Friday, May 2, 2014

Equality, Monkeys, and Bananas

You may have heard this story. I got it from John Kennan, who claims it originated with computer scientists. There are some monkeys in a cage. Someone puts a banana at the other end of the cage. When one monkey goes for the banana, the other monkeys are sprayed with a fire hose. This is repeated a few times. Then, whenever one monkey makes a move for the banana, the other monkeys beat him/her up. After a while, one monkey is replaced. Of course, that monkey goes for the banana and the others beat him/her up. This keeps happening, and eventually none of the monkeys in the cage have ever seen a fire hose. But whenever a new monkey arrives and makes an attempt to get the banana, he/she gets beaten up. Asked why, the monkeys say: "That's the way we've always done it."

37 comments:

  1. Replies
    1. Sarcasm will get you nowhere. I thought it was good. You can go tell Kennan he's shallow and see where that gets you.

      Delete
  2. You are really good at this propaganda stuff so you really should work for CATO or any other libertarian / right-wing "think "tank. They pay real money for packaging the big lies well.

    ReplyDelete
    Replies
    1. Propaganda stuff? One reading of the story (the one I think the computer scientists had in mind), is that it's an organizational behavior lesson. There's some low-hanging fruit (literally) in the organization, and the firehose is the punishment the other people in the organization get when they don't show initiative. But, the incentives that the management is providing don't have the intended effect. Instead of working harder to get the low-hanging fruit, people in the organization punish each other for showing good performance so they won't get punished. Then the bad behavior just propagates itself. There were stories like that recently about Microsoft. Conclusion: You need to provide incentives to reward the monkeys for cooperating.

      So, what was the interpretation you had in mind?

      Delete
    2. Yeah right, that's why the title is "Organizations, Monkeys, and Bananas" and not "Equality, Monkeys, and Bananas".

      Delete
    3. dredging up an old post to tell the OP anonymous: you're a trailer trash monkey. read a fucking book; you're so dumb it hurts

      Delete
  3. Sorry, I thought your post was supposed to be sarcastic as well...

    ReplyDelete
  4. It seems to me that it is a story about Chesterfield's fence and the wisdom of social institutions.

    None of the monkeys has ever been sprayed by the hose and no one really knows why they beat up the monkey who tries to get the banana, but they've probably saved themselves a lot of spraying with this social custom.

    On the one hand, you might say that the custom is useless and counterproductive if the fire hose is gone forever. But -- on the other hand -- real human institutions respond to dangers that are much more persistent.

    ReplyDelete
  5. I don't understand. What is to prevent the initial monkey who makes the first move from eating the banana unperturbed. Since it is only the other ones who get hosed...

    ReplyDelete
    Replies
    1. It seems the first monkey gets the banana. That's part of the story.

      Delete
  6. The question is, why didn't the monkey share his banana! When you put others at risk, you are morally, and usually legally, obliged to compensate. The monkey that got the banana didn't compensate the others. It's furthermore completely rational for a society to impose rules to minimize downside risk.

    ReplyDelete
    Replies
    1. It is because the monkey-oligarch bough the allegiance of the Monkey Old Party which, serving the interest of monkey-business, blocked the passing of a law that would require the top 1 monkey to compensate the other monkeys when he went bananas.

      Delete
    2. That's right Jefftopia, Bill Gates, Paul Allen, Steve Jobs, etc. are morally obliged to compensate postal employees because, due to their inventions, people can now use email rather than snail mail. They should also compensate clerks and typists because their inventions have made it so much simpler for people to write and file documents, and accountants because without their inventions people would not have access to Turbo Tax. This should teach other ambitious inventors to think twice before they act, wouldn't it. Schumpeter was on to something, I believe.

      Delete
    3. Interesting that you name Allen and Gates, i.e. people whose incomes consisted to some degree in monopoly rents.

      Delete
    4. Where they born with these rents? How did they get them?

      Delete
    5. I think you guys missed the point where the parable is just a bad analogy. Bill Gates et al actually brought benefits to society. In the story, one monkey got the banana, while the others bore the burden of its consumption. Hence why I asked why it didn't share - sharing it would represent social benefits to innovation. I mean, imagine if Bill Gates invented an OS that worked perfectly for himself, but then allowed his friends to use his technology to spy on everyone. You'd think he would owe society. Innovators should enjoy the benefits of their success but pony up when they harm. Not a radical concept.

      Sheesh.

      Delete
    6. Jefftopia, but to say that Gates' action benefited society is not the same as saying that it benefited everyone. Often times the gains are unequally divided, with some people ending up worse off than before. This applies to innovation as well as international trade. I do think that people who get hurt should be helped to stand back on their feet, but the responsibility to provide assistance should fall to everyone, not just Gates.

      Here is another example that involves positional externalities, a favorite topic of Steve's favorite economist (just kidding), Robert Frank. Suppose that a teacher scales her exams. If everyone studies less they will get the same grade as if everyone studies more, since each student's performance is graded relative to the performance of others. Now suppose that the students are lazy except one who is unusually smart and also hard working. To deter him from performing so well and therefore making them look back, the other students start calling him a nerd and bullying him. This is not a hypothetical situation, it happens in real life all the time. Should the good student underperform to satisfy the others?

      Delete
    7. That's an interesting but fundamentally different situation from the monkeys for a few reasons. One notable difference being that students in a classroom aren't a legislative body. I do actually wonder whether whether students under some veil of ignorance would vote to redistribute grades so as to maximize average utility or perhaps the utility of the lowest, etc.

      Of course, there's another important difference, that we already know that all students save one are lazy - not disadvantaged per se, but lazy. Were the other monkeys lazy? No, they were just pissed.

