Weekend Reflections is on again. Please comment on any topic of interest (civilised discussion and no coarse language, please). Feel free to put in contributions more lengthy than for the Monday Message Board or standard comments.
- Comment
- Reblog
-
Subscribe
Subscribed
Already have a WordPress.com account? Log in now.
John Quiggin 
Most of the time the decision theory we use attaches subjective probabilities to events that are uncertain and then evaluates them using expected utilities. What happens if the subjective probabilities themselves influence the outcomes?
As examples:
(i) We know about placebo effects in health – they are surprisingly effective. When people believe some stupid medical treatment will cure their ills it may.
(ii) Also neuroscience suggests religion is an evolutionary stable response relying on placebo effects. When you hold religious views you may be happier and more materially prosperous because religious trust facilitates trade. Thus when people hold daft religious views (my atheist presumptions intrude here) they may be better off.
Suppose beliefs are translated into subjective probabilities. How does this change our view of the world?
My guess is that the effects of such probabilistic misjudgments are to promote social stability and social welfare by promoting (for example) altruism. Religion promotes, at low cost, social cohesion among society’s dullards by avoiding nasty prisoner dilemmas. Those with intelligence can reason philosophically and are unharmed by such foolishness. The positive effects here arise because religious belief has significant placebo effects in promoting well-being. Religion IS the opium of the people – for example, the current New Scientist sees religious belief as having similar effects on the brain to taking a shot of opiates.
Adam Smith’s Moral Sentiments shows that common illusions connected with sympathy effects produce good outcomes. For example it is nonsense to feel compassion for the dead because they are being eaten by bugs. They feel nothing. But having a fear of death is socially sensible – it inhibits self—destruction. Similarly having the illusion that accumulating wealth drives happiness gets the economy going and makes the workers better off since their wages rise and they spend their money on food not trinkets. Finally, having the foolish belief that your values are objective rather than subjective and approximately those of the other guy helps promote cohesion in society. My guess is that illusions about favorable religious and health outcomes promote the social advantage.
Harry
It’s pretty obvious that there’s an evolutionary explanation for altruistic behaviour. Given the prevalence of prisoner’s dilemmas in human interactions, a gene that predisposes us to cooperate rather than pursue the rational dominant strategy, would confer a distinct survival advantage.
But I’m not sure where religion fits in. Does altruism prosper better when it’s supported by a belief structure? Isn’t it enough that we get pleasure from being kind? A preference for sexual intercourse also confers a survival advantage, but generally lust is a sufficient motivation. You don’t need a religious justification to hop into bed – if anything religion is often a brake on lust.
The empirical happiness literature does support a weak connection between happines and religious belief (though this doesn’t accord with my personal experience – I seem to know an awful lot of mixed up and melancholy believers). But again, do happy people have a higher probability of producing children and raising them to child-bearing age?
This is not to say that religious belief might not have some biological basis. I continue to be struck by the apparent fact that susceptibity to superstition is uncorrelated with education. But I suspect it’s more like hair colour, in that natural selection is indifferent between the variations.
I suppose I should read the New Scientist article.
Harry, could you explain why you believe this (if you do)?
(I should state that this picture of how people make decisions accords neither with much of the empirical psychology that I’ve read, nor with my personal experience. But I’m more interested in hearing your reasons than objecting).
Running a small biz one realises that (1) for the most part one is a price takeR rather than a price maker in terms of both buying and selling and (2) that all banks are bastards.
Over the past couple of months we have received three letters from Bestpac WC advising us that various fees charged to us om our accounts are going up. These letters are signed by some bank johnnie rejoicing in the title of “General Manager, Business Financial Services”. Naturally this gutless turkey doesn’t put his contact no. on the letter he sends out with his glorious title but with some effort I did manage to get him on the phone.
When challenged as to was such and such a fee was going up to the amount in the letter all he could say was that was the figure was “in line with what the other banks charge”. He could not say what the amount of increase was and when told admitted that the increase was far in excess of inflation.
When is a political party going to get up and defend us and break up this private enterprise “single desk”?
^^^^
When challenged as to why such and such a fee was going up to the amount in the letter all he could say was that was the figure was “in line with what the other banks charge�. He could not say what the amount of increase was and when told admitted that the increase was far in excess of inflation.
When is a political party going to get up and defend us and break up this private enterprise “single desk� cartel?
Harry —
There are many problems with applying classical decision theory to real-world decisions, including the one you identify in your post. I have listed a few of these problems on my web-pages, here:
http://www.csc.liv.ac.uk/~peter/complex-dec.html
On your specific issue, the fact that assessment of probabilities of outcomes may influence the outcomes is well-known to real-world decision-makers, where the beliefs that people have may greatly influence outcomes. If we believe we are going to win (a military battle, a commercial contest, etc) we throw our all behind it, and thereby increase the likelihood that it will happen. If, in contrast, we believe we are not going to win, we put less effort in and, lo and behold!, lose is what we usually do.
Not only that, but we often have the ability to define the outcome ourselves, after it happens. There are often only subjective definitions of success, as in many business or military ventures.
James the argument runs that most of the time we require evidence for belief but not with respect to religion. So why do so many believe? The New Scientist argument is that religion has the same effects on the brain as endogenous tranq
Peter, optimism biases can work against you too. I think for example the empirical literature on firm takeovers suggests that most don’t work out and turn out to be expensive for the aquiring firm. The reason might be that they are based on the overly optimistic expectations of entrepreneurs. On the other hand as Adam Smith pointed out its probably better generally in business to be optimistic than pessimistic. The example you give on warfare is interesting.
