THE ECONOMICS OF CASTE AND OF THE RAT RACE AND OTHER WOEFUL TALES*
There is a standard model of economic behavior, the Arrow Debreu general equilibrium model of perfect competition.
It was nonsense. For a market to exist, market-makers must exist. AD assumes they don't. Prices are set by magic so all markets clear. Thus, suppose I lost my sweater yesterday on my day-trip to Paris and Monsieur Dubois found it, I can buy it from him for 100 Euros- the price he is willing to accept because my sweater doesn't fit him, and which I am prepared to pay because it has sentimental value for me. The fact that I don't know Dubois & he doesn't know me does not matter. The price vector contains all the information in the universe. There is no Knightian uncertainty- i.e. Evolution is a false theory. We live in an Occasionalist universe.
While this model may not be entirely adequate as a description of economic reality, it is most useful as a standard of comparison. For in equilibrium in this model, subject to the careful qualifications of Pareto optimality, peoples' lives are as pleasurable as they possibly can be, given their tastes and productive capabilities.
Thus Dubois can get 100 Euros for my sweater because, by magic, the market wafts it away from him to me while simultaneously taking 100 Euros out of my wallet and passing it to Dubois.
Consequently, to understand why peoples' lives are not as pleasurable as they might be (in the Pareto sense), it is necessary only to know why the real world fails to correspond to the Arrow-Debreu utopia.
It's because there is no market unless some people create it. But markets don't have magical powers and cost money to run. There are 'hedging' & income effects such that there is no unique equilibrium. Indeed, speaking generally, markets don't clear.
In the real world, contrary to the assumptions of Arrow and Debreu, information is neither complete nor costless.
But arbitrageurs have an incentive to acquire more information & make it cheaper to access.
On the contrary, given the cost of information and the need for it, people typically make predictions about the behavior of the economy and the behavior of individuals based upon
the opinion of those with greater knowledge or a better track record of making predictions
a limited number of easily observable characteristics.
Nonsense! If the decision matters, we ask a smart dude. He can see beyond what is 'easily observable'.
We say that such a prediction is based upon an indicator; an econometrician would call it a prediction using the method of instrumental variables.
Only if causation is impossible to establish and thus correlation is all we have to go on.
This paper shows the distortions caused to examples of the A-D (Arrow-Debreu) model by the introduction of indicators.
You can't distort that which is incompossible with our universe.
Other approaches to the difficulties encountered by the A-D model in explaining labor markets are given by the "new" labor economics.
Which was inferior to knowledge possessed by people who actually did the labouring.
There are two types of examples of the use of indicators in the models that follow. One sort of indicator owes its existence to the potentially useful economic information provided. In the example of sharecropping the output produced is used as indicator; it serves the useful function of differentiating between farmers who have expended different levels of effort in tilling the crop.
No. A smart guy may put in less effort & get a bigger crop than a stupid donkey.
In the example of work conditions the speed of the assembly line predicts the ability of workers on that assembly line,
Not if quality deteriorates.
and therefore differentiates workers of different ability.
That would have already been done before they were put on the assembly line.
In contrast, in the following two examples the indicators owe their existence purely to social convention. In the example of statistical discrimination, under conditions described, all persons of the same race are predicted to have equal ability.
Nobody has ever made any such assumption. We may say 'African Americans are better at basket ball than Asian Americans'. We don't say any random black dude is equal to Michael Jordan.
In the example of caste, the behavior of one member of society toward another is predicted by their respective caste statuses.
Not in India. Money & power are what matter.
In this second type of example, introduction of indicators into the A-D model brings with it a second previously missing aspect of reality, the panoply of cultural characteristics used by anthropologists and sociologists to describe a society.
i.e. stupid shite.
For, by definition, culture consists of "regularities in the behavior, internal and external, of the members of a society, excluding those regularities which are purely hereditary."
Regularities change very quickly if there's money on the table.
Since culture concerns regularities in behavior and since subcultural membership is easy to observe, members of society, as well as visiting anthropologists and sociologists, can predict individual behavior from subcultural membership.
They really can't. That's why Hitler was defeated. Even if Germans were more 'noble' than Americans, Americans had way more money. Japan may not, as he said, have been defeated for 3000 years, but Uncle Sam made them whimper & plead for mercy quickly enough.
The indicators by which men judge each other may warp their values and distort their goals.
