Tim Harford The Undercover Economist

Articles published in March, 2015

Highs and lows of the minimum wage

‘The lesson of all this is that the economy is complicated and textbook economic logic alone will get us only so far’

In 1970, Labour’s employment secretary Barbara Castle shepherded the Equal Pay Act through parliament, with the promise that women would be paid as much as men when doing equivalent jobs. The political spark for the Act came from a famous strike by women at Ford’s Dagenham plant, and the moral case is self-evident.

The economics, however, looked worrisome. The Financial Times wrote a series of editorials praising “the principle” of equality but nervous about the practicalities. In September 1969, for example, an FT editorial observed that “if the principle of equal pay were enforced too rigorously, employers might often prefer to employ men”; and the day after the Act came into force on December 29 1975, the paper noted a new era “which many women may come to regret”.

The economic logic for these concerns is straightforward. Whether because of prejudice or some real difference in productivity, employers were willing to pay more for men than for women. That inevitably meant that if a new law artificially raised women’s salaries, women would struggle to find work at those higher salaries.

The law certainly did raise women’s salaries. Looking at the simple headline measure of hourly wages, women’s pay has gradually risen over the decades as a percentage of men’s, although it remains lower. Typically, this process of catch-up has been gradual but, between 1970 and 1975, the years when the Equal Pay Act was being introduced, the gap narrowed sharply.

Did this legal push to women’s pay cause joblessness, as some feared? No. Women have steadily made up a larger and larger proportion of working people in the UK, and the Equal Pay Act seems to have no impact on that trend whatsoever. If any effect can be discerned, it is that the proportion of women in the workforce increased slightly faster as the Act was being introduced; perhaps they were attracted by the higher salaries?

The lesson of all this is that the economy is complicated and textbook economic logic alone will get us only so far. The economist Alan Manning recently gave a public lecture at the London School of Economics, where he drew parallels between the Equal Pay Act and the minimum wage, pointing out that in both cases theoretical concerns were later dispelled by events.

The UK minimum wage took effect 16 years ago this week, on April 1 1999. As with the Equal Pay Act, economically literate commentators feared trouble, and for much the same reason: the minimum wage would destroy jobs and harm those it was intended to help. We would face the tragic situation of employers who would only wish to hire at a low wage, workers who would rather have poorly paid work than no work at all, and the government outlawing the whole affair.

And yet, the minimum wage does not seem to have destroyed many jobs — or at least, not in a way that can be discerned by slicing up the aggregate data. (One exception: there is some evidence that in care homes, where large numbers of people are paid the minimum wage, employment has been dented.)

The general trend seems a puzzling suspension of the law of supply and demand. One explanation of the puzzle is that higher wages may attract more committed workers, with higher morale, better attendance and lower turnover. On this view, the minimum wage pushed employers into doing something they might have been wise to do anyway. To the extent that it imposed net costs on employers, they were small enough to make little difference to their appetite for hiring.

An alternative response is that the data are noisy and don’t tell us much, so we should stick to basic economic reasoning. But do we give the data a fair hearing?

A fascinating survey reported in the most recent World Development Report showed World Bank staff some numbers and asked for an interpretation. In some cases, the staff were told that the data referred to the effectiveness of a skin cream; in other cases, they were told that the data were about whether minimum wages reduced poverty.

The same numbers should lead to the same conclusions but the World Bank staff had much more trouble drawing the statistically correct inference when they had been told the data were about minimum wages. It can be hard to set aside our preconceptions.

The principle of the minimum wage, like the principle of equal pay for women, is no longer widely questioned. But the appropriate level of the minimum wage needs to be the subject of continued research. In the UK, the minimum wage is set with advice from the Low Pay Commission, and it has risen faster than both prices and average wages. A recently announced rise, due in October, is well above the rate of inflation. There must be a level that would be counterproductively high; the question is what that level is.

And we should remember that ideological biases affect both sides of the political divide. In response to Alan Manning’s lecture, Nicola Smith of the Trades Union Congress looked forward to more ambition from the Low Pay Commission in raising the minimum wage “in advance of the evidence”, or using “the evidence more creatively”. I think British politics already has more than enough creativity with the evidence.

Written for and first published at ft.com.

The pricing paradox: when diamonds aren’t on tap

‘Diamonds are costly because we desire them. But what if that isn’t true? What if they are desirable because they are costly?’

