Tim Harford The Undercover Economist

Articles published in March, 2019

Why happiness is easy to venerate, hard to generate

In 1972, the teenage king of Bhutan, Jigme Singye Wangchuck, declared that “gross national happiness is more important than gross domestic product”. The sound bite has been echoed approvingly down the years, although the king may just have been making excuses. Bhutanese GDP per person was then the grinding poverty of about a dollar a day. If I were king of such a country, I’d be tempted to change the subject, too.

Clearly he had a point. Most of us would rather be poor and happy than rich and depressed. If so, gross national happiness seems a fine goal. But it is one thing for a monarch to announce that happiness is important. It’s quite another to make people happy. Shangri-La does not move from fiction to reality just because we desire it.

Bhutan has not always lived up to its own hype. Same-sex intercourse is illegal, which suggests a country with a less-than-expansive view of whose happiness matters. Three decades ago, around 100,000 of the Nepali-speaking Lhotshampa minority fled Bhutan to escape military persecution during a campaign of ethnic cleansing on a colossal scale. One-sixth of the entire population of Bhutan ended up in refugee camps in Nepal.

Even setting aside this enormity, it’s hard to see that Bhutan paid much more than lip service to gross national happiness. They hosted conferences, but according to a recent IMF working paper, nobody in the government collected systematic indicators on happiness until 2005. The World Happiness Report ranks Bhutan at 97th out of 156 countries, down from 84th a few years ago. Happiness is easy to venerate, but hard to generate.

Perhaps I am doing a disservice to a small kingdom wedged between sparring regional powers. It can’t be easy. And Bhutan has a lesson to teach us all — maybe we should think a little less about over-arching goals and a little more about specifics.

Consider some of the issues that are notoriously bypassed by GDP, the most common measure of economic activity: digital services are hard to value, while by design GDP omits any consideration of inequality or environmental damage. Unpaid work — of which men do a great deal, and women a great deal more — is also left out.

But if our aim is (for example) to reduce carbon emissions, we don’t achieve it by moaning about GDP. We achieve it with specific policies such as carbon taxes and investments in public transport and a renewable-friendly electric grid. Neither gender equality nor respect for unpaid work would be automatically improved by any change in the way national income accounts are computed.

And when Bobby Kennedy movingly commented that GDP “does not include the beauty of our poetry or the strength of our marriages”, he was quite right. Nevertheless I fail to see how any solution follows, whether for connubial harmony or American verse.

The specifics matter when it comes to happiness, too. Broad research into the causes of national happiness has tended to produce banal conclusions: we tend to compare ourselves to others, unemployment makes us miserable, and we hate being ill. There is nothing here to suggest that we need to overhaul commonplace policies such as redistributive taxation, the avoidance of recessions, and support for public health.

Just as with GDP itself, it is only when we move to the specifics that gross national happiness becomes useful. Richard Layard, one of the leading happiness researchers, argues that mental illness is a leading cause of misery, and that it can be treated very cost-effectively. That seems useful enough to me.

What is not useful is the sense that measuring GDP is the problem, and measuring gross national happiness is the solution. Few societies have ever really focused on either. We should all be happy about that.


Written for and first published in the Financial Times on 1 March 2018.

My book “Fifty Things That Made the Modern Economy” (UK) / “Fifty Inventions That Shaped The Modern Economy” (US) is out now in paperback – feel free to order online or through your local bookshop.

Free email updates

(You can unsubscribe at any time)

Just because you’re paranoid, doesn’t mean the algorithms aren’t out to get you

If you do not like the price you’re being offered when you shop, do not take it personally: many of the prices we see online are being set by algorithms that respond to demand and may also try to guess your personal willingness to pay. What’s next? A logical next step is that computers will start conspiring against us. That may sound paranoid, but a new study by four economists at the University of Bologna shows how this can happen.

The researchers allowed two simple artificial intelligence algorithms to compete against each other in a setting where they simultaneously set prices and reaped profits accordingly. The algorithms taught themselves to collude, raising prices from the cut-throat competitive level towards what a monopolist would choose. Price cuts were met with price wars, after which collusion would return. Just because you’re paranoid, it doesn’t mean the computers are not out to get you.

