Nudge lullaby

Libertarian paternalism (Thaler and Sunstein 2003), using peoples own heuristics and biases to steer them to better choices while preserving choice freedom, has shown to be an effective and popular policy paradigm. It has also sparked many philosophical debates (e.g. Mitchell (2005) and Sugden (2008)) on the role of the government in consumer choice issues. In this paper we address a more practical possible side effect of libertarian paternalism in an experimental study. We have participants perform a number of difficult multi-attribute choice tasks where each option chosen yields a particular payoff. We examine whether participants who are provided with good defaults in the first half of the experiment perform differently compared to a control group, when they perform the task
with random defaults in the second half of the experiment. We indeed find that subjects who first were nudged perform worse than the control group. Interestingly this effect holds even if we control for a proxy for effort. We find a significant treatment effect for men but not for women. Attitudes toward defaults thus seem not to be independent of previous experience. Our result has possible implications for government and nongovernment consumer choice policies.