A glass of water costs very little; a diamond costs a lot. Yet there is nothing more useful than water, while the most prized uses of diamonds are decorative. This apparent paradox has tested some fine minds. Adam Smith’s answer to the paradox was that diamonds were expensive because it was hard work to find them and dig them up. That seems to strike close to the truth but it’s not the way that modern economics approaches the problem.

The usual name for this puzzle is the “paradox of value” or “the water-diamond” paradox but I now prefer to call it the “Button Gwinnett paradox”. (I hadn’t heard of Button Gwinnett until his life was described in a recent episode of the WNYC radio programme Radiolab.) The British-born Gwinnett moved to the colony of Georgia in the mid-1700s. He was a failed businessman, a serial debtor and a B-list politician in the independence movement. But, as it happened, he was one of the 56 signatories of the Declaration of Independence.

Gwinnett might seem a minor figure compared to some of the other men whose names sit beside his: John Hancock, Thomas Jefferson, John Adams and Benjamin Franklin. Despite that, a Button Gwinnett signature is vastly more valuable than a Jefferson or a Franklin. The simple reason for this is that collectors naturally wish to own the complete set of 56 signatures. Ben Franklin lived into his eighties and was a prolific correspondent, so there is no shortage of Franklin signatures.

Gwinnett died in a duel the year after signing the Declaration of Independence. His signature was recently discovered on the parish register of St Peter’s Church in Wolverhampton, where he was married. Most of the other signatures he left behind were on IOUs.

Benjamin Franklin may have been one of the most remarkable human beings in history but when collecting your set of Independence signatures, it’s the Button Gwinnett that will prove the final piece of the jigsaw. Anyone selling a Gwinnett will find few other sellers and many eager buyers.

Which brings us back to water and diamonds. Diamonds are expensive because at the point at which the supply of diamonds dries up, there are plenty of buyers willing to pay handsomely, and they compete with each other. Water is cheap in temperate climes because after satisfying our demand for drinking and cooking, then for washing and for irrigation, and finally for swimming around in, there is still plenty left. The value of the first litres of water may be incalculably high but the marginal value of one more litre is very low, and it’s this value that sets the price.

Everything so far has assumed that our desire for an object — a diamond, a glass of water, a Button Gwinnett signature — is a given. Diamonds are costly because we desire them, and not the other way around. But what if that isn’t true? What if diamonds are desirable because they are costly?

The economist Thorstein Veblen coined the term “conspicuous consumption” to describe situations where an object is attractive merely because it is expensive. The designer watch or car is valuable because, like a peacock’s tail, it is a credible indicator that you have resources to spare. What was the point of spending so much on that diamond engagement ring otherwise?

Another possibility is “pricing bias”. If we don’t really know a good suit or a good bottle of wine from a bad one, we tend to use the price to give us a clue. This is not strictly logical — after all, anyone can double the asking price of anything they are selling, so price is not by itself a reliable clue to quality. But pricing bias exists. Studies show that people will rate a wine more highly in a taste test if they think it is expensive; even placebo painkillers are more effective if the patients believe they are costly new drugs rather than cheap new drugs.

. . .

The final word on this should go to a team led by Laurie Santos at Yale’s Comparative Cognition Laboratory. Santos has spent some time teaching capuchin monkeys how to use money, to exchange it for food and to understand the idea that food can have a price that is high or low. In recent work with Robin Goldstein of UC Davis, Santos’s team has been trying to figure out whether the monkeys also display pricing bias.

It seems not. After a series of trials where monkeys were allowed to buy cheap or expensive jelly and ice lollies, they were then let loose on a free buffet to see if they gravitated towards the once costly items. They didn’t; unlike humans, the monkeys couldn’t care less what the item typically cost. They liked what they liked. In this, they differ not only from humans but also from starlings: Alex Kacelnik and Barnaby Marsh, zoologists at Oxford, have found that starlings prefer more costly food.

My guess is that the monkeys would have little interest in a Button Gwinnett signature. And those glossy advertisements for diamonds and designer handbags? They are evidently far too sophisticated for capuchin tastes.

Written for and first published at ft.com.