This is not a surprising result for anyone who — like me — squandered their youth studying the theory of industrial competition. Robert Axelrod’s book The Evolution of Cooperation (US) (UK) published in 1984, described a tournament in which computers played a “prisoner’s dilemma”, a scenario analogous to two competing sellers. The best approaches used the threat of punishment to sustain co-operation. They were also simple: not something that a machine-learning system would struggle to discover.

An obvious question is, who — if anyone — should be prosecuted for price fixing when the bots work out how to do it without being told to do so, and without communicating with each other? In the US, where the Federal Trade Commission has been pondering the prospect, the answer seems to be no one, because only explicit collusive agreements are illegal. The bots would only be abetting a crime if they started scheming together. Tacit collusion, apparently, would be fine.

This is a reminder that algorithms can misbehave in all kinds of intriguing ways. None of us can quite shake the image of a Skynet scenario, in which an AI triggers a nuclear war and then uses Arnold Schwarzenegger as the model for a time-travelling robot assassin on a mission to suppress human resistance. At least that strategy is refreshingly direct. The true scope of algorithmic mischief is much subtler and much wider.

We are rightly concerned about algorithms that practice racial or sexual discrimination, by accident or design. I am struck by how quickly tales of racist algorithms have gone from novelty to cliché. The stories may fade but the issue is not going away.

Algorithms that simply magnify human errors now appear almost quaint. In 2012, the Financial Times had a headline, “Knight Capital glitch loss hits $461m”; those were innocent times.

Then there were those T-shirts selling on Amazon a few years ago, offering offensive slogans such as “Keep Calm and Hit Her”, and bizarre ones such as “Keep Calm and Skim Me”. Hundreds of thousands of slogans were assembled by an algorithm and, if any appealed, the vendor would print them on demand. “We didn’t do it, it was the algorithm,” was a weak defence in 2013, but at least it was novel. That is no longer true.

We are also realising that the algorithms can amplify other human weaknesses — witness recommendation engines on YouTube and Facebook that seem to amplify disinformation or lead people down the dark tunnels of conspiracy thinking or self-harm.

By no means are all malevolent programs an accident; some are designed with mischief in mind. Bots can be used to generate or spread misinformation. Jamie Bartlett, author of The Dark Net (US) (UK), warns of a future of ultra-personalised propaganda. It is one thing when your internet-enabled fridge knows you’re hungry and orders yoghurt. It’s another when the fridge starts playing you hard-right adverts because they work best when you’re grumpy and low on blood sugar. And unless we radically improve both our electoral laws and our digital systems nobody need ever know that a particular message was whispered in your ear as you searched for cookies.

Obviously, both the law and regulators must be nimble. But ponder, too, the challenges for corporate public relations and social responsibility departments. The latter is about being a good corporate citizen; PR is about seeming to be so. But who takes corporate responsibility for a harmful or tasteless decision made by an algorithm?

It is not an entirely new problem. Before there was tacit collusion between algorithms, there was tacit collusion between sales directors. Before companies blamed rogue algorithms for embarrassing episodes, they could blame rogue employees, or their suppliers. Can we really blame the bank whose cleaning subcontractor underpays the cleaning staff? Or the sportswear brand opposed to sweatshop conditions, whose suppliers quietly hire children and pay them pennies?

The natural answer is: we can and we do, but subcontracting is a source of both deniability and complexity. Subcontracting to algorithms complicates matters, too. But we are going to have to figure it out.


Written for and first published in the Financial Times on 22 February 2019.

My book “Fifty Things That Made the Modern Economy” (UK) / “Fifty Inventions That Shaped The Modern Economy” (US) is out now in paperback – feel free to order online or through your local bookshop.

Free email updates

(You can unsubscribe at any time)

Is life a bet, or an experiment?

Since the official Brexit policy of the UK government now seems to be “accidents happen”, it is a nourishing age for those who dine on uncertainty. Most of us, however, started feeling queasy long ago. If you’re looking for guidance as to how to digest the unpredictabilities of life, why not turn to a poker player for advice?

Annie Duke, author of Thinking In Bets (UK) (US), says that “wrapping our arms around uncertainty and giving it a big hug will help us become better decision makers”. Fair enough. If we’re going to have to eat our uncertainty broccoli anyway, it may be best to swallow it down before it goes cold and limp.

As the winner of several million dollars as a professional poker player, Ms Duke has some credibility on this point. But is she right?

