Patterico's Pontifications

8/23/2015

“Human Action” and Robert Murphy’s “Choice,” Part 2: What Is Action?

Filed under: Economics,General,Human Action and Choice — Patterico @ 10:31 am

This is Part 2 of my ongoing series of posts summarizing Bob Murphy’s excellent book Choice: Cooperation, Enterprise, and Human Action. The idea is to popularize and spread the word about Austrian economics and educate the public. Part 1 is here.

So what do we mean when we say economists study “human action”? First, the action must be purposeful behavior; reflex actions do not count, for example. It can even be inaction. Rush fans will recognize the quote: “if you choose not to decide, you still have made a choice.” Right?

Remember yesterday, I said: “Mises thought of economics as a deductive discipline, in which one divines, through analysis and reflection, the fundamentals of why humans act, and derives the necessary logical implications of these fundamentals using a deductive and logical chain of reasoning.” So what can we deduce from purposeful behavior?

Well, some of these are going to seem obvious. But it’s still necessary to state them — because it’s all foundational for what comes in the future.

First, obviously, if there is human action, then there must be a mind behind it. Human action is different from the mindless falling of a rock that we see in physics. Second, the actor must have goals, or preferences, in taking an action.

So far, pretty obvious. Here is another implication that is perhaps less obvious, but is very important: preferences are subjective and not objective. A related concept is that preferences are ordinal and not cardinal.

OK, I’m going to have to explain this one.

Cardinal numbers are “counting numbers” that measure things in units. In the equation 2+2=4, the numbers two and four are cardinal numbers.

“Ordinal” numbers arrange things in a series. Joe is first in line. Chocolate is my 4th favorite flavor of ice cream. (It’s actually my favorite; this is a theoretical discussion!) These preferences can’t be measured in units. I can say chocolate is my fourth favorite flavor, and vanilla is my third favorite. But I can’t measure the difference between these preferences in units. I can’t coherently say I prefer vanilla three times as much as chocolate. More importantly, there is no way to compare, in units, one person’s preference to another’s. You can’t say “Murray likes Buicks twice as much as Joe does.” (You can say Murray is willing to spend twice as much on a Buick as Joe is, but that’s a separate discussion for a separate day.)

The bottom line is: preferences are ordinal and not cardinal. You prefer one good or service to another, but — just like Frank is a better friend to you than Pete — this can’t be expressed in units (cardinal numbers).

Finally, an acting man believes he can influence the future. This is a better way of expressing the concept at issue than saying “man acts rationally” — because the word “rational” is a term that is loaded with connotations not used by Mises. For example, Murphy argues that to Mises, a rain dance is a “rational” act. Why? Because the person doing the dance has a goal (making it rain) and is engaged in action that is directed towards achieving that end.

The fact that we know the action will not accomplish the purpose does not mean the action is not “human action” (or “rational” as Mises uses the term). That’s why I like Murphy’s formulation that the man engaged in “human action” must believe that he can influence the future. This way, we can avoid judgments about whether that action is based on correct information, or whether it actually will achieve the purpose he believes it will achieve.

That summarizes Chapter 2, and is enough for today. Remember, we are still in foundational mode. Tomorrow, we will address the issue of Mises’s economics being an a priori discipline.

UPDATE: Let me say, as I should have at the outset, that any errors in these summaries are mine and not Murphy’s. I am restating his points in my language, and that opens up the possibility of inaccuracy. Should that happen, blame me, not him.

39 Responses to ““Human Action” and Robert Murphy’s “Choice,” Part 2: What Is Action?”

  1. I hope the hay is still within reach of the goats.

    Patterico (3cc0c1)

  2. So far, yup — for this goat, anyway.

    Beldar (fa637a)

  3. I always like an excuse to talk about algebraic constructions like linear orderings.

    Be alert for the attempt to add units to ordinals by embedding them in an ordered field. In the ice cream example, one might try to do so by saying “I would prefer one scoop of chocolate to one scoop of vanilla, but I would prefer two scoops of vanilla to one scoop of chocolate”. One could use that to create the units of preference v and c, and use that statement to determine that

    v < c but 2v > c

    and thus that one’s preference for vanilla is some real-valued multiple of one’s preference for chocolate, with said real number being between 1 and 2 exclusive. The reasons that this construction is not coherent are related to the “separate discussion for a separate day” about Murray, Joe, and Buicks.

    That said, I’m not aware of an argument that preferences are inherently not densely-ordered — that for any human preference among possibilities, it is always possible to find x and y where x < y such that, for every other possibility z, either z < x < y or x < y < z. This can lead to a mix-up if you keep using the term “ordinal”, since logicians reserve the term for well-orderings which are (among other things) never dense.

