Tim Harford's Blog, page 104

How a prison camp recession explains the economy

For more about these ideas, why not buy my new book?

Let’s have some real-time economics

‘It would have done the Fed no harm to have had more people with a habit of making snap decisions’

What would it take to make economics more useful in a crisis? Not more rigorous or more realistic – although that would be nice – but simply better equipped to deal ad hoc with the financial equivalent of a burning building?

It’s sobering to read the recently published transcripts of the Federal Reserve’s Open Market Committee meeting held on September 16 2008, the day after Lehman Brothers filed for bankruptcy. Fed chairman Ben Bernanke and his colleagues knew AIG was also in trouble but not that the worst recession since the 1930s was under way.

The transcripts induce, at times, the frustration of watching Titanic. The ship is doomed, yet our heroes suspect nothing! The Fed committee raises the possibility of a sharp cut in interest rates but inertia wins out. They are unwilling to act until the dust settles.

The rearranging-the-deckchairs moment comes as the committee discusses the right words to use in its press release. Should it say it is watching developments “closely”, or “carefully” or just watching? Nobody really knows.

In retrospect, the Fed was slow to act – as subsequently evidenced by three later, large rate cuts in an attempt to catch up. But it would be unfair to suggest that the committee was clueless. The meeting begins with a crisp discussion of the impact so far of the Lehman collapse. That’s followed by an agreement to swap currency, without limit, with other major central banks.

The overwhelming sense, however, is of a group of men and women who are rooted to the spot in the face of uncertainty. One of the staff economists, Dave Stockton, presents a detailed economic outlook before admitting, “I don’t really have anything useful to say about the economic consequences of the financial developments of the past few days.” With hindsight, what that meant was that he didn’t really have anything useful to say at all.

Bernanke himself sums up the mood perfectly as he reflects on the rapidly evolving bank bailouts and the risk of moral hazard: “Frankly, I am decidedly confused and very muddled about this.” There is wisdom there – but not of a very reassuring sort.

Hindsight is a wonderful thing and nobody should envy policy makers in such a situation. But is there a better way to conduct emergency policy? I have three suggestions.

First, increase diversity. Despite the reputation of the US for having a revolving door between big business and government, the Fed’s board looked weighted towards government insiders such as Timothy Geithner, then head of the New York Fed, and academics such as current Fed chairwoman Janet Yellen and Bernanke himself. Not many board members had high-level business experience. A variety of perspectives tends to generate a more honest conversation, and it would have done the Fed no harm to have had a few more people with a habit of making snap decisions.

Second, overhaul the economic data available. The Fed was flying blind: it knew surprisingly little about who was exposed to a collapse of Lehman and, immediately after that collapse, a vast tangle of contracts sat in limbo while the picture was slowly sorted out.

In a speech in New York two years ago, Andrew Haldane of the Bank of England argued that financial regulators and risk managers should draw inspiration from the development of supply-chain standards. These standards turned the humble barcode into a way of tracking products as they moved around the world. Because firms could follow products through the production and logistics system, bottlenecks could be bypassed and supply crunches spotted in advance.

What we need now are barcodes for financial products and companies – and they are on the way. The Financial Stability Board, which tries to co-ordinate financial policies across borders, has been developing the building blocks of a system designed to identify specific financial products and legal entities. That last point sounds trivial but it isn’t. Lehman Brothers was a Gordian knot of corporate vehicles. An up-to-date network map of who owned whom would have been invaluable.

Once better data are available, they also need to be displayed in a clear, robust and timely manner. Emery Roe of UC Berkeley, author of Making the Most of Mess, studies high-reliability systems such as electricity networks, whose operators must keep the system up and running despite a constantly evolving set of constraints and setbacks. Roe argues that one key feature of such systems is a clear visual display of trustworthy information. Electricity network operators have this but finance is way behind. The ultimate goal should be for regulators to glance at a computer display and spot stresses and vulnerabilities in the financial system, in real time – not easy.

