Rat choice theory

The magnificent edifice of modern microeconomics is built on a simple model of human behaviour: rational choice theory.

The rational person has goals; things she wants and values. She makes choices; she acts to achieve her goals. Saints can be rational, and so can sinners. What matters is having preferences and making choices, whatever those choices happen to be. 

The assumption of rationality captures something so essential to survival that it explains non-human behaviour as well. Indeed, with a few simple experiments, it is possible to test the rationality of a wide range of animals.

The first element of rationality is the ability to make choices. This, in turn, requires that an individual knows what he or she wants. The first assumption of rational choice theory is that individuals have complete preferences, that is, given two alternatives, A and B, either A is preferred to B, B is preferred to A, or the individual is indifferent between the two.

The completeness of a rat's, cat's, dog's or other animal's preferences can be tested using this simple experiment:

• Take two pieces of food from the fridge, for example, some left-over chicken and a cube of tofu. Simultaneously present the two pieces to the animal. Take care to position the two pieces so that they are equally accessible, and not to distract the animal with movement.

If the animal is always able to make a choice, be it tofu or chicken, he is demonstrating that he has complete preferences. 

(When offered a choice between say lettuce and tomato, the animal might choose the third, implicit option - none of the above - by ignoring the experimenter and walking away. But that is rational too.)   

If the animal consistently chooses tofu over chicken, then he prefers tofu to chicken. If the animal sometimes chooses tofu and sometimes chooses chicken, he may be indifferent between the two. Indifference means "no difference", that is, chicken and tofu are equally delicious and yummy, so one just picks one at random.

The only way to fail the "ability to make choices" test is to sit there pondering the chicken and the tofu, thinking "which one do I really want? what will really make me happiest? I don't know, I don't know, I don't know." Humans do that, but not most other animals.

To be rational, choices must be coherent. The second assumption of rational choice theory is that preferences are transitive, that is, if A is preferred to B and B is preferred to C, then A is preferred to C.

A test of the transitivity of an animal's preferences requires three objects, for example, three toys, or three pieces of three types of food, or some combination of toys and food.

• Present animal with each possible combination of the three objects, for example, chicken v. tofu, throw toy v. chicken, throw toy v. tofu. Follow the experimental protocols described above, that is, present each pair of two objects simultaneously.

Transitivity means that, if the animal chooses tofu over chicken, and chicken over a throw toy, then, given a choice between tofu and the throw toy, he should choose the tofu.

If a dog fails the transitivity test, does it mean that he is irrational? Not necessarily. He might simply be indifferent between the various options. Alternatively, he might feel full after so much sniffing and snacking, and want to play a game instead. The assumption is that "all else being equal" preferences are transitive - but a hungry animal is not equal to a well-fed animal. Finally, it is extremely difficult to avoid framing the animal's decision by, for example, making one object easier to reach. Such framing, however, can distort an animal's choices, making them appear to be less than fully rational.

Textbooks usually add non-satiation, which means "more is preferred to less" to the list of assumptions economists make about preferences. The non-satiation test is simple:

• Give an animal a choice between a small amount of something delicious, such as chicken, and a large amount of something delicious. The rational animal will choose the super-sized option.

Researchers have found that animals as diverse as parrots, horses and salamandars "go for more" when presented with two options.  John Henry Kagel, Raymond Charles Battalio, Leonard Green, in their book Economic Choice Theory: An Experimental Analysis of Animal Behavior provide extensive documentation of the rationality of animals in general. Indeed, with phrases such as "Labor supply curves (number of lever presses) for two rats working in a closed economy...." they give a whole new meaning to the phrase "rat choice."

But if rational choice theory is so simple and universal that it can explain the behaviour of monkeys, rats, salamanders and pigeons, why do some people find it controversial? 

First, it does not have as much predictive power as perhaps one might like. Rational choice theory says that people will act instrumentally, trying to achieve their goals and desires. But it is silent on what those goals and desires actually might be. So while almost any behaviour can be explained after the fact as a rational choice, predicting behaviour before hand is a little trickier.

Second, like Newtonian mechanics, rational choice theory isn't a precisely accurate representation of how the world actually works, as can be demonstrated by a simple experiment. WARNING: do not try this with any animal who is aggressive, or whose temperament is unknown.

• Take two identical desirable objects, for example, two identical bones. Give one object to the animal and let her enjoy it for a while. Then take the first object away from the animal and attempt to replace it with the second object. Observe the animal closely. Does the animal exhibit "loss aversion", that is, snarl, growl, or otherwise object to having the bone taken away? Does the animal exhibit "status quo bias", that is, prefer the bone she has to the new bone?

The whole field of behavioural economics is dedicated to documenting ways - like loss aversion and status quo bias - in which individuals' behaviour fails to correspond to predictions of intermediate micro level rational choice theory.

A third issue with rational choice theory - at least when applied to people - is the normative content of the theory. Economics typically argue that what people choose is what makes people most happy. Some go further, arguing that the role of government should be limited to maintaining the institutional framework that allows people to make choices: providing information, ensuring competition, and monitoring product quality.

When it comes to dogs, cats and rats, we don't think this way. An average dog might prefer, say, chocolate over dry kibble. Yet an average dog owner has no qualms about ignoring the dog's preferences and feeding the dog kibble over chocolate. Chocolate can kill a dog.

In the same way, an average human might prefer, say, soft drinks over water. Excessive soft drink consumption leads to a variety of health problems, such as increased risk of diabetes. Yet any attempt to encourage people to consume water rather than soft drinks through, for example, soda taxes, or bans on super-size soft drinks, is extremely controversial.

Why is it acceptable to limit animals' choices, but not humans'? The obvious response is "humans and animals are different." Yet as we learn more and more about ourselves, and about other animals, the more we discover that humans and other animals are not that different after all.

Comments

[nemi]

"But if rational choice theory is so simple and universal"

I do not know what literature you have been reading, but in every experimental test I have seen about half of the subjects violate “rationality” – even when there are very straightforward assignments such as “choose which basket of goods that you want from these baskets” (repeated maybe 5-10 times).

