Monday, June 1, 2015

More Mathiness

Dave Andolfatto has a nice take on Paul Romer and mathiness.

11 comments:

  1. Probably better if economists just stuck to reporting the facts. Eg: There was a massive financial crisis in 2008. Banks were bailed out by the Federal Reserve, and this was followed by an extended period of subpar growth, stagnant wages, low inflation, low interest rates and low employment.

    Everything else is just speculation, and no amount of math can help. Steve's speculation is as entertaining as Krugman's but lets not kid ourselves either they or anyone else actually knows anything about the causal mechanisms that underlay this sequence of events.

    history guy

    ReplyDelete
    Replies
    1. Is this your opinion of all science, or do you confine your disdain to economics? Maybe you think history is OK?

      Delete
    2. (Different history guy here)

      Documenting exactly what happened, we can actually understand a lot about the causal mechanisms and remove a lot of speculation. I would not go as a fas "Everything else is just speculation, and no amount of math can help". But I agree with the history guy in that there are major limitations in what maths and abstraction can achieve, the balance is a bit out and more time should be spent on concentrating on the facts and evidence.

      For example an ISLM analysis does not tell me how a country gets into a liquidity trap. In the Japanese case this was caused by deflationary psychology and risk aversion and restructuring in the banking sector which led to an increase cash deposits with the central bank, while small businesses once reliant on smaller FIs were denied access to funds. This is well-documented and we can be sure about this. A gadget is simply not going to tell us how and why something happened and how we are going to deal with it. You have to have a thorough examination of the historical record. This is where 90 per cent of the analysis should be. Not in a fictitious model. It should be based on a lot of fieldwork (literally going out to banks and businesses etc) and historical documentation. What I disagree with, for example, is the Krugman approach where whenever there is a crisis you just dust off a model off the shelf.

      Delete
    3. @ History guy: To be blunt, what you say first fairly unrelated to the methodological discussion the Romer blog posts have triggered and your speculation claim is plain bullshit.

      Krugman actually predicted what you label as facts, namely that unemployment and low interest rates will persists for quite some time. Krugman on the other hands uses a mess of theoretical models and knows nothing about finance. This is where economists like Williamson step in.

      Now economists fight with each other all the time about which theory is best and how to construct/interpret data well (economics is a social science so the data is often ambiguous at best). I might consider the theory of another economist to be useless but I would never call it "speculation". Because unlike some guy who just speculates the theoretician at least lays out his thoughts and while maths is indeed overdone in econ it does provides the best way we have to formailze or thoughts and check them for errors.

      Sorry, history guy, but pre-Samuelsonian, math-less econ would be, as you as historian could hopefully tell, a step backwards.

      Delete
    4. Typical -- because history professors are scientifically illiterate, they think everyone else is too. Go back to reading Howard Zinn, angry little man, and leave the hard stuff to the grownups.

      Delete
    5. Stephen, further to your last blog, a social scientist that had been in a subject that had gone through the Critical Theory revolution would never say "use a model/device/theory to organise my thoughts". Of course a post-modernist would demolish that in a second.

      Sure one of Krugman's gadgets or Sargent's stuff might be able to 'explain' the facts (and if they don't they can be made or interpreted to do so). But probably a lot of other things can as well. And almost certainly you have missed the most crucial thing.

      Theories are reference points, AFTER you have done a comprehensive and ground-up investigative analysis. The facts initially organise your analysis, nothing else. Models and theories should not dictate the analysis, because if they do, they and not the historical facts have predetermined your answer.

      Delete
    6. It should be based on a lot of fieldwork (literally going out to banks and businesses etc) and historical documentation. What I disagree with, for example, is the Krugman approach where whenever there is a crisis you just dust off a model off the shelf."

      OK, history guy #2, tell us first what you intend to find out at all those banks and firms you intend to visit (do yourself a favour and read some statistics before you embark on this journey lest you spend the rest of your life visiting hundreds of companies) and second what you intend to find out about MACRO by talking with CEOs and taking a look into their books or whatever.

      About Krugman, he actually did write a paper when the Japanese crisis hit. Now you can disagree with that paper (first you have to read it though) but you cannot make stuff up and pretend that he used an old model.
      And this is not just so in the case of Krugman. Mainstream macroeconomists added financial stuff to DSGE models after the financial crisis. We can again debate the merit of that but what you claim is simply a lie. Economists do new things all the time.

      Delete
    7. "For the point he makes about the implications even of perfectly well-informed and rational consumers was and as far as I know still is totally misunderstood by freshwater economists, who throughout this debate have given every sign of not understanding their own models...but aside from exposing the intellectual decline and fall of the Chicago School... (Paul Krugman Jun 3, 2015)"

      Your response seems uncharacteristically muted?

      Delete
    8. For the point he makes about the implications even of perfectly well-informed and rational consumers was and as far as I know still is totally misunderstood by freshwater economists..but aside from exposing the intellectual decline and fall of the Chicago School..."

      Krugman has really thrown down the gauntlet. There must be a reason he is so confident in doing so.

      Also where are Lucas and Sargent? I guess they think the issues are too difficult for the public to understand. How convenient. Perhaps the problems are with their theories and assumptions that do not wash for too many people. Einstein said a good theory was one which your grandmother could understand and be convinced by.

      Delete
  2. Your preference for starting assumptions based on Victorian philosophy is well known, but as you now have introduced some form of regulation in terms of screening, perhaps you could also apply this to some tiring and uncordial interjections (a la Anon 7.24), even if you agree with them. This is a good case of how free-for-alls or a very the selective application of friction can lead to diminishing returns.

    ReplyDelete
    Replies
    1. You should be the first to be edited out.

      Delete