Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This paper studies the effect of voter loss aversion in preferences over both candidate policy platforms and candidate valence on electoral competition. Loss-aversion over platforms leads to both platform rigidity and reduced platform polarisation, whereas loss-aversion over valence results in increased polarisation and the possibility of asymmetric equilibria with a self-fulfilling (dis)-advantage for the incumbent. The results are robust to a stochastic link between platforms and outcomes; they hold approximately for a small amount of noise. A testable implication of loss-aversion over platforms is that incumbents adjust less than challengers to shifts in voter preferences. We find some empirical support for this using data for elections to the US House of Representatives.