Was Kaushik Basu always shit at Game theory? Sadly, yes. This came out in '83.
On Why We Do Not Try to Walk off without Paying after a Tax-Ride Kaushik Basu
Thugs may do so as may some drunks or crazy people. They have no reputation to lose. Moreover, for those who regularly use taxis, the fare represents a trifling expense and so no great saving is made by dodging it. On the other hand, fear of prosecution or a beating would deter the niggardly from seeking to deprive the cabbie of his due payment.
In some jurisdictions, the police may take no action. Cab drivers develop screening methods while potential passengers may acquire signals in response. Customary morality has evolved in the context of such screening and signaling devices. A habit of paying what is due makes for a more peaceful and prosperous life.
THIS almost facetious question throws light on two conflicting claims about the order that prevails in society.
No. It throws light on Basu's stupidity. In most cities, cab-drivers are licensed and have a working relationship with the police. Moreover, they can rely on support from passers-by unless they are dealing with a notorious gangster. But, in that case, they may seek to ingratiate themselves with the fellow. He may be inclined to tip generously or, if the cab-driver insists his respect for the Don is too great to permit him to accept payment, he may say 'if any rascal gives you a problem, let my people know.'
The paradoxical claim (P) asserts that the pursuit of selfish aims by individuals is sufficient to generate social order.
This is not a paradoxical claim. It is the truth. Cab-drivers don't really love you very much and wish to take you where you want to go because they hope you will give them hugs and kisses.
The unparadoxical claim (U) asserts that human adherence to certain commonly accepted values
like the value of money- not hugs and kisses- as the means of settling accounts in commercial transactions
is necessary. Unless we define selfishness so loosely that it is difficult to conceive of unselfishness, (P) and (U) are conflicting claims.
No. A cab-driver may convey an injured person to hospital even if there is no prospect of his gaining payment. That is an unselfish action. True, it has a reputational benefit but, the fact is, it yields utility. Few of us would not consider the chance to save a human- or indeed, a dog's, life- at the price of a few hour's worth of wages as a bargain which only the Grace of God afforded us. But, many of us believe God rewards such actions in a munificent manner. At the very least, we know that by doing something like this, we think better of ourselves and this enables us to become better people.
This is worrying because both occur within the social sciences: the former in economics and the latter in sociology.
No. In Econ, it is recognised that we gain utility by being able to think better of ourselves. Sociology recognises 'Tardean mimetics'- i.e. imitation of a superior- as itself a source of upward mobility.
The purpose of this essay is to demonstrate with an example, which is representative of a class of real life situations, that it is the unparadoxical claim which is valid.
In which case, Basu has chosen a bad example. Cab-drivers are mercenary though, in exceptional circumstances, they do good deeds. The right example would be voluntary ambulance services- e.g. those in Karachi and Bangkok. The Karachi one is better. In Bangkok, if two voluntary crews turn up, they may get into a fight over who gets the 'religious merit' of taking the injured person to hospital. I mentioned this example to a Pakistani game-theorist back in the Eighties. Once he understood I was praising his religion and deprecating 'Karmic arithmetic', he resolved never to drink Whiskey with me again. Win win.
When we travel by taxi, we do not usually make an effort to run away without paying.
Cab-drivers tend to screen out those who have some trait- e.g. being bleck or (in the old days, in London) living 'sarf of the river'- which correlates with that propensity. Oddly, I never had a problem getting cabs in London. Dark skinned 'Asians' were assumed to be meek and docile. Also we tipped well.
Some economists have tried to model crime and marriage in terms of individual rationality,
They succeeded. Certain sorts of crime- e.g. mugging- did indeed fall when when people started carrying less money in their wallets. On the other hand, as Rishi's father-in-law is telling him, Indian billionaires are annoyed that they can't wear their nice watches in Mayfair the way they can in Mumbai. Also, your Range Rover is safer from theft in Karol Bagh than in Knightsbridge.
As for marriage- look at that of our Prime Minister! That boy knows which side his bread is buttered on right enough!
so they would no doubt explain our scrupulousness in terms of the probability of being caught and the agony of being jailed.
The fact is Basu had already lived in cities like London and New York where, notoriously, cab-drivers didn't pick up passengers on the basis of observable traits which might plausibly correlate with a higher propensity to skip the bill. Ultimately, these guys have to make a judgment call. True, a chap who comes across as a 'Hooray Henry' might actually be a con-man. He may get the better of you- but you live and learn.