You can try an experiment yourself with your students, This website will help you administrate it for free: http://www.econport.org/content/experiments.html

take ten goods and ten repitations and I bet that most of your students will violate rationality. Now imagine if these choices were to be made over a long period of time, over millions of very different goods and services, over complex issues such as which career to pursuit, who to marry etc.
I do not think rationality holds even approximately, not even without consideration of limited information and cognitive biases – but give your students the test and evaluate the results by yourself.

[Janice Compton]

I don't think the obvious response to your question is "humans and animals are different" but rather that you are comparing a human limiting the choices of a different animal to a human limiting choices of another human. Another adult human, to be more precise, since we don't have qualms about limiting choices of children. I'm sure if one dog was trying to limit the choices of another dog, he might not be too happy about it.

[Janice Compton]

Sorry, should be is NOT "humans and animals are different"....

[Ryan]

Why is it acceptable to limit animals' choices, but not humans'?

Because limiting human choices is an act of aggression, and most of us believe it's wrong to initiate force against our fellow human beings, except in self-defense.

Limiting an animal's "choices" (if they have any) is done to either protect or enhance the standing of the human race on planet Earth. In other words, animals are our property, whereas other human beings are definitely not.

To be honest, I'm a little stunned that you'd even ask the question.

[Frances Woolley]

nemi - thanks for the link, that looks great. Do you find that it's easy to use in a classroom situation.

On your specific comments:

A preferred to B and then B preferred to A isn't a violation of rational choice - it could indicate indifference, or a taste for variety. Sometimes I prefer red wine, sometimes I prefer white - so what?

On the universality of rat choice: keep reading to where I say: "Second, like Newtonian mechanics, rational choice theory isn't a precisely accurate representation of how the world actually works, as can be demonstrated by a simple experiment."

Janice - but why is it acceptable for humans to limit the choices of different animals and *not* to limit the choices of other adult humans? I'm smarter and more knowledgeable than my dog, but there are other people who are smarter and more knowledgeable than I am. What is the essential difference between these two situations?

Ryan - "In other words, animals are our property, whereas other human beings are definitely not."

I can see the argument that my dog is my property - in fact I refer to him at times as a slave. Without me, he would die, and arguably that gives me some right to treat him as my property.

But what gives me the right to enslave an orca (killer whale) or a chimpanzee? Why does Sea World have legal standing as a person, but not the whales who work there? From a moral point of view, the distinction is between humans and other intelligent animals is far from clear cut.

[Wonks Anonymous]

Dogs can't vote, or do a lot of other things. Conversely, in the United States at one time there were people treated like non-human animals. Then slavery was abolished.

By the way, Mike Moffat had a piece I found through Felix Salmon rather than WWCI:
http://m.theglobeandmail.com/report-on-business/economy/economy-lab/getting-a-grip-on-eggo-nomics/article4377089
Didn't he used to post here?

[Frances Woolley]

Wonks - Mike still does post here sometimes, but I think the Globe and Mail is able to offer him a variety of things that we can't by way of readership, and terms/conditions of employment, so from what I understand he's made a medium-term commitment to them.

[Ryan]

But what gives me the right to enslave an orca (killer whale) or a chimpanzee? Why does Sea World have legal standing as a person, but not the whales who work there? From a moral point of view, the distinction is between humans and other intelligent animals is far from clear cut.

Frances, you're changing the subject. The issue is not who owns whom. I may as well concede that humans don't own wild animals, and that they, too have choices. But that has nothing to do with Rational Choice Theory.

The issue is, as you asked, "Why is it not acceptable to limit the choices of humans?" I provided an answer that more than deals with that question. Unless we fancy ourselves dictators, it is morally reprehensible to do so.

If you're Hayek, you call this "the pretense of knowledge." If you're Mises, you call it "vain." In any case, voluntary action is a precondition for anything we might call "economic" action. You can force me to do what you want me to do, but in doing so you prevent me from exercising both the Law of Demand and the Law of Diminishing Marginal Utility. It might be human action, but it's not longer economics.

[Ryan]

...Unless of course I've completely missed the point, and what you are actually getting at is that animals, too, exercise rational choices.

If that's what your real point is, then I guess I don't have much to add. Animals may be sentient beings in the human sense of the term, or they may not be. We'll never know. We have no way of knowing what the feline, canine, or pachyderm experience is like.

[Sina Motamedi]

I became a much better economist the day I concluded that we're all just robots.

[Janice Compton]

If one could know for sure that the people making policy decisions were smarter and more knowledgeable than the rest of us (as we know (hope) that they are smarter and more knowledgeable than 5year olds or dogs), and that their policy decisions were made to maximize welfare and not to benefit special interest groups then perhaps there would be less controversy.

[Frances Woolley]

Ryan: "The issue is not who owns whom." If, as you said, "animals are our property," then the issue is precisely who owns whom. On the other hand, if animals are not our property, the question becomes: why do we have rights over them, but not other humans?

"We have no way of knowing what the feline, canine, or pachyderm experience is like." Yup, that's the morally consistent stance for a libertarian to take. Actually, there's a blog post to be written on "why libertarians should be vegetarians" - though I guess hunting would be o.k., too, since that isn't imprisoning an animal, just playing one's part in the dog-eat-dog (or human-eat-deer) world.

Janice - fair enough. Though in the case of the arguments against the soda tax, couldn't you say that a freedom-of-choice agenda is being used to enrich special interest groups, i.e. soda manufacturers?

[nemi]

"Do you find that it's easy to use in a classroom situation."

I have only done it once (but will do it again). The problem is in coming up with relevant goods and prices. The administration is extremely simple. You simply give them the password. They log in whenever and do the test. You get all the results summarized and individually. You can then present them to the class (but you probably shouldn’t show individual results – it might hurt to be called “irrational”).

"A preferred to B and then B preferred to A isn't a violation of rational choice - it could indicate indifference, or a taste for variety. Sometimes I prefer red wine, sometimes I prefer white - so what? "

No - that would not be a violation. GARP allows for “flat” indifference curves (unlike SARP - if i remember correctly). I think this specific test was created by Varian.

