Thursday 17 February 2022

Did Amartya Sen ever know any Economics?

It is tempting to think that Sen turned senile some thirty years ago- which is why he got a Nobel Prize. The truth, however, is that he never knew any economics. The following makes this clear

Back in 1985 Sen told Arjo Klamer 

I was (as as research scholar at Cambridge) ... persuaded that to see value theories—either the labor or utility theory of value—just as an approximation for predicting relative prices is a mistake; one could more fruitfully see them as ways of describing, interpreting, and understanding some features of society.

Stuff which men make and sell has some productive labor embedded in it and must possess some utility to those who buy it. Nobody does not understand this. Economists know prices are affected by supply and demand. But not all supply is based on labor- it is merely one of many inputs- nor is all demand based on utility- there is 'derived demand' and 'speculative demand' etc. Thus no theory of value 'predicted prices'. All it could do was explain Demand. But prices depend on both Demand and Supply. What Sen has said is stupid. Did he misspeak or was he genuinely stupid? Let us see-

There are also elements of normative evaluation in the value-theoretic perspective, helping to assess what is being achieved by society. This is straightforward in the case of utility theory.

No. Utility theory does not distinguish between what is repugnant or meritorious. Such considerations can be introduced afterwards under the rubric of 'externalities' or 'social or legal constraints'.  

Preferences, in themselves are not normative. Meta-preferences- i.e. the stuff one wishes one liked- may be normative. But meta-preferences have no unique representation. One may simultaneously wish one had the preferences of a Saint, an aesthete, a bon viveur, and a guy who actually likes the lifestyle you happen to be stuck with. 

It can be broadened by moving away from the narrowly psychological interpretation of utility to the classical concepts of use value and the fulfillment of needs.

Which are still 'psychological'. Why do some nutters bang on about 'use value'? Is it coz their mummies didn't breast feed them? Or is it just coz they iz misery guts? As for fulfillment of needs- do they genuinely stand around on street corners giving hand-jobs to the needy? 

In the case of the labor theory of value, the notion of exploitation invites one to compare what could have happened—all the rewards going to human efforts—with what is actually happening—much going to owners of non-human resources.

But there can be no exploitation if both supply and demand are elastic. Kantorovich had a well developed theory of shadow prices by then. Sen is talking nonsense even from the ultra Marxist p.o.v.  

There are difficult ethical issues implicit in these approaches,

No. Legal or other constraints may be ethical or they may be mere shibboleths. But they are exogenous.  

and they had been taken seriously by classical political economists such as Smith, Marx and Mill, not to mention Aristotle, and also much discussed recently by moral philosophers of different schools of thought, such as Rawls, Nozick, Scanlon, Dworkin, Hare, Harsanyi and others.

All of whom accepted that Legal or other constraints are exogenous to Economics.  

I was convinced that there was something quite big that needed to be explored. The extant approaches to welfare economics at that time included:

Cost Benefit analysis to determine if a particular project should go ahead. Also there were 'Pigouvian taxes' or quantitative controls on 'negative externalities' and attempts to calculate 'dead weight losses' in this context and also that of fiscal redistribution. This was all part and parcel of standard Public Finance courses- which Sen remained ignorant of. He just wrote crazy shit about magical Social Choice in a world without Uncertainty- i.e. indulged in stupid fairy tales.  

one, the positivist rejection;

there was no such thing. Everybody agreed you had to tax drink and ban certain types of drugs. The question was how best to do it and what 'unintended' consequences might be.  

two, taking welfare judgments out of welfare economics

in the mind of this cretin. The truth is that Judges and Tribunals and Civil Servants were making 'welfare judgments' all the time when deciding compensation claims etc. 

—such as compensation tests, which make normative evaluation turn essentially on possibilities,

Hicks-Kaldor type tests weren't actually done. What happened was one guy would say 'let us do x which will have benefit y at cost z' and then the other guy would say 'in that case Tax revenue will fall (because it is not a Hicks Kaldor improvement) and so we'd have to axe some other spending program'. Thus the politicians took the ball away from the economists. 

no matter how good or bad the actual situation is;

Sen, being Bengali, thought 'normative evaluation' was required of 'buddhijivis'.  If they didn't say 'boo to Fascism. Government is very evil because it is not saying boo to Fascism. Boo, boo!' then they might turn into Hindutvadis and suddenly remember they had been chased out of East Bengal by Muslims.  

three, concentrating only on non-conflict comparisons—Paretian welfare economics, which is limited by its silence on distributional issues;

There is no 'Paretian welfare economics'. The fact is, if transaction costs are low and information is cheap, Pareto optimality will be achieved anyway. This is the essence of Coase's theorem. Who owns what does not matter. People will trade till neither party can be made better off without the other being made worse off. True, there may be a 'holdout problem' if the market is small and shallow but the solution there is ideographic.  

and, four, examining what social welfare functions we can use without much of an informational base,

Only the Government could use a SWF and its information base was one common to the Treasury and the Office of National Statistics. Nobody was going to give some stupid Bengali economist any money with which to acquire an information base. No doubt, the fool could give lectures on some theoretical shite but it was shite everybody knew to be pure mental masturbation.  

in particular, not using interpersonal comparisons of utility, nor making any foundational use of non-utility information—Arrow's impossibility theorem shows the informational poverty of such a framework.

But Arrow's theorem is stupid. It says 'if we define a person who aint a Dictator as a Dictator then cats are dogs'.  

I felt that the informational issue was quite central, not only to avoid Arrow-type impossibility results,

Which are avoided by pointing out that they are nonsense. Anyway 'Preference' is a Tarskian Primitive. As Tarski's student, Arrow had no business giving it a definition and ascribing properties to it.  

but also to make adequate room for concepts such as justice, equality, freedom,

Love, God, Moksha, Beauty, Saintliness, Ahimsa, Sodomy, Strength through Joy, Killing Witches, abolishing Vampirism, Transgender Rights, Cat Worship, Satori, achieving a non-repressive desublimation, ending Racism and Slut shaming, improving the quality of Music, 

even efficiency. But I did not then understand the main issues clearly enough

What were 'the main issues'? The answer is that if human beings evolved by natural selection and the future is unknowable, then the main issues of welfare economics can only be understood by those who are actually economizing on the use of scarce resources to increase the welfare of actual people. No doubt, a pedant can come up with stuff for bureaucrats to do but that won't enhance actual welfare. The thing will be a type of hypocritical fraud.  

and was not able to pursue these things much at that time. On my plan to work on this, Dobb was quite sympathetic; Sraffa was skeptical in a friendly way; Dennis Robertson felt that it was a sensible thing to do, but he was clear that utility provided enough of a basis for all welfare economics. Joan Robinson thought that this was just not an interesting subject at all.

Since post-War Governments actually were increasing welfare it was not an interesting subject. When they ceased to do, Sen-tentious shite came into its own because this type of mathematical masturbation was occurring in the context of a widespread acceptance that there were no more 'low lying fruit' for bureaucratic welfare enhancement for the majority, or dominant, community in advanced economies.  Indeed, the bureaucracy might have to be pared back. 

I did not manage to get back to welfare economics and its informational base until much later. 
A lot of what you do is critical of what other economists do.

But academic economists were increasingly disintermediated or else said whatever they were paid to say.  

Well, every economist is somewhat critical of other economists because that is the way in which you assess your position—in the coordinates established by others. But it is true that I have complained about the narrowness of modern economics. I was going to say neoclassical economics, but it is also true of straightforward Marxian economics, straightforward neo-Keynesian economics and so on.

Not to mention the 'narrowness' of modern Medicine which focuses just on people's health and pays no regard to immoral trajectory choices in the asteroid belt.

Most of modern economics

academic economics- sure, but Law Courses too concentrate heavily on the Law and neglect aesthetics and culinary science. It is also the case that Maths Instruction at University tends to avoid areas of interest to Archaeology.  

tends to concentrate too heavily on very narrow things, leaving out  enormous areas of what are seen as political and sociological factors on the one side, and philosophical issues on the other.

Unless, something useful can be said about them.  

But these issues are often central to economic problems themselves. Taking an interest in them is part of our own heritage. After all, the subject of modern economics was in a sense founded by Adam Smith, who had an enormously broad view of economics.

He was a moral philosopher. He did not found modern economics.  

Smith would be an absolutely ideal example of giving economics its due.

The subject did not exist. Some Nations became wealthy by doing smart things. Others- like India- found it was smarter to pay a lot to get ruled and schooled by enterprising people from those Nations.  