Perhaps we should also treat such endless firefighting with more respect. In economics, ecology and other disciplines, Roe argues, those making tough decisions in the field are disparaged as practising “agency science”. Yet somewhere there is an ecologist who needs to decide how to respond immediately to the latest toxic spill, and there is an economist who needs to decide how to respond immediately to the latest bankruptcy.

We need people with the art of real-time economics – an art that shouldn’t be dismissed just because it cannot match the rigour of the ivory tower.

Also published at ft.com.

Sorry decline of English social housing

We are getting rid of homes that are compact, cheap, fully used and warm, writes Tim Harford

‘In 2012-13, the private rented sector overtook the social rented sector to become the second largest tenure in England.’ English Housing Survey Headline Report 2012-13, Department for Communities and Local Government, February 26

What on earth does it mean to be the ‘second largest tenure’?

You know you’re reading a government report when the top line of the “key findings” section is incomprehensible without an explanatory note. What they mean is that there are more English households renting private housing than renting social housing. Social housing is provided at a discount by local authorities or housing associations, and the number of households renting social housing has been in slow but steady decline for three decades despite a rising population.

And as traditional social housing has been in decline, privately rented housing has been rising to take up the slack?

Over 15 years, yes – the private rental sector has boomed. But the longer-term story is that home ownership is on the rise, from about 55 per cent to about 65 per cent. The private bit of the rental sector has a much larger slice of a smaller pie.



Well, that’s good. Private markets deliver prosperity.

Sometimes. But there are reasons to mourn the decline of social housing. A quarter of the people now privately renting are receiving a government handout in the form of housing benefit, so this isn’t an argument about needing help from the taxpayer – it’s an argument about which form of help might work best. Social housing is smaller, cheaper and less likely to be “under-occupied” than private housing. Yet social housing is also the best insulated by some measures – and private rentals are the worst insulated no matter how you look at them. There’s also a thing called the “decent homes” standard, meant to eliminate the worst damp, drafts and hazards. Social housing is the most likely to pass that standard, and is improving quickest. Private housing is the most likely to fail, and is improving slowest. Social housing is compact, cheap, fully used, fast-improving, safe and warm – and we’re getting rid of it.

But at least homes in general are more likely to pass safety and energy efficiency standards. That’s surprising: I keep hearing we’re in a housing crisis.

The crisis is that we’re not building enough houses. The quality of the ones we have is indeed improving.

You seem grouchy about social housing, but the big ‘right to buy’ sell-off did have one positive consequence: far more people now own their own homes.

I hadn’t realised that counted as a positive consequence.

Isn’t it? Isn’t owning your own home what prosperous people do?

You’re getting confused. It’s true that many people want to own their own homes, and that these days you need a good sock of cash to do so. But that doesn’t mean that it’s a good idea for a society as a whole to have very few rented properties. Have you heard of Oswald’s Hypothesis?

Didn’t they just put out a second album?

Oswald’s Hypothesis is that home ownership is a cause of unemployment. The idea was put forward by Andrew Oswald of Warwick University in the mid-1990s after he observed some worrying correlations. First, home ownership and unemployment had been on the rise across the OECD for many years. Second, countries such as Spain had high home-ownership rates and high unemployment, while countries such as Switzerland had low home ownership rates and low unemployment. There are similar correlations across US states.

It must be a coincidence. Surely you’re not saying that home ownership is hazardous to your job?

In the social sciences, who really knows anything for sure? But Mr Oswald’s story has some plausibility. If you own your house it’s not easy to move, so home ownership reduces the number of jobs you can consider.

So maybe the rise of owner-occupation isn’t good news at all?

It has its risks, for sure. But consider something that Mr Oswald wrote in 1999: “A key part of the problem is young unemployed people living at home, unable to move out because the rental sector hardly exists.” That’s no longer the reason young people live at home; the private rental sector has boomed. Maybe that helps explain why – despite all the economic woes of the past six years – unemployment in the UK was never as bad as economists feared.

Also published at ft.com.

