Medieval guilds existed for this reason. Trust is always more difficult to gain in cities because of the anonymity they afford, and guilds help to offset this problem. If it is costly and time-consuming to join one, the only people who enter are those with a serious commitment to a craft. Guilds are also self-policing; the upfront cost of being admitted adds to the fear of being ejected.
Commitment devices of this kind exploit the fact that, once you have made a commitment – whether in time, money or effort – there is no way of reversing it. To put it another way, would I really trust a black cab driver with my daughters’ safety if they could obtain a black cab licence merely through attending three or four evening classes and shelling out for a second-hand TomTom?
Reciprocation, reputation and pre-commitment signalling are the three big mechanisms that underpin trust. You can repeatedly use a small local firm that needs your loyalty, you can use a larger company with a brand reputation, such as Addison Lee or Green Tomato Cars, or you can trust someone who has made a big investment in getting a badge and stands to lose everything if he is caught cheating. If you don’t believe this, go to Athens, where foreign taxi passengers are on average taken on a 10 per cent longer route than Athenian passengers. Try Seville, where I was menaced to pay an imaginary €20 ‘suplemento aeropuerto’. Or Rome, where a colleague of mine was mugged by his taxi driver.*
Uber is a taxi company with different mechanisms to promote trust. There is a digital record of every journey, a rating system and, increasingly stringent checks on driver background. I’m not arguing that the Knowledge is the only solution to this problem, but what I am saying is that it is only partially of navigational value; a large part of its value is as a signalling device. It also means that black cab drivers tend to be highly experienced, because there would be little point in undergoing a four-year initiation if you were only planning to drive taxis as a stop-gap. In that respect, the Knowledge is an upfront expense that is proof of long-term commitment.
3.2: A Few Notes on Game Theory
Many things which do not make sense in a logical context suddenly make perfect sense if you consider what they mean rather than what they are. For instance, an engagement ring serves no practical purpose as an object. However, the object – and its expense – make it highly redolent with meaning; an expensive ring is a costly bet by a man in his belief that he believes – and intends – his marriage to last.
Now, you might expect a book of this kind to have a chapter about the Ultimatum Game* and other experimental, game-theoretic investigations into the nature of trust and reciprocation. This book contains no such chapter. The reason for that is that the Ultimatum Game is stupid, and so is the Prisoner’s Dilemma: these games exist in a context-free, theoretical universe with no real-life parallels. They both posit the idea of the one-shot exchange, a transaction involving two strangers with no knowledge of the other’s identity. In the real world, such transactions never take largely place – we choose to buy things in shops, not from random strangers in the street.*
When we engage in transactions, we are generally aware of the other party’s identity and can see clues to their commitment. For example, if I enter a shop in a town I have never previously visited, there is a slim chance that the shopkeeper may take my money and refuse to hand over the goods – he might be an imposter. However, let’s now assume that the shop is called H. Jenkins & Sons and above the door is a notice that reads ‘Established 1958’. The shopkeeper has clearly invested in his premises and stock, and it’s unlikely he will have survived in business for several decades if his business model depended on ripping off the local population.* He is at a point where the loss of his reputation will be far more costly than whatever he might gain by refusing to honour our exchange, which allows me to derive trust from the context in which the exchange takes place. Upfront investment is proof of long-term commitment, which is a guarantor of honest behaviour. Reputation is a form of skin in the game: it takes far longer to acquire a reputation than to lose one.
If you wish to make an Ultimatum Game work so that everybody would cooperate, allow me to propose a simple mechanism: you simply demand that, before anyone is allowed to play, they are required to memorise 25,000 streets and 20,000 landmarks in London. This will take four years of their life. At that point, all you need is a simple mechanism to ensure that cheaters can be expelled from the game. Under such circumstances no one will wish to cheat, since it would place them at risk of being thrown out of the game before they have recouped the effort they invested in gaining entry in the first place.*
There remains one problem, which is the possibility that people may cheat immediately before they leave the game.* In theory, there is nothing to stop a London taxi driver from taking an appallingly roundabout route when carrying the last passenger of his working life. If he loses his badge then, he sacrifices no future revenue, since there is no future revenue to lose. Whether intentionally or not, London taxi drivers seem to have a solution to this. A few years ago I took a £15 ride in central London and at the end of the journey proffered a £20 note to the young driver. ‘Can’t take it, mate, you’re my first ever ride,’ he said. ‘What? Seriously? Why not?’ ‘It’s a tradition – you don’t take money for your first ever fare.’ I loved this at the time, but it now occurs to me that this tradition is an unbelievably elegant one – if, in the course of my life in London, I also end up being picked up once by a cab driver as his last ever fare and he rips me off a bit, me and the London cabbies will effectively be quits.
3.3: Continuity Probability Signalling: Another Name for Trust
I mentioned that before the prospect of repeat custom is something that keeps businesses honest, but there is another conclusion we can draw – that you can signal that you are an honest business by showing that your business model relies on repeat custom.