Among the classical economists whom I have enjoyed reading, I have got more joy out of reading Smith than anyone else.

Smith didn't want people to just read him to get 'joy'. He was saying there were sensible things people could do so their Nation became Wealthier and more Secure. This might involve ruling over- for a hefty fee- stupider people.  

Marx comes very close to it.

But Marx's disciples didn't make their Nations Wealthier. They created tyrannies. 

Someone who is regarded as a major figure, David Ricardo

who made himself very wealthy before using economic arguments to champion the class interest of those who triumphed in 1832 

—my gurus Piero Sraffa and Maurice Dobb

who were utterly useless and ignored by their own countrymen 

spent their lifetime editing his works—I find on the whole more boring to read, although I see the excellence of his purist analyses. He is quite narrow, not really deeply interested in social or political aspects of economics.

This is crazy shit. Ricardo was all about how the Aristocrats were parasitic. It was the industrial middle class which should take a leadership role. This is precisely what happened in countries which grew wealthier. The power of the landed class decreased as the population shifted to manufacturing and services. Scientific knowledge, nor Religion or endless Moralizing, became the key to rising affluence.  

I find it odd that neoclassicals and the neo-Keynesians, who battle with each other, tend to agree on the unique importance of Ricardo.

He was a guy who got rich and then produced a good argument why people like him should decide policy so the country as a whole kept getting richer. Ultimately this wealth did trickle down to the proletariat. Marx was adapting Ricardo to give an argument for why the industrial working class should have more say in how the Nation was run. If this class concentrated on things like good compulsory education and collective insurance for Health, Pensions, etc- then everybody gained because a virtuous circle of rising productivity and increasing market size came into existence.  

Ricardo was the first economist in the narrow sense among the classical political economists. He was not going to look at many things, but he was going to assess rigorously a few things well.

That's the key to success. Focus on one thing. Try to be the best at it. If this aint making you or your country richer, stop doing it. That's it. That's the whole story.  

It is an approach that has borne many fruits in our discipline, but it is also limited and narrow, and ultimately counterproductive.

The opposite is the case. The theory of specialization and division of labor on the basis of comparative advantage- i.e. opportunity cost ratios- is still very broad and very productive. Sen-tentious shite is useless when it is not actively mischievous.  

By narrowing the focus sharply—possibly quite arbitrarily—difficult problems can be made nicely neat. Think of Ricardo's corn theory of profit rates, later extended by Sraffa.

Don't. We have a better model which we learn in High School. Sraffa's extension was horseshit. That's why we don't teach it.  

But this is also the way to lose sight of many crucial influences. An economic analyst ultimately has to juggle many balls, even if a little clumsily, rather than giving a superb display of virtuosity with one little ball.

The opposite is the case. Football stars or Basketball stars make a lot of money. Jugglers don't. Why? Competitive team sports are like life. If we don't act collectively to do smart things then others eat our lunch. Juggling may pass the time, but it doesn't really raise productivity.  

On Rationality and Adam Smith You have given lots of thought to the rationality assumption in neoclassical economics. Do you agree that it is a fundamental assumption?

It isn't. Only if regret-minimization under Knightian Uncertainty converges, as a limit case, to Arrow Debreu Expected Utility maximization, does the assumption hold. 

Yes, I think it is a fundamental assumption. Not only to neoclassical economics, but also to most other schools of modern economics. Of course, the standard "rationality" assumption is defined in two quite different ways. It is sometimes defined in terms of the pursuit of self-interest, sometimes as consistent behavior.

It is defined as the consistent pursuit of self-interest. This holds provided preferences don't change over the period in question. This is spelled out in the relevant models.  

The defenders of the former approach seem to refer frequently to remarks by Adam Smith on the so-called economic man,

But those 'defenders' don't matter at all.  

but then one overlooks much of Adam Smith's other writings including The Theory of Moral Sentiment, and, indeed, a good part of The Wealth of Nations, where Smith took a broader view of human motivation in society and did not see self-interest pursuit as uniquely rational.

It would be foolish to waste any time reading a rabid racist who died a long time ago.  

The other definition is based on the idea—which is a rather modern idea—that rationality is just being internally consistent in choices.

Provided your preferences don't change. If you prefer to spend today at the beach, you will choose to wear swimming trunks. If you prefer to go into the office, you will wear a suit. It would be inconsistent for you to put on a suit so as to swim in the sea or put on a bikini so as to attend a meeting with your Banker.  

In this view, you are rational if you choose consistently, even if you happen to consistently choose the opposite of what you want and value!

This is because choice is not the same as preference. This is obvious. I prefer lying on the beach to going to work. I choose going to work because I am afraid of what will happen to me if I don't have a paying job.  

If the pattern of consistency in choice has certain properties, which we may call binary properties of choice, then we can represent a person's choice function by a binary relation;

Nonsense! A choice function is a binary relation between the set of available options and the set of chosen outcomes. This exists whether or not there is any pattern of consistency.  

with a stronger set of assumptions, the binary relation would be an ordering relation; with somewhat stronger assumptions, the ordering relation would have a numerical representation; and with an even stronger assumption the ordering relation with a numerical representation would have cardinal properties.

A person could choose to order options in more and more rigorous ways without ever making a choice. Instead of just ascribing a cardinal utility function to everything on Ebay, I could go further and put in a complex valued function capturing the type of fart I'd like to emit while buying it. 

However, nobody actually bothers to do any such ordering. Generally, we just look at the next best alternative- the opportunity cost- or adopt a simple 'stopping rule' e.g. in the 'Secretary problem' where we interview three applicants and then hire the first who is at least as good.  

The mathematical structure is neat and useful, and can certainly be used to simplify economic analysis in many contexts, when so-called rational behavior is assumed.

But it is even more rational to just do what smart peeps are doing and not bother with any fucking 'mathematical analysis'.  

But I am very skeptical of the adequacy of both of these ways of seeing rationality. 
That becomes clear in your "Rational Fool's" article (reprinted in Sen, 1982).

Sen was such a cretin that he thought it rational for a guy to entrust an envelope with money in it to some stranger he just met. Sen is an irrational fool.  

But the defenders of the assumption maintain that thinking of economic behavior as the maximization of an objective function under constraints makes sense.

Only absent Knightian uncertainty etc.  

And it does, doesn't it? 
I think it does and it doesn't.

Sen's worthless 'Capabilities' approach is still Arrow-Debreu type nonsense. If perfect information and frictionless futures markets for everything existed- why would we need language or price signals? If there is no Knightian Uncertainty, we couldn't have evolved by natural selection.  

One point to be clear about is that maximizing behavior does not entail maximization specifically of self-interest.

But it is specifically of whatever the self is interested in. But only provided there is no Knightian Uncertainty. If there is, regret minimization is the way to go.  

The objective function can include other goals and commitments. I believe it is true that self-interest must be a major motivation among the various motives that we have. No one would deny that and, of course, Adam Smith rightly saw it as such. As he put it, prudence,

Why be prudent unless the future is unknowable? But, in that case, why maximize utility rather than minimize regret?  

of which the intelligent pursuit of self-interest would be a part, is the quality which is most helpful to the individual, even though humanity, justice, generosity, and public spirit, he argued—and I think I am now using his words—are the qualities most useful to others.

and thus to oneself. Indeed, there is a reward for having a good reputation in this respect. 

I think it is important to see precisely what role we give to this part of our motivation, that is, to the prudential pursuit of self-interest. If it is given a role which excludes all other motivation and all other modes of behavior, we get a modelling of human beings that can scarcely accommodate rationality in general.

Nonsense! Accountants are plenty rational. Prudence is their watchword. Don't book profits till they are realized. 

If people have other goals and motivations, why should they be compelled by economic theory to pursue self-interest?

They aren't. Whatever their goals and motivations might be those are things their selves are interested in. Thus that is their self-interest. Why did nobody explain this to Amartya in his first year at College? 

People may truly want to promote causes which are not identical with their own welfare, and which they don't perceive as their own self-interest.

But they won't deny that is what interests them.  

There is no reason why rational human beings should not pursue those other causes. For example, if we wish to promote what we see as the interest of some group, such as a family, a community, a class, a political party, a social group, even at some personal sacrifice, what's irrational about intelligently pursuing that goal?

What's more, the reward for doing so is often greater- if, like Amartya, you are a good self-publicist.  

Furthermore, it is not clear that the relentless pursuit of self-interest is a good description of how people actually behave. There are these two distinct problems: What is rational behavior, and how do people actually behave?

