I develop a Downsian model of electoral competition in which candidates with policy and office motivations use platforms and campaign spending to gain the median voter’s support. The unique equilibrium involves randomizing over platforms and spending and exhibits the following properties: (i) ex ante uncertainty about the winner, (ii) platform divergence, (iii) inefficiency in spending and outcomes, (iv) polarization, and (v) voter extremism. I show that platform polarization and spending move in tandem, since spending is used by candidates to gain support for extreme platforms. Factors that contribute to both include the candidates’ desire for extreme policies and their capability at translating spending into support for them. I also show that strong incumbents parlay an advantage into more extreme platforms, consistent with the classic marginality hypothesis but contrasting with a large theoretical literature in which candidates with an (exogenous or endogenous) valence advantage tend to moderate their platforms.
Read full abstract