PS: Off course you still have to assume that rationality holds – at least in absence of identified cognitive biases – what else could you do if you want to derive normative results? However, I do think that many, many, economists are very dissatisfied with it. Thus the very quick (in academic time) embrace of behavioral economics. Not that I know of any cognitive bias that would make people fail this kind of test (the one I linked to). This is really as simple as it gets.

PS2: One study shows that your ability to choose rationally increases with age until you are 15 or something like that (there were only three groups, and those who were 15 (or something like that) made significantly less violations than those who where 7 (or something like that), but the group of university students didn’t outperform those who were 15 ). Other things that generally will affect the probability that you will make rationality violations is education and intelligence.

[Frances Woolley]

nemi - are the findings of behavioural economics enough to convince Ryan and Janice that it's o.k. for one human to restrict another human's choices?

[Janice Compton]

I was convinced long ago. I just don`t think this is the same as restricting the choices of children or dogs. For children and dogs, I agree with one (adult) human restricting the choices of one child; for adults I agree not necessarily with one human restricting choices of another human but of of society (through government, when the government reflects society`s wishes) restricting choices of itself. I have no problem with a soda pop tax - if I can be convinced that this will reduce the consumption of pop, and that this will reduce obesity. (I`m already convinced that obesity is not good).

[Ryan]

At the risk of going too far...

I know how a human being can own a dog - I do not know how a dog can own a human being. First, I have no evidence that dogs (or any other animal) adhere to the concept of private property. Second, dogs (or any other animal) do not manage what little property could be ascribed to them. In our legal framework, a human "owned" by an animal would either be eaten (as by a tiger), or left alone (as by a turtle).

So one of the reasons animals can be considered property is because we can own them and they cannot own us. (The same could not be said of slaves. A human slave has all the intellectual means to own his master; it is but a cruel twist of fate that their roles are not reversed.) Ownership in the economic sense seems to be a uniquely human trait. Dogs fight over scraps, as tigers may fight over a human carcass, and in that sense I suppose humans can be "owned" by animals - but only in that sense.

Not surprisingly, human ownership of animals reflects the same mechanisms: security of food. We own cattle that we may eat beef. We own sheep that we may either eat lamb or "trade" room/board in exchange for wool. We own dogs and birds that we may train them to help us hunt.

At some point, it comes down to food, and whether it is "fair" to eat the flesh of a species that tastes good, or use our bonds with other animals to obtain meat through cooperation. I'm not sure economic or libertarian reasoning could be applied to the survival of the fittest. Humans are the only animals that seem to care about the feelings of the other species. For all others, it's eat or be eaten. It would take a lot of convincing for me to see this completely natural state of affairs as being opposed to my libertarian ethics.

[Frances Woolley]

Ryan - think of Nozick: start off with a state of nature, and then proceed through a series of just transactions. Resulting state is just.

With dogs, there's growing evidence that they domesticated themselves, i.e. hung around human settlements and volunteered their hunting services because humans were just so hopeless when it came to catching squirrels. So our relationship with dogs arguably satisfies Nozickian principles of justice.

It's not clear that our relationship with, say, the dolphins at Sea World satisfies those same Nozickian principles - sure they willingly perform now in exchange for food, but they're at Sea World not because they want to be, but because they were captured and enslaved.

Janice - it sounds as if you and I are on the same page. Though it's interesting, in Ontario, people are quite willing to impose their views on others - e.g. the obese person who drives past me in a car and tells me off for not wearing my bicycle helmet (which is more likely to cause premature death - my dangerous cycling, or your sedentary lifestyle?). Nothing like some random stranger coming up and giving you a long lecture on some subject to turn anyone into a rabid libertarian!

[Determinant]

I know Ryan's position, and in short I believe he assumes that humans are individuals and that individuals can arrive at a socially and individually optimal solution.

I believe that he would disagree with the idea, though, that there are emergent properties of groups and therefore a socially optimal solution will involve a tradeoff from a purely individually optimal solution. One person's tradeoff is another's tyranny.

This idea covers much of modern life like the efforts against smoking, tobacco taxes, sin taxes generally and pop taxes.

[nemi]

I do not think you need behavioral economics for that, at least not if you are a parent. Also, most of us do not think it is wrong to limit one individuals choices if it increases another individuals choices/utility “more”/”in a more important way”. Furthermore, there might even be in everyones best interest that choices are limited (e.g. prisoners dilemma, public goods etc.). This is all standard economics.

Personally, I still only read behavioral economics for “fun” (not really the hedonistic kind of fun but you know what I mean). As far as I understand it, most behavioral economists think it is to early to use most of it yet (pretty much all of it could be used to argue for paternalistic policys – obviously the soda tax for example - but even though e.g. Krugmans trade teory is a very good argument for certain tariffs, Krugman do not seem to think that it is very likely that you would be able to apply it). On a conceptual level, I think it somehow have increased the acceptance of old classics as positional and status goods (I am not sure why – but behavioral economists seem to talk about them much more than neoclassical economists – maybe they aren’t as constrained by theory). I also see a rather big interest in different forms of soft paternalism – i.e. different ways for people to constrain their own future choices (which of course wouldn´t make any sense given standard notions of rationality). They also, increasingly, use insights from behavioral science to get people to do things by e.g. reframing voluntary choices (i.e. put a cross in this square if you DO NOT want to participate in x rather than … if you DO want to).

PS: Funny story.
I´m a PhD student at an economics department where neoclassical (micro) economics totally dominates. When I started we had this selection of candy that you could buy. It wasn’t a machine or anything, instead the system relied on people simply leaving the money that they owed (and you didn’t even have to pay when you picked your candy – you could do it once a week or whatever) .

Today, the candy is long gone. Not because people cheated (as I think economic theory would weakly predict), but because a lot of people complained that it made it to accessible and made them eat a lot more candy than they otherwise would have. I do not think a behavioral economist would be very surprised.

Even funnier, when some companies introduced bans on tobacco use during working hours and, among other arguments, said that it would help some employees to do something that they wished to do anyway, those same economists talked about it as it was the most absurd thing they ever heard.