No. There is only one problem. Academics have shit for brains and go on recycling their worthless lecture notes till they drop dead.  

There is, of course, also Milton Friedman's question, namely, what behavioral assumption leads to the best predictions?

None. The best predictions just look at the Time Series and simulate what will happen next.  

This is an important issue, even though Friedman takes too narrow a view of the nature of economics in concentrating only on this predictive efficiency. Prediction is only one of the purposes of economic theory—there are many other purposes, evaluation is one of them, description is another.

There is only one purpose of economic theory- viz. to create jobs for cretins and credentials for aspiring sycophants who are prepared to do very very boring shit so as to make a little money.  

But even within the predictive perspective, there is little evidence that assuming self-interest maximization would yield the best predictions. There is not much to indicate that you get great predictive success by assuming that people relentlessly maximize profits and nothing else.

 But there is a lot of evidence that people who make their money by making successful predictions about a particular market do a lot better than stupid Professor endlessly recycling their lecture notes.  

If you have to explain economic success in countries such as Japan, which have strong social norms of behavior, it is not particularly helpful for making predictions to ignore those norms and assume that people simply maximize profits. It is inefficient for prediction, in addition to being a bad description. 

Maximizing profit is a norm. So is committing hara kiri at the drop of a hat. People who need to make predictions about what Japan will do so as to maximize their profit will find a good enough way to do so. This won't involve talking to Sen- because he has shit for brains.  

We could incorporate these other motives to which you are referring in the objective functions.

No we shouldn't- unless that makes us more money.  Don't gas on about objective functions unless that is what you are getting paid to do. Also don't listen to guys who do gas on in this manner unless they are making mega-bucks. 

Yes, up to a point, but there are problems. Let me clarify by making some distinctions.

Sen's distinctions are always fucked in the head. 

There are three separate but interrelated ways in which we could be self-interested or self-promoting. (He hesitates.)

We will soon see that this is utterly false. There is only one way in which selves are interested in stuff.  

I'm not sure if either of these terms are right. Well, the term that Adam Smith used—self-love—may be the best.

Nope. Masturbating while gazing deeply into a mirror is self-love.  

There are three components of self-love, related respectively to the composition of personal welfare, personal goals, and personal choices.

Nonsense! Self-regard has no components. It concerns only your welfare. It does not concern your 'personal goals' or 'personal choices' or 'personal attitudes' or preferences which, at best, are merely strategies to promote your welfare and, at worst, are stupid shit you heard others talk about and then thought you yourself ought to repeat. Thus, I may say it is my 'personal goal' to achieve Nirvana while earning mega-bucks as a Beyonce impersonator but I don't actually endanger my welfare by taking any costly steps in that direction.  

The first component is your welfare being self-centered.

No. Your welfare is your welfare. It may be 'self-centered' if you are a Johnny-no-mates but in that case your life is pretty fucking miserable.  

In this case your welfare, as you see it, depends only on things you yourself consume and possess: there are no externalities in the individual welfare function—your well-being does not go down if others suffer.

Sen does not understand what the word 'externality' means. A cost or benefit received from a market transaction by people not a party to that transaction is an externality. It is not the case that your suffering coz your mummy got hit by a truck represents an externality.  This is because your mummy didn't pay the truck to run over her. 

There is no jealousy of other people either.

Nonsense! Self-regard does not preclude envy. Indeed, such envy motivates mimetic effects of a self-regarding kind. Sen is simply parading his ignorance of Economics here.  

This component of self-love we could call self-centered welfare.

There is no difference between the two. One is not the component of the other.  

It constrains the composition of individual welfare,

No it doesn't. The self may want its mummy to thrive coz mummy packs its lunch and gives it lots of kisses. Sen is just pulling this shit out of his arse. He doesn't get that he is talking about an Arrow-Debreu world with future markets for everything- including mummy packing your lunch and giving you plenty kisses. In that world, your utility function does indeed cover the case where the truck hits your mummy. Essentially, you and others who love your mummy would pay the driver not to hit mummy. Arrow-Debreu is an artificial world with perfect information and zero transaction costs. Sen's own capabilities shit is based on it.  

but does not itself say anything on what your goals may be,

They are irrelevant. Only actions count- not 'goals' or 'intentions' or 'ideology' or the fact that you are wearing a T-shirt which says 'Fuck Capitalism!' . 

or what your choices may depend on.

Choices do matter. What matters is whether they are coerced or are free. True, you may choose to suck off a gangster coz otherwise your welfare will decrease by reason of having had your head kicked in, but that is not a free choice. Speaking generally, you'd be happier if the gangster was locked up far away from you.  

The second component concerns the content of your goals.

If changing the contents of your goals could make you happier, then just do so already. It costs nothing. I've become much happier now that I've set myself the goal of becoming the world's leading Beyonce impersonator by the time I reach the age of 120.  

This component of self-love entails that you are only interested in promoting your own welfare, no matter what that welfare depends on.

Nonsense! It entails nothing at all. It is not the case that my goals have led to any change in my behavior or my welfare.  

Whether or not that welfare depends only on what you yourself consume is a separate question, namely the question that defines the first component.

Only in an Arrow Debreu world where there are markets for everything- including comets predicted to come crashing into the earth in the year 2033. 

This second component—self-welfare goal—only speaks about your exclusive goal being the pursuit of your own welfare.

No. It speaks about nothing at all. It is mere puffery. Saying 'my goal is to live my best life' doesn't change shit. Nor does saying 'my goal is to have a fucking miserable year'.  

The third component of self-love—this is the hardest one to articulate—is concerned with your choices.

Self-regard is expressed by choices but choices are not a component of self-regard in the same way that saying 'Sen is a cretin' is an expression of intelligence even though no cretinous Sen is a component of intelligence.  

Self-goal choice requires that your choices be based on using all the instruments in your control to pursue your own goals, paying no attention to the goals of others with whom you live, except insofar as their goods influence your own goals.

Nonsense! You can pick any goal you like for yourself. It costs nothing. I suppose Sen could say 'your true goal is not the one you picked. It is the one I ascribe to you.' But you can reply 'you eat dog turds. You may think you are eating luchi but I ascribe to that luchi the property of actually being a dog turd.' 

Self-goal choice does not constrain your goals,

So your goals don't constrain your goals. This means goals are a 'free good'. They have no opportunity cost. I can have the goal of being a Beyonce impersonator as well as that of becoming a walrus. But this means goals don't matter in the slightest- at least for economics. 

or the determinants of your welfare

in which case, why have goals? The fact is I feel good about my goal of being a walrus Beyonce impersonator. It makes me feel warm and gooey inside.  

—only your choices are completely tied to your own goals in every act of your choice.

But choices are never tied to goals. We all know this. Our goal is to lose weight. Our choice is to pig out at the all-you-can-eat buffet. Fuck is wrong with Sen?  

You can have any one of these three—self-centered welfare, self-welfare goal, self-goal choice—or any combination.

You can have none of them in the real world. Your welfare depends on your not getting run over by a truck. Goals are meaningless. Choices are not uncoerced. There is no entailment property tying these three things together.  

This is the structure that I have tried to spell out in a paper called "Goals, Commitment and Identity."

It is shit.  

When you have all three, you have an extreme form of self-love.

No. You have nonsense. The fact is, you can refuse to pay for mummy's hospital treatment- indeed, you may not even buy her some flowers- while telling everybody your goal is to eliminate poverty and end global warming.  

This extreme form is, of course, standardly assumed in mainstream economic theory, for example, in standard general equilibrium theory.

Nobody assumes this stupid shit- save Sen.  

You can drop the first two by changing the objective function by, say, bringing in sympathy for others—others' miseries and happiness may influence your own welfare—and commitment to goals other than the pursuit of your own well-being—in that case, your goals will not depend only on your own well-being.

In which case, there is no 'objective function' because neither the domain or  the co-domain are sets. They are inchoate.  

But self-goal choice cannot be dealt with in this way. A violation of self-goal choice will require acting against, in an immediate sense, your own objective function—whatever it is—perhaps in recognition of other people's goals and aims, or perhaps in pursuit of some socially useful strategy that promotes the goal of all. 

If there is such a recognition then there was no 'objective function' because the domain and codomain were inchoate- they were not sets. You can't act against something which does not exist. Thus no 'violation' is possible.  

So are you telling me that at times we make choices that have consequences that conflict with our own goals?

