Skip to content
Home » The Lost Genius of Irrationality: Rory Sutherland (Transcript)

The Lost Genius of Irrationality: Rory Sutherland (Transcript)

Here is the full transcript of Rory Sutherland’s talk titled “The Lost Genius of Irrationality” at TEDxOxford conference.

Rory Sutherland’s talk, “The Lost Genius of Irrationality,” explores the power and effectiveness of non-rational solutions in influencing human behavior and societal norms. He argues that arbitrary laws, like the Sabbath or the French working hours directive, can have profound impacts on behavior due to their simplicity and universality, regardless of their rational basis.

Sutherland highlights the importance of understanding human psychology and social behaviors, such as the flock behavior observed in traffic patterns, to design better policies and systems. He introduces the concept of “heuristics,” simple, socially contagious rules or practices that can significantly improve societal behavior without the need for compulsion. One example he gives is the “Minnesota Zipper Merge,” which efficiently manages traffic flow through social norms rather than enforced rules.

Sutherland also discusses the power of naming behaviors, like the “designated driver,” to make them more socially acceptable and widespread. His talk advocates for governments and institutions to leverage these insights into irrationality and social influence to enact positive changes in a more nuanced and effective manner.

Listen to the audio version here:


Introduction to the Debate on Religion and Society

There’s an Oxford resident called, name of Dawkins, who’s written an influential and worthwhile, I think, book called “The God Delusion,” and I have no particular beef with it except I think with the Southern English edition. This is because, whereas there are parts of the world patently where the meme of religion and the divisions that it generally inculcates in society can lead to extraordinary dangerous and divisive effects, I don’t really see the problem in Britain to the same extent.

I mean, when I traveled up from Kent this morning, I didn’t really pause for a second or more to consider the possibility that my train might be hijacked by suicidal Methodists. When I walked from the station to the theatre here, I didn’t actually, you know, circumspectly glance around me all the time in terror that I might be pistol whipped by Quakers.

But it’s highly unlikely, even if Britain and parts of continental Europe are actually less adversely affected by some of the worst excesses of religious belief, that there are other delusions which we ought to be investigating far more. And my great criticism of Dawkins’ English edition of the book is that it really should have been retitled “The Football Delusion.”

Now, I’m probably, I’m from that part of the country where there is no football, I’m from the Welsh borders, and we just don’t have the soccer gene to the same extent.

The Fascination with Football

But it does strike me as interesting, not so much football, because I have no beef with people watching football, enjoying football, attending football matches at all. It doesn’t seem to be a problem. If you choose to derive a large amount of pleasure from watching a series of completely random events play out, effectively pinball on grass, and then choose to construct around those events a narrative entirely of your own devising, well, it’s not for me to say you’re wrong. It’s not for me really to find out that the statistics of football make it actually, in terms of individual matches, not league play, the impurest of all sports. Actually, the second.

Baseball is actually even worse. But if you’ll notice, they never play individual knockout matches in baseball; they will always play, the World Series is decided on the best of seven. No, but I have no problem with this at all. It’s obviously patently an enjoyable sport from which people derive a huge amount of pleasure. What I do have a problem with is football commentary.

Critique of Football Commentary and Human Agency

Because it seems to be a case where every single thing that happens in soccer is attributed to intentional human agency, and not to luck. So if a team loses a match, there are immediately calls for the resignation of the manager, despite the fact that statistically the single result is more or less irrelevant. Notice too that it’s not a game like tennis, where the play takes place in lots of individual compartmentalised sections. And where actually you can say that each point played in tennis is more or less independent of the preceding one.

Although there’s an interesting separate debate about tennis, which a friend of mine, Richard Thomason, observed, which is to say that the interesting thing about tennis is that the entire drama and exciting narrative of tennis is created by the scoring system. If you played tennis, but you scored it like basketball, it would be unwatchable.