One quibble is that games have a more tightly-defined spectrum of uncertainty than reality does. Edward Thorp, a mathematician who achieved considerable success at blackjack and as a hedge fund manager, found (UK) (US) that the true risks he faced in the casinos were not an unlucky turn of the card, but crooked dealers and poisoned coffee. Not for nothing does Nassim Taleb, author of The Black Swan (UK) (US) warn of the “ludic fallacy” — treating the unknown risks of life as though they were the known risks of a game of chance.

Still, poker is a more instructive game than many others. John von Neumann, the brilliant mathematician who laid down the foundations of game theory in the late 1920s, was a poker player. Poker was no mere computational problem like chess, he said: “Real life consists of bluffing, of little tactics of deception, of asking yourself what is the other man going to think I mean to do, and that is what games are about in my theory.”

So what does poker teach us about “wrapping our arms around uncertainty”? Ms Duke offers us several lessons.

One is that, since luck matters as well as skill, bad decisions can have good outcomes, and vice versa. If you drive drunk, you will probably get home without killing anyone, but that would not make drunk driving a good decision.

A second lesson is that we should always be willing to ask ourselves, “do I want to bet on that?” — it’s easy to be overconfident if there are no obvious consequences for being wrong. A bet forces us to think about the odds and the possibility that someone else may know better.

So there is much to be said for thinking in bets. Yet we should not overlook an alternative approach to uncertainty: thinking in experiments. An experimental thinker views the uncertainties of the world as something to be resolved through tentative trial and error. Try something modest or reversible; an experiment doesn’t need to be a double-blind, randomised controlled trial to yield useful information. If it works, do more of it.

Some decisions are by their nature irreversible. Each big poker hand is a one-shot proposition that does not offer much scope for experiment. The same could be said of some investments: if you think shares in Tesla are cheap, there is little to be said for buying just one share and watching it to find out what happens to its price. An experiment cannot help; instead you must figure out whether the odds are in your favour, and take the plunge.

But other decisions are more experimental. In these cases, the choices can be made in stages, with each step designed to reveal some information. From Marvel’s decision to publish the Spider-Man comics (that went well) to Google’s launch of the G+ social network (that didn’t) a company can see what works and then either redouble its efforts or abandon the project. For an individual, anything from a new hobby to a new career can be treated as an experiment.

Many of the decisions we make are reversible. Only our stubbornness makes them permanent. Thinking in bets forces a commitment, which is sometimes helpful but sometimes not. Thinking in experiments allows us to learn.

I thought of all this when reading of poker professional John Hennigan’s $30,000 bet that he could move to Des Moines, Iowa, and live there, just for a few weeks. As Ms Duke tells the story, Mr Hennigan was bored out of his mind within days, and paid $15,000 to buy himself out of the bet and move back to Las Vegas. That’s thinking in bets at its worst: an idle thought (“should I move to Des Moines?”) became a high-stakes zero-sum game. Fun — if you like that kind of thing.

A few years ago, my family were agonising over a similar question (“Should we move to Oxford?”). We vacillated and made lists of pros and cons. What resolved the uncertainty was realising the decision could be an experiment: rather than selling up, we could rent a place in the city for a year to see how things worked out. Running this experiment created some additional costs, but we have never regretted doing it.

Thinking in bets is a rigorous and admirable habit, but not everything has to be a high-stakes poker hand. If you can make it work, thinking in experiments is less painful.



Written for and first published in the Financial Times on 15 Feb 2019.

My book “Fifty Things That Made the Modern Economy” (UK) / “Fifty Inventions That Shaped The Modern Economy” (US) is out now in paperback – feel free to order online or through your local bookshop.

Free email updates

(You can unsubscribe at any time)

Why inflation is good for us

Not long ago, I wrote a column in defence of central banks. Some readers were quick to disagree. Central banks had failed to maintain “the real value of our fiat currencies”, wrote one, urging me to ponder how inflation had eroded the true value of savings over the decades.

A glance at Venezuela, where inflation over the past year has been more than 100,000 per cent and the economy is breaking down, reminds us that this is no idle complaint. Central banks must keep inflation under control. But what does “under control” mean? How much inflation is too much? And — a question only an economist could ask — how much inflation is too little? It goes without saying that hyperinflation is an economic catastrophe, so the first thing to check is whether hyperinflation is likely in an advanced economy — or indeed a competently governed country of any sort. It is not.

