundation report finds a couple with two children need £36,800 – a third more than before the recession.”

The Guardian, July 9

You could buy a lot of beans on toast for that.

I see – you’re an “absolute poverty line” kind of guy?

What do you mean?

There’s a long, honourable tradition in the field of poverty analysis of drawing a poverty line by trying to figure out what it might cost to feed, clothe and shelter somebody. If you earn more than that, you have food, clothes and a place to stay, and can hardly count yourself poor. Your “beans on toast” comment suggests you share this vision.

Is there anything wrong with that?

There’s nothing logically wrong with the idea of an absolute poverty line. Seebohm Rowntree, of the wealthy Rowntree family, calculated one based on the price of such essentials. Pease pudding and bacon for Sunday lunch was included, if I recall correctly.

When was this?

It was 1899. I think we can cut him some slack as a Victorian philanthropist, but today the idea of “absolute poverty lines” – meaning some well-fed expert trying to figure out what we should be eating – seems a little discomfiting.

Has the idea fallen out of fashion, then?

The US still has a poverty line – it’s $30.52 a day in 2012 for a single-person household. That number is thanks to Mollie Orshansky, a government researcher who put a price tag on the cost of following official nutritional advice back in 1963, then used the rule of thumb that food consumed one-third of all household spending. Goodness knows what people were advised to eat back in 1963, but the advice is fossilised in the poverty line; it’s been updated to reflect inflation, and that’s about it.

What’s the US poverty line for a family of four?

Just over $23,000.

A lot less than £36,800, then. Why are the two numbers so different?

Because they are trying to measure different things. The Rowntree foundation number is based on focus group discussions – supplemented by expert advice on, say, nutrition – about what sort of spending is necessary to participate fully in society.

Meaning what?

Well, a television was regarded as necessary to participate fully in society.

That’s nonsense. You don’t have one – are you excluded from society?

I am an economist. I am supposed to be socially excluded. There’s a difference between choosing not to have a television because you’re drowning in books, papers and theatre trips; and not having a one because you can’t afford it.

It still feels very subjective.

Of course it is. The point of the Rowntree methodology is to find out what people think is a socially acceptable minimum. Adam Smith understood this. In The Wealth of Nations, he observed that Greeks and Romans may have survived without linen but, in 1776, “a creditable day labourer would be ashamed to appear in public without a linen shirt”.

It’s all relative, then.

Not so. There are relative poverty measures – Eurostat, the EU’S data-gathering arm, defines the poverty line as 60 per cent of each nation’s median income, for instance. But they are just proxies for inequality, and not very good ones. Smith’s point is not that poverty is relative in some mechanical way, but that it is a function of society’s expectations.

For example?

In the Rowntree panels, parents reckoned others like them needed a computer and internet connection so the children could do homework. But pensioners reckoned they could go to the library to go online. Smith didn’t have a flush toilet; the Queen grew up without WiFi. Now they’re part of anyone’s reasonable expectations.

But why should the rest of us pay to meet these expectations for layabouts and scroungers?

We don’t. The Rowntree foundation says that, while pensioners receive benefits that more than cover the minimum income standard, others receive much less – between 40 and 60 per cent of the Rowntree benchmark. I think you’re confusing two questions. One is: how much money does a household need to enjoy access to what UK?society takes for granted? The second question is what, if anything, the state might care to do about that.

And why has this minimum income standard risen so much in the past few years?

Simple. Rents are rising rapidly, partly as a matter of government policy. Childcare costs have also risen sharply.

Also published at ft.com.

The lessons that flow from Bali’s water temples

As with Bali’s ancient irrigation system, we cannot be confident that there is time to adapt in a fast-changing world

In the mid-1970s, a young anthropologist, Steve Lansing, arrived in Indonesia to study the ancient rituals of the water temples on the volcanic slopes of Bali. In a less spiritual realm, the island’s agronomists were stumbling into an agricultural mess. Bali once had the highest rice yields in Indonesia, and throughout the 1970s, the Green Revolution brought new varieties of rice to the stepped paddy fields. Yet despite intensive training and expert advice, by the early 1980s the new crops were overwhelmed by pests.

The narrow technical problem was a trade-off between pest-control and irrigation. Farmers made careful use of the 170 rivers and streams flowing down the slopes of Bali’s sacred volcanoes; water conservation was most easily done with different communities – subaks – planting at different times. Pest control, in contrast, was best achieved by every farm going fallow at the same time, so that the pests were starved of food. The ideal compromise is a complicated and ever-changing schedule of regionally clustered rotation – and one which must be agreed or imposed in a system where one selfish farmer can damage many others. With the new rice crop, the government was centrally co-ordinating the rice planting, and the new system was failing.

For years, economists showed little interest in the subtleties of such resource-control problems, accepting the perspective of Garrett Hardin, an ecologist who coined the term “the tragedy of the commons”. Hardin argued that commonly owned resources, such as common grazing land or fish in the ocean, would inevitably be overexploited as each individual gobbled up resources like two children sharing a carton of popcorn.