Got cut off at end of para 1 while editing it.
tranquilisers so religion makes evolutionary sense. Smith and others suggest it promoted initial trades rather than violence etc.
Cristin I agree. I am reading some empirical psychology myself and think that standard decision theory doesn’t take you that far in understanding behaviour. Peter’s web page above gives other examples. The current Journal of Economic Perspectives has an article on Adam Smith as a behavioural economist which shows that Smith understood that narrow theories of self-interested behaviour don’t take you far. I think Smith was a perceptive psychologist as well as economist/moral philosopher.
Generally a case for behavioral economics and for studying psychology I think.
Are the cartoons the butterfly’s wings?
I am very intereted in the implications that recent discoveries in psychology for economics and other social sciences.
We know that humans are inherently social animals that evolved to live in groups, usually consisting of 150 or so individuals. We also know that alot of our behaviour is genetically hardwired in. For example we now know that gender behaviour is a result of physical differences between the brains of men and women. And research indicates that homosexual behaviour has a genetic component. This suggests to me that we also have hardwired into us a number of behaviours that make living in a social group more successful but that are contradictary to the interests of a rational individual.
If we are can learn with some certainty what those irrational behaviours are then the model of individuals as rational utility maximising agents can be replaced by models of the individual that incorporate those irrational behaviours. Things such as being irrationally trustful or overly optmtistic (might explain lotteries).
This will depend on advances in psychology. This is encouraging because it seems that pyschology is making these discoveries. Almost every week I read about something new that has been learnt about how our minds work that have implications for social behaviour.
Those offensive cartoons could well be the ‘butterflys wings’.
What good will they do, to be reproduced here?.
None.
Will Australian papers choose NOT to publish the cartoons?
I do hope so.
Urgly Dave: A few more observation for your data bank on bank behaviour:
1. About 6 months ago I received a MasterCard statement with $37and a few cents interest charge for an outstanding amoung of $0.60. The 60 cents was the difference between the total amount of the predecessor statement and the amount I paid. It took a letter to the CEO to get this one cleared up (cost to me, calculated at their charges: at least $250).
2. My current MasterCard bill contains 1 item: Annual Charge $30, payable 20 February 2006. On the same bill there is a note stating :We’ve reduced your annual fee! You’ll be pleased to know that the annual fee on your credit card will be reduced to just $24 (from $30) as of 1 Februry 2006″. I suspect it’ll take another $250 implicit cost to sorted this one out. I’ll start off paying $24 and see what happens next.
It seems to me these large private organisations are worse than the old fashioned bureaucracies. The left-hand doesn’t seem to know what the right-hand is doing in such a way that unless relatively large amounts of resources are spent by the ‘little guys’ the two uncoordinated hands win.
Rationality revisited.
It is falacious to say that individuals who behave in an apparent altruistic manner are ‘irrational’. I have come across this confusion of rationality with a specific assumption about preferences many times. I am not sure about the origin of the confusion. I suspect it stems from undergraduate micro-economic texts, which contain 19th century ‘utility maximisation’ examples.
The notion of rationality does not depend on the arguments in a utility function (eg own consumption bundles versus own consumption bundles and consumption bundles for one or several other individuals). The notion of rationality relates preferences to behaviour.
A utility function is a representation of preferences. A utility function is a derived concept. Preferences is the primitive.
Yes, James, A. Sen has written on the point you raise almost 20 years ago. As I said elswhere, ‘economic rationalism’ has missed out on about 100 years of theoretical research in economics.
“It seems to me these large private organisations are worse than the old fashioned bureaucracies.”
So don’t use them.
Harry Clark or Ernestine Gross or Peter — I have often found myself a bit stuck when trying to delve into reading economics, as I find it irritating to read past the generally crude picture of individual psychology that tends to be assumed. I read on, but can’t help myself thinking it’s all built on such shaky foundations.
You all seem to have more nuanced views. Can anyone suggest any good reading on the fit between economics and psychology (for someone better versed in the latter than former)? How do economists see the relationship between models of normative rationality and how people actually think, judge and decide? Do any economists take any account of the longstanding and fast-expanding psychological literature emphasising the roles of unconscious processes in human thinking and judgement?
Crispin Bennett,
Thank you for your post. First of all I’d like to say that I find the blog site offered by John Quiggin potentially very useful for interdisciplinary questions, which, in one way or another, may well be related to the name of the blog site.
On the fit between economics and psychology:
I don’t think I can even try to offer references for someone better versed in psychology than economics because I know very little about the former. For example, I don’t know what you mean by ‘normative rationality’. The notion of ‘unconscious processes’ is one I’ve been introduced to in a first year psychology course. It makes no sense to me – possibly for the same reason that first year econ. books may not make sense to many readers. More specifically, how would one distinguish between ‘unconscious thought’ and the speed of thinking exceeding the speed of recording the thinking in memory?
To make a start, Kenneth Arrow’s writing may be helpful. Professor Arrow has written some papers on psychology and economics and he is a distinguished mathematical economist. I’ll list his web-site, which summarises his work better than what I could do. Perhaps some titles sound interesting. http://cepa.newschool.edu/het/profiles/arrow.htm.
PS: I’ve read through some of my posts. Too many typos – I think I need eye glasses.
Ernestine, I should add that I really did mean “better versed” in a strictly relative sense. I’m no expert in either.
I’m not sure whether you meant that rhetorically or not: is there something about the distinction between actual cognitive processes and normative models of rationality one you disagree with? Or am I reading too much into your comment? All I mean by normative rationality is models of rationality that are intended to be prescriptive or definitive in some way of how ‘rational thinking’ should proceed, as opposed to empirical studies of how people think, and what’s happening in their brains when they do.