I may say 'Sen's goal is to eat dog turds. You say you once saw him eat something which was not a dog turd. This is explained by the fact that Sen sometimes makes choices that conflict with his own true goals- which only I can properly ascribe'. 

It is fucking obvious that we make bad choices all the time because our best laid plans gang aft agley.  

Well, that can be so, but not necessarily. Even when self-goal choice is violated, we must distinguish between different reasons for rejecting self-goal choice.

What Sen says we must do is the equivalent of eating dog turds.  

There may be good ethical reasons for being non-consequentialist as has been powerfully argued by Bernard Williams and others.

Knightian Uncertainty is a good enough reason. But Sen won't accept this feature of the real world.  

A person will let go of an opportunity to promote his or her goals because of some ethical values not well-reflected in the objective function.

In which case there is no function because the domain and codomain aren't sets.  

This is an interesting subject and I have indeed tried to examine the argument for such a non-consequential ethical position

by eating dog turds 

But there is a different—an indirectly consequential—argument with which non-consequential ethics must not be confused.

We will soon see there it is the same fucking argument. Sen is the only confused person here. 

The reason for departure from self-goal choice can be, in this second view, entirely consequential, but involving more complex instrumental reasoning, including the use of social rules.

In which case it is deontological. Social rules are duties. Consequentialism is out of the window. 'Complex instrumental reasoning' has taken you out of one camp and put you in the other.  

If the pursuit by each person of his respective goals leads to the underfulfillment of the goals of all,

Which must always be the case, unless 'goals' just means 'being happy with whatever we actually get'. 

then there clearly is a consequentialist argument for departing from such act-by-act pursuit of each person's own goal in every isolated act of choice.

No. Only if individual rationality is leading to collective irrationality would it be sensible to adopt 'social rules'. This is Prisoner's dilemma type stuff. The problem is that there has to be costly enforcement or a costly signal based separating equilibrium. In other words, consequentialism underpins the seemingly 'denontological' solution.  

This too is of course an old problem, and relates to the case for so-called rule utilitarianism,

but rule utilitarianism is empty because there can be a halachah vein morin kein type rule forbidding its own application.  

even though what is involved here is the instrumental use of rules, rather than evaluations,

but evaluations are themselves an instrumental use of rules- there is no genuine difference here 

based on utilities as such, which is a very special case.

either utilities are 'evaluated' in a 'law-like' manner- i.e. there is a set of rules which cranks them out- or else they are just stuff you pull out of your arse. That's not a special case. It is the nature of the beast. 

The general point is that the rejection of self-goal choice—that is refraining from taking particular action that would have promoted one's own goal better—can be based on complex consequential reasoning in terms of social interdependence, and this must not be confused with rejecting self-goal choice on the basis of non-consequentialist ethics.

Why not? What great harm is done when we say 'sociopaths find they have to behave in an ethical manner if they want to live long enough to behave at all?' Sen got confused by the term 'complex consequentialist reasoning' because he didn't get that the terminus of such reasoning might be deontological. A hedge fund manager might ask a mathematician whether he should buy put options or call options on a particular stock. The mathematician may submit relevant time-series data to very complex mathematical reasoning. But, at the end of the day, the result he arrives at is not itself mathematical. It is purely commercial.

This is not to say that any deontological system of ethics- however prescriptive it might consider itself to be- neglects what Mill called 'punishability'. There is nothing stopping those who believe in duty, not utility, from using mechanism design just as well- or indeed much more punitively. 

While the analysis of game forms makes it easier to see the nature of the problem, it is of course a question that has been discussed in general terms for a long time, for example by Adam Smith and Immanuel Kant.

Not really. They were employed to gas on in a high minded manner. Meanwhile guys in the Law faculty drooled over all the cool cruel and unusual punishments listed in the Penal Code. 

How do you make the connection with Adam Smith?

Everything reminds Sen of Smith- because Sen did not understand Smith being wholly ignorant of Smith's historical circumstances.  

From Adam Smith we get a clear analysis of the possibility that when goals of different people are partly congruent and partly conflicting, then the pursuit of individual goals in isolation may not be a sensible thing to do for people living in a society.

If people live in a society they can't 'pursue goals in isolation'. Does Sen really not understand that? Smith held that we only acquire goals by mimicking others who have goals. Conflict itself is mimetic. We mirror our adversary and thus begin to fight and get better at fighting. Smith could not conceive of people acquiring goals on their own. Mimicking conflict might cause one to acquire the goals which motivate a smart fighter. But, equally one might just brawl for the sake of brawling. Smith was all about choosing better mimetic targets- get good at fighting but make sure fighting makes you richer. That's how Nations build up wealth. The Scots were no laggards when it came to love of battle. They just needed to make more by it. This meant cultivating useful industries so that commerce could win you greater and greater riches from markets won for you by the sword. 

This leads to the subject of behavior according to social norms, which Adam Smith particularly emphasized.

Smith said these norms are only learnt from experience. They are 'mimetic'. Sen refers to the following passage of Smith's

 When we read in history or romance, the account of actions either of generosity or of baseness, the admiration which we conceive for the one, and the contempt which we feel for the other, neither of them arise from reflecting that there are certain general rules which declare all actions of the one kind admirable, and all actions of the other contemptible. 

Smith is rejecting 'synderesis'- an inborn, 'natural', voice of conscience or Platonic anamnesis of good and evil or hardwired Lockean moral archetypes. He is adopting something close to Hume's associationalism but in a more reflective and 'common sense' manner.

Those general rules, on the contrary, are all formed from the experience we have had of the effects which actions of all different kinds naturally produce upon us.

This empiricism should be qualified by the observation that what we will experience depends on who we imitate. 

An amiable action, a respectable action, an horrid action, are all of them actions which naturally excite for the person who performs them, the love, the respect, or the horror of the spectator. The general rules which determine what actions are, and what are not, the objects of each of those sentiments, can be formed no other way than by observing what actions actually and in fact excite them.

Thus, if goals are related to actions, it must be the case that experience of Society set those goals since 'only by observing actions' do people associate with them with outcomes of a desirable type- e.g. getting love, respect, obedience etc. 

When these general rules, indeed, have been formed, when they are universally acknowledged and established, by the concurring sentiments of mankind, we frequently appeal to them as to the standards of judgment, in debating concerning the degree of praise or blame that is due to certain actions of a complicated and dubious nature. They are upon these occasions commonly cited as the ultimate foundations of what is just and unjust in human conduct; and this circumstance seems to have misled several very eminent authors, to draw up their systems in such a manner, as if they had supposed that the original judgments of mankind with regard to right and wrong, were formed like the decisions of a court of judicatory, by considering first the general rule, and then, secondly, whether the particular action under consideration fell properly within its comprehension.

Sen despite his professed love for Smith, holds precisely the opposite view. He thinks human beings have a tropism towards Justice of an abstract type similar to their tropism towards food or shelter. 

Those general rules of conduct, when they have been fixed in our mind by habitual reflection, are of great use in correcting the misrepresentations of self-love concerning what is fit and proper to be done in our particular situation.

In other words, if we are able to 'correct the misrepresentations of self-love' it is not thanks to any process of abstract reasoning. It just so happens that we have had enough experience to arrive at that happy juncture. But, we may equally say, we can act in our habitual manner without fear of doing something unfitting or improper. 

 The man of furious resentment, if he was to listen to the dictates of that passion, would perhaps regard the death of his enemy, as but a small compensation for the wrong, he imagines, he has received; which, however, may be no more than a very slight provocation. But his observations upon the conduct of others, have taught him how horrible all such sanguinary revenges appear. 

This was scarcely the case in medieval Scotland. Slicing and dicing your enemy was considered hella cool. Smith's audience, however, understood that if the Scots wanted to rise in prosperity like some in the South, then they'd better lay off the Macbeth stuff and act more like lawyers or bankers. 

Unless his education has been very singular, he has laid it down to himself as an inviolable rule, to abstain from them upon all occasions.

Because guys who get Professorships at Universities or else senior appointments with the Treasury are better known for their eloquence and erudition than their deftness with a claymore or dirk. 

 This rule preserves its authority with him, and renders him incapable of being guilty of such a violence. Yet the fury of his own temper may be such, that had this been the first time in which he considered such an action, he would undoubtedly have determined it to be quite just and proper, and what every impartial spectator would approve of. 

Which would have been the case in a ruder age when a University education opened few doors other than into the Church. 

But that reverence for the rule which past experience has impressed upon him, checks the impetuosity of his passion, and helps him to correct the too partial views which self-love might otherwise suggest, of what was proper to be done in his situation.