In order to permit a relatively rigorous discussion let us place the problem at a greater level of abstraction. An individual gets off from a taxi at a place where there is no one sufficiently near to bear witness as to whether he pays the fare or not; and in the absence of a witness, it is pointless contacting the police (with some police forces this is unconditionally so).
This means the individual could rob and sodomise the driver if strong enough to do so. But the reverse is also true.
Further, this is a large city and the passenger does not expect to require the services of this cabman in the future. Would the passenger try to walk off without paying? I think there will be no disagreement that, even in this situation, a vast majority of human beings would not choose to default.
This is because, as Oscar Wilde noted, it only takes one occassion when you refuse to pay the cab-fare- which, because of the subsequent altercation, leads to your strangling and sodomising the corpse of the cab-driver (what? In for a penny, in for a pound)- it is likely that you will soon progress to cheating on the Times crossword puzzle. After that, you might as well just resign yourself to becoming a partner in Goldman Sachs.
How do we explain this, excepting in terms of our sense of values, or our morality, or custom (essentially something beyond selfishness)?
The problem here is that Basu's and my 'values', 'morality' and 'customs' are all known to have a self-regarding origin in our common religion which stresses 'karma'. If you fuck up others today, you will be fucked over soon enough. What goes around &c.
The economist's 'trained incapacity' does not allow him to give in so easily.
Economists are guys who get paid to actually economise on the use of scarce resources- not to teach stupid shite to kids who weren't smart enough to get into Medical Skool.
I posed this question to a number of economists; and barring one or two exceptions, the response was more or less the same: While the passenger's sense of values may indeed be the cause of his adherence to the law, the act could also be explained purely in terms of 'rationality'.
Suppose you have to get to a certain place so as to avoid a large financial loss. Sadly, you have no money. Though your 'values' may be strongly against cheating a taxi-driver, that is what you nevertheless do. This is 'rationality'. If the penalty of skipping the fare is very high and there is a high probability you will be subject to it, you decide not to try to stiff the cabbie.
The workaround is to take the trip and then hand over your driver license, your watch etc., explain the circumstances and hope the cabbie will take pity on you and settle for double the fare at a later date.
Taxi- drivers are often quite strong (which, judging by my small sample, is by no means the most vulnerable assumption). So if the passenger tried to button up his pockets, the taxi-driver would in all likelihood either himself assail the passenger or gather people and try to ensure payment. It is this risk which makes the passenger 'behave'.
But the cabbie 'pre-screens' customers. That's why it is worth having the right 'signal'- e.g. an expensive wrist watch or a docile demeanour or, simply, being the right colour and 'class'.
This is a plausible argument and let us accept it. But the moment we do so we land ourselves in a problem.
Only if we ignore the fact that, notoriously, cab-drivers use screening devices while would-be passengers are careful to emit more or less 'costly signals'. Thus, this is a separating, not a pooling, equilibrium.
The crack however appears elsewhere. it is now the taxi-driver's rationality which becomes questionable. Why do we expect him to retaliate against a passenger who tries defrauding him, and to attempt recovering the fare at the risk or despite the unpleasantness of a scuffle?
Actually, a rational cab-driver who didn't screen or who mis-read the relevant signal will chalk it up to experience. On the other hand, cab-drivers collectively have a 'tit-for-tat' strategy such that those who fail to retaliate are themselves subject to sanctions. Thus, the guy has to make a stink even if he gains nothing by it, because not to do so could cause his fellows to smash up his cab or break his legs.
To me, the most plausible reason seems to be his injured sense of fairplay
In which case the fucker should be a Communist agitator or a member of the National Front. How is it fair the fucking Capitalists/Immigrants are living the life of Riley while he himself is having to ferry around wankers?
or anger at his customer's violation of social norms
In which case this guy should also be beating the fuck out of boys who kiss each other in cabs (back in '83, this was against 'social norms')
(no doubt catalysed by the fact that he is at the receiving end).
of what? Deprivation of money. Suppose I take a cab and say 'my good man, I refuse to pay the fare which is fifty pounds. On the other hand, I found this fifty pound note and believe it to be yours. Kindly take it. As for your fare, you can whistle for it.' What would the cabbie do? He'd shrug his shoulders and drive off. Money is all that matters here. The Cabbie does not care if you violate social norms by trying to eat your own neck-tie. He doesn't give a shit about your cheating on the Time's crossword puzzle. What he wants is money- the more of it the better.