Our ‘thinking’ is a result of various processes happening in the brain. Some of these we are aware of, some we are not. Some of the ones we are not aware of contribute to our judgements and decisions. “What we are aware of” may well amount to no more than that particular ensembles of neural activation in some manner ‘win’ over others. But those others may still take part in cognitive processing. This picture seems to be the strongest contender for the un/conscious distinction coming from current neuroscience (eg. see this survey).
That’s a good question. One answer is just to point out that no model (that I know of) in current psychology or neuroscience suggests that all the processes causally relevant to cognitive functions are somehow ‘transparent’ to consciousness. Much apparent ‘reasoning’ turns out to be confabulation constructed after the fact. Even the simplest cognitive act, that of reporting on our perceptions, is mired in influences we’re not aware of.
But you’re probably looking for an experimental or observational procedure that can tell whether a subject’s failure to report factors clearly-salient to their judgement is a result of their unconsciousness of those factors, as opposed to a mere failure to track their own thinking in memory? I’m not sure if the experimental procedures successfully make that distinction exactly. There’s certainly a lot of literature on how ‘implicit memory’ affects judgement, and in many cases this involves memories that subjects are not able to bring into consciousness at all (ie. no amount of cuing can get them to ‘remember’ memories that are clearly involved in their cognitive processing).
Ernestine — oh, yes, thanks for the Kenneth Arrow suggestion. Have bookmarked his site for a look.
While the West talks about human rights and exporting “democracy”, it seems the rules are a little elastic:
http://antonyloewenstein.blogspot.com/2006/02/we-are-no-better.html
Peter,
Thanks for posting your points on decision theory. It seems to me some of the difficulties you mention arise from the apparent need of some business people to be told what to do while pretending to be decision makers (ie they want specified processes, akin to accounting rules). Others are due to the need to coordinate information among stake holders and there is no ‘manager’ who does this. The size of the coordination problem is at times substantial and complex in a deeper sense of the word. In my experience, complexities (chaos) can arise from what you call ‘democratic approach’ in large organisations. Lack of commitment is only one possible cause for difficulties. Another one is that none of the members of the organisation who is supposed to execute the collective decisions has any idea as to what they are supposed to do because the outcome is something which is neither a boat, because it has wheels, but it is not a car either because it has a sail, etc. …. I would not call these processes ‘democratic’ but rather disjointed decision making processes. Surely, the examples of difficult decision problems you have mentioned do not involve moral or social values, which may be resolved via democratic processes, but rather involve technology, information asymmetries, conflict of interest, and market conditons.
To the best of my knowledge, there is nothing in decision theory which prevents the aggregation of subjective probability values across a set of decision makers (ie to mimick the aggregation of expectations in models of sharemarket economies).
Incidentally, the decision problem of a firm with multiple owners (and without the privileges given by law to corporations) is akin to a public choice problem (Dreze equilibrium). An Australian economist, Frank Milne, previously at the ANU, has contributed to this literature.
I have a question which I often wanted to ask but never dared to, namely, is there any evidence that advertising works in the way it is assumed to work?
Indulge me this.
Andrew Bolt and his recent Iraq trip. Questions abound. So does the abuse:
http://antonyloewenstein.blogspot.com/2006/02/bolt-from-blue.html
“is there any evidence that advertising works in the way it is assumed to work? ”
MacDonald’s in countries with tasty nice cheap food that is served fast. People still eat it, believe it or not.
“More specifically, how would one distinguish between ‘unconscious thought’ and the speed of thinking exceeding the speed of recording the thinking in memory? ”
unconscious thought = things you are not consciously aware of. Like most of the stuff you do walking down the street. It can refer to basic stuff (like breathing, which can become conscious if you think about it), or non-basic stuff, like things in your mental lexicon being actived even though you don’t notice. conscious thought = things you are aware and could tell me about if you wanted (mental reflections)
Speed of thinking refers to both conscious and unconcious processing. You can measure either.
There a gazillions of experiments looking at both. If you want to look at speed of unconcious processing, you can do thinkgs like present things quickly or with a mask so you don’t consciously see them (i.e., can’t tell me what they are), and then look at their effect on something else (like semantic priming — seeing dog and looking at your response time to Cat, the Stroop task etc.). You can present them for different durations to get an idea of the time course. Alternatively, you can present them in people’s parafovea (where you don’t consiously recognize them — i.e., you are unaware of them), and use physiological measures like ERPs, and get an idea too. You see different patterns of electrical activation for things that people are not aware of.
Crispin Bennett,
1. No, I wasn’t asking a rhetorical question. I wanted to know what the term ‘normative rationality’ means. I now don’t know whether you have given me your notion of ‘normative rationality’ or whether there is such a term in mainstream psychology (or the fringe, if you like).
2. Given your description of the meaning of the term ‘normative rationality’, namely “models of rationality that are intended to be prescriptive or definitive in some way of how rational thinking should proceed� are inconsistent with the notion of rationality in economic theory concerned with non-dictatorial resource allocation systems. In a sense, I don’t understand you opposing your ‘normative rationality models’ with neuroscience because the former make no sense to me.
3. Yes, ‘thinking’ is a word associated with brain activity – although sometimes it seems to be helpful to recall Goethe’s “Denke nie gedacht zu haben denn das denken der Gedanken ist gedankenloses denken� (roughly translated: Never think you have thought because the thinking of thoughts is thoughtless thinking.)
