Thursday, November 24, 2011

Degrees of Uncertainty

There is an interesting review here by Michael Emmett Brady of Gerald P. O’Driscoll and Mario J. Rizzo’s The Economics of Time and Ignorance (2nd edn; Routledge, Oxford, UK., 1996):
http://www.amazon.com/review/R25I26NGH79Y3I
Towards the end, Michael Emmett Brady states that
“Let us now turn to p. 9. The authors make the bizarre claim that Paul Davidson’s Shacklean, Post Keynesian views, that there is only certainty and uncertainty, is a continuation of those parts of the [General Theory] … that represent a true subjectivist position. Unfortunately, Shackle never got past chapter 3 of the [Treatise on Probability] … and rejected Keynes’s entire [Treatise on Probability] … approach in toto. Neither Shackle nor Davidson are followers of Keynes because Keynes totally rejected nihilism and the bizarre Shackle-Davidson claim that there are no degrees/gradations of uncertainty which Keynes spelt out clearly in his 1937 QJE article, titled ‘The General Theory of Employment’. Uncertainty is a range that can be specified in the same identical manner as Keynes’s weight of the evidence index, w. Complete and total uncertainty (ignorance) would have a w = 0. The differing gradations of uncertainty would be between 0 and 1 (0 < w < 1). Complete certainty would have a w = 1. The two authors never specify what their term ‘genuine’ uncertainty is. Their discussions of the beauty contest example (p. 156, GT), which is a continuation of a discussion started by Keynes in his introductory chapter 3 of the [Treatise on Probability] … in 1921 (1908) on the measurement of probabilities, does not fit the bill.”
I have just re-read Keynes’s paper “The General Theory of Employment” (Quarterly Journal of Economics 51 [1937]: 209–223). The crucial passage where Keynes talks about degrees of uncertainty is here:
By ‘uncertain’ knowledge, let me explain, I do not mean merely to distinguish what is known for certain from what is only probable. The game of roulette is not subject, in this sense, to uncertainty; nor is the prospect of a Victory bond being drawn. Or, again, the expectation of life is only slightly uncertain. Even the weather is only moderately uncertain. The sense in which I am using the term is that in which the prospect of a European war is uncertain, or the price of copper and the rate of interest twenty years hence, or the obsolescence of a new invention, or the position of private wealthowners in the social system in 1970. About these matters there is no scientific basis on which to form any calculable probability whatever” (Keynes 1937: 213–214).
It is indeed possible to see the concept of degrees of uncertainty, and Barkley Rosser describes Keynes’s views on the probability of future events in A Treatise on Probability (1921: 33), and how uncertainty comes in different forms:
“(i) ... there are no probabilities at all (fundamental uncertainty),
(ii) ... there may be some partial ordering of probable events but no cardinal numbers can be placed on them,
(iii) ... there may be numbers but they cannot be discovered for some reason, and
(iv) ... there may be numbers but they are difficult to discover” (Barkley Rosser 2001: 559).
There is also a literature about degrees of uncertainty, some of it by Post Keynesians (Runde 1990; Dow 1995; Crocco 2002; Dequech 1997).

At this point it is notable that Paul Davidson uses the concept of non-ergodicity to reject the view that there are degrees of uncertainty (Crocco 2002: 19). Nevertheless, there are Post Keynesians who recognise degrees of uncertainty, e.g., S. C. Dow (1994: 437). Moreover, here is Jesper Jespersen:
“Uncertainty is caused by lack of information. Therefore uncertainty might have different intensities or ‘stats of confidence’. You may feel(!) more or less uncertain, but except for rare cases all individual activities are characterized by (different degrees of) uncertainty, because one cannot know nor estimate the exact outcome. Hence, expectations are uncertain due to this inherent lack of information (and a constantly changing environment).”
Jesper Jespersen, “Post-Keynesian economics: uncertainty, effective demand & (un)sustainable development,” Paper, Dijon-conference, Dijon, 10-12 December 2009. p. 8.
It seems to me an exaggeration to say that the Post Keynesian school does not recognise degrees of uncertainty. Clearly not everyone follows Davidson.

BIBLIOGRAPHY

Barkley Rosser, J. 2001. “Alternative Keynesian and Post Keynesian Perspectives on Uncertainty and Expectations,” Journal of Post Keynesian Economics 23.4: 545–566.

Crocco, M. 2002. “The Concept of Degrees of Uncertainty in Keynes, Shackle, and Davidson,” Nova Economia 12.2: 11–28.

Dequech, D. 1997. “Uncertainty in a Strong Sense: Meaning and Sources,” Economic Issues 2.2: 21–43.