But to admit this is to grant the role of commonly accepted values - no matter how indirectly - in the prevention of anarchy.
That commonly accepted value is the value of money.
This leaves only one way out: to explain the taxi-driver's response in terms of his selfishness. To do so, one would have to argue that the agony of ga- thering people and a scuffle may be less than the reading on the meter. This may well be valid. But now comes the main difficulty. If that is. so, why should the taxi-driver not try the same tactic even if the passenger has paid?
This can happen. You pay what is on the meter and the guy says you didn't and won't release your suitcases. Meanwhile you have a plane or train to catch. If you are a foreigner, you may pay up. This is a case of 'screening failure'. Foreigners should arrange their taxi pick up and drop off through the Hotel or- as in India- through the airport itself.
That is, he could take the fare and then pretend that the passenger never paid and go through the same action as he would if the passenger had defrauded him, and thus end up collecting perhaps twice the correct fare, not to mention the tip.
Cab-drivers have been known to rape and rob their passengers. This is why 'screening' and 'signaling' are important. Over and above that, there may be enforcement mechanisms of various types.
Everybody would agree, taxi-drivers do not behave in this way.
Some do. The question is what do they do after they cheat some foreigner? The answer is that they either squander their ill-gotten gains on booze and broads or go cruising to pick up other such pigeons. In other words, because they are aiming at a higher return, they screen for those who can be cheated with impunity. This is a separating equilibrium. If the police want to crack down on the problem they will hire some guy who looks like a loaded Japanese tourist to bait the trap. But, otherwise, ordinary clients are not bothered because the thievish cab-driver is self-screening and avoiding them.
Therefore they must be irrational, because it was supposed, a few lines ago that this behaviour is the one in conformity with their self-interest.
Basu does not understand that 'rationality' in an economic context means 'Muth rationality'- i.e. acting in accordance with the correct economic theory. This involves signaling and screening. We look for signs that the cabbie is properly registered and conscientious. We avoid getting into a 'gypsy' cab or one driven by a guy who gives off a rapey vibe. Cabbies, in turn, screen for dead-beats or gang-bangers and refuse to go to dangerous localities.
(Some defenders of faith would, however, be pleased to know that in the city of the author's residence, particularly in the late hours, taxi-drivers do occa- sionally give evidence of rationality.) Herein lies the crux of the matter. The object of the above exercise was not to show that human beings are not guided solely by selfishness; but to demonstrate that given the order that attends the multitude of economic exchanges in society and the absence of anarchy and fraud, this must be so.
This is not so. If order obtains, it is because there has been a co-evolution of screening, signaling and enforcement mechanisms. Customary morality has itself have evolved such that order obtains in certain areas of commerce but not in all. A city may have reliable licensed cabs but unreliable drug dealers or prostitutes. The reverse may also be the case. It is said that in certain 'ghettos', drug deliveries are reliable but pizza delivery is not. This is because the drug dealer has a gun and is backed up by a ferocious gang. The pizza delivery boy, however, may be robbed and beaten.
The 'invisible hand' would not be able to co-ordinate a multitude of selfish acts to bring order - as it is supposed to do - if it was not aided by the adherence of individuals to certain commonly accepted values.
Only if some other 'invisible hand' is supplying the enforcement mechanism. This may not involve any coercion but consist solely in screening and signaling devices creating a 'separating equilibrium'. This is why, some people can always get a cab for certain journeys- e.g. within Manhattan or Mayfair- but not for others- e.g. within dangerous 'ghettos'.
The example in this paper shows that we can maintain that a subset of human beings conform to the "law" entirely because of self-interest; but that rules out, by implication, the same assumption for all the remaining individuals.
This is a matter of probabilities. Where there is screening- e.g. a licensing system for cabbies- chances are the cabbie won't rob you. Equally a person of the upper class or one wearing an expensive watch is not likely to be a dead-beat.
Thus we have to make room for our sense of values, however small.
We could do so. I might have greater trust in a middle-aged cabbie who appears religious and the reverse is also the case. Some countries or communities are associated with high morality and religiosity and there may be more transactions in such communities because there is higher trust.
In a lot of economic theorising it is presumed that all contracts are en- forceable.
No. It is assumed that the contract is actually a contract. In law, one entered into in bad faith is no contract at all. This does not mean damages may not be payable for a tort or an illegal action even though no contract subsists.