Man v machine (again)

‘The Luddite anxiety has been dormant for many years but has recently enjoyed a resurgence’

I’m writing these words in York, the city in which, two centuries ago, the British justice system meted out harsh punishments — including execution — to men found guilty of participating in Luddite attacks on spinning and weaving machines. By a curious coincidence, I’ve just read Walter Isaacson’s article in the FT explaining how wrong-headed the Luddites were. I’m not so sure.

“Back then, some believed technology would create unemployment,” writes Isaacson. “They were wrong.”

No doubt such befuddled people did exist, and they still do today. But this is a straw man: we can all see, as Isaacson does, that technology has made us richer while employment is as high as ever. (The least appreciated job-creating invention may well have been the washing machine, which helped turn housewives into women with salaries.)

The Luddites themselves had a more subtle view than Isaacson suggests, and one which is as relevant as ever. They believed that the machines were altering economic power in the textile industry, favouring factory owners and low-skilled labourers at the expense of skilled craftsmen. They wanted to defend their interests and they did so violently. As the historian Eric Hobsbawm put it, their frame-breaking activity was “collective bargaining by riot” and “simply a technique of trade unionism” in the days before formal unions existed.

To put it another way, the Luddites weren’t idiots who thought that machines would destroy jobs in general; they were skilled workers who thought that machines would devalue their specific jobs and their specific skills. They were right about that, and sufficiently determined that stopping them required more than 10,000 troops at a time when the British army might have preferred to focus on Napoleon.

The Luddite anxiety has been dormant for many years but has recently enjoyed a resurgence. This is partly because journalists fear for their own jobs. Technological change has hit us in several ways — by moving attention online, where (so far) it is harder to charge money for subscriptions or advertising; by empowering unpaid writers to reach a large audience through blogging; and even by introducing robo-hacks, algorithms that can and do extract data from corporate reports and turn them into financial journalism written in plain(ish) English. No wonder human journalists have started writing about the economic damage the robots may wreak.

Another reason for the robo-panic is concern about the economic situation in general. Bored of blaming bankers, we blame robots too, and not entirely without reason. Inequality has risen sharply over the past 30 years. Many economists believe that this is partly because technological change has favoured a few highly skilled workers (and perhaps also more mundane trades such as cleaning) at the expense of the middle classes.

Finally, there is the observation that computers continue to develop at an exponential pace and are starting to make inroads in hitherto unexpected places — witness the self-driving car, voice-activated personal assistants and automated language translation. It is a long way from the spinning jenny to Siri.

What are we to make of all this? One view is that this is business as usual. We’ve had dramatic technological change for the past 300 years but it’s fine: we adapt, we still have jobs, we are incomparably richer — and the big headache of modernity isn’t unemployment but climate change.

A second view is that this time is radically different: the robots will, before long, render many people economically valueless — simply incapable of earning a living wage in a market economy. There will be plenty of money around but it will flow to the owners of the machines, and maybe also to the government through taxation. In principle, all could be well in such a future but it would require a radical reimagining of how an economy could work. The state, not the market, would be the arbiter of who gets what. Such a world is probably not imminent but, by 2050, who knows?

 . . . 

The third perspective is what we might call the neo-Luddite view: that technology may not destroy jobs in aggregate but rather changes the demand for skills in ways that are real and troubling. Median incomes in the US have been stagnant for decades. There are many explanations for that, including globalisation and the decline of collective bargaining, but technological change is foremost among them.

If the neo-Luddites are right, then the challenge in front of us is simply to adapt. Individual workers, companies and the political system will have to deal with wrenching economic changes as old industries are destroyed and new ones created. That seems a plausible view of the near future.

But there is a final perspective that doesn’t get as much attention as it might: it’s that technological change is too slow, not too fast. The robo-booster theory implies a short-term surge in jobs, as all those lovely new machines are designed and built and installed, followed by a long-term surge in productivity as the robots make the economy ruthlessly efficient. It is hard to see much sign of either trend in the economic statistics. Productivity, in particular, has been disappointing in the US and utterly dismal in the UK. Where are the robots when we need them?

Written for and first published at ft.com.

Boom or bust for bitcoin?

Bitcoin appeals to libertarians on the basis that governments cannot arbitrarily make more of it

In a moment, I’ll gaze into the crystal ball and foretell the future of the world’s most famous cryptocurrency, bitcoin. I should first explain what’s happening now.

