Causal Analysis in Theory and Practice

December 20, 2014

A new book out, Morgan and Winship, 2nd Edition

Filed under: Announcement,Book (J Pearl),General,Opinion — judea @ 2:49 pm

Here is my book recommendation for the month:
Counterfactuals and Causal Inference: Methods and Principles for Social Research (Analytical Methods for Social Research) Paperback – November 17, 2014
by Stephen L. Morgan (Author), Christopher Winship (Author)
ISBN-13: 978-1107694163 ISBN-10: 1107694167 Edition: 2nd

My book-cover blurb reads:
“This improved edition of Morgan and Winship’s book elevates traditional social sciences, including economics, education and political science, from a hopeless flirtation with regression to a solid science of causal interpretation, based on two foundational pillars: counterfactuals and causal graphs. A must for anyone seeking an understanding of the modern tools of causal analysis, and a must for anyone expecting science to secure explanations, not merely descriptions.”

But Gary King puts it in a more compelling historical perspective:
“More has been learned about causal inference in the last few decades than the sum total of everything that had been learned about it in all prior recorded history. The first comprehensive survey of the modern causal inference literature was the first edition of Morgan and Winship. Now with the second edition of this successful book comes the most up-to-date treatment.” Gary King, Harvard University

King’s statement is worth repeating here to remind us that we are indeed participating in an unprecedented historical revolution:

“More has been learned about causal inference in the last few decades than the sum total of everything that had been learned about it in all prior recorded history.”

It is the same revolution that Miquel Porta noted to be transforming the discourse in Epidemiology (link).

Social science and Epidemiology have been spear-heading this revolution, but I don’t think other disciplines will sit idle for too long.

In a recent survey (here), I attributed the revolution to “a fruitful symbiosis between graphs and counterfactuals that has unified the potential outcome framework of Neyman, Rubin, and Robins with the econometric tradition of Haavelmo, Marschak, and Heckman. In this symbiosis, counterfactuals emerge as natural byproducts of structural equations and serve to formally articulate research questions of interest. Graphical models, on the other hand, are used to encode scientific assumptions in a qualitative (i.e. nonparametric) and transparent language and to identify the logical ramifications of these assumptions, in particular their testable implications.”

Other researchers may wish to explain the revolution in other ways; still, Morgan and Winship’s book is a perfect example of how the symbiosis can work when taken seriously.


  1. Some heretical thoughts after posting the above:

    In light of the writings of social scientists like Elwert, Morgan and Winship,
    would Imbens and Rubin be prepared to publically qualify their statement: “In our own work,
    perhaps influenced by the type of examples arising in social and medical sciences, we have not
    found this [graphical] approach to facilitate the drawing of causal inferences,” (2015,upcoming book, page 25)

    I would encourage them to replace “.. influenced by the type of examples arising in social and medical sciences”
    with “… influenced by the type of examples WE SPOTTED in social and medial sciences”

    I would also call on Imbens to qualify his statement (Rejoinder, p. 377):
    “in social sciences applications a graph with many excluded links may not be an
    attractive way of modeling dependence structures,”
    and replace it with “in SOME social science applications..:

    We could then explain graph-avoidance by unfortunate choice of social sciences applications.

    Comment by judea pearl — December 20, 2014 @ 4:12 pm

  2. […] who may benefit from such representation. First, researchers in the Morgan-Winship camp (link here) who are using, interchangeably, both graphs and potential outcomes. These researchers prefer to do […]

    Pingback by Causal Analysis in Theory and Practice » Flowers of the First Law of Causal Inference — December 22, 2014 @ 5:24 am

  3. lustra na wymiar piotrków

    See this lustra na wymiar for yourself.Glass furniture,building and much more in Piotrków Trybunalski

    Trackback by lustra — May 7, 2015 @ 7:28 am

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress