Key points are not available for this paper at this time.
In real-world elections, voters mostly do not face a choice between ’good’ or ’bad’ candidates but rather between candidates who attempt to mobilize electoral support by mixing normatively desirable policies on issues like social protection with normatively undesirable practices like vote buying. In this paper, we study how voters reward or punish candidates who differ in their mix of clientelistic and programmatic distribution. We gather evidence from a conjoint experiment of voter preferences for city councilors in Brazil. Our findings show that clientelism negatively affects voter support, even when candidates add programmatic redistribution to their policy platform. However, clientelism involving offering work is punished less harshly and may even make candidates more electorally viable. Low-income voters, in particular, are more lenient towards clientelistic distribution involving work when it is combined with pro-poor programmatic distribution. Our findings help explain why politicians continue to use certain types of clientelism and how these are paired with programmatic redistribution to mobilize voter support. • Clientelism reduces voter support, but programmatic policies can mitigate this effect. • Low-income voters may accept clientelism involving work when combined with pro-poor policy programs. • Our findings explain why clientelism persists to mobilize voter support.
Justesen et al. (Mon,) studied this question.