If you imagine, you know, effectively Federer leads Murray by 247 points to 163, you’re not going to stick around. The fact that you have compartmentalised sets, you have retaining serve and losing serve, gives it a kind of interesting narrative arc, which makes it watchable as a sport. A very intelligent critic of basketball made the point, he said, “I don’t understand why it takes so long. Why don’t they just start at 100 points each and then play the last four minutes?”

But then the odd thing about soccer is that everything, even though most of the events are random, the scoreline is so small as to be more or less statistically insignificant. Probably only a small percentage of play actually contributes to a goal.

The Narrative Construction in Sports

And where it’s actually philosophically true to say that the entire scoreline would be fundamentally different if the coin toss at the beginning, before play even started, went the other way. And yet people will construct extraordinary amounts of synthetic happiness or indeed synthetic misery around the result. They will construct an entire story of how the match went, which is designed to create a kind of narrative.

And then the commentators effectively make every single result, however statistically minor, I don’t include league results in this, aggregated results seem to have meaning, as though it’s necessary to construct about 500 or 1000 words of description. Everything is the product of human agency, every goal was either brilliantly struck in the case that it actually created a goal, or of course a complete cock-up in the case that it missed.

The only non-human agency is when you get the most peculiar sentence of the lot, which is, “and the crossbar foiled Drogba’s efforts.” So when you do actually attribute agency to something that’s non-human, it’s to a piece of wood. This is strange. And it did lead me to think, watching football commentary, which does seem increasingly nonsensical in many ways, simply because it attributes a kind of simplicity and a narrative to events which are actually highly complex and have a huge degree of randomness within them.

And it did cause me to ask the question, which I suppose is contra Dawkins, which is that, would an ancient Greek football commentator actually be more honest than a contemporary one? The great point I think of Chesterton is that when we cease to believe in God, we don’t believe in nothing, we believe in anything. And the extent to which we tend to attribute to events which are the process of complex phenomena, many of which are completely unpredictable and certainly can’t be designed in advance, the fact that we do that seems to be an inaccuracy.

The Role of Fate in Sports Commentary

An ancient Greek commentator would not attempt to make human agency the single driver of the scoreline in a soccer match. He would use useful phrases which acknowledge the unknowable, the uncertain, and the role of fate. He might say, “Everything was going well for United and then Phoebus Apollo decided to thwart their midfield efforts.” Now, there is an argument that that is actually a more accurate picture of the world, the one in which you see human intentionality and agency in every single outcome.

The Monopoly of Economics in Decision-Making

Now why this seems to be important is for a very interesting reason, which is I went to see a talk by Richard Thaler, the author of “Nudge,” about six months ago, and he said something which I thought was perhaps among the most important sentences I’ve heard in two or three years. And he said that the interesting thing about economists is that they are generally opposed to monopolies, they view monopolies as a very bad thing altogether.

And yet the strange thing about economists is they’re entirely comfortable about the monopoly they themselves enjoy. And economics has gained, and Thaler made this point brilliantly as a behavioural economist, that economics has almost a monopoly, an unconscious monopoly in many cases, over business decision-making and over government policy-making. To achieve this through slightly surreptitious means, which is economics posits a ludicrously oversimplified individualistic model of human decision-making, treating everybody as a rational actor, maximising their own utility, without reference to powerful human emotions such as mistrust or regret.

It assumes that everybody making a decision has access to perfect information, which of course freezes marketing out of the picture altogether, and effectively makes it irrelevant. But it also assumes that people decide individually, and that they’re unaffected either by habit, or habituation, or contagion. In other words, the very powerful human defaults of do what I did last time, and do what everybody else does.

But because it makes these dangerous assumptions, it’s able to construct elegant mathematical models around what should happen. It’s not scientific in the sense of being an empirical science, it’s merely scientific in that you can construct mathematical models and equations which make you appear scientific. And Hayek called this scientism. That you actually, you steal, if you like, the clothes and the language of Newtonian physics, and you apply it to something where such certainty, such sort of linear associations have actually no place.