Once this is granted, the efficiency of markets is ensured barring of course the standard difficulties associated with externalities and returns to scale.
and asymmetries of information. If these are too great there can be no contracts. This is because 'reasonable expectation' is a condition for enforceability. If asymmetry of information is too great- e.g. one or both parties have no reasonable way of arriving at the expectation that the contract will be fulfilled- there is no fucking contract.
It is only when considering markets like the one for loans, which are characterised by a long time-lag between the acts of the two parties involved in the exchange, do we talk of default (i e, the possibility of one party backsliding on his part of the contract).
Nonsense! Basu is not merely ignorant of the law, he also has not observed that many contracts for immediate delivery are not fulfilled. I ordered desert and find desert is on my bill but the waiter forgot to serve it to me. I kick up a fuss till the manager takes it off my bill. If this does not happen, there was breach of contract but I may not bother to launch a small courts claim.
This is what has led to the substantial literature on credit market 'isolation' and 'interlinkage'. What is not always appreciated is that virtually all economic exchanges entail a time-lag.
Some don't. I pay at the till but the item is not brought to me.
Like the taxi-driver, the barber brings the bill after the hair-cut, as does the wtaiter after the meal.
But I could put money into a vending machine and get nothing.
And, as the above example shows, it is not possible to explain the absence of widespread default in these situations without making allowances for our sense of values and norms.
What fucking 'values and norms' does a vending machine have?
Thus while the absence of externalities. etc, is necessary for the efficiency of the invisible hand,
as is asymmetry of information- I don't know if the vending machine is broken.
a more basic assumption is that the agents involved in economic exchanges fulfil their obligations.
Nope. There is merely the assumption that a contract is a contract. It isn't if 'reasonable expectation' is not met. But that is a statistical matter.
And the ultimate guarantor of this assumption is our sense of values and norms.
No. It is statistical and involves things like vending machines which have no values or norms.
As Arrow puts it bluntly in his perceptive essay, 'A Cautious Case for Socialism' (in Howe, I (ed), "Beyond the Welfare State", Schoeken Books, 1982): "The model of laissez-faire world of total self-interest would not survive for ten minutes; its actual working depends on an intricate network of reciprocal obligations. even among competing firms and individuals."
Arrow was wrong. A laissez-faire world featuring only robots is conceivable. They would lack any 'values and norms' and be incapable of forming relationships or acknowledging obligations. Yet, so long as they can process statistical information and form 'reasonable expectations', 'intricate networks' would evolve.
Once this is appreciated, it becomes possible to understand many features of society without recourse to artificial 'economic' arguments.
In other words, economists can stop thinking and start babbling on about values and norms. Why economize on scarce resources when you can scold Scarcity and say it should be ashamed of itself?
Consider, for instance, the threat of violence. It is well known that one way in which a moneylender in a backward agricultural region ensures that money owed to him is repaid is by using the threat of violence.
It is equally well known that moneylending will exist even when moneylenders have zero coercive power. That is why weak communities- e.g. Jews in anti-Semitic Europe or Chettiars in Burma- could flourish. So long as they can process statistical information and access screening and signaling mechanisms, they can form 'reasonable expectations' and contracts can exist. Indeed, the 'folk theorem of repeated games' says that anything that can be done with a coercive mechanism can also be done without it.
This would appear as a paradox to the economist ("Why does he not anyway use such a threat and extort money? Why does he bother to lend the money in the first place? ").
This is exactly the reason that coercive mechanisms tend to be replaced by screening and signaling. A primitive economy may feature the authorities inflicting hideous punishments on those who break contracts. As that economy becomes more affluent and sophisticated, even if religiosity and morality and cultural homogeneity decline, such coercive mechanisms disappear. There are no 'debtor's prisons' in the City of London, yet commerce has greatly advanced since the time of Charles Dickens. One reason for this is that coercive mechanisms can be misused to extort money or to oppress a particular community. Thus the entrepreneurs and arbitrageurs themselves want the thing to disappear.
But as soon as we accept the idea of norms and morality,
we have an excuse to stop doing economics. We can just scold Society saying it is being very mean and not getting rid of Scarcity so everybody can have a butler and a footman and a chauffeur. The very existence of cab-drivers is a damning indictment of Capitalism. Everybody should have their own private helicopter to take them where they want to go.
such behaviour becomes easy to comprehend.
But where do the 'norms and morality' come from? God? Is Basu going to set up as a Swamy preaching peace & lurve?