Golden rules of thumb

‘The human brain is a marvellous thing but it does not seem to have evolved to cope with high finance’

The human brain is a marvellous thing but it does not seem to have evolved to cope with high finance. I’ve written before about the economist and financial literacy expert, Annamaria Lusardi. One of her findings is that one-third of Americans over the age of 50 failed to answer this question correctly: “Suppose you had $100 in a savings account and the interest rate was 2 per cent per year. After five years, how much do you think you would have in the account if you left the money to grow: more than $102, exactly $102, less than $102?”

One answer to this woeful situation is financial education. But can such depths of ignorance be paved over with a few evening classes? Another option is to nudge us into sensible decisions – for example, by automatically setting up pensions for us. Again, this is fine as far as it goes.

A third line of attack, embraced by some financial gurus, is the rule of thumb.

Here’s one: your age in years should be the percentage of your portfolio that is not invested in shares. Here’s another: your debt payments should be no more than 36 per cent of your gross income. A third, widespread in the US, is that in retirement one should draw down 4 per cent of current wealth each year.

Such rules of thumb are clearly limited. For example, the suspiciously precise stricture that debt payments should not top 36 per cent of gross income ignores inflation. If your mortgage interest is 12 per cent, your annual pay rise is 10 per cent and inflation is 10 per cent, then all you need do is survive two or three years and inflation will erode your mortgage. But if interest rates are 2 per cent, your pay is flat and inflation is low, then a difficult repayment schedule now isn’t going to get any easier in the future.

Some economists have been investigating how close these rules of thumb are to the optimal strategies they can compute. In a research paper entitled “Spending Retirement on Planet Vulcan”, Moshe Milevsky and Huaxiong Huang compare the “4 per cent” rule to the optimal drawdown of assets in retirement, as it might have been calculated by Star Trek’s Mr Spock. They conclude that the 4 per cent rule is not a very good one.

Another economist, David Love, looks for more sophisticated rules of thumb – for example, “moving to a (piecewise) linear rule based on the share of financial wealth relative to the sum of financial wealth and a simple approximation of the present value of future income”. Such a rule produces outcomes that closely track the Spock-like optimal strategy – which would be great if I could figure out what the rule actually was.

Love’s research suggests that it is possible to approximate the Spock strategy with decision rules that are simple enough for a website or a trained financial adviser to compute – even if it would be stretching it to call them rules of thumb.

But is the Spock strategy really the one we want to emulate? The psychologist Gerd Gigerenzer has examined a number of genuinely simple heuristics. One example: if you’re buying shares, buy the names you recognise. Another: if you’re distributing a fund between various asset classes, just distribute it equally between them all. Such clumsy-seeming rules suggest themselves naturally to the unschooled investor. Astonishingly, they work splendidly.

Here’s what seems to be happening: economists with powerful computers get hold of a stream of data on asset returns; they compute a complex optimal strategy given this historical data; then the world changes and the old “optimal” strategy turns out not to be quite so optimal.

Meanwhile the naive heuristics work rather well. Perhaps there is some hope after all.

Also published at ft.com.

Testing times for Sochi drug cheats

‘The most famous game of all is the prisoner’s dilemma, and it’s a natural explanation for why athletes take drugs despite the risks’

Testing for performance-enhancing drugs was first introduced by the International Olympic Committee not at a summer games but at the 1968 Winter Olympics. That might seem odd – the winter games aren’t normally associated with performance-enhancing drugs. Yet a dispassionate economic analysis suggests that this disconnection between testing regime and drug prevalence can hardly be a surprise.

The ideal economic tool for thinking about performance-enhancing drugs is game theory, a mathematical approach to understanding co-operation and competition. The most famous game of all is the prisoner’s dilemma, and it’s a natural explanation for why athletes take drugs despite the risks.

The essence of the prisoner’s dilemma in doping is that, regardless of what other competitors do, each individual competitor stands to benefit from taking the drugs – either by ensuring an advantage over a clean field or avoiding disadvantage compared with cheating rivals. And so everyone takes performance-enhancing drugs, even though everyone would prefer a situation where the entire field was honest.

Naturally, sporting authorities wish to change the incentives and so they test athletes and hand out bans to those who break the rules. One might think that such tests largely resolve the prisoner’s dilemma. But an analysis published last year by three economists, Berno Buechel, Eike Emrich and Stefanie Pohlkamp, suggests otherwise.