    CayleyGraph (dfcefe)

  4. Ordinal values, too, can change in order. I used to prefer chocolate, but now prefer vanilla. It’s not that the ice creams have changed, it’s my tastes.

    htom (4ca1fa)

  5. Ok, I’m groping here, but where this seems to be going is that an awful lot of political talk about economics treats preferances as if they were cardinal. As if on person’s preference for so,ething could be added to others and thus make a mass of preference that could be considered measured.

    C. S. P. Schofield (ab2cdc)

  6. CayleyGraph, does this problem disappear if one considers the notion that a preference may depend upon more than the flavor? For example, a person’s ice cream preference might depend upon quantity, flavor, texture, temperature, and color, and simply mashing this down to flavor doesn’t capture the complexity of the problem. A two dimensional graph of flavor versus quantity might, for example, create understandable and consistent preferences. And, of course, you could always fall back on the idea that the price of the item would tend to capture all these things, and that might be the ordinal that is sought.

    bobathome (6f310e)

  7. Ok, I’m groping here, but where this seems to be going is that an awful lot of political talk about economics treats preferances as if they were cardinal. As if on person’s preference for so,ething could be added to others and thus make a mass of preference that could be considered measured.

    Not really. I’m just summarizing basic concepts. We’re not really headed here towards some giant political revelation involved cardinal vs. ordinal.

    Ordinal values, too, can change in order. I used to prefer chocolate, but now prefer vanilla. It’s not that the ice creams have changed, it’s my tastes.

    Absolutely. Because (as noted in the post) preferences are subjective, they can change at any time. They are revealed in action, and one might prefer one flavor today (as demonstrated by his action of choosing it today) and another tomorrow. This is “demonstrated preference.”

    Patterico (3cc0c1)

  8. preferences are ordinal and not cardinal

    this makes me think of that silly half-dead kitty-cat in a box

    more and more we’re bombarded with metrics… oodles of cardinal data

    if you tell me the trade deficit is bad and as an input into the GDP formula a high trade deficit makes it harder for our food stamp fascist douchebag president to establish a good economic record, pikachu is gonna stop buying domestic wine and spirits, take the opportunity to become super familiar with cheeses of the whirl, and heavy up on the imported cleaning robots

    and this is just one example

    me I aim to misbehave

    my preferences aren’t cardinal they’re not ordinal they’re attitudinal oppositional

    happyfeet (831175)

  9. CayleyGraph,

    This is getting ahead of ourselves a bit, but units are very important when we get to the concept of exchange between two people.

    In your example of two scoops of vanilla, you can set up an ordinal preference rank for Murray as follows:

    1. Two scoops of vanilla
    2. One scoop of chocolate
    3. One scoop of vanilla

    In your example, the consumer (whom I am calling Murray) prefers two scoops of vanilla best, followed by one scoop of chocolate, followed by one scoop of vanilla.

    Now, pretend Murray has one scoop of chocolate.

    Now imagine a second actor, Milton. Milton has two scoops of vanilla. Milton’s preference rank is:

    1. One scoop of chocolate
    2. Two scoops of vanilla
    3. One scoop of vanilla

    Note that there is no talk here of units of preference — in other words, we have no idea how much more Milton prefers one scoop of chocolate to one scoop of vanilla. And we have no idea how much more intensely Milton wants two scoops of vanilla, as compared to how much Murray wants two scoops of vanilla.

    However, the concept of units (how many scoops) can be expressed in the preference rankings as shown — and the relative preference rankings show that Murray can trade his single scoop of chocolate for Milton’s two scoops of vanilla, and both will be better off. Murray will gain something (two scoops of vanilla) that he ranks higher in preference than what he has (a single scoop of chocolate). Milton will gain something (one scoop of chocolate) that he values more highly than what he has (two scoops of vanilla).

    So importing units into the analysis in this way is not only permissible but critical. But it doesn’t change the fact that the preference rankings are still ordinal. We can’t compare the intensity of Milton’s and Murray’s relative preferences, either as compared to each other or as they value one thing over another. But we do know that each is giving up something lower in their preference rank for something higher in their preference rank. So, clearly the exchange makes both better off.

    This is all fodder for a future post. No problem if readers aren’t following it now.

    Patterico (3cc0c1)

  10. I have created a category for all these posts, called “Human Action and Choice,” so that all these posts can be read (in reverse order) with a single click.