      So in the classroom case, the differences are in social make up. Gates helps himself and others, but only because others are willing to accept those benefits. In the classroom example, this is not also true, which is an absurdity. It's an absurdity because there's no a priori incentive to punish or remove any and all achievement; there's just an incentive for everyone to try to take a cut from the achievers. There's no such thing as natural law; if that's what constitution designers want, then that is what is right.

      And that's my answer: with the way the world is, the student should do as well as he wants. In another world, scores would be recognized as high or low based on how a given student does given their demographic characteristics, so there would be no need for a student to underperform. Relative performance makes sense, but only if relative to the right control.

      More generally, abstracting away from your example, we can recognize the adverse effects of low SES, and offer programs to alleviate the burdens.

      Delete
    8. Let me be more concrete: there's no reason to redistribute grades if we're redistributing income, and it makes far more sense to redist income. We don't want to send idiots to the top, we just want to ensure that whatever inequities that are present in the system are economically accounted for.

      Delete
    9. I am a bit confused. Suppose the students are not lazy, but just not as smart as the "nerd". Should he play dumb to avoid their wrath? Or should grades be adjusted according to IQ? I am not sure either one is a good option.

      By the way, there is another, perhaps the main point in the story that has been neglected. Eventually the behavior of beating up the monkey that goes for the banana (or calling someone a nerd) becomes engrained in the culture. No one knows why they beat up the monkey that goes for the banana since no monkey has seen a fire hose. They do it because that's what they have always done. It becomes institutional memory.

      Delete
    10. Yeah, I was pretty confusing at the end there. I'll try this again. But first, I with you on institutional memory; things should be done for good reasons, not because it's the way things have been done.

      Redistributing grades in a world where we sortof redistribute income is like double taxation - doesn't make much sense. The question then is, which are you going to do? I say target income, or wealth for that matter. It's not a punishment; punishment assumes one was entitled to that reward in the first place - an assumption, not a truth.

      But in the bigger picture, the student should aim as high as he wants and get that top grade. You say it brings everyone else down, I say not necessarily. Smart policy analysts look at performance relative to SES (socioeconomic status). So instead of, "Wow, Jimmy had a GPA of 3.8!", it would be more like, "So Jimmy got a 3.8, which was the expected result for a student with characteristics x,y, and z".

      It's not needed to redistribute grades because our analysis should already be conditioned to expectations. This is why the affirmative action case in Michigan opens the door for more interesting admissions policies. We're getting off topic now, but I posted a comment about this and got a great response from Miles Kimball over at this blog: http://blog.supplysideliberal.com/post/83614366672/michigan-university-and-state-occasion-a-landmark

      Delete
    11. Good point on affirmative action, it is a topic I go back and forth on. It is complicated.

      On the example I gave, you are changing the parameters. There is no choice of redistributing income. Let's keep it simple. How do we deal with the student who does so well, therefore making the rest look bad, either as a result of effort or because he was lucky enough to be smarter than the rest of us? Personally, I do not think there is an easy answer here. I believe we all agree that a student shouldn't be allowed to get a good grade by cheating. But beyond that, it is not clear to me what the appropriate response is.

      Delete
    12. It's almost impossible for me to properly be a consequentialist about student grades when I don't have knowledge about the rest of the world! How I respond depends on what other policies are in place. In general, I do not believe the student should change his grade, but this is because I further assume that policy controls for demography! :P If that isn't true, then I think the student should bow to their pressure and offer a lower score, knowing that society may very well have screwed them over in the first place.

      Delete
    13. Well, here is what I do. I give some points for effort to help weak but not indifferent students get a reasonable grade. But among the better students, I reserve the A for the top performers. Weak students appreciate this, the best students like it, but many decent students who are used to get easy As are not that happy. Nevertheless, I think this is a good rule, and not just for grading.

      Delete
  7. By the way, Steve, did you have these in mind:
    https://www.aeaweb.org/articles.php?doi=10.1257/0002828054825655
    http://ideas.repec.org/a/eee/moneco/v52y2005i7p1227-1244.html

    ReplyDelete
    Replies
    1. It's funny how you can always take apart Alessina's papers via merely reading the abstracts. I like his notion that luck (i.e. probability distribution) depends on people's believes. Always entertaining when an economist goes postmodern.

      Delete
    2. Postmodern? Was Keynes postmodern for talking about animal spirits? Or Akerlof with his identity economics?

      Delete
    3. Nope. Keynes was far to vague to be considered a scientific economist while Akerlof tried to connect economics with sociology and psychology. But luck aka probability distributions are not subject to belief, either an event is more likely to occur or less likely. Unless you wanna regress to non-adaptive expectations and claim that people can have wrong believes about a distribution the entire time ... but this seems kind of pre-Lucasian backwards into the stone ages.

      Delete
    4. Now here is a fine example of a postmodern comment. Thanks!

      Delete
    5. You are welcome. Always funny when the halfwits comment on economics.

      Delete
    6. Angeletos and Alesina halfwits. Hmm... I wonder how your publication record compares to theirs. By the way, it's beliefs, a noun, not believes, a verb!

      Delete
  8. Why did you choose that particular title for this blog post? It seems tailor-made for generating hostile responses.

    John Kennan doesn't have a wikipedia page. Is there a faculty page for him somewhere?

    ReplyDelete
    Replies
    1. Exactly. Kennan told me about this, and it's clear that the people who thought up the story had organizational behavior in mind. But we could think of this as an economy - albeit a very primitive one. There's an obvious inefficiency and there appears to be equality. But can we draw any useful economic lessons from it? I wasn't sure. It's at least funny, so I thought I would just throw it out there and see what people might say. Some people are very excitable apparently.

      Delete
  9. Jesus Christ, the quality of commentors on this blog....

    http://lmgtfy.com/?q=john+kennan

    ReplyDelete