4. My question, ‘how would one distinguish between ‘unconscious thought’ and the speed of thinking exceeding the speed of recording the thinking in memory?’ remains unanswered. (Your description of an experimental design seems to rest on a ‘normative thought model’.)
5. It seems to me somewhere in all this there is a line between people who are not permanent clients of psychologists – of interest to economics – and those who are.
6. So far I have not received anything which would support the suggestion that there is something crucial missing in the foundation of economic theories of non-dictatorial resource allocations systems which take preferences of individuals as axiomatically given. There is no claim that such models represent ‘a theory of everything’.
Conrad,
1. MacDonald: Yes, I’ve observed that too with children. Those in my sample grew out of it. It seemed as if these children just wanted to experience what ‘the world’ has to offer – they didn’t wish to miss out. And, in 1995 I’ve seen former East Germans queuing (time) in front of a MacDonald shop in Prague while around the corner one could have exquisite Tschek cuisine and silver service for very little more money. But, how much of the effect of advertising is merely due to exposure, how much is due to the sampling (compare image with direct experience – a form of learning) and how much is ‘psychology’?
2. As for the intersection of economic theory concerned with non-dictatorial resource allocation and psychology, the MacDonald examples would, in my opinion, not constitute examples of ‘irrational behaviour’.
3. “Speed of thinking refers to both conscious and unconscious processing. You can measure either� . Thank you. Which one is faster – on average? What is the dispersion among people? Are there people who are so fast in conscious processing and so disciplined that there is no discernable difference between conscious and unconscious processing?
4. Is there anything known about the ‘unconscious processing’ and learning? It seems to me some fleeting impression or ‘undigested’ bit of information can grow (‘pressure cooker’ we used to call it) into a question that requires thought.
5. ERP: In the more recent management jargon, ERP stands for ‘enterprise resource planning’ (previously ‘planning’ or ‘budgeting’ or ‘state contingent planning’ or ‘financial planning’). I trust this is not what you mean.
6. Now, what happens when a I drop, say, a glass (it breaks) and I had not (consciously) planned to break the glass? Would you say that there is some ‘unconscious thinking’ or is it simply an ‘error’ – a malfunctioning of either my brain or my limbs or both?
Ernestine Gross, Earlier you wrote
“Rationality revisited.
It is falacious to say that individuals who behave in an apparent altruistic manner are ‘irrational’. I have come across this confusion of rationality with a specific assumption about preferences many times. I am not sure about the origin of the confusion. I suspect it stems from undergraduate micro-economic texts, which contain 19th century ‘utility maximisation’ examples.
The notion of rationality does not depend on the arguments in a utility function (eg own consumption bundles versus own consumption bundles and consumption bundles for one or several other individuals). The notion of rationality relates preferences to behaviour. ”
When I first read this it didn’t seem right and now I see why. This is equivalent to the argument that the man who drinks crankcase oil and kills himself is rational since he must like crankcase oil (the example is due to Robert Frank). But this is a totally empty viewpoint since any behavior can be rationalised by simply positing a taste for it.
The self-interest model is wrong. People do act charitably, give tips etc. But the reformulation you are suggesting doesn’t help. And the problem in trying the extend things in this way is the reason traditional economists have been reluctant to consider more plausible behavioral underpinnings. Its hard to refute hypotheses based on these more general frameworks.
Isn’t the relationship between economists and psychologists on this question of individual behaviour somewhat analogous to the relationship between quantum physicists and physicists of the “macro” world?
Many agree that there ‘ought” to be an elegant marriage between the branches of physics. Trouble is, no one’s found it yet, and maybe never will.
Similarly, group behaviour can intuitively be conceptualised as the sum of the behaviours of all the individuals in that group. Trouble is, it is readily observed that the size of the group is often critical to the behaviour of the group.
Perhaps the size of the group, as perceived by the individuals in that group, influences behaviour.
The smaller the perceived group: married couple, family, the stronger the influence of impulses that have traditionally been labelled “psychological”.
The larger the perceived group: economic role in an impersonal economy, perceived social status, etc., the stronger the influence of impulses that have traditionally been labelled “economic”.
This question of crowd dynamics nuzzles up alongside historical researches of, among others, Rude and Thompson, and anthropologists such as Geertz.
Harry Clarke,
Thank you for your reply.
1. “ If a man drinks crankcase oil and kills himself is rational since he must like crankcase oil� No. However, if a man wants to kill himself and he knows that crankcase oil does the job, then drinking crankcase oil is rational behaviour (equally applicable to females). An observation of this type would invalidate the assumption of ‘non-satiation’ but not the assumption of rationality. (But non-satiation can easily be refuted in some special situations. For example, give someone 10 bananas and increase the amount to 10 tonnes. I bet my bottom dollar that the person will refuse to eat the 10 tonnes of bananas, even if he or she ‘likes’ bananas.)
2. Unfortunately, I cannot claim originality for altruistic preferences.
3. Charity is one example of altruistic behaviour. A more obvious example is child rearing.
4. Yes, it is difficult to design empirical tests (‘experiments’) of rational behaviour. But this does not mean it can’t be done at all. However, I would suggest there is no gain at all from people designing tests where the researcher confuse their preferences with rationality (eg the assumption that a person who drinks crankcase oil wants to live while it is the researcher who wants to live).
In his book “The Selfish Gene” I think Richard Dawkins did an excellant job putting together a comprehensive thesis for the origins of altruism in otherwise competitive survival machines. For anybody seriously interested in the topic of “altruism” it should be near the top of your reading list.