The Need for a Broader Understanding of Human Behavior

Human behaviour is far, far closer to meteorology than it is to bridge building, or train building, or track laying, in the kind of science you need. And yet strangely, economists have rather enjoyed this completely unwarranted control they have. So nearly all businesses, for example, and all governments will have a chief economist, or a chief economic officer. But no business I know of has a chief anthropologist, no business I know of has a chief game theorist.

I don’t know many businesses, Southwest Airlines is one exception, which actually have a behavioural economics team. But there are multiple sciences which are needed to understand the workings of mass groups of people. And economics on its own, with its assumption of individual action unaffected by the actions of others, this is to make the assumption that actually you don’t really have fashion in clothing. Everybody gets up every morning and decides personally, without reference to anybody else, what is fashionable. And it’s extraordinarily dangerous.

A very interesting man called, well, you’ll know, Max Planck, who was effectively the founder of quantum physics, had a friend who was an economist who said, “Max, why don’t you become an economist? There’s good money, you get to go to conferences, advise governments, that kind of thing. You know, probably a bit more fun than doing the maths gig.”

And Planck, who was probably the most brilliant mathematician of the century, implied, “I couldn’t do economics,” he said, “the maths is too difficult.” Because what he understood is that once you acknowledge the fact that people are actually, there are feedback mechanisms at work, that people are influenced by the behaviour of other people around them, those simple mathematical models that make economics convincing and give it its influence, actually start to fall down.

Understanding Social Constructs and Individual Agency

So, looking for delusions or strange behaviors, other than soccer, which I think is a huge product of narrative bias and tribalism, elsewhere in society, other than religion, seems to me a fruitful area for activity. And the big problem seems to be, in many, many cases, we assume individual agency. We assume that everything we do is a product of individual choice and that we do it because we like it. The extent to which we outsource our decision-making processes to other people and actually default to norms or default to past habit, tend to be downplayed because this economic model still predominates in the understanding of how a business, how capitalism is efficient, why competition is good, and so on and so forth.

The Peculiar Popularity of Wine

So, here are a couple more. Wine. I have always thought that wine, like football, is slightly strange in that it doesn’t deserve its runaway popularity, especially as a drink without food, in which conditions French people and Italians very rarely drink it. It’s an atrocious drink in many respects. It’s atrocious in that it breaks the rules of all successful consumer products, where, first of all, some sort of consistency of quality is usually considered a good thing. You would find it very difficult operating a pub where one pint in every three of beer tasted like piss.

And yet, pubs can sell wine under those conditions where when you buy a glass of wine in a pub, the odds that it’s actually not disgusting are I’d place it slightly better than 50-50, and yet people happily pay four pounds fifty for it. Most people, I think it’s fair to say, most of the time would prefer a gin and tonic to a white wine, but they ask for a white wine.

Social Biases and Choice Architecture in Beverage Selection

What’s going on here? I think there are lots of complicated things all working in effectively together. One of which is there are effective social biases. It is very appealing for places like this, I’m sure there’ll be a drinks party afterwards in which we’ll be uttered to you what King’s name is described as the three worst words in the English language, which are “red or white.”

One of the reasons is that it’s extraordinarily profitable to sell wine, or indeed to give it away, because there’s no known price anchor. You can’t charge 25 quid for a Gordon’s and tonic because we know what Gordon’s costs and we know what Schweppes costs. When you have an obscure wine of no known provenance, you can actually achieve egregious markups simply because people go well it sounds kind of posh and this is an expensive restaurant, so I suppose it’s vaguely plausible that I’m getting my money’s worth. It’s extraordinarily inconsistent.

Michael McIntyre, the comedian, makes the point that it’s the only thing served in a restaurant where they ask you to smell it before you drink it to check it’s not wrong. Now, I don’t want to be wrong, but isn’t that their job? You don’t get a waiter bringing you the milk to go with your coffee at the end of the meal and going, “I think this might be slightly off, I was wondering if you could just check.” There may be a degree, I’ve got friends who know a great deal about wine who work in the restaurant business who are actually convinced that most people pretend to like it, that even if you serve a bottle of wine that’s fundamentally corked, people will nearly always drink it.

