Preferences, Goals, and Implications for Paternalism

This paper proposes a conceptual model of decision-making tying specific preferences to broader individual goals. In particular, the model considers two hierarchically ordered types of goals: Terminal goals, which represent fundamental objectives (e.g., health, social connection, etc.), and instrumental goals, which serve as complexity-reducing intermediate steps towards achieving terminal goals (e.g., healthy eating, meeting people, etc.) and which are used to derive eventual preferences (e.g., salad instead of cake, sports club instead of reading, etc.). Notably, originating from the simplification of a complex optimization problem, the hierarchical goal structure allows for contextual misalignments between different instrumental goals (directed towards different terminal goals). As a consequence, in some contexts, it may lead agents to make suboptimal decisions – as evaluated from an outside perspective. Thus, applied to the discussion about paternalism and nudging, the model is compatible with arguments in favor of external interventions as it assumes agents to be boundedly rational, which can be addressed, instead of having generally inconsistent preferences, which are hard to rank without further assumptions on the agency of decision makers. In that sense, the model provides methodological justification for (some aspects of) paternalistic interventions.