What do the following things have in common?
The fact that large, carnivorous fish desist from eating useful fish such as wrasse, which clean the parasites from their bodies
The posh rope-handled carrier bags you get when you spend an appreciable amount of money on clothes or cosmetics
The free extra scoop of fries they give you at Five Guys
The fortune spent on a wedding
Your modest minibar charges that are waived by a hotel
The marble and oak lavishly used in bank branches
That fancy training course your company sent you on
A lavish advertising campaign
A free glass of limoncello given to you by a restaurant after your meal
Investment in a brand
Looked at from the perspective of simple, short-term economic rationality, none of these behaviours makes sense. A bank could conduct its business perfectly well from a Portakabin. Those rope-handled bags are expensive but they aren’t even waterproof. Limoncello isn’t cheap, and a lot of people don’t like it. Was the training course really worth £5,000?
All these things only make sense if we assume that some signalling is going on – they are examples of a behaviour which is costly in the short term and which will only pay off, if at all, in the long term. They are thus – if nothing else – reliable signals that the person, animal or business engaging in that behaviour is acting on the basis of long-term self-interest rather than short-term expediency.
This distinction matters a great deal. Unlike short-term expediency, long-term self-interest, as the evolutionary biologist Robert Trivers has shown, often leads to behaviours that are indistinguishable from mutually beneficial cooperation. The reason the large fish does not eat its cleaner fish is not because of altruism but because over the long-term, the cleaner fish is more valuable to it alive than dead. The cleaner fish in turn could cheat by ignoring the ectoparasites and eating bits of the host fish’s gills instead, but its long-term future is better if the big fish becomes a repeat customer.* What keeps the relationship honest and mutually beneficial is nothing other than the prospect of repetition.<
br />
In game theory, this prospect of repetition is known as ‘continuation probability’, and the American political scientist Robert Axelrod has poetically referred to it as ‘The Shadow of the Future’. It is agreed by both game theorists and evolutionary biologists that the prospects for cooperation are far greater when there is a high expectation of repetition than in single-shot transactions. Clay Shirky has even described social capital as ‘the shadow of the future at a societal scale’. We acquire it as a means of signalling our commitment to long-term, mutually beneficial behaviours, yet some businesses barely consider this at all – procurement, by setting shorter and shorter contract periods, may be unwittingly working to reduce cooperation.
Yet there are, when you think about it, two contrasting approaches to business. There is the ‘tourist restaurant’ approach, where you try to make as much money from people in a single visit. And then there is the ‘local pub’ approach, where you may make less money from people on each visit, but where you will profit more over time by encouraging them to come back. The second type of business is much more likely to generate trust than the first.
How might we distinguish the second type of business from the first? Well, the scoop of extra fries you get at Five Guys is one such gesture – an immediate (if modest) expense with a deferred pay off, and a reliable signifier that the business is investing in a repeat relationship and not milking a single transaction. Likewise, when your company pays your salary each month, it says you are worth that money for now; when it sends you on an expensive training course, it signals that it is committed to you for at least a few years.*
If fish (and even some symbiotic plants) have evolved to spot this sort of distinction, it seems perfectly plausible that humans instinctively can do the same, and prefer to do business with brands with whom they have longer-term relationships. This theory, if true, also explains some counterintuitive findings in customer behaviour: it has long surprised observers that, if a customer has a problem and a brand resolves it in a satisfactory manner, the customer becomes a more loyal customer than if the fault had not occurred in the first place. Odd, until you realise that solving a problem for a customer at your own expense is a good way of signalling your commitment to a future relationship. The theory of ‘continuation probability’ would also predict that, when a business focuses narrowly on short-term profit maximisation, it will appear less trustworthy to its customers, something that seems all too plausible.
Remember that hanging over every human interaction is an unspoken question: ‘I know what I want from this transaction. But what’s your interest in this exchange? And can I trust you to fulfil your promises?’ We do not need to know that the other party is honest; we simply need to know that they will behave in the transaction as an honest person would. In a closer community, you may simply be able to establish a reputation for all-round honesty, and that’s that. No bank manager could risk cheating a single customer in the 1950s, because if one customer found out they’d been cheated, the bank manager’s reputation was destroyed through the entire town.
There are many forms of expenditure – of money and of effort – which make sense within the context of a relationship, but which make no sense in a single transaction. Small acts of discretionary generosity, such as waiving a charge when a passenger has bought slightly the wrong kind of train ticket or a complimentary chocolate at the end of a meal are regarded by customers as reassuring indicators of trustworthiness; we correspondingly see the absence of such signals as being a cause for concern.