Hardin had a point, but he was wrong to think that common resources cannot be managed by communities. They can; that was the discovery for which Elinor Ostrom became the first woman to win the Nobel memorial prize in economics, in 2009. (Professor Ostrom died in June, and her husband and colleague, Vincent Ostrom, died shortly afterwards.)

Lin Ostrom had a curiosity for how the real world works that too few economists share, and studied examples of communal resource management from lobster fisheries in Maine to irrigation in Spain. In Nepal, she found villagers upstream helped maintain canals downstream, while villagers downstream helped maintain dams upstream. When aid donors funded modern dams that needed less maintenance, this bargain fell apart.

And so to Bali, where Lansing discovered a particularly beautiful system of communal management: highly sophisticated, more than a thousand years old, and in the process of being short-circuited by the new crops and their new planting cycles. Those water temples were no superstitious curiosity: the network of temples, their true function hidden from outsiders in plain sight, ran the irrigation system.

In a computer simulation developed with colleagues, Lansing showed that the rotation developed by the water temples was the most efficient the computer could find; the simulation also showed how the temple system could have evolved quickly through trial and error.

Although Bali’s rice farmers still have their problems, the situation has greatly improved since the 1980s. This month, the entire system was recognised by Unesco’s World Heritage Committee.

Lansing’s ideas have reinforced my own prejudices that complex problems tend to be solved through trial and error, often on a local level. But the contrast between the water temples and our own institutions left me uneasy. Problems such as overfishing and climate change are regional or global, and moving targets. Bali’s temples evolved in a stable environment. In a fast-changing world we cannot be confident that there is time for our own institutions to adapt.

Also published at ft.com.

Why the government should play house

Building new homes could play a role in creating useful jobs, but the planning system remains a hindrance

I’m not a macroeconomist myself and so when I meet proper macro folk, I try to draw them on the topic of the day: how to stimulate the British economy. Should George Osborne reverse his tax rises and postpone his promised spending cuts? Or would any stimulus that resulted be lost in the turbulence of lost credibility and higher borrowing costs? With a £1.5tn economy, any stimulus spending that was big enough to make a difference should hardly be taken lightly. Hence my interrogation of any macroeconomist I could find.

To my surprise, on a couple of recent forays into this territory, I have found my inquiry sidestepped completely. The real issue, I am told, is microeconomic: the government needs to find a way to get more houses built.

This could certainly work in principle: there is a robust demand for houses at well above the cost of building them, and house building would be a good source of employment for semi-skilled or unskilled young people, a group badly affected by this recession and one that has not been in good shape for some years.

Could a house-building splurge make a difference? Surely. The UK has recently been building a little over 100,000 new homes a year, but the country is acquiring more than 200,000 new households annually, largely as a result of its internal demography, but with net immigration also playing a part. The shortfall has been substantial for many years; there is no reason to expect the UK couldn’t find a use for 300,000 or even 400,000 new houses a year for the next few years – and that means, very roughly, a million new jobs in construction, the entire number of unemployed people under the age of 25.

Building houses is an occupation that could plausibly play a substantial role in creating useful jobs and stimulating demand for several years. How, then, to make it happen?

The chief obstacle to house building in the UK is the planning system, which, 65 years ago, did away with the idea that if you owned land, you could build on it, and replaced it with a system where planning permission was required. Permission to build houses is severely rationed, and such rationing can be seen clearly in the gap between the value of agricultural land without planning permission (a few thousand pounds a hectare) and the value of such land once permission has been granted (a few million).

The difficulty is that local authorities have the ability to grant planning permission but have little incentive to do so, because it tends to be unpopular with existing voters. The huge windfall from winning planning permission falls to whoever has managed to speculate on land and navigate the tangle of planning rules. These serve as nice barriers to entry for existing developers, while driving up the price of building land and so driving down the size of new homes.

Tim Leunig, chief economist at CentreForum, a think-tank, has proposed a two-part system of land auctions to get around this problem. Local authorities would buy land at auction, grant planning permission on it and then sell the land on to developers – with some strings attached, if they so choose. The profits would be enormous, and enjoyed by existing residents in the form of lower taxes or better public services. This isn’t the only way to liberalise planning, but it retains local control and democratic accountability – while dramatically increasing the incentive to develop.

The Department for Communities and Local Government said last year it would “pilot elements of the land auctions models, starting with public sector land”. That is like practising a dinner party with a doll’s tea set. The government has been in office since 2010; the financial crisis is five years old. A bit of urgency wouldn’t hurt.

Also published at ft.com.



  • 1 Twitter
  • 3 RSS
  • 5 Podcasts
  • 6 Facebook


  • Fifty Inventions That Shaped the Modern Economy
  • Messy
  • The Undercover Economist Strikes Back
  • Adapt
  • Dear Undercover Economist
  • The Logic of Life
  • The Undercover Economist

Free Email Updates

Enter your email address to receive notifications of new articles by email (you can unsubscribe at any time).

Join 188,408 other subscribers.