I agree with Ernestine regarding rational. Altruism is not irrational. Rationality is about the linkage between preferences and behaviour. I have a slight bias towards the terminology of functional, versus disfunctional behaviour. Functional behaviour is where your actions tend to achieve the things you want. Disfunctional behaviour is where your actions tend not to achieve the things you don’t want.
Wanting to lose weight and eating cake three times a day would be an example of disfunctional behaviour.
Wanting people to like you and being genereous and nice to people would be an example of functional behaviour. Entirely rational.
I believe that most people, most of the time know what they want and act accordingly. I refer to this belief as “economic rationalism” although I know that the term gets used in a multitude of ways (and not all are endearing).
Ernestine. If you observe someone drinking crankcase oil you can always rationalise it by saying they had a preference for it and were acting rationally. This explains zilch. Simply adding things to a utility function explains nothing. Moreover the point under discussion has has nothing to do with non-satiation (always ‘wanting more’) at all. Non-satiation is one of the easiest assumptions in consumer theory to test and irrelevant here.
Moreover no-one said you were the originator of the idea of altruistic preferences – interdependent preferences in economics have been discussed since at least Veblen and are one of the simplest externality ideas in economics.
I didn’t say it was difficult to test the postulate of rationality as you define it. Its impossible to test it as you define rationality since the idea is empty. Any behavior observed is consistent with rationality in your terms since your definition is completely circular. This is old hat too.
Terje. Again no-one is saying altruism is irrational. There are so many books and articles on why altruism arises that no-one would ever say such a thing. But, again, to simply say, as Ernestine does, that altruism is ‘explained’ by a preference for altruism is to explain nothing. The whole point of behavioral economics is to provide an explanation. One that has been discussed is this thread that altruism is fostered by religion which has favorable effects on brain chemistry. There are others though I hardly think Richard Dawkins is someone you would want to cite as a prime mover in discussions of non-self-interested behavior.
Harry,
1. It seems to me you have mis-interpreted what I wrote.
Contrary to what you claim, I did not say that: “If you observe someone drinking crankcase oil you can always rationalise it by saying they had a preference for it and were acting rationally”.
I said, if a man wishes to kill himself and he knows that drinking crankcase oil does the job then drinking crankcase oil is rational (also applies to females).
In other words, one cannot draw any conclusion from the observation of a man drinking crankcase oil regarding ‘rational’ or ‘irrational’ behaviour because one has insufficient information to distinguish between:
a) the person does not know the effect of drinking crankcase oil on health (‘accident’ in everyday language)
b) the person has information regarding (a) but wants to kill himself (rational)
c) the person has the information regarding (a) but wants to live (irrational).
Your statement that one can always ‘rationalise it’ ….. is not true. If one wishes to test rationality then one has to set up an experiment to get the information required to distinguish between rational and irrational behaviour and accidents, errors, etc.
2. ‘Non-satiation’: No. I don’t agree that the assumption of ‘non-satiation’ is irrelevant here. Obviously, if someone wants to die than he or she does not wish to consume more. Hence my reference to non-satiation is not irrelevant. That is, the plan to end ones life is equivalent to being ‘satiated’.
3. I can’t agree that the assumption of non-satiation is easy to test empirically if there are say 1 million different ‘commodities’. (I deliberately chose an example of 1 commodity where only introspection is required to reach agreement.)
4. May I suggest that having discussions is different from having a theoretical result. Veblen’s notion of ‘conspicuous consumption’ is not an example of altruism. I would agree that it is an example of interdependent preferences.
5. It seems to me our notions of ‘explanation’ differ.
6. It seems to me Terje Peterson understands the notion of rationality in general equilibrium theory.
7. Strange, you agree that altruism is not irrational. How would you prove it?
8. What is it that behavioural economics wants to ‘explain’? I am not arguing here that gathering information in a more or less systematic way about people’s behaviour at a particular time, in a particular place, in a particular situation is not important. I just don’t know what you want to ‘explain’?
Sorry, I forgot to ask one more question of Harry:
What is the difference between behavioural economics and detailed economic history? By detailed conomic history I mean the collection of historical accounts on the behaviour of individuals under specified conditions.
Ernstine :
“Which one is faster – on average? What is the dispersion among people? Are there people who are so fast in conscious processing and so disciplined that there is no discernable difference between conscious and unconscious processing?”
It depends what you are unconsciously processing. There is a huge variety of things you can process unconsciously. Lots of really low level perceptual stuff (like early visual processing) is processed unconciously almost immediately by almost everybody (XXX->cat sequence, presumably via unconscious semantic association).
“4. Is there anything known about the ‘unconscious processing’ and learning? It seems to me some fleeting impression or ‘undigested’ bit of information can grow (‘pressure cooker’ we used to call it) into a question that requires thought.”
Yes, a huge amount is known (a huge amount of work has been done in perception of low level things). Alternatively, I’m not sure how much is known about the deja vu experience where you recall a complex concept after only being exposed to a small piece of simpler information. I presume people looking at it would have a huge amount of problems with false memory and so forth (although its surely possibly)
“5. ERP: In the more recent management jargon, ERP stands for ‘enterprise resource planning’ (previously ‘planning’ or ‘budgeting’ or ‘state contingent planning’ or ‘financial planning’). I trust this is not what you mean.”
Sorry, there is jargon everwhere. It just a techinque in cognitive-neuroscience where you stick electrodes to people’s heads and measure the electrical activation produced at the skull.