Game theory is a clever tool but the risk is that the theorist misses some bigger game. Buechel and colleagues argue that the bigger game in this case isn’t just between the athletes and the sporting authorities, but involves sports fans. The sad truth is that sports fans aren’t necessarily providing the right incentives.

Fans, understandably, do not like drug scandals: news that another sprinter has been caught taking steroids or another cyclist has been found using the blood-enhancer EPO does nothing to enhance the reputation of these sports. That typically means lower box-office takings, smaller broadcast revenues and less money from sponsors.

But what constitutes a drug “scandal”? It’s not the doping itself: that’s the offence, not the scandal. The scandal doesn’t break until the offence becomes public knowledge – that is, when the drug-taking coincides with an effective drug test. What, then, is the incentive for sports administrators to beef up their drug tests? The immediate impact will be more scandals, fewer spectators and fewer sponsors.

Of course, a totally foolproof drug-testing regime would solve this problem because nobody would cheat if detection and punishment were guaranteed. But even if such a regime were possible on some plausible budget, the costs of getting there might be prohibitive. Only after a huge scandal such as the discovery of industrialised cheating in the Tour de France do the authorities have an opportunity to improve their testing programmes.

This is a problem that may never be fully resolved. But the drug-testing regime at the Sochi Olympics has had two clever features. First, samples are being kept for 10 years. The usual reason given for this is that it may allow new and improved detection methods to catch cheats after the event, but there is nice side benefit: sporting authorities have more incentive to catch historical cheats than current cheats, because doing so presents an attractive image of a sport that is making progress.

Second, the IOC took great pains to describe the increased numbers of drug tests that would be conducted during the games – up by more than 14 per cent from the tests during the Vancouver Winter Olympics. We’re invited not to think of the failed drugs tests but all the tests that have passed. The real target for such announcements is not the athlete. It’s the sports fan.

Also published at ft.com.

Low inflation can be a disease not a cure

Deflation seems unlikely – but even a low risk is worth losing sleep over, writes Tim Harford

‘UK consumer inflation has fallen below the Bank of England’s 2 per cent target for the first time in four years.’ Financial Times, February 19

Hurrah!



Well, perhaps.

Why are you so grumpy again? What wrong with low inflation?



Nothing, so far – but you can have too much of a good thing. Inflation is also low and edging downwards in the US, Japan and the eurozone. Correction: Inflation is also low in Japan, and low and edging downwards in the US and the eurozone. Yet in all these places, interest rates are low and central bankers have printed enough money to get the tinfoil hat brigade screaming about hyperinflation. Such low inflation might be an indication of trouble ahead.

Are you saying that low inflation is a bad thing, or are you saying that low inflation is merely a harbinger of doom?



A bit of both – but mostly I am concerned that low inflation is a bad thing in itself. One issue is that unexpectedly low inflation redistributes from borrowers to creditors.

About time too, most savers will be thinking.



I hear you. Still, borrowers are more likely to be cash-constrained (that’s why they are borrowers) and are more at risk of bankruptcy. That means lower-than-expected inflation may damage the economy as a whole rather than just moving money from one person’s pocket to another’s. And there’s another problem with deflation: the “lower-bound” problem.

What’s that?



It’s a fancy way of saying that nominal interest rates can’t fall below zero. If inflation is, say, 4 per cent then a central bank can give an economy a shot of adrenalin by cutting interest rates after inflation to minus 3 or minus 4 per cent. If inflation is 0.7 per cent – as it’s currently estimated to be in the eurozone – then that’s impossible.

Why would anyone want real interest rates of minus 4 per cent?



Usually we wouldn’t. But, against the backdrop of a slack economy, such rates would be a strong incentive to spend. And if outright deflation took hold, effective interest rates would rise: people would earn money simply by sitting on their cash and waiting for prices to fall. Sounds great but for an economy it’s a disaster. If nobody buys anything there will be a recession and more deflation – a vicious spiral.