One of the reasons why customer service is such a strong indicator of how we judge a company is because we are aware that it costs money and time to provide. A company which is willing to spend time after you have bought and paid for a product to make sure you are not disappointed with it is more likely to be trustworthy and decent than one which loses all interest in you as soon as the cheque has cleared. The same applies in interpersonal relations; being rude isn’t so different from being polite, but it requires less effort. Politeness demands that we perform hundreds of little rituals, from opening doors to standing up when someone enters the room, all of which are more effortful than the alternative. By such oblique means we convey that we care about their opinion – and about our reputation.
3.4: Why Signalling Has to Be Costly
Twenty years ago, a colleague and I were working on a small-butimportant advertising brief. Our task was to send a letter to a few thousand senior IT professionals, asking them to try out the Microsoft Windows NT 32-bit server software in advance of its wider release. We could have simply sent them all a letter by second-class post telling them what the product was and what we were offering them – that would have conveyed information, but no meaning. Instead we produced an elaborate box containing a variety of bits and pieces including a free mouse-mat* and a pen, inside gratuitously expensive packaging.
We did that to convey to them not only that the product existed, but that it was a highly significant launch, to which Microsoft had committed a great deal of money. We also needed to convey the fact that very few people were in the privileged position of being able to test the software for free. We could have said that in the letter, but it would have meant nothing. It’s what’s called ‘cheap talk’, something that anybody selling anything can tell you – it is merely a claim, not an item of proof. Indeed to send an ‘exclusive invitation’ by second-class post (even worse, by bulk mail) would have been self-contradictory – ‘We’re sending this exclusive invitation to loads of people.’ (This is why no truly exclusive club can ever advertise in the mass media.)
So the mail pack we produced was elaborate, and of a kind that would have been uneconomical to produce in bulk – and it also won an award. But I most recall it because we worked on it with a Midwestern account director called Steve Barton, who said something telling when he briefed the project. ‘Look’, he said, ‘I’d like you to produce a stand-out piece of creative work here. But if you can’t, what I’d like you to do is write them a really nice one-page letter – and we’ll send it out by FedEx.’ Steve was effectively describing what biologists call ‘costly signalling theory’, the fact that the meaning and significance attached to a something is in direct proportion to the expense with which it is communicated.
Imagine you are the recipient – could you throw away a FedEx envelope unopened? I think it is safe to say that you could not. What we wanted from the recipient was not simply the ingestion of information: it was attention, conviction and a sense of import, something that an economically rational 50p stamp could never obtain but that a £10 FedEx envelope could. In the end, the campaign was hugely successful. Almost everyone opened the package and read the contents – and over 10 per cent tried the product, which required a significant amount of effort. In 2018 a digital rationalist would suggest that the way to reach several hundred senior IT people would be via Facebook, or by email – two options which were mercifully not available to us in the mid 1990s. He would be rationally right but emotionally dead wrong.*
Bits deliver information, but costliness carries meaning. We do not invite people to our weddings by sending out an email. We put the information (all of which would fit on an email – or even a text message) on a gilt embossed card, which costs a fortune. Imagine you receive two wedding invitations on the same day, one of which comes in an expensive envelope with gilt edges and embossing, and the other (which contains exactly the same information) in an email. Be honest – you’re probably going to go to the first wedding, aren’t you?*
3.5: Efficiency, Logic and Meaning: Pick Any Two
‘Credo quia absurdum est’, said Saint Augustine, supposedly – ‘I believe it because it is ridiculous.’ He was talking about Christianity, but it is equally true of many other facets of life: we attach meaning to things precisely because they deviate from what seems sensible. It is hardly surprising that we have evolved to invest more significance in unusual, surprising or unexpected stimuli and signals than to routin
e, everyday ‘noise’. As a result, like any social species, we need to engage in ostensibly ‘nonsensical’ behaviour if we wish to reliably convey meaning to other members of our species.
The psychophysicist Mark Changizi has a simple evolutionary explanation for why water ‘doesn’t taste of anything’: he thinks that the human taste mechanism has been calibrated not to notice the taste of water, so it is optimally attuned to the taste of anything that might be polluting it. If water tasted like Dr Pepper, it would be easier for sensory overload to drown out the hint of ‘dead sheep’, which would alert us to the fact that a carcass was decaying in a pool five hundred yards upstream. Water ‘tastes of nothing’, so we notice the smallest thing which deviates from this. You can try a similar experiment with young children. Feed them their favourite food, but add a subtle herb or spice. They will find it revolting, because the slight deviation from what they expect alarms them into believing it is somehow unsafe.*
My contention is that our perception is calibrated more widely in this way. We notice and attach significance and meaning to those things that deviate from narrow, economic common sense, precisely because they deviate from it. The result of this is that the pursuit of narrow economic rationalism will produce a world rich in goods, but deficient in meaning. In architecture this has produced modernism, a style with a marked absence of decoration or ‘spurious’ detail, and a corresponding loss of ‘meaning’.* My secret hope is that, with the 3D-printing of buildings becoming possible, a certain Gaudí-ness may reappear in twenty-first-century buildings.*
Alchemy Page 14
