Why Behavioural Policy Needs Mechanistic Evidence

Abstract

Abstract:Proponents of behavioural policies seek to justify them as ‘evidence-based’. Yet they typically fail to show through which mechanisms these policies operate. This paper shows – at the hand of examples from economics and psychology – that without sufficient mechanistic evidence, one often cannot determine whether a given policy in its target environment will be effective, robust, persistent or welfare-improving. Because these properties are important for justification, policies that lack sufficient support from mechanistic evidence should not be called ‘evidence-based’.

Modeling model selection in model pluralism

“…Second, there is ambiguity with respect to the role of non-critical assumptions. Verification supposedly only focuses on critical assumptions, which were identified through their impact on some input-output pairs of a given R. Nevertheless, it is obvious and is often evidenced in economics that the same output can be produced from the same input in conjunction with the same critical assumptions through a number of substantially distinct mechanisms (for an example from behavioral policies, see Grüne-Yanoff, 2016). Two models M i and M j might be equivalent in their critical assumptions with respect to R, for instance, but differ in their non-critical assumptions Ai1 and Aj1.…”

Reflections on the 2017 Nobel Memorial Prize Awarded to Richard Thaler

“…This evidence says little about whether these interventions can be transferred to some other setting. For this, evidence for the mechanisms through which the interventions operate is required (Grüne-Yanoff 2016). Yet Thaler and other behavioral economists have largely eschewed investigating such mechanisms.…”

From Libertarian Paternalism to Nudging—and Beyond

“…A policy is not persistent for example when its effect wears off with time. Yet to understand when to expect such wear-off effects-and how long to run field experiments for in order to test for such effectsevidence about the mechanisms that produce the behavioural effects is required (Grüne-Yanoff 2015).…”

“…This argument is even more explicit in Felsen and Reiner (2015, this issue), who argue that neuroscience, by identifying how decisions are made by the nervous system, also illuminates how nudges may affect the rationality of a decision, its correspondence to fundamental goals, and the presence of undue external influences-all factors relevant for determining the degree of autonomy of a decision. It is also pursued by Grüne-Yanoff (2015), who argues that the welfare-improving capacities of a nudge depend on how the nudge contributes to the formulation and realization of the agent's reflected preferences.…”