Doing stupid shit has this effect. Being stupid isn't a handicap provided you imitate what smart peeps are doing.
The anthropologists give accounts such as those of the Kwakiutl Indians, among whom the chief at feast-time who burned the greatest number of blankets, as the mark of the most conspicuous consumption, received the greatest honor.
There was a political & religious angle to this. 'Potlatch' was banned.
II. SHARECROPPING
The first example of indicators deals with the simplest phenomenon. Several economists have asked why sharecropping is a common form of land system.
Only where there is only limited monetization.
After all, since the sharecropper is much poorer than the landlord and much less liquid as well (not owning land that can be mortgaged), it would be more natural for the landlord rather than the tenant to bear the risk of crop failure.
He does. He pays the land tax.
This would be accomplished if the landlord paid the tenant a wage and sold the crops (perhaps even selling some of it back to the tenant).
That's not a tenant. That is wage labour with accommodation provided- e.g. on a Tea plantation.
There is also evidence that fixed wage payments are more "natural" than sharecropping.
In a monetized economy- sure.
A recent study of sharecropping in the United States South concludes that immediately following the Civil War "the wage payment system was, from all indications universally
attempted."
Conditions were pretty fucking unnatural back then.
Travelers' accounts seem to show that at the end of the Civil War sharecropping was viewed as an "experiment."
Not beating and raping slaves was the experiment imposed on the South.
There is, however, a very simple reason for a preference for sharecropping over a wage-payment system. There are two components to the sharecropper's input: the time he puts in and the effort expended. While the first is easy to observe, and can be paid a fixed wage, the second cannot be observed without careful supervision of the labor.
Fuck off! Just ask around & you will be able to identify the slacker quickly enough.
Suppose that the input of the sharecropper depends upon his time at work and his effort; suppose further that his effort can be measured and called e. With a wage system the sharecropper should receive an income w dependent on e and t:
w = w(e, t).
Without supervision the landlord cannot determine the effort put in;
Unless he asks around. Also, when it comes to agriculture, it is easy to determine which farmer is hard-working because the weeding and drainage etc. will be better. Just ride around your estate a couple of times a year. Stand a round of drinks & you will hear all the gossip- X is a hard worker but as stupid as shit. Y is smart. He puts in fewer hours but gets a better yield.
and the wage paid to the individual worker will depend on
transfer earnings. How much can he earn in his next best job?
the average effort of the average worker, e: thus
w = w(j, t).
This leaves no incentive to the worker for any effort beyond the minimum necessary to be paid for his time.
If the guy looks like a slacker, warn him he will get the sack. But it is enough to say 'you are lucky I don't listen to gossip. I'm sure you are working as hard as ever. Still, this is not a good time to lose your job.'
If he dislikes effort, he will minimize it.
He will end up a beggar.
In contrast, in sharecropping, the farmer is paid for the effort that he puts in as well as for his time;
his payment is his share of the crop. The question is whether he can get the best price for it.
but this effort and time are estimated imperfectly from another characteristic-the output produced.
Nobody gives a shit about effort & time. How much of the crop you get and the price it sells for are all that matters.
The equilibrium is distorted by this procedure, since the risk-averse farmer remains unprotected from the natural randomness inherent in agriculture.
He is better protected than the village artisan or landless labourer. The Bengal famine showed this. Akerloff was writing about something he had zero knowledge of. Also what he was saying was utterly stupid. No wonder he got a Nobel Prize.
The basic stylized facts of this model conform with the conditions of sharecropping. In traditional agriculture the hard-working farmer usually receives yields that are considerably greater than the yields of the average farmer. A Punjabi peasant, who prided himself on yields greater than those of his neighbors, once listed for me "the seven
things which a good farmer does, which a poor farmer does not do."
It is significant that many of these seven things involve arduous work and much patience; many are also difficult to observe.
They were so easy to observe that you could see from an aeroplane that East Punjabi farmers (mainly Sikhs) were more hardworking than West Punjabi farmers (mainly Muslim) On one side of the border the land was much more green.
A similar story has been told by John Mellor in his study of farms in a village of Uttar Pradesh.
Many parts of UP have an indigenous class of crop estimators. These are guys who can walk through an orchard & tell you how many tons of fruit will be produced.
Hard work generated significantly higher yields even with the use of only traditional farming methods.