But that isn’t going to happen. Is it?



Don’t ask me, I’m an economist. We never know what’s going to happen to the economy. But it is a serious enough problem that even a low risk is worth losing sleep over.

Is there a constructive response?



In the case of the US and the UK, there’s always “hope for the best”. Both economies have been growing in a more-or-less encouraging fashion. One could try cutting taxes and raising government spending but it may be too late.

What about this forward guidance business?



Yes, an awkward affair. In principle forward guidance makes sense: the idea is to promise to keep interest rates very low until some condition is met, even if the economy might benefit from higher rates down the track.

Why make such a promise?



Because promising low rates for a long time is the next best thing to cutting rates below zero. Imagine buying a house: a mortgage rate of minus 3 per cent might be nice; but a promise that interest rates will be held artificially low for a while is almost as good, if you believe the promise.

That all makes sense.



Yes, but recent experiments with forward guidance haven’t been a huge success. Bank of England governor Mark Carney tied his forward guidance to unemployment rates, and unemployment rates promptly plummeted, so his attempt to commit to low interest rates ended up meaning very little. Still, it was worth a try.

This all seems flimsy. Are there any other approaches?



Most of these central banks are targeting inflation of 2 per cent, or something along those lines. But four years ago IMF chief economist Olivier Blanchard proposed a sharp break with that tradition. He floated the idea of a 4 per cent inflation target next time round. The thinking makes some sense: as long as interest rates and salaries tend to adjust, a higher target should cause little harm in good times and provide a great deal more room for manoeuvre in severe recessions.

That sounds radical.

Agreed. It’s never going to happen, which is a shame. But from the standpoint of today’s sluggish growth all Mr Blanchard’s advice would amount to is: “Next time let’s start from somewhere else.”

Also published at ft.com.

The seductive appeal of cultural stereotypes

Is it obvious that Jewish or Chinese Americans have superior willpower? asks Tim Harford

‘It may be taboo to say it, but some cultural groups starkly outperform others.’ Jacket blurb of ‘The Triple Package’ by Amy Chua and Jed Rubenfeld

Which cultural groups are they talking about? I vote for the baby boomers. They run the world and retire early on fat pensions; the rest of us are still listening to their idea of good music. They are outperforming the stuffing out of the rest of us.

I don’t think the boomers are a cultural group as far as this new book is concerned.

Why not?



I’m not sure. The authors take it as obvious that the logical way to divide up the US is on ethnic and religious lines: Nigerians, Iranians, Jews, Mormons and others. Hipsters don’t seem to count. Neither do punks, lesbians, hackers or hippies.

So by ‘cultural groups’ they mean ‘racial groups’?



Draw your own conclusions. We find it very tempting to carve the human race at those particular joints. Such generalisations have a seductive power: they have tempted otherwise humane and sophisticated individuals into the grossest racism.

You don’t think Chua and Rubenfeld are guilty of the grossest racism, do you?



No, I do not. And it must be legitimate to ask questions about ethnic and cultural differences – with care. The Triple Package is sold alongside titles claiming the secret to success is focus, or salesmanship, or psychoanalysis, or finding your own element, or being a psychopath. Or even that the secret is luck and we’re all being fooled by random patterns we interpret as causal. The book should be held to a higher standard of evidence, though: it says the secret to success is being raised by Jewish, Chinese, Indian or Nigerian parents. Chua and Rubenfeld don’t meet that higher standard. They offer heaps of anecdote with carefully selected academic evidence.

Everyone argues by anecdote. Is it really a particular problem here?



Yes. In a rare instance where they cite research from experimental psychology, they describe studies about the power of “stereotype threat”. For example, you can put women off their chess game by reminding them that the top players are all men. White kids struggle with a mathematics test if told it is research into the mathematical prowess of Asians. These stereotypes can cause real psychological, social and political damage. They should not be casually slung about to sell books. Insisting on care isn’t mere political correctness.

Are the authors careful?