The only clue is that they usually don’t order a second bottle, but they very rarely send it back because no one is actually confident enough to say there is something wrong with this wine. So, simple fear of social embarrassment and the pretense that you enjoy something plays a part. Choice architecture, we’ve seen a bit about choice earlier. Wine, by an extraordinary accident, comes in two colors, which means that if you just buy six bottles of white wine and six bottles of red wine, you’re deemed to have offered your guests a choice.

You can’t offer your guests one alcoholic drink because that’s kind of rude, you’ve got to actually offer them the courtesy, I think the previous talk was almost talking about this, the illusion that they are in control. The only problem is if you wanted to serve spirits, that would require you to buy five different bottles of spirits, four or five different mixers, and three different kinds of fruit. It’s troublesome. Bizarrely, if you just say red or white, tea or coffee, still or sparkling, those two things basically tick the box marked have offered a choice.

The Influence of Supermarket Layout and Consumer Choices

What you’ve actually done is you’ve given them two drinks they don’t like very much, but they get to choose the color. Nonetheless, for the purposes of social events, that is deemed to be sufficient. There are tons of other biases as well. Because there is an insane number of types of wine, when you go to the supermarket, it dominates the shelves. And we have a very big heuristic in our shopping in a supermarket, which whatever there’s a lot of in the shelves is probably pretty good.

So, wine will occupy something like 10 feet of shelf space to every one foot occupied by spirits. It’s also, if you notice, very easy to buy because you get to the end of the supermarket and you go, “oh look, there’s wine, I’ll have red and white and I’ll remember to put the white in the fridge,” and you usually don’t. If you want to buy spirits, suddenly you realise at that point the past dependency is working against you. You have to double back, go and find some mixers, and then, in the case of pins, of course, you have to return all the way to the beginning of the store to buy seven different types of fruit against the direction of traffic in the store, so you default to wine.

An experiment, interestingly, in past dependency was done by a group of retailers and spirits manufacturers who created a cocktail pod right at the end of the supermarket. And it had a freezer full of ice, it had a whole rack of mixers, and it had two baskets of lemons and limes, and it just said cocktail pod. But it was at the point in the supermarket where it was effectively alongside your what-shall-I-drink-tonight, what-shall-I-offer-our-guests point, rather than requiring you to double back.

When they introduced that, sales of spirits in those stores overall went up by about 8%. It was an extraordinary effect, simply the product of past dependency. Now, when this concatenation of various stuff all comes together, there’s also the thing that we actually like alcohol, and wine is the acceptable middle-class dressing in which to consume it. Stand around with a glass of wine at lunchtime and you’re seen as a bit of a bon viveur but basically a respectable bourgeois chap.

Tequila shots in the office at lunchtime, on the other hand, do not convey anything of the same thing. So, at some level, we have no choice because, of course, we buy wine not for the effect it has on us but the impression it has on other people. And that is actually a social construct over which we have very little control. But the important thing about this is the overstatement of individual human agency and the power of self-control when faced with external network forces.

It’s very, very widely known in the software industry that to some extent market share is a huge product benefit. That there is a great advantage once you create the iOS ecosystem and you have more apps than everybody else, that’s one ecosystem, and you also have more accessories, cases, chargers, and strange Bluetooth sort of accessories than any other operating system, then you enjoy a disproportionate advantage over everybody else. It’s known widely in network externalities and known widely in the software industry. I would just contend that they’re actually much more prevalent than we think but the fact that they work on us unconsciously causes us often to discount them.

The Challenge of Learning Foreign Languages

It also causes us, I think, to expect too much of individual effort in terms of behavioral change and I’ll come to that in a second. I’ll give you one third example of network externalities. It is widely believed and universally stated by Brits that the English and English language speaking countries in general make very bad linguists. We just don’t make the effort to speak and learn foreign languages. And that’s generally attributed to our own failings or our own overconfidence, smugness, lack of interest.