Traditional methods are fucking hard work.
The division of crops between those grown on a wage-payment system and those grown on shares is also consistent with our explanation.
No. It is consistent with the property regime which in turn depends on the land revenue system. But equally important is coercive power. Why was some land 'self-cultivated' with wage labour while other land was given to tenants who gave it to sub-tenants? The big landlord needed muscle men but also wanted allies of particular castes/ creeds.
Where supervision is needed for reasons other than determination of effort, the model predicts that wages rather than shares will be paid. In India, for example, as an excellent rule of thumb, capital-intensive plantation crops are grown on a wage-payment system.'
No. Either workers who were brought in from outside- even if this happened several generations ago- and housed & fed & looked after by the employer or they sold their produce or handed over a fluctuation portion in lieu of rent. What mattered was the property regime which in turn was determined by fiscal convenience to the Government.
Speaking generally, a 'law of increasing functional information' operates such that useful information improves in quality and falls in cost. This happens because it pays to specialize in gathering & providing it. Alternatively, it happens because the alternative is death or demographic replacement.
Akerloff next turns to
indicators of social origin (which) may lead the economy into a low-level equilibrium trap.
in which case it is at risk of invasion or insurrection.
We begin with Arrow's model' of statistical discrimination given in "Models of Job Discrimination," and "Some Mathematical Models of Race in the Labor Market'.
The model here is different in important detail from the original by Arrow, who does not consider the two equivalent. I am sure that he would agree that, however the mathematics differ, the economic spirit of the two models is the same.
In this example, under some circumstances, employers use the average quality of a given race to predict the quality of individuals of that race.
If you are being judged on the basis of race, chances are you are below average.
It is easy to see that if such an indicator is used, it will destroy all incentive for self-improvement for that race,
Nonsense! It incentivises occupational, geographic & social mobility. In particular, a discriminated against group is likely to chose entrepreneurship & self-employment to a greater degree.
since all individuals of the race are judged the same and therefore paid the same wage irrespective of individual merit. In this way prejudice may produce a lower level equilibrium trap: if a race is deemed by prejudice to be unqualified, no incentive is given to become qualified, and the prophecy is self-fulfilling.
Akerlof's mother was German Jewish- i.e. of higher class & education than most Polish or Russian Jewish immigrants. Thus he may had no personal knowledge of the manner in which poor Yiddish speaking Americans clawed their way up. The second generation went to shitty inner city schools and were kept out of Ivy League by the 'Quota'. That's the reason they worked so hard & took such great risks to rise & rise.
The Model
In this model there are just two types of jobs, one requiring qualified labor and the other requiring either qualified or unqualified labor.
If you are qualified in some manner it is worth your while to spend a little money on getting some sort of accreditation. That is how Professional Associations come about.
It is costly to test workers individually to see whether or not they are qualified.
Which is why they bear the cost of accreditation & pocket the reward.
The change in proportion of qualified workers depends upon the incentives for self-improvement, which are differences in wages for qualified and unqualified workers of that race.
I suppose Arrow means that a black accountant had less incentive to pass the relevant exams to become a CPA because he still wouldn't get White clients. But the thing can work the other way. People might say- this 'boy' must be a genius! Moreover, he can't charge as much as a White dude. We are getting a bargain!'
V. CASTE AND GROUP ORGANIZATIONS
...It may appear that the tastes of persons in discriminating societies are so overwhelmingly biased in favor of discrimination that, relatively, the positive or negative effects of economic incentive are of only minor moment. But this ignores the broad historical perspective, which attempts to explain the stability (or disappearance) of institutions over a long period of time. For there are a fair number of cases where opportunities have arisen for deviants to break the caste code and make economic profits, with consequent rise in their social position and erosion of the caste taboos.
Class status can depend on what 'mimetic target' a Society chooses.
Consider three diverse examples of this phenomenon. In Japan
when Confucian China was the model, Merchants were downranked. Once the Brits defeated the Chinese, the 'nation of shopkeepers' became the mimetic model.
as merchants have become more economically successful,
because the military government had been overthrown & markets were displacing bureaucrats drawn from the Samurai class
so too have the taboos against trade and manufacture been reduced.
Because, in England, the Prince of Wales was happy to stay in the Country mansions of his grocer- as the Kaiser disapprovingly remarked.