Thus the proud Highlander, quick to avenge an insult, would grow genteel and politic if transplanted onto a University campus after which he could accumulate wealth for himself and his people by doing deals and only fighting when the likely profit justified the risk. 

Here... I should find this passage from Smith—I am so keen on that (he looks up the passage), "These general rules of conduct, when they have been fixed in our mind by habitual reflection, are of great use in correcting misrepresentations of self-love concerning what is fit and proper to be done in our particular situation."

Being ignorant of Christianity, Sen does not understand that Smith is substituting the old notion of 'synderesis'- i.e. an inborn, natural, inclination to the good- with a theory of an 'impartial spectator' which, by mimesis- i.e. by imitating smart people around us- we introject into our own psyche to govern our behavior. One reason Smith wanted to replace 'synderesis' was because progress had become rapid and things like slavery or bonded labor in the collieries and salt-works which previously had been considered 'natural', were clearly no longer anything of the sort. The problem with synderesis is that it hadn't stopped people burning witches and banning 'usury' and so forth. That's why Smith needed a smart 'impartial observer'- a kind of internalized genteel man of business who was getting rich in money and reputation- to take the place of the still small voice of conscience. Smith is saying, 'move about in society and attach yourself to the people who are rising in business and public esteem. Imitate them. Try to think as they do. Habitually reflect on why they did this or said this under such and such circumstances. They themselves will speak of the rules of conduct that they observe- because by making this known their behavior becomes more predictable and they themselves more serviceable to those with wealth and power. Find out those rules- spoken or unspoken. Think deeply on your own emotions and motives and learn to control them so that you too present a genteel and estimable face to the world. By following the right rules, you make yourself reputable and place yourself advantageously to thrive in business and rise in society.'

All this is plain enough. Not to Sen. Let us now look at the dog's breakfast he makes of the canny Scotsman's sound common sense.  

Smith is trying here to capture the idea that in a society there are certain things that are fit and proper to be done.

No. In every society there are dirty jobs- executing people some of whom may be innocent, collecting debts with menaces from debtors some of whom may have been swindled, sucking off soldiers so as to feed your wee bairns, etc, etc. However, if you want to rise in Society and thrive in Business, don't do those dirty jobs. Do the things which are fit and proper. That way the rich will give you some of their money to invest. A wealthy merchant will decide to give you his only daughter's hand in marriage. The Rector of your kirk will speak well of you. The lord lieutenant of your County will invite you to dinner.  

We may base our behaviour not only on our own goals, but also on the goals of others.

Which is why, on seeing a pretty girl, we don't start masturbating. We go up to her and find out what her goals are and then, hopefully, jizz inside her. 

The ultimate objective is that all of us manage to achieve our respective goals better.

This is nobody's 'objective' anymore than it is to breathe in and then breathe out or to obey the law of Gravity.  

It would be nonsense to say, "If you arrive at this kind of compromise, then whatever you seem to be pursuing are in fact your real goals."

But that is type of nonsense Sen specializes in. He is objecting to it here because it would prevent him from making  a distinction between self-interest and self-goals and self-choices and self-sneezing and self-farting and self-pooping yourself when all you meant to do was self-sneeze.  

No, the argument involves socially instrumental use of departures from the immediate pursuit of your own goals in isolation.

Nonsense! It involves understanding that you aint isolated. Your actions are strategic or game theoretic- i.e. outcomes depend on how others react. True, on an open market, you don't need to bother with strategic considerations. But a lot of social interaction isn't on open markets. It is about coalition formation and bargaining and mechanism design.  

You, like others, want to pursue your own goals better, but you know that others strive to pursue their own goals better too, and that you will get in each other's way. In a situation of interdependence like in the Prisoner's Dilemma, each can do more harm to the other person's goals than he can do good to his own by following self-goal choice. In that context the fit and proper rule to be followed, as Adam Smith would put it, may involve practical recognition of other people's goals; it involves norms for living in a society.

Why do cretins like Sen think 'recognition' is so important? Did their parents often mistake them for pieces of furniture? The truth is we easily recognize that other people are not furniture. They are people with goals of their own. If we wish to thrive and rise in Society we must try to get some of them to do the dirty jobs while gaining a fair name and a burgeoning fortune for ourselves by showing we are above that sort of thing. Gentlemen, however, aint so stupid as to think that their own 'norms'- which add value to their own aspirational 'High Society'- should also apply to the collier's brat or the tupenny whore.  

Had Sen read Smollett before he read Smith, he wouldn't make himself so ludicrous. 

It is a matter of social living, social intercourse, of social cooperation, through an acceptance of instrumental use of fit and proper behavior.

Fit and proper for a 'bildungsburgertum' on the make. Smith was the favorite author of the 'beamtenliberalismus' German bureaucrat.  

Indeed, as we know from some experimental results with the Prisoner's Dilemma, people often depart from narrow personal goals, even when they are instructed otherwise.

We also know that those 'experimental results' are not duplicable.  

They often take note of other people's interests and goals in the game.

They often get bored and just tick any fucking box whatsoever.  

While there have been interesting theories by Axelrod, by Kreps and others, to explain this kind of behavior in finitely repeated Prisoner's Dilemma by assuming some ignorance, such as not knowing how many times the game will be repeated, or not knowing what the others really prefer, the explanation may well lie in the Smithian direction, involving a different line of reasoning altogether, rather than any ignorance.

But Smith was writing for a narrow but rising Scottish class. He wasn't writing for the hereditary debt-slaves in the collieries or for the whores needful to service the Army garrison.  Precisely for that reason, Smith wanted the colliers to be free to contract with other employers and for whores to get jobs in pin factories. One reason for this was that Smith, as a Scotsman, was suspicious of the Crown- which might reassign collieries and salt-works to its own favorites. As for the Red Coats, let them bum each other rather than take up the time of Scottish lasses who could supply cheap labor to export industries. 

Can you give some concrete examples? We are not only competitive with each other; much of the time we can be primarily cooperative with each other. If workers in a factory, for example, were to pursue their narrowly perceived interests or goals, I don't think that you would get a very high achievement of productivity.

No doubt, Sen is thinking of factories in Left Front Bengal.  

Many systems flourish precisely because people have codes of conduct; there are certain things to be done. This may, of course, involve loyalties, influencing the goals they wish to promote, and thus violating self-welfare goal, at least to some extent. But in addition, people may not relentlessly pursue their own goals—whatever they are—and may be guided also by the recognition of the strategic interdependences—the social instruments in the form of demands of fit and proper behavior.

But all this has always happened without people worrying their heads about it. Why not speak instead of the necessity to self-regardingly breathe in air and then altruistically expel that air from one's lungs while being mindful of strategic interdependence and the need for social inclusion of sexually disabled minorities within a broader context of combatting Fascism by saying 'boo to Narendra Modi!'.  

This involves practical recognition of the goals of others,

as opposed to what? Theoretically ignoring them as goal-posts because we are playing cricket not football?  

of the enterprise, of their colleagues, and the fact of living in situations of social interdependence.

While breathing air into our lungs and then expelling it in a mindful and compassionate manner such that disabled sexual minorities enjoy social inclusion and saying 'boo to Narendra Modi' in an environmentally sustainable manner.  

Similarly, when an accident has happened or someone is in danger, you are not going to figure out how your helping will affect the promotion of your goals.

Sadly, Bengalis, as a whole, showed precisely this type of behavior during Famines and Floods and so forth. Mujibur Rehman complained that British soldiers had to bury people killed during the hurricane. The local people couldn't be arsed. 

You do certain things immediately because there are certain rules of good behavior which you follow.

In which case you don't do it immediately. You do it only after the salience of the rule has been established. Sadly Bengali good behavior did not extend to burying their own people. Janam Mukherjee writes that during the War, White soldiers had to bury those killed in a Japanese air-raid. A question was asked about this in the Legislative Assembly so the Army put out a fake press release listing Indian personnel who had supposedly participated in this.  

These rules of behavior give people confidence about what they can expect from each other.

No. Actual behavior affects expectations. But this is also true of animals and plants. We expect pussy cat to purr and come sit on our lap. We don't expect a tiger to do so.  

Self-goal choice may be, thus, rejected, but it is a very productive rejection.

If by productive you mean utterly useless- sure.  

Rules of conduct may create a situation which is superior for all.

Observance of rules- maybe. But that is just behavior.  

The norms that emerge in society are sensitive to the issues of social instrumentality.

