WP 2020-19 Strategic Theory of Norms for Empirical Applications in Political Science and Political Economy
AUTHORS: Don Ross, Wynn C. Stirling, and Luca Tummolini
ABSTRACT: We adapt Bicchieri’s (2006, 2017) account of norms as aligned descriptive and normative expectations for application to identification and estimation of laboratory choice data. This involves a number of steps. First, borrowing from Kuran (1995), we amend the form of the utility function suggested by Bicchieri so as to abstract away from specific social preferences. Second, we incorporate variable risk preferences so as to relax Bicchieri’s assumption, applied in experiments by her and various co-authors, that subjects are risk neutral. Third, we allow for subjects to optimise rank-dependent instead of expected utility. Fourth – the work that takes up the bulk of the paper – we show how to directly represent Bicchieri’s concept of conditional preference dynamically, so as to allow normative preferences to influence one another and spread through populations. This is done using Stirling’s (2012, 2016) Conditional Game Theory, enriched to incorporate influence cycles, risk aversion, and probability weighting. We simulate three Investment / Trust games to verify that the theory allows for identification of the relevant parameters in hypothetical data.