The Political Ideologies of Cabinet Officials

In this earlier post, I posted the comment of a Monkey Cage reader, who suggested that political science has not devoted enough attention to the executive function of the president, and especially presidential appointees:

At least as importantly, NOMINATE as an assessment of a President continues the common fixation on the President as a legislative actor and thus obscures the importance of their appointments, regulations and foreign policy…

Some interesting new research speaks to the ideological of presidential appointees.  I’ll have two posts about this.  The first concerns cabinet secretaries.

In this forthcoming paper (ungated), Anthony Bertelli and Christian Grose examined thousands of pages of congressional testimony by cabinet secretaries to identify their positions on key pieces of legislation.  Having done so, they can then calculate the ideological “ideal points” of each secretary, much as the NOMINATE approach does for members of Congress who actually vote on legislation. Bertelli and Grose generate ideal points for for 46 cabinet members during the Clinton and George W. Bush administrations.  (UPDATE: The total time period examined is 1991-2004.) Here is one graph, depicting the ideal points of the secretaries, presidents, and key Senate and House members:

Obviously, Clinton (in the graph for the 103rd Congress) is well to the left of Bush (in the graph of the 107th Congress).  Moreover, so are many of Clinton’s cabinet secretaries.  But note the variation as well: presidents do not appoint ideological clones.

Some other interesting findings:

  • There is more ideological diversity in cabinet appointees under unified government (1991-2004 1994-95, part of 2001, 2003-2004) than divided government (2005-2008 1991-92, 1995-2000, part of 2001-2 ).
  • The strongest predictor of the ideal point of a cabinet appointee is not the ideal point of the president.  It is the ideal point of the median voter in the House, followed by the ideal point of the median Senator.
  • Ideological similarity to members of Congress matters.  The closer a cabinet appointee is to the House or Senate median, the larger the discretionary budget authority that Congress gives to that cabinet department.

Finally, below I will paste the big table with the estimated ideal points and confidence intervals for all of the cabinet secretaries in their data.  Click for a larger view.  Much more detail on the methodology and analysis is in the paper.  Recommended.

9 Responses to The Political Ideologies of Cabinet Officials

  1. PM May 16, 2011 at 2:28 pm #


    “under unified government (1991-2004) than divided government (2005-2008).”

    Would this be Counter-Earth? Don’t we normally consider 1995-2000 (at least) to have been divided government? (I recall some sort of kerfuffle about impeaching someone.) Etc.

  2. John Sides May 16, 2011 at 3:03 pm #

    PM: My mistake. Working too quickly. And I messed up the dates of the study’s data as well. All is fixed now. Apologies

  3. Andrew Gelman May 16, 2011 at 3:16 pm #

    Nice graph, excellent choice to not make it alphabetical (or a table!). My only recommendations are:

    1. Get rid of half of the dotted vertical lines.

    2. Use color to distinguish appointees in the different administrations.

  4. PM May 16, 2011 at 3:25 pm #


    No worries! I thought that I’d probably missed something.

    Fantastic paper, though. Between this and Twitter-NOMINATE, ideal-point stuff is getting very, very fun.

  5. WMR May 16, 2011 at 5:09 pm #

    Interesting! One implication of this paper is that presidential administrations are not as extreme or polarized as suggested by NOMINATE scores.

  6. PM May 16, 2011 at 6:44 pm #

    But also that Congress has many ways to affect executive policy…

  7. Andrew Rudalevige May 17, 2011 at 2:06 am #

    This looks worth a more detailed read – it’s a neat approach. One immediate reaction is that all Cabinet secretary testimony is cleared (or supposed to be) by OMB before it is given. If this holds, then we should assume that the president, by proxy, has approved it (this is the point of central clearance). And thus, any differences (via this measure, at least) between a Cabinet secretary’s ideal point and the president’s are by permission of the president. I’m a bit nervous about NOMINATE being pushed too far generally – this may be a place where it’s asked to do too much.

  8. Christian Grose May 17, 2011 at 1:12 am #

    Andrew R., good point. This is something we were concerned about, but we have addressed this critique.

    The publicly revealed “roll call” positions that underlie the cabinet secretary ideal point estimates come from both prepared testimonies and the Q&A sessions between Cabinet secretaries and members of Congress. Crucially, many positions emerged during the Q&A parts of the testimonies, and some of these positions were distinct from presidential positions. Obviously, OMB can’t clear extemporaneous Q&A. I think this addresses your criticism, and we discuss it in the “supplemental information” in the paper’s appendix.

    Thanks to all for your comments.

    • Andrew Rudalevige May 17, 2011 at 10:04 pm #

      Cool. I look forward to reading the paper in more detail.