Replication Data for: Three Is a Crowd: Information and Electoral Coordination in Argentina

Successful coordination around a Duvergerian equilibrium requires accurate and consistent information about parties’ expected electoral support. In practice, such information is often unreliable and rarely available at the local level, thus hindering voters’ coordination. In this paper we leverage Argentina’s Open, Mandatory and Simultaneous Primary Elections as a large-scale survey of voter preferences. Using data from 135 municipalities in the province of Buenos Aires (2011-2023), we show that a narrower margin between the top-two placed parties in the primary increases both turnout and the proportion of positive votes in the general election, while decreasing electoral fragmentation. In line with Cox’s (1997) expectations, we further show that the second-placed party in the primary is substantially more likely to win the election than the third-placed one. Consistent with theoretical predictions, these effects are more pronounced (a) in concurrent elections; (b) in smaller municipalities; and (c) when the second-placed party is closer to the first-placed one.围绕迪韦尔热均衡（Duvergerian equilibrium）实现有效协调，需掌握关于各政党预期选举支持率的准确且一致的信息。现实中，这类信息往往不可靠且在地级地方层面难以获取，进而阻碍了选民的协调投票行为。本研究将阿根廷的公开、强制同步初选（Open, Mandatory and Simultaneous Primary Elections）作为选民偏好的大规模调研工具。基于布宜诺斯艾利斯省135个市镇2011至2023年的选举数据，研究发现：初选阶段前两名政党的得票差距越小，大选的投票率与有效票占比越高，同时选举碎片化程度越低。契合考克斯（Cox, 1997）的理论预期，本研究进一步证实：初选阶段位列第二的政党赢得大选的概率显著高于位列第三的政党。与理论预测相符，上述效应在以下三类场景中更为突出：(a) 多项选举同步举行时；(b) 市镇规模较小时；(c) 初选第二政党与第一政党的得票差距更小时。