If you look at that from a point of view not of individual agency but from the point of view of network effects, you know, read, by the way, a very, very good book by Paul Ormerod called “Positive Linking” which has just come out. What you suddenly realize is it’s a bit more complicated than that.

First of all, if you imagine a Dutchman considering whether he should learn English versus an Englishman deciding whether to learn Dutch, the gains to the Dutchman in learning English are forcibly a hundred times greater than vice versa. In terms of English will be useful to him wherever he goes, even speaking to people for whom English is also not their first language. It will be of use to him in France. It will even be of more use to him in former Dutch colonies like South Africa, for example, or Indonesia. He also has a head start. He’s consumed something like one or two, probably 10,000 hours of English language TV, wisely with subtitles not with the idiot dubbing habit, over the course of his life.

So, English idiom and tonality and so forth, everything in fact other than the pronunciation of the letter S will come fairly naturally to him. A friend of mine had to fly over from Stansted to Amsterdam regularly on a plane called the City Hopper, which unfortunately had Dutch pilots, which led to very amusing announcements I have done. But now let’s look a little further.

To the English person wondering whether to learn a foreign language, which one? The choice architecture is a complete mess. You can now just about make a case for Mandarin, Chinese, maybe Spanish, Portuguese conceivably, but it isn’t a clear cut, absolutely obvious thing. It’s extremely unlikely that there will be any social effects.

Network Effects and Language Learning

If the Dutchman says to a few of his friends, “I’m thinking of improving my English,” he will find four or five other friends who say, “I agree with you, let’s go along and do something about it.” If I suddenly conceive the urge as an English speaker to speak Turkish, I’m pretty sure my friends will basically give me the reaction, “You’re on your fucking own, mate.”

Now let’s take it a bit further still and look at further network events, which is let’s say I put a huge amount of effort into learning Dutch. My Dutch only becomes useful to me in real terms, not in gradations, bit by bit, word by word, verb by verb, adjective by adjective.

It only really becomes of any use whatsoever when my Dutch is better than the average Dutchman’s English. And that’s really, really difficult, unless you live there. Even if you do, they will try and reply to you in English in any case. So, the point that whereas a Dutchman actually gains from learning English incrementally, an Englishman only gains from learning Dutch once he reaches some vital tipping point of extraordinary fluency, makes the whole thing completely unequal.

Understanding Across Language Barriers

So it’s not entirely our fault. You can add an extra dimension of confusion, which is in one respect, and someone wrote to me about this, the English are actually very good linguists, which is we’re very good at understanding English when spoken by non-native English speakers. So we have actually honed our linguistic skills quite readily to understand English when it’s spoken or accentuated or intonated in a very unusual way. If you actually get the gender wrong in French, they don’t understand what on earth you’re talking about.

If you ask for “la limonade,” they haven’t got a clue. And then about six minutes later, they go, “Ah, la limonade,” now it’s suddenly, you know, clarity has now descended upon me. Now, it’s possible, it’s possible this is just frankly willful perversity on the part of French speakers. But since there have been no recorded cases of willful perversity in France in the last 200 years, I think this explanation can be discounted.

The Challenges of Uncommon Language Learning

So I think it’s more likely to assume that in fact, we get a lot of practice that we don’t even realize of understanding people speaking English in a weird way. I mean, it’s very interesting in the United States where people aren’t exposed to unusual English accents, asking for a cup of water is almost impossible, because I can’t pronounce “water” in a way that an American can understand it.

So actually, some of these utterly bizarre differences, the fact that even worse was a friend of mine who learned Hungarian, by the way, which is, the problem is that no Hungarian has ever come across a foreigner who’s learned Hungarian before. So he assumed that they would go, “Hi, I see that you’re an Englishman who has learned Hungarian, I’m so gratified and flattered that you have bothered to learn my virtually intractable language.”

No, they assumed he was a Hungarian, but just very, very stupid. So what I’m saying here is that we make a mistake by not looking at the world nearly enough from a network perspective. In many cases, I think that technological problems are often marketing problems and perception problems in disguise. The electronic cigarette, one fairly good scientist believes, if adopted by most American smokers, would reduce the smoking-related deaths by two orders of magnitude.

