Abstract

Why does ideological debate sometimes become so polarized, both sides apparently finding their own view morally superior? We propose an explanation based on four main assumptions: i) People have self-image and social image preferences; ii) if normative views move towards one of the polar cases, image costs increase for some individuals but decrease for others; iii) over time, people may self-select into social groups where views are more favorable to themselves; and iv) within social groups, biased social learning takes place. This causes normative views to converge within social groups but diverge between social groups. We present a formal model where incomes differ exogenously, image depends on one’s voluntary public good provision, while individuals’ redistributive views evolve endogenously over time. We show that the steady state is highly polarized and segregated: In equilibrium, the poor (rich) stick together, holding completely (non-) egalitarian views. Furthermore, since everyone is surrounded by peers confirming that one’s own responsibility to contribute is limited, actual public good provision is minimal.

Reference

Karine Nyborg, “Image Preferences as a Driver of Normative Polarization”, IAST Lunch Seminar, Toulouse: IAST, November 26, 2024, 11:30–12:30, room Auditorium 4 (First floor - TSE Building).