    Patterico (3cc0c1)

  11. Happyfeet, you will be pleased to know that as an acting man who believes he can influence the future, von Mises declares you to be rational! Congratulations! But don’t let this go to your head.

    bobathome (6f310e)

  12. i should at least get a participation trophy i think

    happyfeet (831175)

  13. Oh Sh*t!! I’m already almost two assignments behind. All-nighter coming, I can feel it.

    Gramps, the original (bc022b)

  14. This old goat also reaches the hay.

    Tomorrow, we will address the issue of Mises’s economics being an a priori discipline.

    As opposed to empirical, yes?

    felipe (56556d)

  15. As opposed to empirical, yes?

    Precisely so.

    Patterico (3cc0c1)

  16. I might have to reacquaint myself with my copy of The Black Swan

    Leviticus (48a857)

  17. racist

    happyfeet (831175)

  18. Er. I hate to be a dweeb, but once these technical notions get into popular usage with a wrong meaning, they never go away, and I’m seeing some suggestions in the comments that it has been misunderstood. I still cringe every time I see “epistemic closure” used as if it means “closed-minded” when it really means that your beliefs are logically sound.

    So, just to clarify, cardinal numbers are discrete just like ordinal numbers are. That is, the cardinal numbers are 1, 2, 3, etc. They do not include 1/2, pi, or any other non-whole number. I assume that what Murphy is getting at is that you can’t add, subtract, multiply, and divide ordinal numbers because it makes no sense to ask “what’s third plus fifth?”

    However, the difference between cardinal numbers and ordinal numbers doesn’t really have anything to do with measurement or how things are compared. If you want to use mathematical jargon to make that point it would be better to say that preference relationships are a discrete partial order rather than a metric space. But of course it would take longer to explain what that means than it does to explain that you can’t use arithmetic to compare “amounts” of preference.

    Cugel (e574ce)

  19. Gramps, I just got Mark Steyn’s A Disgrace to the Profession, got the Kindle version of Murphy last night, and I’m still digesting several others. So I’ll be up late tonight also! The good news is that the Murphy chapters are fairly short. Skipping the introductory matter which is very interesting, the first three chapters are just 20 pages. The bad news is that Steyn annotates each chapter putting the remarks in context, and I’m finding it fascinating. If you listen to Steyn, he’s been talking about two trees in Quebec’s Gaspe Peninsula that account for the steady temperature in the hockey stick between 1400 and 1447. It turns out this nugget is from a seminar McKitrick gave at U. Texas in ’05 (based on the citation.) Now I’ve got to download that!

    bobathome (6f310e)

  20. I recently reread the Captain Blood stories. Very enjoyable. I don’t remember, was it The Black Swan or The Sea Hawk which was dark?

    (I’m kidding.)

    nk (dbc370)

  21. Cugel,

    I should have said in the post (and will say in future posts) that any errors in these summaries are mine and not Murphy’s. I am trying to restate things in my own words, and when I do that, I may get things wrong.

    That being said, I don’t understand your objection. You say “cardinal numbers are discrete just like ordinal numbers are” as if you are contradicting me, but I did not use the word “discrete” anywhere in the post or in previous comments.

    When you say this:

    I assume that what Murphy is getting at is that you can’t add, subtract, multiply, and divide ordinal numbers because it makes no sense to ask “what’s third plus fifth?”

    I think that’s exactly right. In fact, Murphy does say almost exactly that on page 30. But then (unless I am misunderstanding you) you seem to contradict yourself when you say:

    the difference between cardinal numbers and ordinal numbers doesn’t really have anything to do with measurement or how things are compared.

    Well, doesn’t it? Cardinal numbers, as you just said, can be used in mathematical operations, which (when used to describe preferences) implies the possibility of measurement. Ordinal numbers describe the numerical position of an object (or, here, of a preference) and cannot be used in mathematical operations or used to measure “units” of preference.

    In short, I don’t understand your objection.

    Patterico (3cc0c1)

  22. UPDATE: Let me say, as I should have at the outset, that any errors in these summaries are mine and not Murphy’s. I am restating his points in my language, and that opens up the possibility of inaccuracy. Should that happen, blame me, not him.

    Patterico (3cc0c1)

  23. I believe Cugel, in using the word “discreet”, made you think you he was putting words in your mouth, when I think he really meant to say “integral”, as in integers. The word “discreet”, then, refers to the integral property of both the cardinal and ordinal number systems.

    felipe (56556d)

  24. It’s interesting: the concepts I discussed in comment 9 above, I have now reached in my series of posts. I am up to chapter 7 (and therefore post 7) in the book, meaning that when I am done I will have a post a day at least through Friday. Right now I am wondering whether to use Murphy’s example in the book, or to strip it down to an even simpler example as I do in comment 9 above.