“6. Now, what happens when a I drop, say, a glass (it breaks) and I had not (consciously) planned to break the glass? Would you say that there is some ‘unconscious thinking’ or is it simply an ‘error’ – a malfunctioning of either my brain or my limbs or both? ”
It could be both. There is certainly the motor signal that can go wrong (“open hand”, and there is certainly a process that can controls it, which can presumably go wrong as well. THere is a study by Chris Frith (currently at UCL?) where they used TMS = electrical stimulation of brain areas, which cause a motor movement. However, depending on the time-frame people could either a) stop the movement; b) not stop the movement; and most interestingly; c) realize the movement had happened, but not stop it anyway. THe last of these suggests that there is some point of time that the motor signal has been given and some higher level process told to make you aware of it, but that the actual action can occur after you are aware that the action will occur in some circumstances.
Some of this seems to have got killed (I assume due to the use on an operator)…
Ernstine :
“Which one is faster – on average? What is the dispersion among people? Are there people who are so fast in conscious processing and so disciplined that there is no discernable difference between conscious and unconscious processing?�
It depends what you are unconsciously processing. There is a huge variety of things you can process unconsciously. Lots of really low level perceptual stuff (like early visual processing) is processed unconciously almost immediately by almost everybody (i.e., less than 20 ms). Alternatively, there is stuff that can take ages to get to conscious awareness, even though it is processed early. A good example of this is “change blindness” where a picture that changes quickly is put in noise (flickery crap). It can take over a minute for people to notice the change, but if you use cognitive neuroscience measures, you get a response to a change very much earlier (if I remember correctly, I’m not a vision guy).
Unconscious is really just a term refering to things that people are not aware but process anyway, so speed is not really relevant. THere are huge differences in some tasks in terms of the speed that things get into awareness, alternatively. A good example of this is priming, where you show a word but mask it. At around 40 ms (depending on font size, luminance…) some people recall all the words, some people none, and some people some. However, almost everyone is affected by the word, which you can measure, as people respond to cat in a (dog->XXX->cat) sequence faster, presumably via unconscious semantic association).
“What is the difference between behavioural economics and detailed economic history? By detailed conomic history I mean the collection of historical accounts on the behaviour of individuals under specified conditions.”
EG,
Whatever HC may conclude about the differences between behavioural economics and “detailed economic history” I believe that your (doubtless necessarily telegraphic) characterisation of economic history points the searchlight of enquiry in the wrong direction.
Economic history is much more than “the collection of historical accounts of the behaviour of individuals under specified conditions.”
But let’s restrict ourselves to that historical project.
The first thing that most historians note is the limitation placed upon the freedom of action of actors by structural, cultural and political constraints. Thus, actors may perceive themselves to be free agents, but historical perspective reveals that their actions were patterned by forces beyond their control. This applies even to the classic archetypes of economic egotism, the great tycoons who emerged duringthe Gilded Age in the United States (Morgan, Frick, Rockefeller, etc.) Why did they rise in the United States? How did Rockefeller rise to dominate the oil economy when he was just one of hundreds who may have? In other words, only one oil man made decisions that resulted in the destruction of the plans and projects of hundreds of others who hardly differed from him in ambition and methodology.
The vast majority of humans throughout history (including myself) have had only a tiny fraction of the freedom of economic decision making enjoyed by the now anonymous victims of Rockefeller’s prudence, vision and ruthlessness.
And my freedom of economic action is much greater than my father’s was, as was his compared with his father’s.
Most economic actors stay unthinkingly or timidly (I know which one applies to me) within the cultural, legal, political and moral constraints imposed by the institutions that protect the status quo. As a result, their perception of “economic rationality” is almost entirely culture-bound.
Even a man like John D. Rockefeller was very much the product of his upbringing. But his economic success required him to breach many unwritten and written laws. His success rewrote the rules of the economic game in the US and in the rest of the world. For some, this rewrite presented new economic opportunities. For others it closed off options.
I think it would be difficult to imagine a research design constructed by an economic behaviourist that would take robust account of the effects of culture.
Perhaps it is a trivial example but I’ve noticed that the most aggressive players of the game Monopoly are often the most cautious people when making economic decisions in the real world.
Ugly Dave,
I can only re-emphasise what Yobbo said – if you are not happy with the banks, leave them. In WA we have a good and viable system of credit unions – Police and Nurses (the accept deposits from anyone, not just police and nurses), United Credit Union and many tiddlers. StateWest, the second biggest, is demutualising and merging with Home Building Society – but will keep their fees structures. I am the Credit Unions are just as good over East.
.
Move – it is the only way the banks will get the message. If you keep paying the fees, who is the real fool here?
Antony Loewenstein Says: February 4th, 2006 at 12:11 pm “While the West talks about human rights and exporting “democracyâ€?, it seems the rules are a little elastic”
Indeed.
While Islam talks about a religion of peace, the fanatical Islamic Right torch sovereign Danish and Norwegian territory in Damascus, and shout death to just about everyone all for……the death of innocents in Darfur? No. End to suicide bombings? No. End to honour rapes? No. Oh yes, it’s all about some politcal cartoons.
Conrad,
Thank you for your detailed replies and for the reference to Chris Frith. I’ve ordered a book with a title that sounds interesting. One has to start somewhere.
May I ask one more favour. Can you point me to an author who specialises in the processing of sound stimuli?