Even in caste-bound India caste status rises with the economic success of the caste,
& vice versa
although, typically, newly successful castes also adjust their social customs, at least partially, to reflect their higher status.
This happens everywhere. The American meat-packer soon hired an English butler.
The best example of economic success reducing taboos is, most probably, the elimination of the sanctions against collection of interest. The usurer of the Middle Ages
was the banker of the Middle Ages. Lombard street is named for Lombard money-lenders who arrived in London in the 12th century.
has turned into the banker of today.
Goldsmiths turned into bankers because they had strong-rooms. Fractional reserve banking arises from them.
This section introduces a new class of models in which, as in Arrow's statistical-discrimination equilibrium trap, those who break caste customs suffer economically.
They may face social ostracism. It is said that Jinnah's ancestors were ostracized after they took to dealing in fish.
This class of models depends upon an important facet of caste societies missing in previous models of discrimination. In previous models current transactions (so long as they are legal) do not result in changed relations with uninvolved parties in subsequent transactions. For example, if farmer X makes a contract for sale of wheat to speculator Y, his subsequent dealings with speculator Z will be unaffected.
i.e. in this model, no one has market-power. Everybody is a price-taker.
On the contrary, in a caste society any transaction that breaks the caste taboos changes the subsequent behavior of uninvolved parties toward the caste-breakers.
in other words, there is market power.
To take an extreme example, consider what would happen if a Brahman should knowingly hire an outcaste cook: the Brahman would be outcasted,
No. He would merely have to spend a bit of money on 'prayaschitham' (expiation). Plenty of Brahmins hired Muslim or Christian cooks who could prepare Mughalai or European food. True, they generally maintained a separate kitchen.
and the cook would find subsequent employment almost impossible to obtain.
Only if he was a shitty cook. Why the fuck would a non-Brahmin care if the cook had worked for a Brahmin? If anything, the thing would be a recommendation.
True, a particular sub-caste, or ethnic group, could operate like a Trade Union. If even one non-member is employed, the entire sub-caste boycotts the employer. But actual Trade Unions also exist.
The possible intervention of third parties in a transaction allows for a richer class of indicators than that given by Arrow's statistical discrimination- typically, the use of indicators in caste societies being less narrowly technological.
This is foolish. What a guy did in medieval societies was fucking obvious from the way he was dressed & the tools he carried.
Generally, in a caste society if a member of caste A relates to a member of caste B in a given way, he can predict from knowledge of the relations between caste A and caste B how members of all castes will relate to him in future transactions.
Very true. If a Brahmin fucks you in the ass, all Brahmins will fuck you in the ass even if they are straight.
Such predictions can lead to an equilibrium in which
if one Brahmin is gay, every Brahmin is gay.
all expectations are met and economic incentives favor obedience to the caste code-
There is no such code. Brahmins are priests by caste but most don't practice priest-craft
even in the extreme case where tastes are totally neutral regarding the observance of caste customs.
Because the model is shit.
The following three conditions describe marriage customs in
India.
1. Society is divided into mutually exclusive groups (called castes).
No. There are castes whose ritual status changes at different times of the day (afternoon Pariahs) or depending on their geographical location or occupation.
2. A code of behavior dictates how members of these castes should behave.
That depends on religion not caste.
Regarding marriage there are complicated rules as to who may marry whom, payment of the dowry, the timing and performance of the marriage rites, etc.
The Brits thought so. Brahmins knew otherwise. One reason they got behind Democracy was because they wanted elected legislatures to get rid of British era laws- e.g. the one by which Indira's marriage to Feroze (or her Aunt's to a Jain) were per se illegal.
The caste rules dictate not only the code ofbehavior, but also the punishment for infractions: violators will beoutcasted; furthermore, those who fail to treat outcastes as dictated by caste code will themselves be outcasted.
Unless they pay a bit of money to do prayarschitham or beat the shit out of a couple of the more vocal of their critics.
3. Caste members predict that those who do not follow the caste code will be made outcastes and will receive the treatment of the average outcaste.
No. They get that the dude will move to a different village or stay in the same village but join a different sect.
An outcaste in India is permitted to hold only scavenging (or other polluting) jobs.
What was permitted was determined by the King, Governor, of other official. True, there might be a village level boycott but the solution was to move to a different village. Akerlof simply didn't understand how the real world works.