    Patterico (3cc0c1)

  25. Gahh! made you think you he was…

    I blame the absence of participation (a)trophies in my life.

    felipe (56556d)

  26. I think you will be well served in continuing your practice (#9). Good advice to any lawyer!

    felipe (56556d)

  27. If you find Murphy “heavy going.” try Pr. George Reisman. He has plenty of practice forcing basic economic principles into thick skulls (I attended his seminar in graduate school) and will repeat a concept several times in as many different ways. I have five of his books on my Kindle and still learn something new with each reading.

    Michael Keohane (9d2d81)

  28. Now, a jaded fella might take my previous comment as way of saying “don’t quite yer day-job.” That fella would be mistaken! I clearly meant that you should continue using your own examples.

    felipe (56556d)

  29. If you find Murphy “heavy going.” try Pr. George Reisman. He has plenty of practice forcing basic economic principles into thick skulls (I attended his seminar in graduate school) and will repeat a concept several times in as many different ways. I have five of his books on my Kindle and still learn something new with each reading.

    I have read sections of “Capitalism” (available online for free) and enjoy it quite a bit. I think he has some unique ideas about classifying profits and such, and he has helped teach me about the crucial importance of capital.

    Patterico (3cc0c1)

  30. Now, a jaded fella might take my previous comment as way of saying “don’t quite yer day-job.” That fella would be mistaken! I clearly meant that you should continue using your own examples.

    I already incorporated one of his charts in a draft post. I agree with you, though — I think I will try to use my own examples for this one. In part, what I am doing is assuring myself that I have absorbed the material in the book, by restating it. If I do my own examples, that will help me know that I am truly internalizing the concepts.

    Patterico (3cc0c1)

  31. Patterico, I didn’t have any real objection to your description; it was some of the comments that made me think people were conflating cardinal numbers with the real numbers or something similar.

    Cugel (e574ce)

  32. (You could — and should — use your own example in post m, and if the discussion shows great confusion because of your example, post m+1 be the more elaborate example. I doubt that would be needed, though, you do a good job of these things.)

    Cardinal numbers are a subset of the integers. Ordinal numbers are in some ways not numbers at all.

    htom (4ca1fa)

  33. I finished part 7 just now, and am very happy with the examples (I did use my own) and with the post as a whole. I think it is my favorite post of the series so far.

    Patterico (3cc0c1)

  34. I am certainly no goat, but I am keeping up nicely. I like your examples, Patterico. And I remain unclear on your point, Cugel.

    Dana (86e864)

  35. This is great that you are doing this. You are correct in saying that reading Human Action is a difficult undertaking. I did a year and a half of back ground reading before I thought I could even tackle it. But it was well worth the time. One point of clarification. Human action studies the action itself. It is not psychology. Human action does not study the inner forces that make a man act. It studies human action as purposeful behavior toward a specific end, and not why the person chose the action. A man acts because he thinks his action and the means he chooses will produce the desired end. After the fact it can be judged if he was correct or incorrect as long as you know what his end actually was. If the persons desired end is known before he acts, the means he attempts to use to bring about this desired end can be analyzed before they are implemented for their ability to bring about the end. An example is when a politician wants to help low wage workers by raising the minimum wage. We know before he acts that this end will not be brought about using the means the politician has chosen. I’m looking forward to reading the rest of your posts on Murphy’s book.

    Fred (3e71fa)

  36. Fred — that assumes that the actor knows what he’s doing. Sometimes one does a thing thinking that that is the step towards the expressly desired result, while actually it’s a leap away from that result and towards a result that is unknowingly desired. I suspect that’s why psychology is ignored; we look at the outputs of the psychology, the actions, discarding both the motives and the expectations. What people do is hard enough.

    htom (4ca1fa)

  37. … I can’t coherently say I prefer vanilla three times as much as chocolate. …

    What is wrong with the utility function framework in which you would say something like “I am indifferent between one scoop of chocolate and an one third chance of one scoop of vanilla”?

    James B. Shearer (84bde5)

  38. I think that most social scientists would say that preferences are ordinal, and neither interval (i.e., one can’t measure their differences) nor ratio (i.e., one can’t measure their quotients).

    David Pittelli (b77425)

  39. One point of clarification. Human action studies the action itself. It is not psychology. Human action does not study the inner forces that make a man act. It studies human action as purposeful behavior toward a specific end, and not why the person chose the action.

    Can you explain what I said in the post that caused you to issue this “point of clarification”? I’m pretty sure I didn’t say Mises was talking about psychology.

    Patterico (3cc0c1)


Powered by WordPress.

Page loaded in: 0.2915 secs.