Does Presidential Approval Influence Congressional Approval? An Order Effects Experiment

Download data and study materials from OSF

Principal investigator:

Krista Loose

Massachusetts Institute of Technology



Sample size: 2080

Field period: 05/07/2013-08/28/2013


Congressional approval has been historically low for the past several years. While it is certain that the public does not think much of Congress, it may be that many surveys artificially depress congressional approval ratings due to the order in which they ask these questions. This experiment evaluates whether asking the public about their impressions of Congress is influenced by first asking for their impressions of the President. Moore (2002) identified several ways in which order effects could influence attitudes toward paired objects. After randomizing the order in which respondents were asked not only about presidential and congressional approval, but also approval of the Supreme Court, I conclude that attitudes toward all branches of the federal government are stable with regard to order effects. Low congressional approval is not a function of comparison to the President or to the Supreme Court, the branch that is typically most highly regarded.


Are attitudes toward Congress influenced by the fact that respondents are typically asked about the President immediately prior?

Experimental Manipulations

Randomly varied the order in which respondents were asked about their approval of the job the following branches were doing: (1) the President, (2) the Congress, and (3) the Supreme Court.


The primary outcome of interest for me was congressional approval, though order effects could have influenced approval of the other branches as well.

Summary of Results

There was no evidence of order effects on any of the three branches.