[Alice Woolley]

I think one of the problems here is the use of the word "rational". A dog may choose, and choose consistently, but the one thing the dog is *never* doing is reasoning through that choice. It is rather a process of smell-want-act. And the fact is that there is overwhelming evidence to suggest that humans in large part decide that way too, at least at any given moment. Now that doesn't make this any less a matter of choosing, but when thinking about public policy and the relationship between the person and the state, it might be helpful to recognize that our reason and our actions don't always correlate very well.

Further, even if we confine ourselves to the world of rationality I'm not sure I understand why a person couldn't rationally choose to live in a state which helped her achieve an optimal life, the life that she would rationally choose if her choices were always purely rational? That doesn't have to be fascism or communism or any other 'ism' but simply a recognition that life as a person in that society might be happier and better than one where I'm left to my own idiotic decisions. I recognize, e.g., that I might not save enough money for my pension, or to look after my health needs when I'm older, and I'm quite happy to have the state do that saving for me. That's even leaving aside the low hanging fruit of the fact that singular "choices" often have collective implications.

[Simon van Norden]

Another problem with rational choice theory is that it typically assumes that preferences are fixed and exogenous. However, I've never quite been able to reconcile that world view with the existence of disciplines like marketing or psychology. (I'm inclined to think of clinical psychiatry as a profession devoted to changing preferences.) Even within economics, I'll run into papers where modellers happily explain behaviour as due to "preference shocks."

[Frances Woolley]

Alice - interesting point. It's easy, as someone who's been in this business for a long time, to throw around words like 'rational choice' without really thinking about the connotations of using 'rational' to describe a particular type of behaviour.

Simon - rational choice theory is compatible with some amount of endogenous preferences - there's a huge literature on preference formation. But it does assume that people have, deep down, something that's fixed - so preferences over preferences. E.g. I might choose to to cultivate a taste for yoga because I believe that, in the long run, enjoying yoga will make me happy.

True, marketing and psychology don't assume preferences are in any way exogenous - but they're also not deductive disciplines. They're more inductive 'run an experiment and see what happens' disciplines. It's hard to have a magnificent deductive edifice - like modern micro - without something that's exogenous, that can be used as an explanatory variable. If not preferences, then what?

[Alice Woolley]

Frances I think that's not limited to economics. Moral theory, law and regulation are all premised on the idea that human behaviour is reasoned and susceptible to the kinds of limits we think that reason can impose to control our desires and impulses. And clearly humans can make decisions that are rational to an extent. But the question is what follows from realizing that we don't actually make decisions that way most of the time, and what drives our decisions may not be susceptible to our controls or even our awareness. We do have things we prefer, and prefer consistently (although as the behavioural economists show, that is pretty malleable too), but if those preferences are affective rather than rational, then I think that has to have policy implications, and in particular has to have policy implications in relation to "paternalist" sort of decisions. Of course there are obvious dangers to that too - even if we aren't rational as actors it might be better to have a legal/market system that assumes we are, because otherwise one might end up with one of the "isms".

[DavidN]

Frances,

Economists should’ve called RCT axiomatic choice theory or something else because 'rational' is pejorative.

[rsj]

The odd thing here is that in macro, seeing as how all goods are collapsed into a single good, preferences is primarily about time preference. Is there any evidence that animals exhibit time preference? Do dogs forego chewing on a bone today and bury it to enjoy tomorrow or do they bury bones in the ground because they like to accumulate bones?

[Frances Woolley]

rsj: "Do dogs forego chewing on a bone today and bury it to enjoy tomorrow or do they bury bones in the ground because they like to accumulate bones?"

One could ask the same question about why people save. I suspect it's a bit of both.

DavidN - is rational really pejorative?

Alice - I'm teaching intermediate micro next year, and debating between using a conventional textbook and Frank's book that places greater emphasis on behavioural economics. I'm not really seeing, though, how to integrate the behavioural econ insights in a fundamental way, as opposed to a "here's a little box with a fun story in it" way. Your research on moral intuition (here) is, I guess, struggling with precisely this issue in the legal domain - I hadn't realized that.

[Chris J]

People are not rational, but can use rational thought to correct for some irrationality. We let the important be undone, while letting the urgent (if unimportant) hold our attention. We can correct for that through RESPs, RRSPs, designated drivers (I won't be rational after a few beers but I've got a friend who will be and we will switch off next week).

I don't know if this fits into B.E. or rational choice theory but it seems to me a key issue is the time scale of the choice. A classic example is the netflix/zip queue: when I did DVDs by mail I would put up "great films" on my queue when I wasn't due to receive a new DVD. But as I mailed back a DVD and thus would be sent a new one, I would put a fluffier movie to the top of the queue. I'll have my "Citizen Kane" next time; just this Saturday night I need a silly laugh, gratuitous HK action flick, or old episodes of MI-6...

I don't know if any of this is "irrational" but it is nicely summed up by Augustine: "Lord, give me chastity, but not yet."

[J.V. Dubois]

Very nice article. I would just add that there are several differences of the meaning of “rational” when studying human beings. First two are in economics language:

1) People choose means to achieve their goals not only on some cognitive "wicksellian" forward-looking process, but they often use past experience and account for past behavioral response of other people prior to action.

2) People are driven by emotions, they feel shame, disgust or joy. This in short means that while self-interest of "homo economicus" can be a powerful driver of decisions even for homo sapiens sapiens, it is not all there is. People are often other-regarding as is shown in experiments when many participant in one-time dictator games are willing to sacrifice their well being to punish other people for what they feel is not just.

Of course Frances preempted these objections by saying "so while almost any behaviour can be explained after the fact as a rational choice, predicting behaviour before hand is a little trickier". I would jut say that such extensive explanation of the word "rational" is not what it means for most people. This is after-fact explanation: we did something and since we think about ourselves as rational beings every act has to be rational. In this line of thought even patellar reflex could be explained as a rational response, it is only your consciousness that does not always understand that you really "desire" to move your leg.