Katz,
Thank you for your reply. You are perfectly right that I used a telegraphic style. But this is no excuse really – I should have phrased my question to HC better (specialized branch of micro-economic history?). Whatever a better phrase might have been, I certainly did not wish to suggest that there is not more to economic history. I’ve spent about one year reading only on the history of corporations and multinational corporations – Fernand Braudel alone could have kept me busy for a lot longer – and this only for the purpose of extending an existing theoretical model of a ‘competitive private ownership economy with complete commodity markets’ to the case of a partially segmented economy with multinational firms. The history of money is another subject of interest. But, I don’t have the skills of a historian nor the memory for institutional and historical detail. This is quite frustrating. For example, at present I am unable to recall where I read (many years ago) that in the USA there was a time period when corporations had a time restricted ‘license’ (not the exact word), issued by local governments (can’t remember which level). Initially the period of the ‘license’ was 10 years. Then it was extended to 50 or 100 years. Now it is unrestricted in time (not only in the USA). (I don’t understand why anybody worries these days about a hereditary monarchy within a parliamentary democracy.) If you happen to have at your fingertips a reference on the changing corporations laws in the USA prior to say 1950 at your fingertips, I’d be most grateful getting it. Your point on ‘perceptions of economic rationality’ makes sense to me against the background of the past 20 years or so. I find it quite interesting observing the reactions to theoretical models of ‘competitive private ownership economies’, published in the finest and most respected economics journals, written in a language that doesn’t change its meaning for as many generations into the future one wishes to contemplate (‘applied-postmodernism-proof’ so to speak). You mentioned restrictions on the freedom of economic actions. I can’t but agree. Even the simplest possible economic situation (exchange of ‘things’ among people) entails restrictions (the things they have to exchange and the preferences of the trading partners.) Obviously, even the simplest form of trade is a social situation. But this is masked if one examines ‘consumers’ choices’ for given prices.
Conrad,
Thank you for your detailed replies and for the reference to Chris Frith. I’ve ordered a book with a title that sounds interesting. One has to start somewhere.
May I ask one more favour. Can you point me to an author who specialises in the processing of sound stimuli
JQ,
I’ve posted a short message to Conrad, essentially thanking him for his replies and the reference to Chris Frith. This message has been ‘awaiting moderation’ since last night around this time. Something seems to be ‘stuck’.
Ernestine: You said (quite a few posts back):
“To the best of my knowledge, there is nothing in decision theory which prevents the aggregation of subjective probability values across a set of decision makers (ie to mimick the aggregation of expectations in models of sharemarket economies). “
Au contraire, here’s what stops this:
1. The fact that for many decisions, individuals do not know or cannot calculate subjective probability values and preferences over outcomes, due to ignorance, lack of resources (time, processing power, attention) or lack of experience.
2. The fact that for many decisions, it only makes sense to calculate probability values and preferences over outcomes in the process of making the decision (eg, after listening to the arguments of others), or even afterwards.
3. The fact that, to calculate such values one has to assume that reality can be represented as propositions (statements which have a truth value).
4. The fact that to calculate these values one has to assume that these propositions can be assigned single numerical probabilities and preference values. This process reduces a multi-faceted reality to a single number, and so something must be lost in such a representation.
5. The fact of the impossibility of any decision process amalgamating a set of Bayesian preferences into a single compromise Bayesian preference, when the process satisfies a set of reasonable assumptions. (Theorem of Kadane, Schervish and Seidenfeld.) This is an analogous result to Arrow’s impossibility theorem for voting procedures.
Sorry, but classical decision theory just doesn’t cut it for real-world group decision-making.
Peter,
I’ll try to respond by working backward
“Sorry, but classical decision theory just doesn’t cut it for real-world group decision-making”.
Response: Your initial post read:
” There are many problems with applying classical decision theory to real-world decisions, including the one you identify in your post [Harry’s].” [added]. (You did provide a link to something you wrote which I did take into account in my reply. I’ll get to this later.)
Note, the word ‘group’ is not in your initial post.
The term ‘group’ itself introduces several questions for ‘real-world’ decision making (in the sense of going from theoretical results to practice):
a) A group of executives in a corporation who have one objective as stated in the constitution of the corporation and they do not compete with each other internally (the picture I had in mind).
b) A group of executive in a corporation who follow their own interests (ie they play a non-cooperative game within the corp, which could entail a breach of contract.)
c) The size of the group may matter. You refer to Arrow’s impossibility theorem. Note, if ‘the group’ consists of 2 people (eg CEO and Finance Director) then Arrow’s impossibility theorem is irrelevant (because it assumes there are 3 agents).
You refer to Kadane, Schervish and Seidenfeld. Are you referring to Kadane, J. B., Schervish, M. J., and Seidenfeld, T. (1996),
“Reasoning to a Foregone Conclusion”, Journal of the American
Statistical Association 91: 1228-1235.? I understand in this paper the authors showed that a Bayesian decision maker may rationally pay not to see the outcome of a certain cost-free experiment, when only finite additivity of the probability measures is assumed. Would you mind telling me how this result is related to the decisions of Corporation A to bid on a stock exchange to acquire ownership rights of fraction 0
Peter,
It seems part of my post has been lost, possibly because I used a symble.
Continuation:
Would you mind telling me how this result is related to the decision of Corporation A to bid on a stock exchange to acquire a strictly positive fraction of ownership rights of Corporation B? Just to be clear, I do consider such a decision a ‘real world decision’ problem.
Now, going fast backward to my reply to your initial post. I made guesses about the meaning of the material you attached. I am now updating my prior: I know nothing about your ‘real world’.
Hi Ernestine —
1. I used “group” in my later post because I was responding to your statement, which talked about group preference aggregation. I believe that decision theory also does not cut it even when there is only a single decision-maker.
2. The theorem I mentioned is in the following paper:
T. Seidenfeld, J. B. Kadane and M. J. Schervish [1989]: “On the shared preferences of two Bayesian decision makers”. Journal of Philosophy, 86 (2): 225-244. This is reprinted as Chapter 1.1 of their book:
J. B. Kadane, M. J. Schervish & T. Seidenfeld (Editors) [1999]: Rethinking the Foundations of Statistics. Cambridge University Press.
