Leaders, the Public, and the Domestic Effects of International Law

Download data and study materials from OSF

Principal investigator:

Geoffrey P.R. Wallace

Rutgers University

Email: geoffrey.wallace@rutgers.edu

Homepage: http://www.geoffreyprwallace.com/

Sample size: 1734

Field period: 12/11/2013-03/07/2014


A large literature has developed around audience costs theory, that leaders are punished domestically when they fail to follow through on their external threats during a foreign policy crisis. The threats made by leaders are, in turn, thought be more credible due to these prospective costs should they subsequently back down. Past work has manipulated different elements of crisis diplomacy to assess the extent and operation of audience costs, including the level of military escalation, the leader's partisanship, domestic elite consensus, and various dimensions of the broader context of the crisis. This study seeks to evaluate an additional implication of audience costs theory concerning whether the prospects for domestic punishment are greater based on the degree to which a threat is embedded in international law.


Failing to follow through on more legalized commitments is expected to have greater domestic consequences for a leader compared to situations where the country had made no formal promise. Drawing on the existing literature on audience costs, as well as previous work on international law, the hypotheses are as follows.
1) The public should exhibit greater disapproval for a leader who violates a prior commitment than if they had made no initial promise.
2) The level of public disapproval for a leader who violates a prior commitment should increase as the level of legal obligation in the commitment increases.
3) The level of public disapproval for a leader who violates a prior commitment should increase as the level of precision in the commitment increases.

Experimental Manipulations

All respondents were presented with a hypothetical scenario where a foreign country invaded a neighboring country. Respondents were then randomly assigned to additional information based on how the U.S. president responded to the crisis according to the following two treatments.
Level of obligation:
• No Commitment [control]: saying the U.S. would stay out of the conflict.
• Verbal [treatment #1]: Verbal threat that if the attack continued the U.S. military would push out the invaders.
• Soft law [treatment #2]; participation in an international meeting agreeing to guidelines recommending that if the attack continued the U.S. military would push out the invaders.
• Formal treaty [treatment #3]; joined an international treaty promising that if the attack continued the U.S. military would push out the invaders.
Level of precision (only applicable to the three obligation treatment conditions):
• Low [control]: prompt included noting there was a lot of debate among experts about what U.S. actions might be involved for the relevant commitment.
• High [treatment]: no additional prompt introducing ambiguity into the nature of the commitment.
The design involves a total of 7 experimental groups, where "No Commitment" was considered the baseline control group. Irrespective of the particular prompt, all groups were then informed that the attacking country continued to invade and the president subsequently decided not to send U.S. military forces.


There were three main outcome variables of interest.
1) Approval of how the president handled the situation: 1 = Strongly approve; 2 = Somewhat approve; 3 = Neither approve nor disapprove; 4 = Somewhat disapprove; 5 = Strongly disapprove
2) President's competence: 1 = Very competent; 2 = Competent; 3 = Neither competent nor incompetent; 4 = Incompetent; 5 = Very incompetent.
3) Feeling that other countries would believe promises made by the U.S. president in the future: 1 = Extremely likely; 2 = Somewhat likely; 3 = Somewhat unlikely; 4 = Extremely unlikely.

Summary of Results

The findings are generally consistent with the key argument of audience costs theory that leaders are punished more by the public for making a commitment and failing to follow through than if the leader had not made any commitment in the first place. Levels of disapproval were higher across each of the three types of obligation treatments (verbal / soft law / formal treaty) when the president subsequent did not send U.S. military forces compared to the control condition where the president said the country would stay out of the conflict from the beginning. The hypothesis that the size of audience costs should increase with the level of obligation receives weaker support. The public did not appear to differentiate much between failing to follow through on a simple verbal promise compared to a non-binding or formal treaty embedded in international legal principles. On the other hand, the second precision dimension of legalization figured much more prominently. More precise commitments do appear to tie the hands of leaders more tightly, as higher precision was associated with much lower levels of approval, harsher assessments of the leader's competence, and a lower sense that other countries would believe future U.S. promises. Precision further conditioned the operation of obligation, where the level of obligation had much stronger effects at lower levels of precision. This suggests that when facing uncertainty in the terms of commitments, domestic actors appear to attach much greater weight to the commitment's legal obligation, but do not do so with less ambiguous promises. The findings thus have a number of implications for understanding the limits and conditions under which international law is expected to have consequences for domestic politics.