Home > Uncategorized > Polling and nonergodic stochastic processes

Polling and nonergodic stochastic processes

The latest polls showing Clinton would win the presidential election is more evidence to support my argument that trying to predict human behavior  — whether it involves economic decision making or politics decision making — involves dealing with a nonergodic stochastic process.

Pollsters believe that if you take a RANDOM SAMPLE OF THE VOTING POPULATION ON DAY X  BEFORE THE ELECTION  and calculate the probability distribution of voting for the various candidates, then this x day before election probability distribution is equivalent to the probability distribution you will get  from any sample drawn from the same population universe on election day.  This presumption would be correct if the stochastic process generating voters preferences is ergodic.

But as I have argued regarding economic decisions — drawing a sample on day x provides no reliable probability distribution of what economic decision makers will do on day x + y.


When will I convince professional economists and political scientists, that the future is uncertain regarding their discipline???



  1. November 10, 2016 at 9:55 pm

    Isn’t what you demonstrate here a form of the Problem of Induction (as per Hume, Sextus Empiricus, Popper…)?

  2. Paul Davidson
    November 10, 2016 at 10:55 pm

    It solves the problem of Induction. If the stochastic process generating the data [the facts] is nonergodic, then it means, by definition, the probability distribution that fit the past facts is not representative of they probability distribution that will fit the fact in the future.l

    • November 11, 2016 at 4:15 am

      So for nonergodic data no induction is possible, by definition. Correct? But presumably some data can be ergodic, for which induction is possible. If I am right so far then the key problem is how to differentiate between the two, since only ergodic data is useful for planning the future.

      • paul davidson
        November 11, 2016 at 5:27 pm

        there are some statistical tests time series data. Since Stationary is a necessary (but not sufficient) condition for ergodicity, unit roots test can test if data is nonergodic.

  3. Alan
    November 11, 2016 at 1:40 am

    Anyone who bothers to get off their rear end and actually talk to people would make more sense. It’s what Adam Smith did (not that he was a bad mathematician but there’s a line in Wealth somewhere where he mocks people who take arithmetic too seriously.) It’s what those supposed pseudoscientists, ethnographers and qualitative sociologists, much mocked by neoclassical economists do. If you want to understand how to understand human behavior, start with Gilbert Ryle. Then take in a little Wittgenstein if you are feeling brave.

    It’s also presumably what allows someone like Michael Moore to call the election correctly: Five Reasons Why Trump Will Win

    From Green Bay to Pittsburgh, this, my friends, is the middle of England – broken, depressed, struggling, the smokestacks strewn across the countryside with the carcass of what we use to call the Middle Class. Angry, embittered working (and nonworking) people who were lied to by the trickle-down of Reagan and abandoned by Democrats who still try to talk a good line but are really just looking forward to rub one out with a lobbyist from Goldman Sachs who’ll write them nice big check before leaving the room. What happened in the UK with Brexit is going to happen here. Elmer Gantry shows up looking like Boris Johnson and just says whatever shit he can make up to convince the masses that this is their chance! To stick to ALL of them, all who wrecked their American Dream! And now The Outsider, Donald Trump, has arrived to clean house! You don’t have to agree with him! You don’t even have to like him! He is your personal Molotov cocktail to throw right into the center of the bastards who did this to you! SEND A MESSAGE! TRUMP IS YOUR MESSENGER!

  4. Alan
  5. November 11, 2016 at 5:30 pm

    I note that it is not rational, in the economic sense, to vote if you are confident that your preferred candidate is going to win anyway. So a poll can only reflect current intentions, which may change.

    This effect may have been operative for BREXIT, where young people favoured remain but relatively few voted. On the other hand, Trump and others seem to think that if they say they are ahead then they are likely to get more votes. Either way, the relationship between polls and outcomes is not one of straightforward approximation..

  6. patrick newman
    November 12, 2016 at 12:55 pm

    I do hope this self styled personal antidote to the excesses of neo-liberism and globalism doesn’t disappoint the done downs by turning out to be a highly skilled purveyor of quackery!

  1. No trackbacks yet.

Leave a Reply to Michael Kowalik Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.