3. You wrote:
“a) A group of executives in a corporation who have one objective as stated in the constitution of the corporation and they do not compete with each other internally (the picture I had in mind).
b) A group of executive in a corporation who follow their own interests (ie they play a non-cooperative game within the corp, which could entail a breach of contract.)”
I have two decades of experience working in and for Fortune 500 companies, and I have not yet seen a group of executives with one objective. Even without any assumption of breach of contract or competition on the part of the executives, this is an abstraction which simply does not exist in reality (like so many assumptions in economic theory). People have different beliefs, different preferences, different values, different goals, and different time horizons (e.g., time-discount factors), even when they are just 2 people and when they all work for the same company and they are working co-operatively and sincerely. Reasonable people can disagree, even executives, and with justification.
Consider your statement (a): If the the constitution of the organization says that the organization aims to maximize shareholder value, then two executives can agree that they both want to increase shareholder value, but profoundly disagree about what this means and how best to do it. For example, over what time horizon is value to be increased? Is it for all shareholders or only those with large holdings? What if shareholder interests diverge? Is it absolutely, or only relative to some average index of stocks? Is to be achieved via profit distribution or increase in share value? Is it to be achieved via market dominance or through cultivation of market niches? Etc, etc.
The only way we can say that an organization has a single, over-riding objective is to make it so abstract and generic as to be useless when it comes to obtaining agreement on any actual decision.
Ernestine —
I have just seen your continuation post.
Your example also strikes me as a real-world decision. But I’m not sure what your question is.
EG, thanks for your interesting comments. I’d like to return to them at a more leisured moment.
In the interim, here’s a useful web source on historico-juridicial issues of US corporation law. It also has a select bibliography which may serve as a starting point.
(Current New York law is both interesting and, for me, a little surprising.)
http://www.corporatepolicy.org/topics/LL.htm
Hi, Peter,
Thanks for your reply. I’ve only glanced through it. It seems we’ve overcome initial communication problems. I’d like to reply but I won’t be able to until next week-end.
Regards
Ernestine
Crispin,
A good book on behavioral economics that looks at the implications of relaxing self-interest assumptions is: C.C. Camerer, G. Lowenstein & M. Rabin, Advances in Behavioral Economics. Its a convenient set of readings and quite hard going in places – it is used as a text at University of Melbourne for a unit in Behavioural Economics.
Harry,
Thanks for the book reference (only just seen because I’m rather limiting time spent on the ‘net right now): added to my reading list. Have had a glance through the table of contents. Looks most interesting.
Good morning Peter,
Sorry about the delay in replying. As mentioned in my last post, I was too busy during the week to give proper attention to the series of posts. I am not sure I’ll manage to get all the possible points of discussion down in one post. Let’s start and see what happens.
1. There is one point of agreement (or successful communication), namely the example I gave of a ‘real world’ decision problem is a ‘real world decision’ problem. It was: Corporation A bids on a stock exchange to acquire a strictly positive fraction of ownership rights of Corporation B.
2. Setting aside (i) the topic ‘share holder value maximisation’ as an objective of a corporation, (ii) the question of what you consider ‘facts’ in an earlier post, (iii) the publications by Kadane, Schervish and Seidenfeld and their relevance to at least two other questions. Lets focus only on the statement ‘strictly positive fraction of ownership rights’ in (1) above.
3. Let the fraction be denoted by ‘alpha’. I propose that the decision problem (and hence the suitability of a particular conceptual framework) depends on the ‘size’ of alpha and on the purpose of the acquisition. Moreover, the type of decision problem determines who within a ‘large corporation’ (Fortune 500 companies are ‘large enough’) will make the decision. To illustrate:
a. ‘alpha’ is decided upon by A’s treasury department for the purpose of short-term cash management. ‘alpha’ is ‘small’
b. ‘alpha’ is small at time t but alpha is part of a strategy to gain control but less than full ownership over a time period t, t+1, …., t+n. Let S1 denote the strategy to which alpha t belongs. S1 is proposed by the CEO, on the advice of his or her staff, and decided upon by the Directors. The decisions on the timing of the future acquisitions may involve the treasury department or an outside agent.
c. ‘alpha’ is 100%. Let S2 denote the strategy to which alpha belongs. S2 is proposed by the CEO, on the advice on his or her staff, and decided upon by the Directors.
d. Suppose A and B are listed on the NY stock exchange.
e. Against your 20 year experience, as described in your post, are you saying that treasury staff do not use theoretical models in Finance which are based on decision theory (eg CAPM; in this model the security prices are determined by aggregating subjective probability values across a set of decision makers)? I’ll wait for your answer before proceeding with the rest.
4. (i) Shareholder value maximisation as an objective. I am not surprised that there is disagreement about this one because, even though many Finance texts state this as the objective, it is known in Economics that the conditions under which this objective can be implemented (complete securities markets) don’t hold in reality. If you want, I’ll provide references.
5. (ii) Can we postpone this one for the time being?
6. (iii) Kadane, Schervish and Seidenfeld. You did not provide a reference. I did not know anything about these authors. So I searched and found the reference which I gave. I found this reference was referred to by others in an area more familiar to me. I could not see how the result by these authors has anything to do with the topic of our discussion. Hence my question. I could not get the reference you gave by remote access. However, the result sounds interesting. Would it be much trouble to state the actual theorem. Thanks in advance.
Regards
Ernestine