Social stuff is sensitive to issues of Social stuff- what a great discovery!  

Of course, people need not always behave in this kind of way.

Nor need they be actual people. They may be figments of the imagination of a stupid professor. 

Nor is such behavior the only way of seeing rationality in group contexts.

Nor is rationality an apt description of Sen's behavior- more especially in group contexts of dog turd devouring.  

But it is one way of seeing the demands of reason, involving what may be called social rationality.

Or rationed sociality. The trouble with writing tripe is that any other type of tripe can be substituted for it. Where sociality is rationed to the ultimate degree- the individual is isolated. Where it is wholly unrationed- the individual is indistinguishable in his goals and objectives from the great mass of society. Everything Sen cretinously says about social rationality can be said about rationed sociality.  

(I am currently trying to write a book on the subject.)

If by write, you mean, 'shitting out'- sure.  

The claims to rationality of such practical reasoning are quite strong,

indeed, they are as strong as the claims to sociality of such practical rationing 

but there is no claim that this must be uniquely rational.

or uniquely social. 

There are different ways to reason and different possible conclusions as to what would be rational

or social 

to do under these circumstances. The acceptance of this permissible plurality within the general idea of rationality

or sociality 

is extremely important.

No it isn't. It is nonsense. Rationality is not permissive. It may be plural or inchoate but it can't permit itself to be other than it is. Otherwise, Rationality could be a walrus Beyonce impersonator in the year 2033. But, in that case, it is pointless to speak of Rationality. The thing doesn't solve a coordination problem. It has no pragmatics. It is just a Humpty Dumpty word which means anything you like.  

Are these new elements to be introduced in economics? These are not new elements. For example, Weber, Tawney and others have discussed the role of values in the successful emergence of capitalism.

They were wrong. Plenty of people had 'protestant values' but if they stuck to cultivating the soil- like the Amish- they played zero role in the emergence of capitalism. Meanwhile plenty of Catholics and Jews and even Zoroastrians got rich in commerce and invested in industrial enterprises.  

Ethics was seen as playing a part in it. What kind of value systems made capitalism successful?

Fuck value systems. Capitalism only succeeded if the country didn't get invaded and enslaved. But this was also true of other isms.  

Did simple virtues like honesty, truthfulness, keeping promises and honoring contracts, even if they went somewhat against one's immediate interests, play a part?

Did breathing in and then breathing out play a part?  

These are old questions.

They are stupid. Only pedagogues bother with them.  

We need a broadening of such questions to understand present-day economic problems better.

No. Solving present-day economic problems- like getting vaccines to peeps quickly- involves getting very very rich. Ask Adar Poonawalla.  

It would be difficult to explain the varying successes and failures of different countries in their production achievements without bringing in variations of behavior norms.

Nonsense! Countries succeed when they do smart things. They fail when they do stupid things.  

Countries with certain types of strong social norms, such as Japan,

India's 'social norms' are stronger than Japan's. That's why our system of untouchability is way more complex and graded than theirs. Even when they banned the meat of four legged animals, some hierarchical Hindu dietary restrictions were much stricter- e.g. no root vegetables.

have had considerable advantage.

Japan had never been conquered. Also it had no Muslims. Thus it was more homogenous than India. Bengalis were nationalistic. However the Brahmo Samaj was set up by Hindus who wanted the Brits to rule so as to keep the Muslims in check. Sen was a Brahmo. At Independence his family had to flee their native East Bengal.  

These issues relate to the analyses presented recently by Michio Morishima (1982), Ronald Dore (1983 and 1984) and others, using empirical insights drawn from studies of Japan and other countries.

Back then, people thought Japan would continue to rise. They would soon discover their mistake.   

Many neoclassical economists may become nervous of all this. It sounds interesting but what do we do with it? How can we bring these ethical factors into our models?

How can we have sex with super-models? That's the real question.  

I think that one has to give up first the unitary conception of one ordering standing for, one, your view of your own welfare, two, your goals, and, three, your choices. I think that's absolutely hopeless.

Why? We can certainly ascribe an ordering to our welfare- e.g. I was happier yesterday because I had pizza and went for a swim, both of which I like, whereas today, I am miserable coz I have to go to the dentist and will only be able to eat soup for dinner. Similarly we can say 'my goal of being a Beyonce impersonator' is higher than my goal of being a walrus. The same applies to choices we have made. My choice to wear a thick overcoat was a mistake because it was unseasonably warm outside. I should have checked the weather forecast. On the other hand, my choice of footwear was sound. Proper walking boots really do relieve the strain on one's arches.  

In my paper "Rational Fools," which you kindly referred to, I tried to show the limitations of a model in which economic agents don't know the differences between their welfares, their goals, and the bases of their choices.

Though, they could be identical in which case this does not matter. But even if a guy mistakenly thinks that his goal of becoming a Beyonce impersonator means he must choose gender reassignment surgery- that too self-performed with plastic cutlery- today, it is not the case that any 'limitation of a model' has been shown. We are simply dealing with a nutcase.  The guy was bound to end up in a padded cell sooner or later. 

One way of structuring a departure would be to bring in at least three distinct binary relations

or non-binary relatives of my cat owning neighbor.  The effect would be the same. Why? Because none of my neighbors have non-binary relatives. Binary relations can't exist over what is incohate because no domain or codomain can be specified. Welfare, Goals, Ideals, Aspirations and so forth are 'Tarskian primitives'. They are undefined. What they mean depends on the context. For some specific purpose, we can specify 'extensions' or domains and codomains, but only for that purpose. It is not the case that we have hit upon an 'intensional' concept. All we have is a limited pragmatics of extension without any 'intensionality'. Sen has made a career out of not understanding Logic 101. 

—if you have to do it in binary terms, which you need not of course, but I shan't pursue that issue now. You can ask how these distinct binary relations correspond to each other.

The answer depends on the context- i.e. the specific purpose for which it is useful to do this. 

The correspondences may vary. You may well expect that the correspondence may differ between Mrs. Thatcher's England, with its increasing emphasis on self-interest, and say, Japan with a powerful hold of social norms of behavior.

Thatcher's England saw unprecedented social mobility even for ethnic minorities- that's how come Britain now has a Chancellor and a Home Secretary who are Hindu. Japan remained a 'dual' economy with a bunch of aristocrats running things as the population aged and more and more of the young gave up not just on having babies but having sex. 

Thatcher was about increasing choice for everybody- even those at the bottom. This translated into increased life chances, higher aspirations, and a culture of self-improvement. 

The work that you end up having may possibly be called sociology.

Which is shit. 

But it is economics as economics has been understood until very recently.

It was understood to be gloomy shit. 

Adam Smith, Marx, Mill, even Edgeworth, Wicksell and Marshall regarded these types of inquiries as perfectly legitimate parts of economics.

And everybody else regarded them as shit. Newton is important. Darwin is important. Economists are glorified Accountants- except in the view of Partners of Accountancy firms who look down on them as poor, stupid and smelly.  

It is only in recent years that this kind of an exercise would be placed outside economics. Let me give an example. Mrs. Thatcher is constantly talking about the importance of wealth creation, and she sees this being best done through the pursuit of profits and self-interest.

Which is why Labor had to become Thatcherite to get elected. Wealth is a real thing. Capabilities are bullshit.  

But it is quite possible that Britain's long run problems have something to do with the tremendously narrow motivation structure in British industry.

Instead of trying to make a profit, British industry should be focusing on giving gratuitous rape counselling to the environment.  

In any kind of small, independent enterprise in which cooperation over large groups isn't involved, the British are marvelous.

Very true. The British Raj was a 'small, independent, enterprise'. It didn't involve cooperation over large groups. There were just three members of the ICS, not over a thousand.  

They run the best pubs in the world, with the labor of two or three people, possibly a family.

British gastropubs are now a lot better because they employ more people. A pub where one guy is cleaning glasses while his wife pulls a pint of warm beer is not 'best in the world'. It is shit. 

Good British pubs even in the Eighties had plenty of staff. Indeed, some were already gastropubs avant la lettre. Sen was too stupid to notice. By contrast, there were excellent Italian and Greek restaurants which were family run. This was because their margins were higher. Low margins means high volume which means greater labor intensity- at least at peak times. 

There is not much shirking. A lot of efficient work. In contrast, whenever there is a situation involving large groups and team work, requiring an understanding of, and response to, each other's interests  there seem to be problems.

Not if the boss cracks the whip- which she will do if she owns, or has a stake in, the enterprise. Otherwise there will be 'shrinkage' and slacking off and guys pissing in the soup.  