Dow, S. C. 1994. “Uncertainty,” in P. Arestis and M. Sawyer (eds.), The Elgar Companion to Radical Political Economy, Elgar, Aldershot. 434–438.

Dow, S. C. 1995. “Uncertainty about Uncertainty,” in S. C. Dow and J. Hillard (eds), Keynes, Knowledge and Uncertainty, Edward Elgar, Aldershot. 117–127.

Keynes, J. M. 1921. A Treatise on Probability (1st edn.), Macmillan, London.

Keynes, J. M. 1937. “The General Theory of Employment,” Quarterly Journal of Economics 51: 209–223.

Runde, J. 1990. “Keynesian uncertainty and the weight of arguments,” Economics and Philosophy 6.2: 275–292.

13 comments:

  1. I think that Brady is more critical of their lack of a proper understanding of uncertainty itself. He has criticised Jochen Runde as a "Keynesian fundamentalist" in this excellent working paper, which argues that Keynes should be recognised as a modern, founding pioneer of the first non-linear and non-additive approach to probability.

    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1546726

    Only a small minority of specialists in mathematics understand how George Boole's Last Challenge Problem (which Keynes based his theory of probability upon) works.

    According to Brady, an organisation of mathematicians, SIPTA (Society for Imprecise Probability - Theories and Applications) has replicated Boole and Keynes with a "Choquet integral" type of approach.

    As for the uncertainty bit, Jesper Jespersen also does not provide a technical analysis. He may distinguish between different degrees of uncertainty, but because the "Keynesian fundamentalist" position rhetorically overstates the weight of evidence being nil, Post Keynesianism kind of shoots itself in the foot.

    ReplyDelete
  2. There is a very fascinating comment here from Brady:

    both Adam Smith and J M Keynes would have expected such a result to occur for certain since it can be completely predicted with a weight of evidence of 1 and a probability of 1, given that speculative behavior always generates a bubble that will always collapse. The result is perfectly predictable once speculators come to dominate over entrepreneurs in any economy, be it past, present or future. It is simple to predict that the Obama bubble will also inevitably collapse, although not before the 2012 election. Paul Samuelson's correct prediction, made in the 1996 edition of Kindleberger's classic book on Manias, Panics and Crashes,that readers would be kicking themselves within 5 years if they ignored Kindleberger's analysis, is an example of the accuracy that can be obtained once it is realized that speculative excess is ergodic and the collapse is perfectly predictable. Samuelson appreciated Mandelbrot's work and studied it very carefully. Samuelson's understanding of Mandelbrot's work stands in great contrast to the ignorance of Paul Davidson. Samuelson correctly realized that financial markets are ergodic once speculators start to bid up prices. The crucial point is that the distributions are not normal or log normal. The correct policy is to prevent the speculators from starting to bid up prices in the first place by preventing them from having access to credit/bank loans.

    http://www.amazon.com/Maynards-Revenge-Collapse-Market-Macroeconomics/product-reviews/0674050460

    That bears further investigation.

    The claim that

    (1) once speculators come to dominate over entrepreneurs in financial asset markets, then

    (2) there will be a "weight of evidence of 1 and a probability of 1" that a bubble that will occur and always collapse.

    (3) therefore "financial markets are ergodic once speculators start to bid up prices."

    Intriguing.

    And presumably under other circumstances financial markets are non-ergodic?

    ReplyDelete
  3. You'd have to ask the man himself, he has a better knowledge of mathematics than I do, but from what I've read of Kenneth Rogoff and Carmen Reinhart's "This Time is Different" (which Brady has also reviewed), it looks like speculation-induced slumps are highly commonplace. But I can't say for other financial circumstances, given my understanding of mathematics is not as strong as Brady's.

    ReplyDelete
  4. The correct policy is to prevent the speculators from starting to bid up prices in the first place by preventing them from having access to credit/bank loans.

    It is impossible to separate speculators from non-speculators, since all individuals in the market are speculating about the uncertain future when they act towards attaining future ends.

    Brady only seems to vaguely grasp the real problem, which is not that "speculators" receive bank loans and credit, it's that the bank loans and credit are being created ex nihilo, and are the very fuel that generates malinvestment and credit expansion induced economic bubbles, for example the housing bubble and collapse that just took place. Over $2 trillion was created out of nothing from 2001-2006, and much it poured into the housing market, bidding up both housing prices and derivative securities prices that depend on housing prices.

    Brady is unable to understand that regulations and rules cannot stop credit expansion bubbles, because of the fundamental uncertainty and subjective expectations on the part of the individuals asked to be regulators.