Technological Solutions and Social Acceptance

The problem actually is one of networks, that it’s a weird thing to use, people don’t feel comfortable using it. So all those rock stars and aging cool rock stars who’ve been briefly implicated in the Jimmy Savile scandal, one way you can redeem yourselves is by smoking electronic cigarettes heavily in public, lending them an atmosphere of cool and general acceptability.

But here you have a technology which individually should be an easy choice, or at least an easy thing to experiment with. The problem is actually the extent to which most of our behavior is actually mediated by the behavior of, and affected by the behavior of other people.

Electric cars are no longer a technological problem, it’s how you get people to accept them. But in one thing I think this is really important, and I’ve talked to some very interesting people who know much more about this than I do, so credit to them, which is that when we design behavioral change programs, or when we tell people what the ideal behavior is, we design it in a way which is perfectly rational from an individual standpoint, but from a collective or social standpoint is virtually impossible to adhere to.

Now let me give you an example. If you want people to cut down on their drinking, because people are worried about excessive drinking, the general assumption is you look at lots of epidemiological studies and you come to a conclusion that people who drink up to, let’s say, 28 units, 21 units of alcohol a week, don’t seem to come to any long-term harm, so you make that the recommendation.

The Complexity of Behavioral Change

You make it a numerical recommendation. Similarly, calorie-restricted diets tell you, you should eat this much every day. Now there’s a couple of problems there, one of which is that it is much harder to stick to a quantitative restriction than a binary one. If you notice, most of us, when we drive, break the speed limit every time we drive.

Not by much, but by a bit. It simply doesn’t feel that bad or wrong going 34 in a 30 limit. Very, very few of us, even at 3 in the morning when there isn’t a person in sight, and actually the person who pressed the button on the pedestrian crossing has long since walked away, very few of us will run a red light. It simply feels wrong in a way that isn’t true of what you might call eating a few calories too many.

Secondly, in terms of behavior, if you think about that limit of 28, let’s say, units of alcohol a week, there’s a fundamental problem there, which is that it often requires you to stop drinking when you’re already drunk. “Oh, I’m fucking having a great evening, oh bugger, I’ve just reached my 28th, right, I must stop right now.” Now, let’s just be clear, it’s not going to happen, is it, right?

But thirdly, in order to make a restriction on how much you drink socially contagious, to actually make it socially self-reinforcing, all your friends, even though you may live in different parts of Britain and only meet up once or twice, all of you have to drink at exactly the same rate.

Because otherwise, you go, “I don’t know, how likely is this to happen? Well, I’m actually, I’ve got 15 units to go actually, I’m planning on a really large one this evening, but I see you’ve just reached your 28th, so I’ll stop and keep you company, okay?” It’s not going to happen again. So this is where what appears to be rational behavioral design comes into conflict both with our own individual ability to stick to things, and we’re better at sticking to things.

If you notice, all religious law is binary. There’s no such thing as 50% kosher food. It either is or it isn’t. And that’s partly because “you can do this, you can’t do that” relies on habituation. I’ve been on a sort of no-carb diet. Whether it’s good for you as a diet or not is completely debatable. It’s worked extraordinarily well, and I’ll explain why in a minute. What’s undoubtedly true is that it works partly because it becomes self-reinforcing.

After a while, you look at Danish pastries with mild disgust, simply because you haven’t eaten them in the last two months and they look kind of weird and different. But there’s another magic to designing things to be binary, which is they’re socially self-reinforcing. And the trick really of my diet, which doesn’t look very successful from where you’re standing but trust me, you should have seen me a few months ago, okay? The real success is I got my wife to go on the thing at the same time.

And the basic point is you don’t eat carbohydrates, you don’t eat grains, you don’t eat sugar, in that sense. So it’s effectively a binary diet. You can eat any quantity you like of X, but you can’t eat Y. Now one of the things is it’s cognitively much, much easier because you’ve only got to exercise self-control once a week, which is when you shop.