This is where Japan, or Germany, or even the United States, may still have an advantage,

Because the Brits are too stupid to manage anything bigger than a tiny country pub. That's why the British Empire was actually run by Bengalis.  

and the nature of social norms and of schooling on behavior patterns may be important here.

British schools are completely shit. That is why Brits can't work in a team bigger than three. Nobody taught them to count to four or five or whatever number comes after five. Is it eight? I wouldn't know. I did my A levels in a British school back in the Seventies.  

Do I understand you to say that there is a danger to the promotion of economic self-interest in a society?

Yes. Sen really is that stupid. Bengali buddhijivis are like that only.  

Well, there are problems here which have to be considered. Of course, a society based on self-interest promotion and no cooperative values may well be culturally unattractive, but I am not concerned with that issue here. What is relevant here is that such behavior may also produce an uneconomic society. Ronald Coase is often cited in defense of the efficiency of the market mechanism. But if you think about it, one of the propositions with which he is concerned is that a firm grows until the externalities are internalized.

to minimize transaction costs- i.e. for a profit motive. So, if Sen had really thought about it, he would see that profit maximization can lead to efficiency. Moreover, the true beauty of Coase's theorem is that who owns what doesn't matter that much. There is still a way forward. We don't have to fix everything before we fix anything. Thus Sen-tentious ideas of Justice can be dispensed with as stupid and mischievous.  

If now within the firm you were to pursue only your own interests and goals, if you were trying to get away with as much cheating as you could manage, the productive efficiency of that firm would be deeply problematic.

The firm would go bankrupt. That's why firms have internal auditors and management information systems and so forth.  

The internalized externalities are internal to the firm, but external to individuals working in that firm.

Not if the internal audit function is effective. Getting canned is a market transaction, not an externality. So is getting a bonus because the remuneration committee understands incomplete contract theory better than smart (i.e. not Sen-tentious) economists.  

Within the firm there is still the question as to how the firm itself may be best organized to deal with these interdependences.

It is a question to which observations as to best practice in more successful firms provides a comprehensive answer. If you find your people are slacking or stealing you bring in a guy to crack the whip or a guy to catch the thieves.  

And that's absolutely the central point of wealth creation. There is no guarantee that the promotion of narrow self-interest in the various forms we discussed would be helpful in this.

It plays no part whatsoever. Why does this cretin think that people need to be told to get rich? Why not suggest that the Government is sadly remis in not constantly encouraging people to breathe in and then breathe out. Without proper indoctrination in respiration theory is there not a danger that the populace will suffocate? 

There has been some discussion of this type of problem among industrial economists. But there is not much of it in the general microeconomic literature.

Because you teach that shit to 18 year olds who are trying desperately to get laid as often as possible coz they know that they will have to work their butts off once they get their sheepskin.  

There have been inadequate explorations of its implications for rationality as well as efficiency.

Because such armchair exploration is useless. If you really are smart figure out a better dynamic programming algorithm or something of that sort.  

When you worry about this, people say that it is a very negative complaint. Well, it's negative mainly because the job of constructing alternative models and presenting different formulations of efficiency problems is still very inadequately addressed.

Because it is useless. Only the context matters. There is no general solution to 'efficiency problems'. The adjoint functor is always ideographic, not nomothetic.  

In some ways we are guilty, I guess. There is a lot to be done. I hope very much that economics will move in this direction.

Maybe it did. Nobody noticed.  

Quite a lot of the high-brow economics, which is impressive and helpful in many respects, assumes that the basic problems have been understood in the central case; it accepts the appropriateness of the standard general equilibrium model,

but that model is 'anything goes' because of income and hedging effects! In any case, Arrow Debreu is shit coz of Knightian uncertainty. Black Sholes volatility surfaces too were misleading.  

with everyone pursuing their self-interest, given tastes and technology. You then skillfully introduce imperfect competition, ignorance, uncertainty; you may bring in learning,

That's fine- Hannan Consistency is regret minimizing 

signalling; you can do disequilibrium dynamics. It is assumed that there is no deep problem with the basic story.

No. The 'basic story' is a limit case. The problem with it is that it is not compossible.  

The extensions and variations may, thus, look like consolidating battles, the main war having been won, and the high ground already secured. But the high ground is not secure at all.

Because it does not exist. Nor is there any actual war. This is all merely a metaphor.  

The most basic element of such modelling, namely the motivation of human beings, is not well addressed.

What would be the point of addressing it? Suppose you could motivate me to take up Algebraic Topology rather than Beyonce impersonation. I'd still be completely shit coz I iz as stupid as I iz ugly and talentless.  

Once we try to understand the challenging issue of human motivation, we enter one of the most neglected areas of economics.

Hypnotism. Get people to think they are chickens. That changes their motivation.  

At this point in our conversation we had reached a natural break in British intellectual life, namely afternoon tea.

Because British cretins didn't take a break for a cuppa- right?  

Sen wondered whether we should continue in the Senior Common room where tea is served, but we opted for sherry and water in his own study. Our conversation continued at a length that, if printed in full, would triple the length of this transcript. We talked, among other subjects, about his recent work incorporating the value of freedom of choice in judging living standards and well-being,

which is easily done. How much would you pay for a larger menu of choice? This is something businessmen know a lot about.  

his attempts to redesign measurements of poverty and real national income,

which were useless. You need smart Accountants with plenty of ideographic knowledge for this.  

Arrow's impossibility theorem ("a surprisingly beautiful result"),

it is sheer nonsense. Define a guy who aint a Dictator as a Dictator causes an ex falso quodlibet type explosion of nonsense.  

and his claim that health statistics and other such indicators should be standard part of the informational basis of welfare economics

It was. The problem was that health statistics improve when welfare goes down and vice versa. Why? We are 'junk food monkeys'- i.e. affluence kills us faster than not getting enough to eat.  

("I once weighed nearly 250 children from two villages in West Bengal to check their nutritional status related to income, sex, etc. If anyone asked me what I was doing, I would have said, I was doing welfare economics.").

He wasn't. The sample size was too small.  

Sen's way of talking is characterized not only by a tremendous seriousness and passion for the ideas, but also by modesty. He is almost eager to downplay his own achievements.

e.g. helping Bengal rise out of poverty? No? Helping Britain rise up by getting its people to work in groups larger than three? No. His achievement was to fool economists into thinking he was a philosopher and fooling philosophers  into thinking he was an economist. 

That showed when I asked him how he regards his own work. "Perhaps the work on famines or the measurement of poverty and inequality is more useful," he answered.

It was useless. B.R Sen, as head of the FAO helped tackle the underlying problem- food availability deficit. Amartya pretended the thing was distributional. This was crazy. Malnourishment has fallen most in parts of the globe where Income distribution has worsened most. As for measuring poverty- Sen was barking up the wrong tree. His type of metric would not have distinguished between a Venezuela type situation- short run gains based on fucking up the country long term- and a South Korea type situation.  

"But I got more fun out of social choice theory... I find it very hard to look at my own writing. It's not a pleasurable activity for me. Going back is like opening an old wound." What is an outstanding achievement according to you? "Being outstandingly creative. Take Arrow's impossibility theorem. That is a shattering contribution." 

It shattered the brains of Bengali buddhijivis who, however, would have achieved nothing worthwhile in any case.  

 

Postscript In the Fall of 1987 Sen became a Professor of Economics and Philosophy at Harvard University, a position which was soon changed to that of a University Professor—the Lamont University Professor. During the preparation of the conversation for publication, we discussed a review of his work by A. B. Atkinson in the New York Review of Books. (Sen: "He was extremely kind, perhaps overkind. He does not criticize me much") The review made me wonder about the connection between Sen's critique of the rationality assumption and his work on social choice theory.

The problem was that no non-arbitrary general standard can exist (at least till the end of the time) for interpersonal comparisons of utility or rationality or anything else we might be interested in. This does not mean the thing can't be done for any specific, useful, purpose- e.g awarding damages for a tort, or deciding scales for disability payments- it is just that the thing will have an arbitrary element and there would be no general method of critiquing it. 