I think that here I would take my hat off to Hume, who sees the relationship between emotions (what he calls passions) and what he calls as "rational deliberation" as former being subordinate to latter. For him, it is always emotion that gives reason to our actions. [Rational] deliberation is just a process that can refine your actions to better fulfill your desires. He explains deliberation as a two way iterative process "means to ends and ends to means" that is used to plan your actions sot that optimally fulfill your goals. Notice that it may very well be so that the reverse process “Ends to means” can be in conflict with other emotion. For instance “I am really angry and I really want to hurt this guy before me. But I cannot use my gun because then I could kill him that would ashame me as I dislike violence”. What seems like rational thought process is really just using rational deliberation both ways – I could use gun to harm someone, but hurting someone this way will evoke some other emotion that I may not like. Deliberation enables us to “empathize” with our future self after he did what we want to do now. But all it really can do is confront two emotions, it is what we “feel” about these emotions that really “decides” what we do.
What people really call "rational" decision is just a difference between passions and emotions that drive our decisions. Such as that your decision is driven by "violent" passions such as anger, fear etc. and "rational" decisions as those based on calm passions such as your moral sentiments. In this sense there cannot really exist any “rational choice” theory as it is always “irrational” emotion that drives our decisions. What can be rational is just a process of how we fulfill our desires driven by these emotions, which is something else entirely.

So if we get to the general comparisons between humans and animals with regard to emotions and deliberation, I do not think that it is as clear as Ryan or other people think. Animals are surely capable of feeling emotions, they can feel anger, fear, mammals are even capable of empathy (very important concept for Hume). Maybe they are not able to grasp concept such as religious faith or other moral attitudes in a human sense - the passions that are most often (according to Hume wrongly) considered as "rational", but they are definitely capable to feel emotions. So are animals capable of what is called "deliberation"? I would say that there is overwhelming evidence that they can do it. Those most intelligent animals are even capable of using tools to achieve their goals, that is evidence enough for me of a higher thought process that can be thought of as a rational deliberation.

And as a special thought for Ryan who says that "First, I have no evidence that dogs (or any other animal) adhere to the concept of private property". I suppose that you never had any dog. Dogs are fully capable of the basic understanding of property in terms of "natural law" - that is a physical possession. You "own" only what you can defend. That animals have only their physical bodies or closest packmates to defend their property and that they cannot speak and defend their property in human court of justice does not mean that they have no natural understanding of a the concept behind. I have it that you never had a dog. Anyone who had dog knows how jealous they are of their "own" favorite toys. Also many other animals have some concept of "private" property, that is mostly "their" territory, "their" females, "their" offspring or "their" nest/lair etc.

[J.V. Dubois]

Correction: It should be "I think that here I would take my hat off to Hume, who sees the relationship between emotions (what he calls passions) and what he calls as "rational deliberation" as LATTER being subordinate to FORMER" - that is rational deliberation is subordinate (is a slave) to emotions.

[Frances Woolley]

JV - interesting comments. There's a parallel here the behavioural econ idea that we all have a "planner" and a "doer" - the planner is the rational deliberative one who buys the gym membership, the doer is the one who decides tonight she'd rather go out for ice cream, and go to the gym tomorrow instead.

[Bob Smith]

DavidN - is rational really pejorative?

Perhaps in the sense that it implies that conflicting theories are irrational, which is generally taken as pejorative (although, in this context, where rationality has a technical meaning, it shouldn't be).

[Ryan]

Frances, FWIW, I depart from many libertarians in that I don't think it's sound reasoning to start at the state of nature or Robinson Crusoe's island. I also depart from the Locke-and-Rothbard crowd in that I don't believe ownership comes from "mixing one's labor with natural resources."

Instead, I think Mises had it right. He actually admitted throughout his works that all private property ownership can be traced back to fighting for it and/or stealing it. In the Misesian framework, market transactions are the only means we have of settling property disputes peacefully. Without the market, if we want something, we have to steal it from the commons or from each other.

I guess this implies that I view wild animals as "the commons" and domesticated or farm animals as commercial commodities. I choose not to opine as to whether animals reason like humans do; to me, this is unimportant. I love my pets, but human beings are simply more important to me because I, too, am a human being. A field full of cute, cuddly puppies is of less value to me than a single human life. We're all in this together, and one of the uncomfortable facts of human existence is that our lives are dependent on the killing and/or "enslaving" of animals. From our clothing to our food to our medicine, we require other animals for use as commercial resources.

Unlike animal rights activists, I have no moral quandry with this. Human life is more valuable than animal life. I accept this as a "Final Cause," exogenous to all economic and moral questions. Humans first, everything else second. I think people who feel otherwise have become too urbanized and too disconnected from how regularly animals must be killed or captured to improve human life.

[Scott P Bacon]

@Francis

Wish I had a copy of this post (or one like it) while studying Econ 4020 last fall... Very good explanation of rational choice theory... Well done!

I couldn't help but notice your suggestive tone throughout this post, perhaps best captured by your final line "the more we discover that humans and other animals are not that different after all."

Am I out to lunch on this one, or are you smuggling in some animal rights argument under the guise of a rational theory post?

[Frances Woolley]

Scott - thanks for your kind words. I'm teaching Econ 2001, intermediate micro for non-majors, next fall, and this post was written for that audience - I was just trying to think of some way of making completeness and transitivity interesting. So I'm glad you had that reaction.

As for the animals rights part - there were two motivations for that. The first is that it highlights the dual role that preferences play in econ - positive i.e. demand curves slope downwards and normative e.g. choice is good, more is better. The second is that I figured it would get people stirred up, and with Nick away canoeing, the WCIers needed something to argue about. Though I do believe the world would be a better place if North Americans on average ate something closer to an Indian or a Chinese diet, with less meat and more beans and vegetables, and if we could give the animals we do eat a reasonable quality of life.

[Shangwen]

LOL a Chinese diet increases animal welfare?? Speaking from lifelong experience, maybe not. There is certainly less consumption overall of meat, but more diversity of species brought to the table. Anyway, I know what you are saying and certainly agree with it--minimalist carnivory is definitely the most enjoyable of diet types. And since this post is really about macro, I will now bow out.