Because if you don’t buy any carbs, there aren’t any in the house, so you don’t eat them. A calorie restriction diet, where you count the calories every day, requires you to exercise self-control every single time you eat a meal. Now that’s almost impossible. I mean, that’s actually like giving up sex by going to lap dancing clubs instead.

You know, it’s effectively saying in order to actually have less of something, we should give you just enough of something to get you really interested and then tell you to stop too soon. Okay? That is not an easy thing. That makes insane demands of our self-control.

So the interesting point I’m making here, and this applies to alcohol, is maybe if you say don’t drink alcohol three days a week and you get all your friends not to drink on Monday, Wednesday, and Thursday or Monday, Tuesday, Wednesday or in January, that has two virtues. It’s absolutely binary. It doesn’t require you to exercise any self-control when you’re drunk, because you don’t get drunk. Secondly, your friends can engage in the same thing at the same time, so you’re never under any social pressure.

And then you realize that actually behind most religious law, there may be very little apparent rationality of the Dawkins kind, but there’s an extraordinary amount of psychological insight. Because it understands that for laws to stick, they need social reinforcement, they need to work in terms of each person’s behavior being visible to each other, they need to make transgression actually apparent, and they need to be binary. Now this may make them less rational. The person who says 28 units a week has tons of science to back up his recommendation.

Three days a week don’t drink, there’s actually no science for that. Theoretically it’s bonkers, because you could be a raging alcoholic on four days a week and not drink on the other three. On the other hand, if you look at religion, it has actually survived for a few thousand years, which suggests that it has some sort of stickability. And religious law is what B.J. Fogg calls crunchy. It’s binary. It’s you can do this, you can’t do that. And the interesting thing, if you look at it, is the French got very worried that people were working too hard, which was causing unemployment, so they created a very French rationalistic thing, which is called, I think, the Working Time Directive, which limits your working time to 35 hours a week.

The Power of Arbitrary Laws

How do you spot someone who’s cheating? He says, “I got up late.” How do you actually spot transition? Three thousand years ago, someone had come up with a better answer. It was called the Sabbath. And everybody does it at the same time. Now it’s not as perfect mathematically as the French solution, but in psychological terms, in terms of crunchiness, and in terms of social self-reinforcement, it is much, much cleverer legislation.

So the understanding that what we may be suffering in part, I genuinely contend this, is the government’s ability to engage in arbitrary universal legislation in defiance of rational criticism. That sometimes, actually, arbitrary law like the Sabbath is vastly more effective at behavioral change than doing everything by some sort of mathematical question of degree.

But unfortunately, the requirement that everything appears to be rational beforehand, and the assumption that individual agency is everything, and we have unlimited sources of self-control, causes us actually, arguably, perhaps, now this is a really contentious point, which I don’t expect you to take seriously, and please don’t, but actually to some extent, the absence of religion and the absence of monarchy have made it more difficult to legislate in some quite intelligent ways. Simply because the requirement is that it’s clear, it’s universal, but it’s seemingly slightly arbitrary. And it’s very difficult for a modern democracy to do this.

Understanding Social Behavior Through Traffic Insights

Just one final thing, which I always felt instinctively, and only realized why this morning. The power of what you might call flock behavior is this. I’d always had a vague feeling that speed cameras, which we can debate about endlessly, and most motorists do, there was something slightly wrong with a speed camera on a dual carriageway. And I couldn’t work out why it was. A speed camera on a single, you know, two-lane road, where cars are going in opposite directions, seemed to me kind of to make sense.

You’re coming into a village, it says 35 limits, if you go 38, the camera flashes and you’re fined, okay. But when we drive on a dual carriageway, we actually drive socially, we almost become like flocks of birds. And one of the rules is, which is more or less in the UK, it applies differently in different countries, is the person in the right-hand lane should at all times go a bit faster than the person in the lane to the left of them.

