Comments on Bad Control

Derek Neal of the University of Chicago comments that our discussion of bad control in section 3.2.3 leaves the impression that more control is always better as long as the controls are pre-determined relative to the causal variable of interest. The leading counter-example is the case of within-family or twins estimates that we discuss as the “baby with the bathwater problem” on p. 226. Here you might indeed increase omitted variables bias even though the controls are not bad in the section 3.2.3 sense:

Hi Guys:

I agree that the issue I am raising is conceptually different, but as a practical matter, the “bad control” issues and “baby with the bathwater problem” both fall under a larger heading of “can more controls ever make things worse.” Your discussion of bad control may lead some students to believe that the answer is “only if the extra controls are endogenous.”

If you ever have a second edition, I think there is an argument for dealing with all aspects of the “can more controls ever make things worse” question all in one place.

Point taken! We hope to fix this in the next edition . . .

Published Tagged . Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.


  1. Jie Ma
    Posted October 24, 2017 at 5:38 am | Permalink


    I’m reading the bad control section again and having trouble connecting the argument with the common practice of including intermediate outcomes as a way to test for the mechanism. Many papers (Nunn and Qian 2014, Chen, Kung and Ma 2017, etc) use the intermediate outcomes as controls: if the coefficient of the treatment variable changes in magnitude or significance when including these intermediate outcome, they claim the intermediate outcome is a potential channel of the effect. For example, by adding occupation to the regression between education and earnings, the coefficient is reduced. This would imply that the occupation is a potential explanation of education’s effect on earnings.

    The method makes sense intuitively but the papers using them do not provide econometric theories to support. I’m wondering what’re your thoughts on this. Thanks!


    • josh
      Posted October 24, 2017 at 10:14 am | Permalink

      Your right to be suspicious, Jie. You can not test for a mediator just by sticking it in as a control. That’s why mediator’s are “bad”! Check out Chapter 6 of Mastering ‘Metrics for more guidance on this difficult problem.

Post a Comment

Your email is never published nor shared. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>