[J.V. Dubois]

Frances: I had something different in my mind. There are two different kinds of rationality. One is deeper philosophical. For instance, most people most of the time do not like pain therefore it is rational for us to expect them to act so that they avoid pain. But this is just a subset of our deliberation process if we want to evaluate likely outcome of our action. We think through our action, use our empathy to evaluate impact on how other people will feel about the outcome of the action (we use how we would feel about it as a proxy) and then we decide if this action is likely to succeed or not. But it is different from basic idea that "it is rational to avoid pain". Pain is just an emotion, there is nothing rational about it. Some people do not feel pain as we do, some people may embrace it and subject it to some other emotion that they value as much more important such as being consistent with their moral principles. We may claim that they are irrational, but it does not have to be because they have somehow damaged cognitive capacities and that they are unable to use deliberation to see impact of their actions. It may just be because they feel differently from us.

So in your example, both decisions - to buy gym ticket and to have an ice cream are driven by emotions. Deliberation is a slave to these emotions. Every second we are being urged by different emotions to do something. We employ our rational thought to evaluate the likely impact of these actions and discard most of such urges and excercise the ones that we feel good about. And by urge I do not mean something sudden and violent. Hume says that this is where most people exchange rationality and emotions. Some people consider adhering to calm emotions (like feeling satisfied with your body) as rational while adhering to violent emotions (like feeling satisfied after eating authentic italian gelato) as irrational. But it may be completely opposite. Sometimes acting upon violent emotion - like expressing your anger in the face of injustice can be generally considered as rational over calm emotion - like feeling safe in conformity.

Then I assume that there may truly be some issues with our rational deliberation stemming from the way how we are biologically wired. It is possible that in the very same way our brain is subject to many optical illusions, our deliberative rational processing can be fooled and influenced by cognitive illusions. This is what I think is mostly assumed to be an outcome of behavioral economics: hyperbolic discounting, impact of framing and many other things. It is as if behavioral economics could provide us with parameters that can be used to tweak our rational choice models and all will be well. I am more skeptical of this approach as I think about irrationality in a different way.

Ryan: "In the Misesian framework, market transactions are the only means we have of settling property disputes peacefully" I must admit that I do not understand this sentence. Let's assume that we both claim that some patch of land is rightfully ours. How can markets help us solve this property dispute? I would say that being subject to the law that is being created in a way that we both assume as just plus consolidated power over violence by the entity that can back up verdicts of any judicial body that has power to interpret the law could help us solve the dispute. But I do not see much space for markets in the process.

[Ryan]

J.V. Dubois - C.f. Ronald Coase on that problem. Generally speaking, the party that more highly values the property and has the financial means to back that valuation will end up in possession, even when the law indicates otherwise. But this is a bit of a tangent from my original point that in a peaceful society, we pay for the property we want to own, and that if this means is denied us, humans tend to result in aggression. Mises was not talking about ownership disputes in this case, he was merely discussing the acquisition of property.

[Ryan]

Oops - "resort to," not "result in."

[Simon van Norden]

Frances:

"True, marketing and psychology don't assume preferences are in any way exogenous - but they're also not deductive disciplines. ... It's hard to have a magnificent deductive edifice - like modern micro - without something that's exogenous, that can be used as an explanatory variable. If not preferences, then what?"

I guess I'm just still puzzled at the number of people who will buy a magnificent edifice despite knowing that the foundation can't bear the weight.

[Alice Woolley]

I think J.V. Dubois has done an excellent job of getting to the gist of the recent work in the moral psychology field about the intersection between reason and emotion in decision-making. I don't think it's fully understood, and what is understood suggests that this intersection is very complicated, but what drives our actions isn't reason in the sense traditionally posited by the Kantian sort of philosophers.

And Frances in terms of how you teach it, I don't think it's a matter of putting out a little info on it, or even looking at behavioural economics in a prescriptive way. I think it's a matter of looking at what you know within your field and then taking into account the relevance, if any, of the idea that people don't make decisions in the way that they have often been understood to - i.e., by reasoning through the implications of facts and in light of the end point they have rationally determined to pursue. I don't know of course what that relevance would be for economics, but in my field there are I think lots of interesting things that follow from it.

[Frances Woolley]

Simon "I guess I'm just still puzzled" - you macroeconomist you!

Seriously, I think this is why some of the more interesting micro-theorists, e.g. Arthur Robson, are working on evolutionary economics - two things that we can take as exogenous givens are (1) the need to survive and reproduce and (2) the natural environment - hence the appeal of models that explain, say, altruism as something that has evolved because it helps people pass on their genes to the next generation.

Alice - yup, one of these WCI moments when the comments are weightier than the post!

[Saturos]

"Why is it acceptable to limit animals' choices, but not humans'?"

Because we are "humans" too. But we are not "animals". There is no way of explaining to a dog that chocolate will kill it. It is possible for us to construct a class, "humans", which in general possesses fundamental properties that the class "(non-human) animals" does not. Not so with telling fat people they shouldn't eat more. "Of course the government ought to ban giant slushies! We need to protect people from their irrationality!" "And this government which will save us from our irrationality, it is composed of Vulcans, is it?"

In fact, can we really conceive of persistent protection of individuals from themselves, without total patent subjugation, as with humans and dogs? Clearly that is a political outcome to which people will not sign on to.

More fundamentally, animals cannot bargain for rights, nor do they truly participate in society with us. If we tried to construct a mutually interactive society with them the way we do with other humans, we would fail. Nor do we have strong kinship motives for favouring them.

[Saturos]

"So while almost any behaviour can be explained after the fact as a rational choice, predicting behaviour before hand is a little trickier."

Personality psychology to the rescue: http://econfaculty.gmu.edu/bcaplan/pdfs/sbvsmb.pdf

[Bob Smith]

But it is different from basic idea that "it is rational to avoid pain". Pain is just an emotion, there is nothing rational about it."

Hold on a second there, pain isn't "just an emotion", it's our bodies way of telling us that something is wrong. That's why our bodies are hard-wired to try to avoid it (think about it, if you accidentally touch a hot pot, your instinctive response is to withdraw your hand). Perhaps in that sense you're right that avoidance of pain isn't a "rational" human response, since it's our default response in the absence of rational thought. On the other hand, the fact that it's our default response no doubt informs our deliberate response to possible pain.

"Some people do not feel pain as we do, some people may embrace it and subject it to some other emotion that they value as much more important such as being consistent with their moral principles."