They’re tactful but that’s not the same thing. One part of their “triple package” of cultural traits, for instance, is that insecurity breeds success. We are told this flouts the entire orthodoxy of contemporary psychology; then, in contrast, that it is supported by “a well substantiated and relatively uncontroversial body of empirical evidence”. If they cited any of that evidence I missed it in a flood of references to biographies, cultural histories, pop science and newspaper articles.

And the other claims?



They do cite scholarly research linking willpower to success in life but that doesn’t seal the deal. Is it so obvious Jewish or Chinese Americans have superior willpower? It’s far from clear that pushy parenting builds willpower, and I can cherry-pick my own research suggesting the opposite.

The third part of the triple package is?



A sense of cultural superiority. But is that particularly Indian or Iranian? Russians are proud to be Russian; Ethiopians to be Ethiopian. It didn’t seem to occur to the authors to look for disconfirmation by examining less successful religious or immigrant groups.



The whole story still sounds plausible to me.

We always find stereotyping plausible. That is why it is treacherous. But there are three big holes. The first is selection bias. If Indian doctors and engineers are given US visas, and they and their children are successful, is that because of Indian culture, or because the children of doctors and engineers also do well? The second is social networks. Chua and Rubenfeld may not emphasise psychological research but their worldview is ultimately about psychology rather than the contacts an immigrant network might bring. Most fundamentally, they fail to justify their basic premise: that the way to understand success is as a phenomenon that applies to religious or ethnic groups. They say that if we couldn’t generalise about Jews or Chinese Americans, “we wouldn’t be able to understand the world we live in”. Treating people as individuals is, apparently, just too much like hard work.

Also published at ft.com.

The murkier side of transparency

‘Publishing clear information is often a way to make the world a better place – but not always. Sometimes it pays to be selective’

On a recent visit to Toronto, I was able to admire the city’s rather fashionable pedestrian traffic signals, which clearly show you how long you have before the lights change. Crossing the street to a flashing “3 – 2 – 1 – 0” is like experiencing your own little rocket launch.

Toronto isn’t the only city with such signals: Manhattan has them, Washington DC has had them for many years, and London is beginning to introduce them. But it’s Toronto that provides the stage for a fascinating new study in unintended consequences.

From late 2006 through 2008, countdown signals were gradually installed across the city. It is important to understand that from the point of view of safety, the order in which they were introduced was arbitrary. (Some accident black-spots are temporary, the result of bad luck. Had the signals been installed at junctions with a history of accidents, they would have looked like brilliant safety measures simply because the run of bad luck ended.)

In effect, Toronto accidentally arranged a randomised trial of the new signals. By examining the accident rate at each intersection and how it changed with the new signals, economists Sacha Kapoor and Arvind Magesan got a detailed picture of the effect of the countdown.

You might well anticipate that the countdowns would make junctions less dangerous, by telling pedestrians whether or not they have time to cross in safety. Toronto’s traffic planners certainly seemed to believe that would be the case. They were wrong. The new signals caused more accidents.

How could this be? Kapoor and Magesan suggest three explanations. One is that the signals cause pedestrians to take more risks – but it seemed that fewer pedestrians were involved in accidents after each signal was installed.

The other two explanations rely on the fact that it’s not only the pedestrians who see the countdown: drivers can too. If a signal is about to turn red for pedestrians crossing at a junction, then drivers who are trying to get across the junction in the same direction are also about to get a red light. Since there was more speeding and more rear-end collisions after the countdown signals were installed, Kapoor and Magesan reckon the natural explanation is that some drivers were accelerating into the junction to avoid being delayed, just as other drivers were slowing down.

It’s not the first time that economists have discovered that what looks like sensible transparency can have unintended consequences. Ten years ago, David Dranove, Daniel Kessler, Mark McClellan and Mark Satterthwaite looked at the impact of mandatory “report cards” in New York and Pennsylvania, which published data on the performance of individual doctors, hospitals or both.

One might imagine that this information would, at the very least, be convenient. At best it should spur physicians to improve their skills because patients would seek out the very best. But the researchers looked at the impact on cardiac surgery, and found a tragic side effect: once doctors and hospitals knew that their success rates would be published, they had a strong incentive to operate on the healthiest patients. The best hospitals had their pick of the sick and selected easy cases. Meanwhile patients with more complicated conditions were more likely to have surgery postponed. The net result: more money was spent, yet more people died of heart attacks.