    "In the sphere of human action there are no constant relations between any factors. There is consequently no measurement and no quantification possible. All measurable magnitudes that the sciences of human action encounter are quantities of the environment
    in which man lives and acts. They are historical facts, e.g., facts of economic or of military history, and are to be clearly distinguished from the problems with which the theoretical science of action—praxeology and especially also its most developed part, economics—deals."

    and

    "They fail to realize that in the field of human action statistics is always history and that the alleged "correlations" and "functions" do not describe anything else than what happened at a definite instant of time in a definite geographical area as the outcome of the actions of a definite number of people. As a method of economic analysis econometrics is a childish play with figures that does not contribute anything to the elucidation of the problems of economic reality." - Ultimate Foundation of Economic Science, Mises.

    ReplyDelete
  5. Out of curiosity Lord Keynes, will you read Benoit B. Mandelbrot? Brady has given the late maverick mathematician and father of fractal geometry high marks. Mandelbrot has also dealt a lot with risk and uncertainty through his techniques and writings.

    http://www.amazon.com/review/R3DE4D5MSQK7GL/

    ReplyDelete
  6. Another thing to add to my "to read" list.

    ReplyDelete
  7. ""In the sphere of human action there are no constant relations between any factors. There is consequently no measurement and no quantification possible. "

    A quotation which shows the sheer level of idiocy to which Mises ascended.

    If there are no "constant relations between any factors" then the whole foundation of Austrian economic collapses like a house of cards. For there would be no credible reason to think that

    (1) entrepreneurs will continue to seek profit
    (2) the price mechanism will continue to function
    (3) the disutility of labour will continue to hold
    (4) entrepreneurs will bother to make capital goods investments at all.

    ReplyDelete
  8. Also, here is Michael Emmett Brady's review of "This Time is Different: Eight Centuries of Financial Folly" by Carmen Reinhart and Kenneth Rogoff.

    http://www.amazon.com/review/RNX0YJ4YH0R0M

    ReplyDelete
  9. "In the sphere of human action there are no constant relations between any factors. There is consequently no measurement and no quantification possible."

    A quotation which shows the sheer level of idiocy to which Mises ascended.

    A response to Mises which shows the sheer level of ignorance to which you are are descending.

    If there are no "constant relations between any factors" then the whole foundation of Austrian economic collapses like a house of cards.

    For there would be no credible reason to think that

    (1) entrepreneurs will continue to seek profit

    (2) the price mechanism will continue to function

    (3) the disutility of labour will continue to hold

    (4) entrepreneurs will bother to make capital goods investments at all.

    The level of your ignorance of Mises is astounding. You obviously have not read Austrian economics.

    This list you posted is all a product of choice, which is not based on any constants. The point of the argument "there are no constants in the field of human action" is that there are no constant relations between factors that can enable you to know that I, or someone else, will even be an entrepreneur tomorrow the way we are one today.

    By saying "entrepreneur", you are ALREADY presupposing the fact that someone is making an investment. The CHOICE to make investments however is what is not based on constant relations. There are no constant relations between factors that determine where, when, and in what type, investments will be made.

    Investments are based on a learning process, and learning is fundamentally unpredictable and cannot be known through analysing any alleged constant relations between economic factors.

    Furthermore, the notion that the price system will fail if there are no constant relations is also completely wrong. The price system is based on choice and values, which contain no constant relations. One cannot learn how to be successful in the market by reading textbooks. If it did, then economists would be the wealthiest people in the world. But they're not, because it cannot be taught like chemistry and physics. It is an art.

    Human action is based on choices, which can and do change, according to the knowledge we acquire, which is itself unknown and unpredictable at the outset, or else we would have the knowledge already.

    It is quite clear that you have absolutely no clue what the argument "there are no constancies in human action" actually means. You have no clue because you have not read the source material.

    You should actually READ the source material before you shoot your mouth off.

    The sources for understanding this argument are the following:

    "Economic Science and the Austrian Method" - Hoppe.

    "The Ultimate Foundation of Economic Science" - Mises.

    "Human Action" - Mises, especially Chapter 12, section 4:

    http://mises.org/humanaction/chap12sec4.asp

    ReplyDelete
  10. "Furthermore, the notion that the price system will fail if there are no constant relations is also completely wrong. The price system is based on choice and values, which contain no constant relations."

    Really? The biological fact that humans require food doesn't must mean that no constant relations exist between human need for food and demand for food. That we have no reason to think that humans will all continue eating/demanding food tomorrow?

    ReplyDelete
  11. Really? The biological fact that humans require food doesn't must mean that no constant relations exist between human need for food and demand for food.

