Better than a bet: good reasons for behavioral and rational choice assumptions in IR theory
| Published date | 01 June 2023 |
| DOI | http://doi.org/10.1177/13540661221137037 |
| Author | James W. Davis |
| Date | 01 June 2023 |
E
JR
I
https://doi.org/10.1177/13540661221137037
European Journal of
International Relations
2023, Vol. 29(2) 476 –500
© The Author(s) 2022
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/13540661221137037
journals.sagepub.com/home/ejt
Better than a bet: good reasons
for behavioral and rational
choice assumptions in IR
theory
James W. Davis
University of St. Gallen, Switzerland
Abstract
Behavioral IR is enjoying newfound popularity. Nonetheless, attempts to integrate
behavioral research into the larger project of IR theory have proven controversial. Many
scholars treat behavioral findings as a trove of plausible ad hoc modifications to rational
choice models, thereby lending credence to arguments that behavioral IR is merely
residual, empirical, and hence not theoretical. Others limit their research to cataloging
outcomes consistent with the basic tenets of behavioral models. Although this expands
the empirical base, it is insufficient for theoretical progress. In this article, I explore
various answers to the question of when rational choice or behavioral assumptions
should guide efforts to build IR theory. I argue that no single answer trumps all others.
Examining the various conditions under which actors reason highlights the importance
of macrofoundations. Macrofoundations condition the effects of microprocesses and
help identify relevant scope conditions for both rational choice and behavioral models
of decision-making. Examining the various purposes of IR theory also provides answers
to the question of when rational or behavioral assumptions are likely to be most useful.
Although many behavioral scholars premise the relevance of their findings on claims
of empirical realism, I argue that under certain conditions, deductive theorizing on the
basis of as-if behavioral assumptions can lead to powerful theories that improve our
understanding of IR and may help decision-makers promote desired ends.
Keywords
International Relations, rationality, rational choice, behavioral IR, theoretical assumptions,
political psychology
Corresponding author:
James W. Davis, Institute of Political Science, University of St. Gallen, Müller-Friedberg-Strasse 8, 9000 St.
Gallen, Switzerland.
Email: James.Davis@unisg.ch
1137037EJT0010.1177/13540661221137037European Journal of International RelationsDavis
research-article2022
Article
Davis 477
Introduction
Behavioral International Relations (IR)—understood here as the empirical study of polit-
ical decision-making by individuals, groups, and institutions with influence on interna-
tional outcomes—is enjoying newfound popularity. The number of published behavioral
studies is increasing as is the number of symposia and special issues devoted to the
subject in leading journals (Bleiker and Hutchison, 2014; Davis and McDermott, 2021;
Hafner-Burton et al., 2017; Mintz, 2007; van Aaken and Broude, 2019). In large part
explained by the growing influence of behavioral economics, this newfound popularity
also is driven by a younger generation of scholars applying a broader range of methods
and data to long-standing questions of interest and leads to questions of how to integrate
their findings into models and theories of IR.1
Two dangers face the further integration of behavioral IR into the larger project of IR
theory. First, behavioral research presents a tempting trove of plausible ad hoc modifica-
tions to rational choice models confronting disconfirming evidence. It is an approach to
incorporating behavioral research that both reflects and justifies arguments that the find-
ings are residual, merely empirical, and hence not theoretical (Hafner-Burton et al.,
2017: S1–S2; Posner, 1997). Nevertheless, even prominent proponents of behavioral
models have fallen into the trap.
Thus, Colin Camerer writes,
The recipe for behavioral game theory I will describe has three steps: start with a game or
naturally occurring situation in which standard game theory makes a bold prediction based on
one or two crucial principles; if behavior differs from the prediction, think of plausible
explanations for what is observed; and extend formal game theory to incorporate these
explanations. (Camerer, 1997: 167–168)
By definition, however, rational actor models cannot accommodate systematic devia-
tions from rational choice and retain the label.2 Incorporating behavioral findings ad hoc,
while retaining a rational choice core, eventually will produce an unfit hybrid. As the
internal contradictions mount, the likelihood of surviving repeated encounters with the
empirical world is likely to decline, with the blame attributed to the baroque behavioral
ornamentation affixed to the clean lines of the original rational choice facade.
The second danger is to succumb to the law of the instrument. Armed with the find-
ings of behavioral research, IR scholars may be tempted simply to catalog the universe
of outcomes that appear consistent with basic tenets. If conducted systematically, the
approach might produce data sufficient for the establishment of strong correlations that
could prove useful for prediction. But merely expanding the empirical base of behavioral
regularities is insufficient for the sort of theoretical advancement that would take us
beyond what we already know (James, 2007: 164).
How should we think about integrating behavioral IR into IR theory? Before trying to
answer the question, a brief discussion of what is meant by the term “theory” is in order.
Theories, in the context of this discussion, are understood not as collections of hypothe-
ses derived from direct observation, but rather abstract models based on a set of interre-
lated assumptions (some of which may be based on observations) that allow the theorist
To continue reading
Request your trial