Rational Choice and Interpretive Evidence: Caught between a Rock and a Hard Place?
| Published date | 01 February 2010 |
| Date | 01 February 2010 |
| Author | Andrew Hindmoor,Iain Hampsher-Monk |
| DOI | 10.1111/j.1467-9248.2009.00776.x |
| Subject Matter | Article |
Rational Choice and Interpretive Evidence:
Caught between a Rock and a Hard Place?
Iain Hampsher-Monk Andrew Hindmoor
University of Exeter University of Queensland
Following Green and Shapiro’s critique, debate about the value of rational choice theory has focused upon the
question of its relationship to what we call ‘external’, largely quantitative, empirical evidence.We argue that what is
most striking about rational choice theory is, however, its neglect of interpretive evidence.Our survey of 570 articles,
published in the American Political Science Review and the American Journal of Political Science between 1984 and 2005
employing rational choice theory,revealed that only 139 made even the most cursory use of interpretive evidence.
Does this matter? We argue that the absence of interpretive evidence undermines rational choice’s explanatory
credentials. However, wealso argue that the admission of inter pretive evidence risks rendering redundant the rational
choice element of any explanation. This is the rock and the hard place between which rational choice is caught.In
the final part of the article we distinguish those cases where rational choice may prove useful, namely those
circumstances in which interpretive evidence either cannot be relied upon or does not subsume that which an
explanation is intended to achieve.
In a celebratory article published to mark the 25th anniversary of the journal Public Choice,
Dennis Mueller (1993, p. 147) predicted not only that public choice theory would
‘dominate political science in a generation or less’ but that ‘alternative approaches to the
study of politics would eventually wither away and die’. Yet barely a year later Donald
Green and Ian Shapiro’s (1994) Pathologies of Rational Choice Theory sparked a backlash
against the use of rational choice theory within political science, culminating in the
submission of a petition to the American Political Science Association in 2001 demanding
greater methodological pluralism (see Hindmoor, 2006, pp. 15–9).
Green and Shapiro did not object to the use of rational choice theory per se. Indeed they
welcomed its attempt to study politics scientifically and, in a response to their critics,
identified circumstances in which ‘rational choice explanations should be expected, prima
facie, to perform well’ (Green and Shapiro, 1994, pp. 94–5). What they objected to was
rational choice’s ‘universalism’; to the misconceived attempt to ‘construct a unified,
deductively-based theory from which propositions about politics – or, indeed, all human
behaviour – may be derived’ (Green and Shapiro, 1994, p.54). Such an approach had, they
suggested, encouraged the development of a ‘method-’ rather than ‘problem-driven’
approach to the study of politics and to what Shapiro (2005) subsequently termed a ‘flight
from reality’.Finally, and more specifically,they argued that a commitment to using rational
choice to explain everything had distracted its practitioners from seeking empirical evi-
dence with which to support their accounts, had blinded them to the ‘banal’ character of
many rational choice explanations and had encouraged them to select biased evidence, to
doi: 10.1111/j.1467-9248.2009.00776.x
POLITICAL STUDIES: 2010 VOL 58, 47–65
© 2009The Authors. Journal compilation © 2009 Political StudiesAssociation
ignore competing explanations and to engage in post hoc theorising in order to reconcile
their arguments with empirical anomalies or discordant facts (Green and Shapiro, 1994,
p. 6).
In examining rational choice’s record, Green and Shapiro focused exclusively upon the
extent to which theorists provided empirical evidence about the ‘outside’ (Collingwood,
1946) of an event:that is evidence, usually quantitative in form, about the behaviour and the
effects of the behaviour of particular actors. They looked, for example,at the way in which
rational choice theorists had sought to test their arguments using quantitative data on the
spatial positioning of political parties, the formation of interest groups and legislative voting
patterns. Evidence,on this view,consists in a ‘fit’ between the deductions of rational choice
theory and observed institutional or behavioural outcomes in any particular case. In what
follows we will refer to empirical evidence of this sort as being ‘external’. However, we
argue that rational choice is also compromised by its failure to provide another kind of
empirical evidence, namely‘internal’ or interpretive evidence about the beliefs of the agents
whose actions comprise the phenomena to be explained. Our distinction between external
and internal evidence maps on to the well-known distinction between a behavioural and
ultimately positivist conception of political science and a hermeneutic or interpretive one.
The former models external ‘behaviour’, seeking explanation by way of statistical or
deductive fit without presuming any intimate relationship between belief and action (hence
external). The latter presupposes that action instantiates belief, and thus that any explana-
tion of action must incorporate statements about the beliefs of actors. Internalist explana-
tions do not claim access to private mental states; they are ‘internal’ only in the sense of
being internal to the world of meanings inhabited by the actor. External explanations, by
contrast, which work via correlations or deductions on the basis of ascribed reasons, need not
concern themselves with actors’ understandings of the world.1In arguing this, we seek to
discriminate among and clarify the roles rational choice might usefully play within political
science.
Our claim about the relevance of interpretive evidence will be more fully elaborated below.
However it may be useful at this point to outline the logic of explanation to which we
subscribe here. We are not merely calling for more weight to be given to agents’ self-
descriptions or qualitative analysis in some eclectic explanatory mix. Our point is that even
deductive explanations, if they pur port to model agents’ actions, must (logically) share
premises with the agents, and that a failure to achieve this, while it might result in an
explanatory conclusion that models the agents’ actions, will only do so contingently. That
is to say, while individuals may well economise in making choices, what they economise
about is available to them via a cognitive repertoire through which they construe how the
world is (and so what choices are available to them), what causal or other chains of efficacy
operate there and how values are to be ranked. Evidence about such cognitive presuppo-
sitions can be derived from that person him or herself whether directly through interviews
or indirectly through speeches, memoirs, diaries or other records, or (more speculatively)
can be ascribed on the basis of educational and cultural background, congruence with other
evident belief features, etc.2The holding of such beliefs – even if they are false – thus
constitutes demonstrable features of the actors, relevant because agents act on such beliefs.
48 IAIN HAMPSHER-MONK AND ANDREW HINDMOOR
© 2009The Authors. Journal compilation © 2009 Political StudiesAssociation
POLITICAL STUDIES: 2010, 58(1)
To continue reading
Request your trial