Joint Publication : loriweb.org - The Reasoner
Change of preferences is a phenomenon that everyone experiences in himself or herself. Yet how can preference change be explained and modelled? This important methodological question - which is far from settled - was the main focus of the Workshop on Preference Change, organized by the LSE Choice Group at the end of May 2009 at the London School of Economics.
The contributions and discussions at this workshop were marked by a shared goal of analysing preference change in new and often unconventional terms, and by a surprising amount of agreement to go beyond orthodox models of rational choice. How exactly to depart from classical models and which of their elements to retain was the object of intense discussion and diverse proposals.
The contributions at this workshop can be largely divided into two groups. Some talks focussed primarily on foundational issues of explaining, modelling and representing preference change. Other talks focussed mainly on theoretical or practical implications of preference change, notably with regard to ethical theories, social choice theory, policy making and political philosophy.
Among the ‘foundational’ contributions, many took the notion of belief change as a starting point, either attempting to reduce preference change to belief change of some standard or (more often) non-standard type, or arguing for a richer understanding of belief changes to cover different types of preference change. In a belief revision approach, Sébastien Konieczny (CRIL - CNRS Lens) introduced improvement operators as a generalization of usual iterated belief revision operators. Brian Hill (HEC Paris) analysed different Bayesian models and problems of disentangling preferences and beliefs from each other in a principled way. In a similar vein, though now in a model of epistemic logic, Sven Ove Hansson (Stockholm) showed that it is not possible to fully maintain a distinction between belief change and preference change. Richard Bradley (LSE) presented a probabilistic framework that extends Jeffrey-type conditioning to hypothetical imperatives. Peter Hammond (Warwick) proposed to generalize extensive form models in decision theory by introducing ‘aberrant’ events and allowing the decision tree to gradually ‘unfold’, which causes behaviour to change.
Other foundational talks proposed to explain preference change without explicitly involving beliefs at all, introducing other types explanations related to the agent or the environment. Christian List (LSE) and Franz Dietrich (LSE and Maastricht) introduced the notions of salient dimensions and motivating reasons to model preferences and preference change, arguing that they can be used to capture limited conceptualisation and limited imagination. Conrad Heilmann’s (LSE) talk on multiple-selves introduced the notion of connectedness between selves to measure the stability of a decision-maker’s preferences over time. Katie Steele (LSE) showed that important questions arise about what in fact are the objects of first-order desire when analysing higher-order desire.
Other talks focussed on various implications of change in preferences. Nick Baigent (Graz) analysed the relations between preference change and consequentialist rational choice and showed that preference change requires new conditions for a well-behaved choice function. Wlodek Rabinowicz (Lund) critically examined the conception of moral deliberation as a process of thought experimentation with concomitant preference change, and discussed implications for preference utilitarianism. Krister Bykvist (Jesus College, Oxford) showed that endorsement theories of well-being are faced with inconsistencies in cases where our preferences (endorsement attitudes) depend on and change with the context. Luc Bovens (LSE) discussed the possibility for the state to affect or improve people’s preferences through ‘nudges’ (intended framing effects), raising the question of whether, and in what forms, liberal paternalism is justifiable.
Three days of stimulating talks and discussion resulted in a widespread agreement among the workshop participants that the phenomenon of preference change calls for extensions to formal theories of rationality.
Thanks to all the speakers and participants for making these days so enjoyable!
Franz Dietrich, Conrad Heilmann and Christian List