I'm not sure that the statement that it is rational to avoid pain detracts from that claim, because it'a statement about pain in the abstract with a discussion of the corresponding benefits. Just because I desire to avoid pain doesn't mean that engaging in, for example, painful medical procedures, or painful pastimes (running, anyone?) are irrational behaviours (ok, running is, but that may be my bias). It may well be that some people derive pleasure from pain in particular circumstances (this is PG site, so I won't go into details), although even then, I suspect there are limits to that (if those people inadvertently touch hot pots, I'll bet they'll withdraw their hand).

I'm also not sure about the purported disconnect between rationality and emotions. As I see it, emotions enter into play by affecting our preferences - just as external factors can affect our preferences. People who act on the basis of, say, violent emotions are not per se irrational (at least as rationality/irrationality is being discussed here, I think we tend to use the term "rational" more loosely in popular conversation). Note that this accepts that preferences are not fixed over time (and I don't think the rational choice model strictly requires them to be fixed over time, ), and I think that the contrary claim is untenable (if only because people's preferences do evolve over time). I don't think this is the least bit controversial - no one would suggest that, because my preference for hot chocolate vs. lemonnade changes from the winter to the summer, my choices are inconsisent with the rational choice model, so why does the fact that my preferences may change depending on whether I'm angry or not make a difference.

If I come home and find a man in bed with my wife, and I kill him, am I acting irrationally? Maybe (no doubt that's what my lawyer will tell the jury while pushing for an insanity defense), but on the other hand my response to the possible choices presented in that situation (kill man, hurt man, don't hurt man) could be entirely consistent with the (very limited) tenets of rationality described by Francis (i.e., I have choices, my preference between those choices are transitive - I'd rather kill him than hurt him, and hurt him than not hurt him - and my preferences are subject to non-satiation - I'd like to hurt him more than less). Killing him may not be an objectively good choice, and it probably isn't in my long-term interest (perhaps because of discounting - at that moment I don't care about the future), but there's nothing about rationality that requires that rational choices be objectively good one or neccesarily welfare maximizing in the long-run. I think it becomes easier to accept the validity of the rational choice model, once you appreciate how limited its prescriptions are.

[J.V. Dubois]

Saturos: Defining "humanity" is not as easy as you think, but I will not delve into this. However even Kant who excersised quite strict criteria of personhood admited that even if animals are not rational and reasonable and thus from his viewpoint can be considered as "means" and not "ends", humans have indirect duties to animals insofar as it can reflect on their behavior to other humans. This is what he says:

"If a man shoots his dog because the animal is no longer capable of service, he does not fail in his duty to the dog, for the dog cannot judge, but his act is inhuman and damages in himself that humanity which it is his duty to show towards mankind. If he is not to stifle his human feelings, he must practice kindness towards animals, for he who is cruel to animals becomes hard also in his dealings with men"

This has profound implications on marginal cases of what being can be considered as persons. During the history it was mostly about human beings that lacked reasoning capacities such as infants of mentally ill, but with transhumanism on the horizon these moral problems can take completely new shape.

Hume on the other hand was more pragmatic, he viewed these moral issues through his concept of "judicious spectator", that is that human beings use a proces of "sympathy", that is we associate the outcomes of some actions to our previous personal experience and then judge their fitness in this light. This also means, that if those outcomes do not relate with our personal experience, we do not develop strong moral feelings about it.

If you ask me, I am more with Hume on this one. Dehumanization and putting distance between what is our own experience and what other beings (even other human beings) feel is the surest way to dull our sense of morality. We may think in a Kantian way how moral principles apply to God or Angels or Martians applying "pure reason" to moral questions, but these will always be distant to what we can relate to our own experience.

Bob: I was talking more about meaning of our behaviour. I used pain as an example exactly because it can be so easily tied to basic bodily functions. We avoid pain just as sure as we seek pleasure or as we feel multitude of other emotions. Hume's point is that there is no act that originates from pure reason. Reason alone is incapable of giving any true meaning to our actions.

As for the disconnect between emotions and reason, I do not see the point in keeping the rational choice model in. First, as you state to predict some choice of some person you need to know his "preferences" or his "emotional state" that can change in time. Then I would also add that even from here most of the time people don't really employ rational deliberation to evaluate the given choices to reach their goals. I mean they do not engage in the actual forward-looking cognitive process, they just use past experience and repeat it under new circumstances without really thinking about it. You may call this process as subconsciousness, heuristic or whatever but using the word "rational" is for me just confusing. If in the end you insist to keep saying that this really is just enhanced rational choice theory so be it.

[nemi]

"I'm teaching intermediate micro next year, and debating between using a conventional textbook and Frank's book that places greater emphasis on behavioural economics. I'm not really seeing, though, how to integrate the behavioural econ insights in a fundamental way, as opposed to a "here's a little box with a fun story in it" way."

Unlike imperfect competition, imperfect information, risk, externalities, public goods etc. etc. etc.?

What book are you currently using that doesn't simply include the above mentioned things in small boxes (or separate chapters), but somehow integrate it and keeps it relevant for the upcoming chapters. Sure, rational choice, perfect competition and everything else that fits with arrow-debreu is integrated - but that is a very very small part of economics.

[Frances Woolley]

nemi: "Unlike imperfect competition, imperfect information, risk, externalities, public goods etc. etc. etc.?"

Exactly. Profit maximization under imperfect competition follows the same basic principles - marginal revenue=marginal cost - as perfect competition does. Indeed one can go through perfect competition, go through monopoly, and then show that all of the intermediate cases of imperfect competition fall neatly in between these two extremes. Externalities are just the standard model with an extra cost added in, etc.

By the way, do you have more info on the GARP games that you played with your students - I clicked around the site, but couldn't find anything that sounded like the game you were talking about.

[Bob Smith]

"As for the disconnect between emotions and reason, I do not see the point in keeping the rational choice model in. First, as you state to predict some choice of some person you need to know his "preferences" or his "emotional state" that can change in time."

I'm not sure why the prospect of changing preferences affects the usefulness (or perceived usefulness) of the rational choice model. I don't think even the most ambitious economist tries to predict individual behaviour at all times in all circumstances.