And so what the speed camera does is it tries to encourage everybody to drive at the same speed in both lanes, which is deeply inimicable to us instinctively, because everything about us says the right lane must go a bit faster. And there are good reasons for this. If you don’t have a difference in speed between the two lanes, it’s impossible for people to change lanes when they need to.

So there’s a very, very good reason for this instinct. But the problem you get with the speed camera is your instinct to actually slow down to the specified limit, and your instinct to go faster than the person to the left or slower than the person to the right come into conflict. And here you have a rule which actually seems to work from the assumption of individual human agency. Once you understand the extent to which our actions are actually determined by the actions of others, it completely falls down. I don’t know how to use this insight to solve problems, but I suspect there are many, many interesting heuristics you can design.

Enhancing Social Behavior Through Simple Heuristics

And the design of better heuristics, even if they’re not imposed by compulsion, they’re simply advanced by governments as good practices, seems to be an extraordinary area for improving social behavior. Simply giving people mentally easy, socially contagious heuristics. And I’ll end with a very good one. It’s called the Minnesota Zipper Merge.

And I don’t know how many of you drive because you’re mostly students. You go around on bicycles and therefore become very environmentally conscious because you ride bicycles, even though you think it’s the other way around. But I’ll park that one. But anyway, the extent to which actually our attitudes are a product of our actions rather than the other way around deserves a bit more investigation.

But the Minnesota Zipper Merge was… I don’t know whether it actually came in when Jesse Ventura, the professional wrestler, was governor of that state, but it’s a beautiful thing. When you come to a point where there’s a contraflow on a motorway or two lanes go down to one, there are huge conflicts going on in people. Am I a late merger, in which case other people in the left-hand lane, the lane that doesn’t close, will accuse me of queue jumping? Do I try and merge too early because I’m terrified of being accused of queue jumping?

Because you have to remember in Britain that queue jumping is on a par with paedophilia on the list of things people don’t want to be associated with. All that happens is completely dissonant behavior from everybody. Those late mergers who are then resented by the people who’ve been in the right lane and then who try not to let them in, versus the people who merge too early, wasting available road and causing the tailback to stretch back too far, effectively are caused by a heuristic conflict.

What they did in Minnesota was beautiful. They just invented a word for it. The Zipper Merge. And suddenly when you’ve got a word for it, it can become a social norm. And the way it works is beautifully simple. You’re told, “Zipper merge ahead, stay in lane, go to the end of the lane. When you get to the merger point, you take it in turns, one lane at a time. End of story.” That’s what Daniel Kahneman called a system one mental process.

Now it would be more efficient, logically, if you had ten cars from one lane, ten cars from the other lane. At that point, people start to cheat and pretend they miscounted and go through as the 11th car. It no longer works. What this does, apparently, is it reduces the accident rate and it vastly speeds the flow of traffic through intersections. And all it is, is not through government in compulsive terms telling you how to behave, it’s simply government creating a word for something which then creates the behavior which is socially beneficial.

Now that’s not my best example, there’s an even better example I discovered a week ago. The entire phrase “designated driver” was invented in marketing. That when you have a name for it, you have a sudden way to explain why you’re sitting in the pub drinking a Coke. And the very existence of the phrase “designated driver” enabled people to say, before we go out this evening, “Who’s the DD?”

In Amsterdam and Holland, they did a separate thing, they did a marketing campaign where the designated driver is always known as Bob. And they asked a slightly surreal question, “Who’s going to be Bob this evening?” Now that is a case where you create effectively a heuristic which is every time we go out on the RAS, one person stays sober. Very useful heuristic, you can take it in turns, it’s easy to do.

Once you’ve named the behavior, the behavior becomes both easier to stick to and more contagious. The interesting question then is, to what extent should government cease to actually aimlessly attempt to actually interfere with us through compulsion and the standard neoclassical economic assumption of effectively incentives and disincentives and punishments.

To what extent can you improve society by the creation and naming of better heuristics? That’s an open brief which I leave to all of you and I’m very pleased to be going last. But I would love to know your suggestions. Thank you very much indeed.

Related Posts