Electoral competition with lying, inattention and targeted narratives

我打算将应用理论与经验相结合，以更好地理解不同政治话语的普遍性和受欢迎程度。我对后真相政治和政治极化现象尤其感兴趣。我目前正在研究一种选举竞争的模式，这种模式缺乏关注，需要第三方进行事实核查。在我的模型中，候选人争夺选民有限的注意力，策略性地选择竞选信息和播出时间分配。注意力不集中会让选民接受谎言，因为事实核查信息可能会被排挤出去。笔者打算进行实证分析和规范分析，研究选民注意力不集中的预期总福利成本。政治经济学文献最近开始扩展规范的选举竞争模型，将行为见解纳入选民对候选人的看法。我的模型在两个主要方面区别于这项工作。首先，选民最佳分配注意力的理性注意力不集中模型不能解释为什么候选人有时会把大量的广播时间花在选民感兴趣的话题上。基于突出刺激在吸引注意力方面的作用的经验证据，我因此假设选民更有可能记住更重要的信息——只看记住的信息的表面价值，并利用他们之前的信息来弥补缺失的信息。其次，这些模型关注的是政策选择，假设候选人致力于政论，因此不会给竞选谎言留下任何空间。我认为候选人会实施他们喜欢的政纲，但可以自由选择误导性的竞选信息。虽然大谎言（前提是被相信）可能会大幅增加候选人的选票份额，但它也有风险，因为大谎言会引发突出的事实核查信息。因此，候选人战略性地选择竞选信息和相应的播出时间分配，以试图挤出不那么有利的信息。因此，广播时间和谎言可能既是战略上的补充，因为大的广播时间可能会挤占事实核查，从而降低大谎言的风险，或者是战略上的替代品，因为广播时间短的候选人可能会使用极端谎言作为最后的求助策略，以触发事实核查，并分散选民对对手的突出言论的注意力。在平衡状态下，对选民注意力的争夺会给候选人之间的谎言动态带来许可，因为播出时间的压力降低了事实核查信息被记住的风险。为了更好地理解候选人选择强调某些主题的驱动因素以及由此导致的选民信息丢失，我打算将我的模型扩展到多主题设置。作为第二个项目，我打算实证研究是什么让某些受众特别容易接受某些政治叙事。经济模型最近开始包括各种背景下的叙述，有效的叙述被理解为改变接受者的信念或对自己目标的感觉。然而，是什么让叙事变得有效，这个问题在很大程度上仍然悬而未决。在政治微目标和社交媒体时代，理解是什么让一些接受者特别容易接受并愿意分享某些叙事，似乎是理解政治叙事传播的关键。由于“叙事”是多方面的，其有效性的可能来源从身份自我信号动机到对悬念的品味，我打算对相互竞争的假设进行实证调查，可能会使用社交媒体数据。

I intend to combine applied theory and empirics to contribute to a better understanding of the prevalence and popularity of different political discourses. In particular, I am interested in the phenomenon of post-truth politics and political polarization.I am currently working on a model of electoral competition with inattention and third-party fact-checking. In my model, candidates compete for voters' limited attention, strategically choosing campaign messages and airtime allocation. Inattention makes voters receptive to lies as fact-checking messages may be crowded out. I intend to carry out both a positive and normative analysis, studying the expected aggregate welfare cost of voter inattention.The political economy literature has recently started expanding upon canonical electoral competition models by incorporating behavioural insights to voters' perception of candidates. My model distinguishes itself from this work in two main aspects. First, rational inattention models in which voters optimally allocate their attention cannot explain why candidates sometimes dedicate a large airtime to topics of limited interest to voters. Building on empirical evidence on the role of salient stimuli in capturing attention, I thus assume that voters are more likely to remember more prominent messages - taking the remembered messages at face value and using their prior to complete missing information. Second, those models have focused on policy choice, assuming that candidates commit to platforms, and therefore not leaving any space for campaign lies. I assume that candidates will implement their preferred platform but are free to choose misleading campaign messages. While a large lie may, conditional on being believed, substantially increase a candidate's vote share, it is also risky as large lies trigger prominent fact-checking messages. Candidates thus strategically choose campaign messages and a corresponding airtime allocation to try to crowd out less favourable messages. Airtime and lies may thus act both as strategic complements, since a large airtime is likely to crowd-out fact-checking and therefore makes large lies less risky, or strategic substitutes, as a candidate with a small airtime may use extreme lies as last recourse strategy to trigger fact-checking and distract voters from his opponent's prominent discourse. In equilibrium, competition for voters' attention gives rise to license to lie dynamics between the candidates, as the airtime pressure reduces the risk of fact-checking messages being remembered. To better understand the drivers of candidates' choice to emphasize certain topics and the resulting information loss for voters, I intend to extend my model to a multi-topic setting.As a second project, I intend to empirically investigate what makes certain audiences particularly receptive to certain political narratives. Economic models have recently started including narratives in a variety of settings, an efficient narrative being understood as something shifting a receiver's beliefs or sense of his own objective. The question of what makes narratives efficient has nonetheless remained largely open. In the age of political micro-targeting and social media, understanding what makes some receivers particularly receptive to and willing to share certain narratives however appears key to understand the spread of political narratives. As "narratives" are multi-faceted and the possible sources of their effectiveness range from identity self-signalling motives to a taste for suspense, I intend to empirically investigate competing hypotheses, possibly using social media data.