Similarly, to explain the larger incidence of default and fraud in economic transactions in some societies, we no longer need to claim an excess or a shortage of rationality on the part of their inhabitants,
thus if we notice that Venezuela or some other such Socialist Utopia has turned into a shit-hole we should not say its people were irrational in blindly following a demagogue. Nor, if we notice that South Korea is much richer than North Korea, should we say that the former country pursued a more rational economic strategy.
but may adduce the more reasonable explanation of differing social norms.
Thus we can say the North Koreans have different social norms- e.g. worshipping a hereditary Dictator- to South Korea. But why is this the case? The answer has to do with North Korea having invested a lot of money in shooting and beating its own people. Why did it do so? The answer is it thought 'laissez-faire' was wicked. Sadly, Arrow showed no inclination to move there.
And with this opens up newer dimensions in policy- making, in which social norms appear as a 'control variable'.
Why is the Finance Ministry bothering with a Budget? It should merely lecture the people incessantly saying 'listen you lazy people, you should have nice nice social norms such that you all become very productive and wealthy and do good deeds and go defeat the enemy and eradicate poverty and cure cancer.'
What is important is to recognise that social norms can alter not only society but even the prices of goods and services.
Moreover, Social Norms can create free, unlimited energy as well as a nice Time Machine so that we visit out distant ancestors and future descendants.
Finally, consider a variant of Sen's delightful application of the two-person game, the Prisoner's Dilemma, to a common social problem (A K Sen: 'Behaviour and the Concept of Preference', Economica,. 1973). Let us assume, as is quite reasonable, that (i) every city dweller prefers his city to be clean rather than dirty,
He would also prefer Death to be abolished
and (ii) one person throwing litter on the streets does not mr.ke a clean city dirty.
Yes it does. One piece of litter is sufficient to make one part of the street in one part of the City to become dirty. If the guy really prefers a clean city, he would pick it up and put it in the bin.
It is easy to see that
this gedanken is fucked in the head
each individual, acting atomistically, would prefer to throw litter on the street
in which case they don't want a clean city. It is a different matter that they want the City to have a more expensive cleaning program financed by somebody else.
rather than go through the trouble of looking for a garbage bin to dispose of it. It being rational for each individual to litter the streets, all citizens - if they were rational
No. If they were irrational, they would want cleanliness while contributing to dirtiness. Rationality involves acting in a manner consistent with your desires. A thirsty man who takes a glass of water and empties it down his pants is not acting rationally. He should drink the water. That is the behavior which is consistent with his desire to slake his thirst.
- would do so. The city would be a dirty one and (given (i)) everybody would be worse off. I find this story convincing
because you are stupid
and therefore believe, though it sounds facetious, that the dirtiness of, for example, Calcutta is a reflection of the rationality of its inhabitants.
Even when the British ruled, 'Black Town' was dirty. 'White Town' was clean. No one has suggested that the Bengalis were more rational than the Brits.
This also shows how much we can gain from a little bit of irrationality.
The Brits were in Bengal to make money. Since they were rational, they kept their part of the City clean. Sadly, their successors preferred to prate about 'norms and values' rather than do sensible things.
Actually there are two ways of solving this problem. One is to impose fines for dirtying the streets; the other is to inculcate in human beings suitable values.
But Bengalis have been boring each other senseless with such lectures on 'shuitable balues' for centuries. The smart ones emigrate to places still ruled by Whites where they will be punished if they litter the streets. Sadly, along with the smart ones, there are some stupid ones- like Sen and Basu- who want to scold Society for tolerating Scarcity. Also, why can't you disgusting desis just stop being so goddam brown? Have you tried bathing in bleach?
The former works by changing what is rational to the individual.
By changing the incentive matrix. This is the 'reverse game theory' which is mechanism design. Economists can improve outcomes for society by doing this. But only if they are smart. If they are as stupid as Sen and Basu, they set up as the Mother Theresas or Rabindranath Tagores of Economics.
The latter works by making people accept a little bit of irrationality.
Basu believes it is irrational not to squat down and take a dump on the street.
It is true that the later would take much longer to implement,
what's a couple of billion years between friends?
but it is ethically clearly more attractive
to imbecilic gobshites
and ought to be the ultimate objective.
What we should aim at in life is to bore the shite out of everybody in the vicinity by scolding Society for not being irrational enough to cease shitting in the streets.
[I am grateful to Andre Beteille, Mrinal Datta Chaudhuri and Amartya Sen for discussions.)
Nobody else is grateful to useless Buddhijivi blathershites like Basu.
No comments:
Post a Comment