It was developed in 2008 by an unknown programmer or programmers. Confusingly, bitcoin is both a payment technology and a financial asset. The asset called bitcoin has no intrinsic value but it has a market price that fluctuates wildly. Like digital gold, it appeals to libertarians on the basis that governments cannot arbitrarily make more of it.

The payment technology called bitcoin is what you might get if you ran the Visa network over a peer-to-peer network of computers. In case that description doesn’t help, it’s a way of sending money anywhere in the world but instead of relying on the authority of a financial intermediary such as Visa or Western Union, it uses a decentralised network to verify that the transaction has occurred. The record of all previous transactions is called the blockchain; it, too, is stored on a decentralised network. The entire process relies on cryptographic techniques to prevent fraud, which is why bitcoin and other currencies like it are called cryptocurrencies.

This may all seem very esoteric but the internet was esoteric once and it turns out to have become important. So what lies ahead for bitcoin?

Here’s one scenario.

Bitcoin has enjoyed many booms and busts in value, and later in 2015, the price surges again. This will be the biggest yet, drawing more and more people into the market. As the dotcom bubble and railway mania proved, even revolutionary technologies can be overvalued; with Bitcoins selling for $2,000, $5,000 and eventually $10,000 each, nemesis is around the corner.

The first sign of trouble will be the scams. A recent research paper by computer scientists Marie Vasek and Tyler Moore identified almost 200 bitcoin scams, in which about 13,000 victims lost $11m. Such scams will only become more common as the stakes become higher and the pool of naive investors deeper. Soon they will be the stuff of mainstream consumer rights phone-ins.

Arguably, scams are a sign that Bitcoin has matured — after all, nobody proposes abandoning the dollar because con artists like to be paid in dollars. But they are just a foretaste of what is to come — Bitcoin will be gutted by predatory monopolists.

The Bitcoin system has always relied on a crowd of people putting their computers to work verifying transactions and writing them into the blockchain, a task which costs money and energy. In a rather confusing analogy with gold, these people are called “miners” and they are compensated in Bitcoins, of course. Yet there is a basic inconsistency at the heart of this system, as the economist Kevin Dowd has observed: Bitcoin mining needs to be done by a decentralised crowd but is more efficiently done by large arrays of computers owned by a few players. Or possibly just a single one.

Even today, Bitcoin mining is a game for the big boys. As the Bitcoin mining industry becomes a tight, self-serving oligopoly, the stage is set for Bitcoin counterfeiting on a massive scale. In 2018, 10 years after the invention of Bitcoin, the system collapses under the weight of its own contradictions.

It’s an intriguing story — but of course, it is just a story. We could give it a name: “BitCon”.

. . .

If you don’t believe that, I have another story for you. The title is “Daisy Chains”. Throughout 2015 and 2016, the price of Bitcoins continues to collapse. Speculators lose interest and some of the big miners sell off their computers at a heavy loss. The spotlight moves elsewhere but the true believers in the power of decentralised blockchain processing continue to develop the system.

Bitcoins aren’t the only things that can be transferred using a peer-verified network, after all — you could transfer the digital lock to a smart car; or a financial contract, with pay-offs and penalties automatically adjudicated and paid for by the blockchain. The question is whether the effort of doing all this is more efficient than the current centralised systems using interbank payments.

The answer is yes but only in certain circumstances. A blockchain is a ledger of every digital transaction ever made on the system. This proves far too unwieldy for a universal means of payment. Yet specialised niche systems evolve: by 2018, block-chain processing is common for remittances; by 2019, block-chain processing pays for and controls self-driving taxis. You can even download an out-of-the-box blockchain app for your local babysitting circle — or your prostitution ring. Blockchain approaches don’t replace Western Union and Visa everywhere but they squeeze margins and make inroads for certain applications.

The only disappointment for the true Bitcoin enthusiasts is that Bitcoin itself, the currency that started it all, fails to catch on. Most people prefer a trusted brand. When a standard of value is used on these disparate blockchain processes, the most popular by far is “FedCoin” — more commonly known by its correct name, the US dollar.

Two stories about the future, and most likely neither one will come true. These are interesting times for cryptocurrencies.

Written for and first published at ft.com.