    Biological needs is not what the "no constancy" argument in human ACTION is referring to.

    The no constancies in human action are related to purposeful behavior, to choices. Not the fact that humans need air and food to live.

    You're dropping the context on the basis of not having read or understand the literature.

    If the no constancy argument is going to relate to food and eating, then the no constancy argument would say that there is no observable or derivable constant relations between factors that can enable one to know at the outset when, where, and what I will eat in the future. You cannot gather statistics and say that there is a constant relation between those statistics and my future choices regarding food.

    That the price system is based on choice and values, which contain no constant relations, is an argument that says there are no constant relations concerning choices or values, which are categories of action, not biological needs.

    You can't know when I will eat, where I will eat, nor what I will eat, based on any scientific laws that concern constant relations between variables.

    For example, if yesterday I ate a ham sandwich for lunch at 12:23 pm, in New York, at the Deli on 34th and Main, and I paid $4.50 for it, if you take all this type of information, the no constancy argument is that there is no scientific law that can connect this historical data, via some constant relation, to my future choices and values concerning food. You cannot take the information about what I did at that time and place, and discern some scientific law that enables you to know that I will choose X later on.

    You are not understanding the argument because you are not educated in the source material I posted. Please read them, and THEN consider yourself informed enough to make a relevant response.

    ReplyDelete
  12. Your original statement:

    "In the sphere of human action there are no constant relations between any factors. There is consequently no measurement and no quantification possible"

    That is pure rubbish, as I said previously.
    People eat, sleep, spend their income, require health care, need child care, dental care, etc. There are any number of commodities the demand for which is relatively inelastic.

    Now you say:

    "The no constancies in human action are related to purposeful behavior, to choices."

    So my decision to buy food for the children today is in no sense a constant relation of human action, huh? We can't predict that the vast majority of parents will continue to buy food for their chidlren in the future? All parents are just as likely to stop buying food for their children tomorrow and let them starve?

    You are a jackass, "David".

    ReplyDelete
  13. Your original statement:

    "In the sphere of human action there are no constant relations between any factors. There is consequently no measurement and no quantification possible"

    Yes, in the sphere of HUMAN ACTION. Biological necessities, and unconscious biological events like heartbeats and breathing, are not in the sphere of human choices and human actions.

    That is pure rubbish, as I said previously.

    It is not pure rubbish just because you say it is pure rubbish. You haven't even read the course material, so your opinion is uninformed.

    People eat, sleep, spend their income, require health care, need child care, dental care, etc.

    You're not getting it. These are choices. These are purposeful actions. These are not determined by constant relations between factors. There is no equation, no constancy relation, that says something like A + B + ... = I act to eat a roast beef sandwich at 12:34 pm in three days, paying $4.50, in New York, at such and such location.

    My actions in three days is not going to be knowable through any constant relations between factors A, B, etc, and the action in question.

    There are any number of commodities the demand for which is relatively inelastic.

    Irrelevant.

    Now you say:

    "The no constancies in human action are related to purposeful behavior, to choices."

    So my decision to buy food for the children today is in no sense a constant relation of human action, huh?

    You know, you'll never learn this if you keep trying to refute straw men all day long. You'll remain ignorant.

    Your decision to buy food today will not be based on any constancy relation between economic factors throughout time. Your decisions will be based on the knowledge you acquire, and your values you decide upon, both of which are fundamentally unpredictable and cannot be known through constancy assumptions the way E=mc^2 is so based.

    We can't predict that the vast majority of parents will continue to buy food for their chidlren in the future?

    You can't predict what a parent will do based on any constancy assumptions, no. The very proof of this is that you have no clue who will be in the "vast majority" category. You don't know if a parent will not feed their children tomorrow based on some constant relations between factors that can enable you to know what they will do at or before they know what they will do. You must wait until after they act before you can know what they do.

    You can guess, you can speculate, you can anticipate, but it will not be scientifically based knowledge based on constant relations of factors that you can put in a formula.

    All parents are just as likely to stop buying food for their children tomorrow and let them starve?

    More straw man. When you say "likely", that means nothing. Either a parent will or they won't. Assigning a "99.9%" statistic to it is just a result of your lack of knowing any constancy relations in human action. You are guessing, by making inferences based on past historical choices.

    You can't know what the demand for sugar will be in a year, by finding constant relations. We can only speculate.

    You are a jackass, "David".

    When you can't make a correct or informed argument, you insult. How nice.

    Please read the course material I posted, because you are not going to be able to understand this through mystical revelation, the way you are going about it now.

    ReplyDelete