He is not allowed to eat with caste members, to touch them, or to touch their food, which in the case of someone outcasted includes his own parents and siblings.
But it is cheap & easy to do the necessary 'prayaschitham' to restore ritual purity.
Of course, his own children will be outcastes and will suffer the same prohibitions.
Unless he moves or changes religion or gets rich.
Why should these three conditions describing marriage customs in India be of interest to the economist?
They weren't of interest to Indian economists because they simply weren't true.
First, note that those who fail to follow, or even to enforce the caste customs do not gain the profits of the successful arbitrageur but instead suffer the stigma of the outcaste.
This is just Trade Unionism or a cartelism. It exists everywhere.
If the punishment of becoming an outcaste
like Spinoza? Conversion overcame the underlying problem.
is predicted to be sufficiently severe, the system of caste is held in equilibrium irrespective of individual tastes, by economic incentives; the predictions of the caste system become a self-fulfilling prophecy.
Caste is a good enough solution to the stable marriage problem for large enough jatis. It probably does reinforce high trust networks.
Formal Model of Caste Equilibrium
This subsection presents a formal model of caste equilibrium.
The correct solution concept is 'evolutionarily stable strategy' which was propounded in 1972/73 by Maynard Smith & George Price.
Caste equilibrium is defined as a state of the economy in which caste customs are obeyed, yet no single individual, by behaving differently, can make himself better off.
This can only arise if there is no Knightian Uncertainty & thus the caste Social Contract is complete. Otherwise, there will always be novel situations where there is doubt as to how to proceed. There will be reactionaries & progressives. If one lot do better over time, they win.
The first concern is, of course, to describe this equilibrium. However, since there are also coalitions of individuals who by acting together can make themselves better off than in equilibrium, it is also of interest to know the relative ease or difficulty of forming such a coalition.
Castes are coalitions. There is a 'natural' way of constructing super-caste groups- e.g. 'left hand' & 'right hand' castes.
For this purpose we also look at the size and nature of the smallest equilibrium-breaking coalition.
Anyone can break it. Otherwise, how could there be 'outcastes'?
Four sets of assumptions describe the economy; those describing technology, market structures, tastes, and the social system. The assumptions describing the social system are laid out in parallel with the earlier description of marriage in India.
Arranged marriage. There were other sorts. It's just that arranged marriages have been more successful demographically. This probably has to do with 'risk pooling'.
In general this model is extremely simple, subject to one complication. By its very nature the caste system involves trade and the division of labor. If outcastes could set up their own economy independent of caste members, the caste system would fall apart.
No. If 'broken men' move somewhere else, clear land & begin cultivating it, sooner or later they will merge in caste with which ever group of villages they habitually exchange brides.
In Akerlof's crazy model
By birth there are just two castes divided into a dominant caste D and a nondominant caste N. Labor of both castes D and N can be outcasted. Outcastes, if any, form a third group.
S3. Persons predict that breakers of the caste code will be outcasted and receive the wages bid for outcaste labor.
The model assumes that no new products which are substitutes for what obtains can be produced. Thus he is describing a stagnant economy immune to invasion or an internal political power shift as demographics change.
Comments on Caste Equilibrium
1. The equilibrium described has two types of distortions due to caste structure. The equilibrium is not Pareto optimal, since in a Pareto-optimal equilibrium N-workers would work in skilled jobs, for which they are fully qualified.
We could say that's a Hicks-Kaldor improvement but it isn't a Pareto improvement because the wages or status of the dominant group falls.
Also, income distribution is skewed along caste lines, since in the absence of caste all workers would receive the same wage.
i.e. caste is a type of wage discrimination.
2. There is another equilibrium, also with fulfilled expectations, in which all workers work in skilled jobs and receive a wage 61sk. The price of all goods is 1.
3. The smallest equilibrium-breaking coalition is
one. Anyone who can subsist & reproduce while defying the caste code breaks the equilibrium. Consider abortion. There was a time when a Doctor could be struck off for performing an abortion. This didn't mean medical abortions weren't available. They were just very expensive.
the smallest group that can set themselves up as a separate subsector and be as well off as in equilibrium while trading with caste members on the terms of trade granted to outcastes.
This is like price discrimination by market segmentation. How are you to prevent 'leakages'- i.e. buying in the cheaper market & re-selling on the more expensive one? More generally, cartels face the problem of 'cheating'.