Publishing clear information is often a way to make the world a better place – but not always. Sometimes it pays to be selective. Doctors could benefit from report cards, provided their patients never find out what they said. And Toronto’s countdown signals would work perfectly if only they could be hidden from drivers.

Also published at ft.com.

The economics of love

If you’re looking for Valentine’s Day inspiration, you could try:

“Everything I ever needed to know about Economics I learned from online dating” – Paul Oyer (Buy in UK; Buy in US) – A guide to economic ideas pegged on the author’s experience trying to get a date online. It’s fun without being revelatory.

“Marriage: A History” – Stephanie Coontz. (Buy in UK; Buy in US) – A weighty but fascinating history.

“Dear Undercover Economist” – My favourite letters from the “Dear Economist” column (Buy in UK; Buy in US). Obviously this book is absolutely marvelous in every way.

 

 

Why opposites shouldn’t attract

‘While it may be natural and familiar, assortative mating also breeds inequality’

Those of you out courting next Friday, do enjoy yourselves – but with a twinge of guilt. Inequality has been rising for a generation in many places, especially the Anglophone countries. Let’s be honest: you and your romantic pursuits are part of the problem.

The issue here is something economists call “positive assortative mating”, a charming phrase that we blame on the evolutionary biologists. It describes the process of similar people pairing off with each other: beautiful people dating beautiful people, smokers dating smokers, nerds dating nerds. All perfectly natural, you might think.

While it may be natural and familiar, assortative mating also breeds inequality. Economists often look at sorting by education level, which is common and easy to measure. If the MBAs and PhDs were sprinkled randomly throughout the population that would spread the wealth around. But, of course, they tend to pair up with other MBAs and PhDs; meanwhile the high-school dropouts tend to end up with other high-school dropouts. Already prosperous people are made more prosperous yet by their marriages.

This is an interesting idea in theory but does it have any practical significance? A recent paper by Jeremy Greenwood and others looks at a large data set from the US Census Bureau through the lens of the Gini coefficient, which is a measure of inequality. It’s 63 in highly unequal South Africa, 40 in the UK and 23 in egalitarian Sweden. It’s 43 in the US Census data set; but if the couples in the data set were randomly paired off, the Gini coefficient would be a mere 34. Assortative mating increases inequality.

But does this pairing-off process matter more than it used to? Does it explain any part of the rise in inequality we’ve seen since the 1970s? The answer, again, is yes – but a guarded yes. Marriage patterns have little or nothing to do with the concentration of earning power in the hands of the richest 1 per cent and 0.1 per cent: women are major breadwinners in the top quarter of the distribution but less so right at the very top – not yet, at any rate.

But assortative mating is having an impact on inequality more broadly. It’s not so much that well-educated people are more likely to pair off – although they are – but that educated women are more likely to earn serious money than a generation ago.

Consider my own mother: she was well on the way to a PhD in biochemistry when I arrived on the scene in the early 1970s. She then dropped out of education and spent most of her time looking after her children. Her academic qualifications had no impact on our household income. Assortative mating has always been with us but it’s only in a world of two-income households that it increases income inequality.

The sociologist Christine Schwartz showed in 2010 that the incomes of husbands and wives in the US are far more closely correlated than they were in the 1960s, and that this explained about one-third of the increase in income inequality between married couples. John Ermisch and colleagues have shown other consequences: in both the UK and Germany, assortative mating substantially explains low social mobility because the children of prosperous parents marry each other.

We should not place too much emphasis on all this. Assortative mating explains only part of the rise of inequality, and perhaps very little at the top of the income scale. The usual remedies for inequality – unionisation, redistributive taxes, minimum wages – still have the same advantages and limitations as ever, even if they need to reflect the reality of the two-income household. It’s a reminder that the most welcome social trends can have unwelcome side-effects.

Happy Valentine’s day.

Also published at ft.com.