Battle for the web’s ‘last mile’

The fact that a few large players have such influence over vital services should make us all queasy

The cable companies who own the wires that plug us into the internet – particularly the “last mile” along your street and into your house – have a great deal of market power. Small wonder, then, that the notion of “net neutrality” is appealing: the term is usually characterised as the idea that all data transmitted over the internet should be treated equally. After all, why should Google get a zippier connection than a small rival? Why should Netflix have to pay an additional fee to Big Cable when customers have already paid handsomely to be connected?

Advocates of net neutrality won a famous victory a few weeks ago, when the US Federal Communications Commission announced plans to regulate cable companies as utilities. The aim of this was to enforce net neutrality rules after a vibrant grass-roots campaign.

Small wonder that the campaign became so popular. The idea that cable companies could partition the internet into slow lanes and fast lanes is infuriating. Customers have already paid for access, and they don’t take kindly to the prospect of “throttling” – deliberately degrading a service to extort money from content providers.

This kind of product sabotage is far older than the internet itself. The French engineer and economist Jules Dupuit wrote back in 1849 that third-class railway carriages had no roofs, not to save money but to “prevent the passengers who can pay the second-class fare from travelling third class”. Throttling, 19th-century style.

But imagine that a law was introduced stipulating “railway neutrality” – that all passengers must be treated equally. That might not mean a better deal for poorer passengers. We might hope that everyone would ride in comfort at third-class prices, and that is not impossible. But a train company with a monopoly might prefer to operate only the first-class carriages at first-class prices. Poorer passengers would get no service at all. Product sabotage is infuriating but the alternative – a monopolist who screws every customer equally – is not necessarily preferable.

It is easy to think of outrageous scenarios in which a cable company might exploit market power – favouring campaign videos from politicians who do its bidding, or shutting down rivals who pose a competitive threat.

But it is also easy to think of good reasons to treat different kinds of content differently. An online back-up service for big data sets might prefer a discount for a connection that will run only at quieter times of day. Stream the World Cup final and you’ll want to guarantee uninterrupted coverage; sell the highlights as a download and you might accept a cheaper, more volatile connection if it saves money.

With a mandatory uniform price, the online back-up might be too expensive to operate, the live stream too slow to satisfy customers, and the video download getting a faster connection than it really needs. (There is a formal economic model of this effect courtesy of Benjamin Hermalin and Michael Katz but it seems intuitive to me.)

What about the idea that customers have already paid for their internet content, so cable companies shouldn’t be able to demand cash from content providers too? That is not how things work elsewhere. In a shopping mall, customers enter for free and retailers pay to be there. (They pay very different rents, too.) At an industry convention, both the delegates and the exhibitors will pay. There is nothing sacred about the idea that one side of the market pays nothing. Customers may even benefit if content providers must pay, since then the cable company might wish to slash prices to attract them and increase its leverage with the content providers.

Should all content providers be able to connect free of charge? This may not be the best rule for consumers nor the best way to promote innovation. The best defence of such a rule is that it seems to have worked well in the past and, with so much at stake, a change would be risky – not a terrible argument but hardly cast-iron.

Nevertheless I am grateful to the advocates of net neutrality, because they have brought into sharp focus the importance of market power on the internet – both of content providers such as Google and Facebook, and the cable companies who connect us to them. The ability to connect to the internet has become a basic part of living a full economic, social and political life. We use the internet to make our voices heard, to spend money, to access services, to find out the news, to connect with our friends. Increasingly our fridges, cars and pacemakers will use it too. The fact that a few very large players have such influence over such vital services should make us all queasy.

Fast lanes and slow lanes are a symptom of this market power but the underlying cause is much more important. The US needs more internet service providers, and the obvious way to get them is to force cable companies to unbundle the “last mile” and lease it to new entrants.

Alas, in the celebrated statement announcing a defence of net neutrality, the FCC also specifically ruled out taking that pro-competitive step. The share prices of cable companies? They went up.

Written for and first published at ft.com.


  • 1 Twitter
  • 2 Flickr
  • 3 RSS
  • 4 YouTube
  • 5 Podcasts
  • 6 Facebook


  • Messy
  • The Undercover Economist Strikes Back
  • Adapt
  • Dear Undercover Economist
  • The Logic of Life
  • The Undercover Economist

Tim’s Tweets

Search by Keyword

Free Email Updates

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

Join 3,476 other subscribers

Do NOT follow this link or you will be banned from the site!