In 2012, economists Steve Hanke and Nicholas Krus assembled a list of every confirmed episode of hyperinflation in history. There weren’t very many: 56 in total, mostly in the 20th century, to which we might add recent outbreaks in Zimbabwe, Iran, Venezuela and perhaps North Korea. France suffered a bout in the 1790s, but most instances of hyperinflation occurred either in central European states after the first world war (including the infamous crisis in Weimar Germany), or during or immediately after the second world war (including Hungary, history’s worst example of hyperinflation), or in the eastern bloc as the Soviet Union disintegrated.

Hyperinflation does not strike at random, and it does not happen because central banks briefly slumber. It must be manufactured by the relentless printing of money, generally as the last resort in the face of political dysfunction alongside a severe fiscal crisis. Perhaps it is rash to say so, but I think we can set aside fears of hyperinflation in an advanced economy today. If it ever does happen, it will be only one element in a far more comprehensive economic disaster.

But the readers who emailed to complain about inflation did not express concerns about hyperinflation. They are worried about low-level ambient inflation, the kind that central banks not only tolerate, but actively seek. Central banks do not try to maintain “the real value of our fiat currencies”, but to erode them, typically by 2 per cent a year. They have often been explicitly instructed to do so by elected politicians.

Are those instructions wise? Even 2 per cent inflation will halve the value of money in 36 years. That seems bad, but let’s try to pin down why it might matter. It may help to remember what it is that we expect any good currency to do.

First, we want it to serve as a medium of exchange, allowing me to pick up a loaf of bread without having to persuade the baker to swap it for a copy of The Undercover Economist Strikes Back. Inflation at 2 per cent a year — or 5, or even 20 — does not prevent money serving as a medium of exchange.

The second, and perhaps most fundamental, role of money is as a stable unit of account. It helps us understand the economic forces around us, whether a particular product is expensive or cheap, without resorting to a calculator. Hyperinflation destroys that. “Are we ruined or in clover?” asks a character in an Erich Maria Remarque novel set in the Weimar hyperinflation. No one knows. But moderate inflation will not boggle minds on a trip to the shop.

Inflation does more obvious damage to money’s third role, which is as a store of value. If you stick your currency under the mattress then inflation will hurt you. It will also hurt if you have a non-indexed pension, or cannot find a high-interest savings account. Normally, however, a well-functioning financial sector offers returns to compensate for inflation.

That has not been the case since the 2008 financial crisis, of course. But the pain savers are feeling is not because central banks have carelessly let inflation take off. It is the result of a deliberate policy of low interest rates to stimulate spending and investment. Perhaps this policy is a mistake, perhaps not. But it would be wrong to view it as a dereliction of duty.

Another source of pain is governments’ fondness for using inflation as a way to grab a bit more revenue, by taxing nominal interest payments. It’s an insult for savers, but let’s be realistic: the tax would be levied somehow anyway. Inflation is sometimes the taxman’s chisel, but he has other tools.

It is surprisingly difficult to find any serious costs to low levels of inflation. In contrast, the benefits are easily stated: more room to stimulate the economy in a recession, and more room for real wages to adjust if they must. Olivier Blanchard, during his time as chief economist of the IMF, even floated the idea that the inflation target should be 4 per cent, not 2 per cent. He is by no means alone in that opinion. Like a low dose of aspirin, a low dose of inflation is unlikely to do much harm — and it can prevent an economic heart attack.


Written for and first published in the Financial Times on 8 Feb 2019.

My book “Fifty Things That Made the Modern Economy” (UK) / “Fifty Inventions That Shaped The Modern Economy” (US) is out now in paperback – feel free to order online or through your local bookshop.

Free email updates

(You can unsubscribe at any time)

How high should the top rate of tax be, and who should pay it?

What should the top rate of income tax be? Should it be 70 per cent, as has been informally suggested by the young star of the US Democratic party, Congresswoman Alexandria Ocasio-Cortez? That instinctively feels too high to me. But, as an economist with sporadic hopes of making logical arguments based on evidence, I admit that “instinctively feels too high” is a weak response.

What about 50 per cent, the official policy of the UK’s opposition Labour party at the last general election? Or zero, the optimal top rate that emerged from a thought experiment posed by the late James Mirrlees, a Nobel laureate in economics?