In situations where this coalition must be large, where trade with the caste economy is necessary, or where the cost of forming a coalition is high, the threat to equilibrium of such a coalition is small. These principles are illustrated in the examples that follow.
Three Examples of Caste Equilibrium
aren't about caste at all. Akerlof is saying 'caste is like something else which exists in non-caste societies'. Similarly, I could write about my life as James Bond by pretending Bond was actually a Cost & Management Accountant.
Example 1. Racial Discrimination. Racial discrimination is implicit in the model, the major difference between the caste model and those of Becker, Welch, and Arrow being in the assumption that persons use race to predict how everyone else will react to hiring persons of different races in different jobs. Their predictions result in a lower level equilibrium trap in which all predictions are met.
You discriminate in favour of your race because of the Price equation- i.e. kin selective altruism. The notion is that your own people will stand with you if the going gets rough.
Example 2. Government-Business Groups. Allegedly many government-business groups, including the military-industrial state, governmental regulator-regulatee nexuses and political machines are held together by a caste-outcaste structure similar to that of our
model.
Which is like saying 'many secret agents- like James Bond- operate in the manner of a Cost & Management Accountant. '
Example 3. Professional Groups.
which are occupational in the (at least notional) manner of caste groupings.
The public often delegates authority to professional organizations to police their own members-the most prominent of these being bar and medical associations. In turn, the members are expected to maintain professional conduct. Since cooperation with others in the profession is a necessary part of the job, the same outcasting mechanism used by caste, races, and government-business cliques enforces a professional unanimity that gives the profession more than its fair share of economic power.
What you are getting is 'interchangeability' and a grading hierarchy. This was a feature of the Shreni (guild) not the jati (caste).
VI. CONCLUSIONS
Our four woeful tales have described the ways in which the use of indicators can distort equilibrium.
There is no equilibrium unless there is a steady state. But, since steady states don't arise out of some magical Arrow-Debreu price vector, Akerlof's woefully stupid tales haven't described shit.
In so doing, we have also answered two challenges to economic theory.
The standard individualistic theories of income distribution and resource allocation are notable by the absence of variables describing social structure, except insofar as these variables affect exogenously given tastes or the initial allocation bundles.
Which captures everything relevant. Stuff about you which is determined at birth is part of your fucking allocation bundle. If you choose to go along with how things have always been- that is your fucking preference.
The absence of these variables poses the first challenge: to construct an individualistic theory in which income distribution and resource allocation reflect, to some extent, the divisions of society as described by the sociologists.
Just cram them into 'allocation bundle' or 'preferences'.
The most common indicators are based upon the standard subcultural divisions of a society. And, as a result, the use of indicators makes equilibrium income distribution and resource allocation dependent on these divisions; and the first challenge is answered.
By something sillier yet.
The second challenge to economic theory concerns the relation between marginalism and social custom.
They change because of what is happening at the margin- i.e. entry & exit.
As long as most persons have positive utility for obeying social customs, and as long as activities are pursued up to the point where marginal costs equal marginal benefits,
but these change if market power exists. Marginal Cost & Marginal Revenue rise if making an extra unit lowers price or raises input cost. Monopolies & Monopsonies, ceteris paribus, restrict output & cause a deadweight loss.
there will be rewards to breaking social customs insofar as they fail to promote economic efficiency.
Sure. At the margin, members of a cartel tend to cheat.
While such rewards occur sometimes, and they may also be spectacular, I would tend to believe that usually the greatest returns go to those who do not break social customs. Archetypically, they join the proper fraternity, work for the proper law firm, and may even marry the boss's daughter.
But then they start losing valuable new types of business- e.g. Mergers & Acquisitions- to Jewish law-firms founded by guys who went to City U.
In a segregationist society, such persons discriminate;
because they don't want to be tarred & feathered or have the KKK burning crosses on their lawn.
in a caste society they follow the caste code.
Unless they are sufficiently respected that they gain even more by reforming it.
While not denying the possible returns to the arbitrageur and social deviant, the models of statistical discrimination and caste explain why economic rewards may favor those who follow prevailing social custom; and in so doing, they give economic reasons why such
social customs may endure.
Speaking generally, fear of being killed trumps greed for a bit more money. Society has coercive powers which even economists need to recognise.