Home > econometrics > “Methodological Mistakes and Econometric Consequences”

“Methodological Mistakes and Econometric Consequences”

From Edward Fullbrook

Asad Zaman has just published an illuminating paper related both to Bryant Chen and Judea Pearl’s recent RWER paper “Regression and causation: a critical examination of econometrics textbooks” and to topics frequently discussed on this blog.  Titled “Methodological Mistakes and Econometric Consequences”, Zaman’s paper appears in the current issue of the International Econometric Review.   Here is an open-access link to the paper and below is its introduction.

1. INTRODUCTION 

The rise and fall of logical positivism is the most spectacular philosophical story of the twentieth century. Rising to prominence in the second quarter of the twentieth century, it swept away all contenders, and became widely accepted throughout the academia. Logical positivism provided a particular understanding of the nature of knowledge, as well as that of science and of scientific methodology. The foundations of the social sciences were re-formulated in the light of this new understanding of what science is. Later on, it became clear that the central tenets of the positivist philosophy were wrong. Logical positivism had a “spectacular crash,” and there was some dispute about who had “killed” logical positivism1. As a logical consequence, it became necessary to re-examine the foundations of the social science, and to find new bases on which to re-construct them. This has occurred to differing degrees in different disciplines. One of the most recalcitrant has been economics. As discussed in Zaman (2011), the foundations of economics continue to be based on erroneous logical positivist ideas, and hence require radical revisions. 

It is not our intention in this paper to describe the whole story, which would be lengthy as well as complex and contentious. We will confine our attention to tracing the harmful impact of a limited number of positivist ideas on how econometrics has developed. We will argue that these ideas have led to a wrong methodology being accepted and used in econometrics. This wrong methodology has prevented progress and advance in knowledge. As evidence for this lack of progress, in a talk on the 100th anniversary of the first published regression by Yule, Freedman (1997:113) writes: “For nearly a century, investigators in the social sciences have used regression models to deduce cause-and-effect relationships from patterns of association. … . In my view, this enterprise has not been successful.”

This paper is structured as follows. Section two below provides a summary of the key positivist ideas, and how they lead to a distorted understanding of science. The third section provides several case studies showing how positivist methodologies lead us to consider the wrong questions. The fourth section shows the contrast and opposition between positivist and realist methodologies. The fifth section sketches some possible alternative methods which could be used to avoid these mistakes and make progress. The sixth section summarized the conclusions.

We will also offer a very tentative and preliminary sketch of alternative approaches that could be more successful.

continue reading here

  1. Rhonda Kovac
    February 17, 2014 at 1:23 pm

    Wonderful post. The link provided to Raman’s paper didn’t work for me (“Page cannot be found”). I located another: http://mpra.ub.uni-muenchen.de/41032/1/MPRA_paper_41032.pdf. Note: This post says that the paper was “just published”. FYI, the paper, of the same title and whose Introduction matches, at my link is dated 30 Aug 2012. Perhaps this is an earlier version.

  2. February 17, 2014 at 1:47 pm

    Highly recommendable article (and one that I actually quoted in this blogpost last year:
    http://larspsyll.wordpress.com/2013/04/19/why-macroeconometrics-hasnt-progressed/).

    • davetaylor1
      February 17, 2014 at 9:39 pm

      From your paper, Lars: “To apply ‘thin’ methods we have to have ‘thick’ background knowledge of what’s going on in the real world”.

      Spot on, unless the purpose of the ‘thin’ methods is to indicate where to look for ‘thick’ background knowledge we don’t yet have.

  3. February 17, 2014 at 5:55 pm

    Not a bad paper. But there are numerous philosophical errors.

    The most striking is that before Kant there was no such thing as idealism (p102). This is just patently false. Obviously George Berkeley first formulated idealism and one could argue that Hume was an idealist (certainly, Leibniz was).

    • February 17, 2014 at 5:59 pm

      Tied to this is the glaring error that it is the fault of Kant’s (supposed) idealism that we get logical positivism. Ironically, George Berkeley formulated idealism precisely to counter the logical positivist error that science does not deal in “unobservables” (Berkeley called them “abstractions”) which the author discusses at length at the beginning of the paper.

      When you actually know the history of thought the whole story becomes entirely the opposite of what the author is trying to insist upon. The paper is, from a history of thought point-of-view, myth-heavy.

  4. February 18, 2014 at 4:30 am

    The second paragraph of the article states that “It is not my intention to describe the whole story (of the philosophy of thought) — rather, we trace the harmful impact of a limited number of positivist ideas.” Definitely the sketch of philosophical history presented in the paper is incomplete and ignores many factors — the paper is focused on the econometric methodology and not on the philosophical background. Nonetheless I do not think it is majorly mistaken as suggested by some commentators. I am always trying to learn, and would be happy to be educated — So I would like to ask:

    1: Where do I claim that there was no idealism before Kant in the paper? I could not find such an assertion implicit or explicit.

    2: What is that I am trying insist upon (the particular claim) that is exactly the opposite of the truth?

    3: What are the philosophical myths that I have subscribed to in my account?

    I would be happy to learn more about this from the commentators

    asad

    • February 18, 2014 at 1:09 pm

      Asad,

      (1) On page 102 you write: “Before Kant, the nature of human knowledge could be described as follows”… You then lay out a diagram that runs: Reality => Perceptions => Mental Image. But in the modern period it was George Berkeley who overturned this and gave perceptions and mental images primacy, not Kant. Indeed Kant, in arguing for a Ding an Sich (‘thing in itself’) and as such argued that there was a reality beyond perceptions.

      (2 & 3) In the first part of the paper you argue that the problem with logical positivism was to do with ‘unobservables’. You say that logical positivists pretended as if they did not need to deploy this in their version of science. Actually, this is precisely WHY George Berkeley undertook the idealist critique. I discuss that here:

      http://fixingtheeconomists.wordpress.com/2014/01/30/abstraction-language-and-modelling-in-economics/

      So, from a history of thought perspective it was precisely the turn that you ascribe to Kant (wrongly) that aimed at alleviating the problem you think to be inherent in logical positivism. That’s why I said that the story doesn’t fit together.

  5. BFWR
    February 18, 2014 at 7:21 am

    One pole cannot be effectively understood without the other. Integration not opposition is wisdom, and the whole is more than the sum of its parts. How long can economics and finance resist the wisdom of integration and wholeness? Scientism and religion are generally the same mindset. Only Grace, not the dogma, but the both/and equillibrating, free flowing experience itself is up to that integration. So let us have the monetary and economic policy of a philosophy of Grace.

  1. No trackbacks yet.

Leave a 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.