Since no one in their right mind would willingly agree to be bound by any 'unrestricted domain' Social Choice mechanism- Social Choice theory is as meaningless as arguments about whether Dracula can beat up Spiderman. Rationality, on the other hand, is usefully studied. But it was John Muth's notion of 'Rational Expectations' and Schellings notion of 'focal points' as solving coordination games and Aumann's notion of public signals promoting better correlated equilibria and John Maynard Smith's notion of uncorrelated asymmetries as dictating 'bourgeois strategies' and so on which represent developments in our understanding of rationality. This is important because there are juristic and professional contexts in which we may have to demonstrate that our actions were prudent, rational and consistent with a higher culpa levis in abstracto duty of care. In this context, having a richer conception of rationality- e.g. regret minimization rather than utility maximization- is the difference between going to jail and getting mega-rich. 

I asked him about this. He preferred to write down his response, because, in his words, "This is an important question about which I probably have no chance to comment in my articles." His response follows. 
There is, in fact, a fairly close connection. One of the main implications of the critique of the so-called rationality assumption used in economic theory is to accept the necessity of having several distinct binary relations representing a person's respective answers to a set of different questions. The answer to the question "Does state x serve your own interests better than state y?" need not be the same as the answer to the question "Which, in your view, is the better state of affairs for the society?"

Sen is trying to get at the notion that people have multiple identities. There is selfish Sen and socially-conscious Sen and spiritual Sen and so forth. But this is nonsense. Asking loaded questions and then answering them yourself enables you to prove that people are cats when they are not Beyonce impersonating walruses. The truth is there is an answer to the question 'what is best for you such that it could also be best for Society?' which sublates the answers previously given. In other words, after Sen cross-examines the witness and congratulates himself on showing that the witness has multiple personalities, opposing counsel shows that the guy is perfectly sane. He has only one personality and only one identity.

It is a fact that my getting a Netflix series to display my Beyonce impersonation would serve my interests. It could also benefit everybody in Society if, while watching my performance, everybody resolves to give up smoking and excessive drinking and farting in crowded lifts. 

There are other questions which, too, require distinct treatment, such as what you will yourself choose, what you ought to choose, and so on.

Nonsense! You are choosing what you ought to choose and whether to beat yourself up about this or not. Meta-preferences are just preferences which may cause you to spend a bit of money on trying to change your preferences.  

In the pioneering work of Kenneth Arrow in establishing the new field of social choice theory, Arrow had represented the attitudes of each person in terms of one "preference ordering," and had sought one "social ordering" as the outcome of the process of aggregation.

But, because 'preference' is a Tarskian primitive, that ordering could never be known 'intensionally'- i.e. you can arbitrarily stipulate 'this is the preference ordering of x'- but there is nothing in it which shows it is truly x's preference ordering.  Equally it could be non-deterministic. (Arrow only stipulated that the social ordering be algorithmic). We  don't know whether the 'natural' preference ordering (only knowable at the end of time) isn't 'entangled' with every other preference ordering such that no 'topological holes' arise. 

This is, in fact, a very useful mathematical format,

it is a joke. Tarski taught Arrow. How fucking stupid are academic economists? What about 'Tarksian primitives are undefined' is so difficult to understand? Why define something Teach told you couldn't be undefined?  

but the interpretation of the aggregation exercise must depend on what the individual preference orderings are supposed to stand for—the person's interests, or ethical judgments, or choice behavior, or what?

 There are people who can look at me and tell my weight, my height, my age, even my gender (male, last time I checked) and sexual orientation (I iz horny for walruses). But nobody could tell my 'interests' because that depends on an uncertain future fitness landscape, or my 'ethical judgments' which depends on the type of jam I get into, or my 'choice behavior' (except maybe my Netflix preferences). The future is radically uncertain but that's why life is plastic and what we want and how we reason changes dramatically and discontinuously over time. 

Back in the Eighteenth Century, there was a brief period when some people- like Hume- thought there could be a 'natural' (non-arbitrary) Religion or Ethics or Ratiocination. This was a pipe dream. Everything changes when circumstances change unless you are no longer actually alive.

Individual preference orderings may exist 'at the end of time' when everything is known because all life is over. However, for any specific purpose- e.g. buying a birthday preference- there is a good enough approximation. 

And similarly, it will depend on the explanation of the social ordering, that is, institutional choice, aggregated judgment, planning decisions, or what?

The explanation is whatever fucking bee the guy gassing on about it has in his bonnet.  It is obvious that the only proper 'social ordering' is one in which my getting elected Pope because of my Beyonce impersonation is on top. Second best is my becoming the Queen. Third best is just the way things are now except maybe moved five inches to the Left and painted orange. 

There is little point constructing 'social orderings' or deciding whether the Taj Mahal would look better next door to the Pyramids as opposed to the leaning tower of cow-shit they were building last time I was in Agra. 

Within the same format, established by Arrow of aggregating an n-tuple of individual binary relations into a social binary relation,

there is no such format. There is merely an impossibility result based on either telling stupid lies about who is or isn't a Dictator or simply assuming N equals NP.  

various different exercises can be accommodated, depending on the interpretation chosen.

No. Nothing useful can be done at all. We may have a preference over which of two wankers we think will jizz less upon us or whether we hate Europe more than we hate each other, but that's about the size of it. On the other hand, lots of the really important Social Choice gets done without any fucking preference aggregation. After 9/11 everybody was like 'fuck 'em all to death!' 

The plausibility of a given set of axioms relating individual and social binary relations will depend on the specific explanations.

No. It is totes implausible that any agent will have knowledge of, let alone preferences over, all social states. This is just mental masturbation. Meanwhile, smart people with some quantitative skills were getting rich off actual social choice- i.e. politics. 

Of course, quite a lot of the mathematical exercise can be performed without particularly worrying about specific interpretations, and indeed a substantial part of my own work in social choice theory, such as introducing cardinality

reintroducing it. Bentham got there first. But he had some sort of physiological theory of a type which may have had some appeal in pre-Darwinian days.  Still, if you have a Great Dictator who knows best, then cardinal utility is the name of the game. 

I suppose one could say how much money you get is the utility Society attributes to you. That's cardinality right there. Perhaps the Chinese 'social credit' system will supplant money. That too would be an example of cardinality. But, getting voted into a prison camp might be less than fun. 

and comparability,

which is easily done. The plaintiff points out that he lost his legs and arms when I drunkenly ran over him. But his suffering is comparable to my intense distress when I found out that the cunt had put a dent in my 12 year old Honda Civic.  

finding conditions for transitivity of majority rule,

why bother? We can't be sure ceteris paribus and that transitivity is violated. Anyway, why not violate it from time to time as part of 'discovery'? As for 'majority rule', fuck that. Most peeps be stoooopid.  

relaxing the requirement of social transitivity or of social binariness, and so on, have been concerned with such general analytical issues

of utter shite. Thus they themselves are shit.  

However, there are other problems in which the exact interpretation is crucial. I have tried to argue that Arrow's own axioms make more sense for aggregating the ethical or political judgments of individuals

Fuck off! One's judgment is improved by talking things over with other blokes and deciding to defer to superior wisdom. Arrow's axioms rule this out.  

than they do for aggregating the interests, or welfares, of different persons, for which an informationally richer structure is essential.

But still woefully insufficient. The thing can only be done, if at all,  'at the end of time'.  

I have also tried to develop some alternative structures, incorporating such things as interpersonal comparisons of welfare,

which are arbitrary and essentially contested 

a person's liberty within his or her protected domains,

isn't really liberty. An animal in a cage may enjoy as much.  

a person's relative deprivation vis-a-vis others.

Is unknowable and easily dissimulated.  

In these exercises, the various distinctions made in the context of the critique of rationality assumptions used in economic theory become specifically relevant.

But they are shit. Shit is just shit. Distinctions about shit uncover no differences unless you deal in shit or it composes a significant portion of your diet.  

The tradition of illegitimately identifying several distinct notions as the same has to be rejected first so as to

talk shite about shite till everybody realizes you are a shithead.  

be able to incorporate sufficient richness in the social choice format.

But that 'richness' will only be accessible, if at all, at the end of time.  

The plurality of the relations sought in a richer structure of rationality is, thus, crucially relevant for the interpretation and use of social choice theory.

This is foolish. We need thinner and lighter 'structures of rationality' because data is getting denser and deeper. Aristotle's categories plugged gaps when facts were sparse. Now we need 'reverse mathematics'- i.e. the most parsimoniously specified axiom system. 

In this sense there is a close connection between these two areas of work.

The close connection is that Sen was taking two types of useless shite and shitting copiously on both because he was a useless fellow who could do nothing else.  

2 comments:

Unknown said...

This is a very brilliant post. I have a few questions. May I contact you via reddit/telegram, discord etc.?

windwheel said...

Certainly. You could email me at polypubs@gmail.com