Tuesday, March 1, 2011

Keynes, time, and uncertainty

Keynes thought very deeply about the significance of two issues for economics: time and uncertainty (which, when you consider them together, we often talk about in terms of "expectations"). What I find interesting is that he was not the first to talk about these problems - it was part of the intellectual milieu of the late nineteenth and early twentieth century. In that sense, one way to think about the "Keynesian revolution" is that Keynesianism could not happen until thinking about time and uncertainty had developed sufficiently to mix with Malthusianism, mercantilism, and various other non-Ricardianisms. The froth of thought on time and uncertainty enabled the Keynesian solution to much earlier problems.

Work on the essence and implications of time was widespread. Leibniz, Kant, and Newton had all passed the buck when it came to thinking about time - but that all began to unravel. Martin Heidegger made time and our existence in time a new basis for ontology. Henri Bergson was talking about the implications of time for free will, and like Heidegger, had a concept of our projected existence in time (Duration). Einstein, of course, deconstructed the Newtonian system by reconceiving of time completely. Einstein had some famous disputes with Bergson over precisely the question of time in the early 1920s. McTaggert was also one that questioned the traditional understanding of time in this period. By the late 1920s our thinking about time had been completely revolutionized. This all coincided nicely with the elimination of localized measurements of time. Much of this discussion occurred between 1884 at the Univeral Meridian Conference and 1928, when Universal Time was adopted to replace Greenwich Mean Time. A very poetic set of book ends.

Thinking about uncertainty was also very common in this period. I've previously gone over pragmatist contributions to this discussion in the late 19th century. C.S. Peirce's work in probability is in many ways a pre-cursor to Keynes (and in many way far exceeded it according to some authors). Later pragmatists would see connections between the implosion of rationalism and the work of the phenomenologists (again - Heidegger) and the pragmatists critique of certainty. Each branch of philosophy was stumbling upon the same problem of fundamental uncertainty that pragmatism had emerged to deal with. People don't want to just stand in a stupor when faced with uncertainty - they want to still act in a reasonable way. So it's not surprising that we also saw the growth of probability and statistics at this time. Least-squares methods had already been developed, but Pearson, Markov, Fisher, etc. Discussion of uncertainty was also fruitful in physics, with the work of Heisenberg and Bohr.

To a large extent, modern science is science that has incorporated problems of time and uncertainty. Darwin got the ball rolling by considering the implications of deep time and randomness. This is also the foundational difference between modern physics and Newtonian physics - the integration of time as something more than a track that matter runs on, and the abandonment of strict determinism. This is also modern economics: time, uncertainty, and expectations. These ideas have been incorporated in many ways, but the man most responsible for re-envisioning the entire discipline along these lines was John Maynard Keynes.

Another early 20th century writer noted this too, of course. A writer that noted both that "conflict with time seems to me the most potent and fruitful theme in all human expression" (1933) and "the oldest and strongest emotion of mankind is fear, and the oldest and strongest kind of fear is fear of the unknown" (1927) is clearly a writer for the modern era.


  1. "These ideas have been incorporated in many ways, but the man most responsible for re-envisioning the entire discipline along these lines was John Maynard Keynes."

    And, you might add, the macroeconomic theory which takes Keynes' ideas seriously and has built on them in constructive ways is Post Keynesianism.

    Both Neoclassical synthesis Keynesianism and New Keynesianism are precisely theories that betrayed Keynes' theory by reverting to the flawed neoclassical axioms:

    (1) the ergodic axiom (which denies radical uncertainty);
    (2) gross substitution axiom, and
    (3) neutral money axiom.

    And also one aspect I think you should mention; it is in what mathematicians call non-ergodic stochastic systems that this concept of uncertainty applies, like the weather, climate, financial markets, economies.

    In these systems, past data is not a useful tool for predicting the future state of the system and the problem of induction is particularly acute.

    But the fundamental point is that it is still possible to reduce uncertainty in these systems, or at least in theory in some of them.

    It is entirely possible that in the future - with a far more advanced human civilization - we could use technology to control weather or climate - preventing an ice age, for example.

    And even today a powerful entity like the government can intervene to reduce uncertainty in the non-ergodic stochastic system we call the economy.

  2. I, too, recommend reading Paul Davidson - in particular:

    Paul Davidson "Is Probability Theory Relevant for Uncertainty? A Post Keynesian Perspective". Journal of Economic Perspectives V. 5, N. 1 (Winter 1991): 129-143.

    And then there is Keynes' 1937 paper quoted here.


All anonymous comments will be deleted. Consistent pseudonyms are fine.