An alternative is to tax wealth instead of income, as US Democratic Senator Elizabeth Warren has proposed. But — at least in principle — there is not much difference between a small annual tax on total wealth and a large tax on the nominal return generated by that wealth.

To make the case for a top rate of tax above 70 per cent, it helps to believe four things.

The first is that taxable income itself won’t evaporate in the face of a high rate, as it did in the UK when the top tax rate was briefly raised from 40 to 50 per cent in 2010, then cut to 45 per cent. Most high earners found it easy to realise income early, or late, and avoid the 50 per cent rate. A permanent increase is harder to avoid; so is an increase that is enforced with determined (or draconian) measures; as is an increase levied by a large economy with global legislative reach such as the US. In smaller economies such as the UK’s, the very rich are more likely to take themselves elsewhere for any given tax rate.

One academic paper produced by Emmanuel Saez (a star in the study of inequality) and Peter Diamond (a Nobel laureate and colleague of Mirrlees) estimated that the combined rate of tax on the income of high earners could be 73 per cent in the US without proving counter-productive. Another paper, published in the same journal, by Gregory Mankiw and co-authors, put the optimal top rate at just under 50 per cent instead. The difference lies in the assumptions.

The second thing one needs to believe is that the rich will barely miss any extra income if tax rates rise. The truth of this is unknowable, although another famous study from yet more Nobel laureates, Daniel Kahneman and Angus Deaton, suggests that money will not improve your everyday mood and wellbeing after an income of $75,000 a year or so. To reach their conclusions about the 73 per cent rate, Professors Diamond and Saez assume that a dollar is 25 times more valuable to a person on about $50,000 a year than to a person on $500,000. That is not an insane assumption, but it’s an assumption nonetheless.

If you accept these first two beliefs, the economic case for a high top rate of tax follows. A high rate maximises revenue if the tax base doesn’t shrink too much, and revenue maximisation is a reasonable goal if it’s true that the rich would barely notice the lost income.

But this argument ranges far beyond economics. If you like high tax rates, the third thing it helps to believe is that inequality is intrinsically corrosive. Perhaps it undermines democracy. Perhaps it causes stress, envy or resentment. The empirical evidence is not much help here; it is sketchy and often seems tendentious. Causal channels are unclear: does inequality lead to a hollowed-out state? Or does a hollowed-out state enable inequality? Perhaps a thought-experiment is more helpful here: how would you feel about a policy that simply confiscated resources from the super-rich and destroyed them? Would such a policy be a criminal waste and a grotesque infringement of liberty, or a helpful rebalancing of the scales?

Then there’s a fourth, often unstated, belief: that the rich have so much money that a high rate of tax will raise serious revenue. That depends on who you regard as “rich”. Ms Ocasio-Cortez mentioned a threshold of $10m. Profs Diamond and Saez focused on the highest earning 1 per cent of taxpayers, implying that the band would apply above around $500,000 a year. The Labour party wanted its highest rates to apply on incomes over £100,000.

These are very different definitions of “rich” and they have very different implications for revenue. For example, Ms Ocasio-Cortez’s income threshold of $10m is higher than that required to get into the top 0.01 per cent of the US income distribution: about 16,000 families. This tiny slice of the US population receives a less-than-tiny 5 per cent of total US income — which nevertheless implies that 95 per cent of income is earned by those making less.

The super-rich are a tempting target, but a serious attempt to raise revenue cannot stop with them. Whether we are talking about income or wealth, the lion’s share lies not with the billionaires but with the comfortably off. It is nice to talk about taxing somebody else’s money, but in a world of chronic budget deficits and worsening demographics, the ethics and economics of higher tax rates are unlikely to remain someone else’s problem.

Written for and first published in the Financial Times on 1 Feb 2019.

My book “Fifty Things That Made the Modern Economy” (UK) / “Fifty Inventions That Shaped The Modern Economy” (US) is out now in paperback – feel free to order online or through your local bookshop.

Free email updates

(You can unsubscribe at any time)


  • 1 Twitter
  • 2 Flickr
  • 3 RSS
  • 4 YouTube
  • 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

Search by Keyword

Tim’s Tweets

Free Email Updates

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

Join 172,203 other subscribers

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