Choice in politics: Equivalency framing in economic policy decisions and the influence of expertise
Introduction
Political decision-making is inherently based on the convictions of individual actors and their choices. Voters choose which topics they consider important and whom to elect as their representatives; politicians choose between alternative policy options and how to react to external events; and public officials again have to make various choices when providing the operational and informational base of the policy-making process. The nature of policy-making and the indeterminateness of the future thereby dictate that many of these choices involve risk. Yet, the actors on the institutional side are expected to be able to consistently find the best policy options. In democracies the public relies on them to take the ‘right’ decisions because they are seen as expert decision-makers. Their education, experience, and access to information give them advantages in their subject areas and in the mechanics of politics (Boswell, 2008, Radaelli, 1999, Strøm, 2000).
This expectation raises two issues: first, it is not always clear what determines expertise and its influence on the suitability and capacity of political actors (Bendor, 2010, Emler and Frazer, 1999, Krosnick, 1990). Second, a long line of empirical research on decision-making has shown that human choices are more susceptible to behavioural influences than the popular rational choice theories in political science allow for. In risky-choice, already small linguistic variations in the presentation of equivalent options can alter decision outcomes (Tversky and Kahneman, 1981, Levin et al., 2002; see Kühberger, 1998 for a meta-analysis).
In this paper, we ask whether policy-decisions are subject to framing effects, which direction these effects take, and whether expertise has an influence. Our research is built on Tversky and Kahneman's (1981, p.453) Asian Disease Experiment (ADE), which demonstrates that framing can have a large effect. As a baseline, the ADE itself is reproduced and we find a similar bidirectional effect as the original study. Because of its relevance to a broad range of political decisions we also reproduce Quattrone and Tversky's (1988, p.727) Employment–Inflation trade-off. While it does not feature any risky choices, it is one of the few questions so far to specifically test economic policy scenarios for framing effects. Contrary to Quattrone and Tversky (1988), we do not find a significant effect.
To test framing in economic policy-decisions, we use relevant hypothetical scenarios building on the same structure as the ADE. A sample of students and professionals is presented with salient issues in (EU) policy-making. Participants make a binary choice between a risky and a certain option with equivalent outcomes, in either a positive or a negative frame. Our results show mostly unidirectional framing effects. In the positive frames of our three scenarios, a significant majority prefers the certain option. This picture changes in the negative frame: in two scenarios no option is systematically preferred; and in the third scenario the participants exhibit a weak preference reversal with a smaller majority choosing the risky option.
To explore the potential influence of expertise on the susceptibility to framing, we focus on the criteria used for the recruitment and evaluation of public officials. To our knowledge, no such attempt has been made yet. Using logistic regression models, we test the effects of education, experience, and other control variables on the observed choice behaviour. We find that respondents with high qualification levels and those without show a similar susceptibility, while the frame is the only variable that is systematically significant across scenarios. The presented results call for caution when assuming that expertise helps people to make better policy decisions under risk, if the available options can be framed differently.
Section snippets
Literature
Decision-modelling and prediction in economics and political science are traditionally based on rational choice theories, such as Expected Utility Theory (EUT) as proposed by von Neumann and Morgenstern (1947). In economics, behavioural approaches have steadily advanced during the last decades and provide an increasingly popular alternative to rational choice approaches. Empirical evidence of decision-biases is well documented in a large literature and is part of the mainstream discourse (
Hypotheses
Equivalency framing effects have been confirmed in a large range of studies with diverse alterations to the experimental settings (e.g. Bless et al., 1998, Druckman, 2001, Kühberger, 1998, Kühberger et al., 1999).5 Frisch, (1993) even finds that the effect holds for participants who, on reflection, agree that the two options should be treated the same. This finding refutes attempts to explain the
Design
We use Tversky and Kahneman's (1981) ADE and three re-phrasings with economic policy scenarios to test our hypotheses and answer our research questions. Like the ADE, our own scenarios present participants with the choice between one certain and one risky option. The certain option offers a fixed outcome, while the risky option contains two possible outcomes and their respective probabilities. The outcome-sizes provided in the options are chosen to match the scenarios. The probabilities are of
Results
One hundred and twenty-seven participants (N = 127) answered the complete survey. Tests of normality revealed that age (in years) and indicators of expertise (years of study, years of work, interest in economics/politics/EU politics, and attitude towards the EU) are non-normally distributed (Shapiro-Wilk test: p < 0.01 for all variables). Consequently, non-parametric test statistics are used where necessary.
Discussion
In this paper we successfully reproduce Tversky and Kahneman's (1981) Asian Disease Experiment (ADE), but do not find a framing effect in Quattrone and Tversky's (1988) Employment–Inflation trade-off. This lack of a framing effect could be interpreted as failure of the external validity of their findings for samples that include professionals. However, splitting the sample by gender reveals very slight framing effects, but men and women seem to be influenced in opposite directions. Because of
Conclusion
While framing research and Prospect Theory have made some advances in the analysis and explanation of political decision-making, the alleged (although reducing) dislike of methods from psychology in the political sciences has left the application of behavioural accounts of choice behaviour trailing far behind other social sciences, especially economics (Mercer, 2005, Schnellenbach and Schubert, in this issue). The continued use of rational choice as favourite tool for normative and prescriptive
Acknowledgements
We wish to thank Benny Geys, Kris De Jaegher, Arun Rattan, Scott E. Page, Björn Kauder, and two anonymous reviewers for their helpful comments and suggestions. The usual disclaimer applies. We also thank the organisers and participants of the Behavioural Political Economy Workshop at the CESifo Venice Summer Institute 2014, where an earlier draft of this paper was presented. Colin R. Kuehnhanss and Katharina Hilken are grateful to the Research Foundation Flanders (FWO) for financial support in
References (79)
- et al.
A note on the robustness of a classical framing result
J. Econ. Psychol.
(1992) Evaluating framing effects
J. Econ. Psychol.
(2001)- et al.
Men, women and risk aversion: experimental evidence
Handb. Exp. Econ. Results
(2008) - et al.
The effects of decision framing on choice of risky vs certain options
Organ. Behav. Hum. Decis. Process.
(1987) - et al.
Risk attitude and market behavior: evidence from experimental asset markets
J. Econ. Psychol.
(2007) - et al.
The novice and the expert: knowledge-based strategies in political cognition
J. Exp. Soc. Psychol.
(1983) Reasons for framing effects
Organ. Behav. Hum. Decis. Process.
(1993)Reference points, anchors, norms, and mixed feelings
Organ. Behav. Hum. Decis. Process.
(1992)(Why) are economists different?
Eur. J. Polit. Econ.
(2005)The influence of framing on risky decisions: a meta-analysis
Organ. Behav. Hum. Decis. Process.
(1998)