Also, I think we have to be careful about equating reason with rationality, because they're distinct concepts. Decisions can be rational without being reasoned. I don't know why I prefer, say, beer over vodka, or pizza over pasta but given that I do, and I can't really say I put much thought into making my decision when I go to an italian restaurant, but given my preferences, my choices for one over the other are rational.

[DavidN]

Frances,

Maybe pejorative isn’t quite the right word though irrational is definitely pejorative. I’ve seen economists and non-economists alike use 'rational' in the technical and non-technical (i.e literal) sense though I’m more concerned with the latter. When used in the latter sense by laypersons it’s often used as a critique ala 'people aren’t rational/are irrational therefore RCT is unrealistic thus econ is BS!'. When the latter sense is used by economists it’s often used to make a normative statement like 'x behaviour is irrational!' as some behavioural economists are prone to do when there is nothing pejorative in exhibiting cognitive biases.

[Alex1]

I always understood rationality somewhat differently.

To me rationality in economics only makes sense due to scarcity. If there is no scarcity then I dont see much point in being rational. If I was to live in a place right next to an infinite water supply there'd be no reason for me to be rational about my water consumption. I could use it, waste it, whatever; there'd be no need for any real decision making process. Now imagine I had a water supply which was limited but still huge. I can't be too wasteful because that'd hurt me at one point. But I can probably be at least a bit inefficient. I could, for example, build some overly complicated device to extract the water and then keep it due to status quo bias even though there would be a better one. My water consumption as a result would be slightly lower than it could be, but it won't matter that much. Qnd now imagine a huge society, with complex social interactions and institutions, lives next to a water supply that is just about enough to keep them alive. Now is the time where you really need to be rational. Irrationality in that case is a luxury this society doesn't have - being calculating robots is all they can afford.

But the thig is that we usually dont deem a scarcity a good thing. We view it as constraint. And for me rationality (not irrationality) is equally some awkward kind of constraint. It is something we have to do sometimes (and even then we sometimes fail). But in that sense I also don't see why it would be a weak foundation for a science. it rather is the strongest concievable foundation, because it describes the assymptotic scenario. It exemplifies the worst case. Yes, in reality many of its tenets do not hold but that is because our earth is very far from being the worst case. But if we want a foundation, we have to start with worst case and then slowly relax some assumptions, rather than the other way around (this is because a worst case of extreme scarcity makes a lot more sense than a best case of no scarcity, given our knowledge of physics with entropy and all).

[Ed Seedhouse]

"Do dogs forego chewing on a bone today and bury it to enjoy tomorrow or do they bury bones in the ground because they like to accumulate bones?"

Well, squirrels observably cache food during good times and find it and consume it during bad ones. The ones that do this most effectively tend to survive and pass their genes on to more offspring. Do they do this because they have foresite? Squirrels have brains, though small ones. Do they have "intentions"? I don't know but it seems to be easier to think of them as though they do. At any event their behaviour seems indistinguishable from what it would be if they did have foresite and intentions. And is it not said that "a difference that makes no difference is no difference"? So on that principal am I not justified in assuming that they do have these things in fact? Name an observation or experiment that could decide one way or another, and we'd have the possibility of deciding, but until then I think it perfectly reasonable to think of squirrels having intentions and foresite.

[EuropeanViolet]

Just few questions:
1. The fact that both humans and animals can be labelled "rational" according to rational choice theory means that we're actually both rational or just that the theory isn't fruitful in explaining why we're different?

2. Is there a theoretical possibility for the rational choice theory to be falsified? I mean: let's say that we have a demarcation criterion which allows us to test an agent and label it exclusively "rational"|"not rational". This is a very simple function which takes as domain the set of all possible agents and output y in (0, 1), but this is maybe of little interest by itself...
Anyway, is there at least one possible case, even if only theoretically, where the demarcation function output can be "non rational"? I mean can we imagine at least one kind of behaviour which will show un-rationality without reasonable doubt?
If there isn't any possible case of this, it means that our function is totally useless, it doesn't explain anything, just arrange words/math symbols in a way they sound always true...
Chances are that applying this kind of theory to particles will give the same output, if not, at least it would be useful in order to demarcate between agents and non agents, so we should rename it as "agent detector" instead of rational choice detector.

3. the same argument as before but substitute "rational choice" with "choice": given at least two options to the agent, is there a way to distinguish between a behaviour which shows a choice and a behaviour that doesn't?

4. Let's say that we were able to come out with a falsifiable criteria for distinguishing
a. choices from non-choices
b. rational choices from non-rational choices
c. rational agents from non rational agents, when and if we find one dog or human which output "non rational", what that means?
a. that our theory is wrong and we need to patch it in order to have it outputting "rational" even in this case
b. that we made the discovery of the century worth the Nobel price
c. that there is at least a non rational agent (and we found it)
d. that there was some bias in our experiment
e. that the agent we have just tested was a Turing machine trying to fake human and so now we've a valid objective criterion for detecting Turing machines faking humans
f. something else

[Frances Woolley]

European Violet - "2. Is there a theoretical possibility for the rational choice theory to be falsified?" That's a long debated issue, on which people take one of three different positions (a) it's not falsifiable (b) it's falsifiable and has been prove false (c) it's falsifiable but has not yet been falsified. Which position one takes depends very much upon how one interprets the theory.

On your other observation: "just that the theory isn't fruitful in explaining why we're different?" It certainly wasn't my intention to explain why animals and people are different - I don't think we are all that different.

[Darren]

Francis: "True, marketing and psychology don't assume preferences are in any way exogenous - but they're also not deductive disciplines. They're more inductive 'run an experiment and see what happens' disciplines. It's hard to have a magnificent deductive edifice - like modern micro - without something that's exogenous, that can be used as an explanatory variable. If not preferences, then what?"

This begs the question. Why should we bother creating a magnificent deductive edifice in the first place when we cannot identify (to borrow Simon's phrase) a foundation that can bear its weight?

I had a similar exchange with Nick Rowe a while ago. He argued that economics was beautiful and fun even if it wasn't useful. I replied that economists weren't being paid to produce "beauty" (if that's what we want, our money would be far better spent on the ballet), they were being paid to produce accurate predictions and useful policy advice. If the field can't do that it should disband itself.