What Do Voters Know about Cabinet Formation?

在美国，选举后分配行政权力的过程通常很简单——得票最多的候选人获胜。然而，在议会制民主国家，情况并非如此，因为单一政党通常不会产生大多数立法席位（所谓的联盟制度）。在这些制度中，每次选举之后都会有一段时间的党派间谈判。这次谈判决定了谁成为总理，以及其他哪些政党拥有行政权力。对于选民如何管理这些更为复杂的系统，人们知之甚少。他们只是投票给他们最喜欢的政党，还是他们的投票策略更复杂？他们是否担心把选票浪费在一个几乎没有机会进入内阁的政党上？他们甚至知道不同党派是否更有可能进入内阁吗？大多数先前的研究都忽略了这些问题，并简单地假设选举后关于内阁的讨价还价太复杂，选民无法理解，因此对他们的投票方式影响不大。在这个项目中，研究者使用全国选举调查，向联盟制度中的选民询问问题，揭示他们对联盟形成过程及其结果的理解和不理解。这项调查的动机是一个特定的理论模型（史蒂文森，1997年发展），该模型表明，选民实际上非常有能力理解联合体制，并利用这种理解来修改他们的投票方式（也就是说，用他们的投票来试图影响谁进入内阁）。他们获得这种理解，不是通过了解联盟形成过程的细节，而是依靠经验规律形式的信息捷径，这些规律往往是他们国家联盟形成过程的特征。例如，虽然一个国家的选民可能不知道如何提出不同执政联盟的提案的规则，但他们可能知道立法机关中最大的政党通常担任总理一职。史蒂文森的理论模型只是表明，这类信息对选民是有用的，并且（连同其他一些经验规律）允许他们理性地投票，旨在影响内阁的组成。当然，这些理论上的主张依赖于这样一种观点，即选民实际上感知到了研究人员可以识别的有关该制度的经验规律（例如，最大的政党担任总理职务，或者总理倾向于选择意识形态相容的政党作为合作伙伴）。然而，这并不一定是真的。选民们可能对这些简单的事实一无所知。研究者在这个项目中使用的调查工具有助于解决这个问题，它揭示了选民是否：1)感知到他们国家联盟形成特征的简单经验规律，以及2)他们是否使用这些规律来构建他们对联盟形成过程可能结果的预期。如果这些问题的答案是肯定的，那么研究人员就可以对研究者提出的联盟系统中投票的具体理论模型更有信心。此外，也许更重要的是，这样的结果可以修正一个普遍的观点，即在联盟制度中，联盟的形成过程过于复杂，无法影响个人的投票方式。

In the United States, the process of allocating executive power after an election is usually straightforward -- the candidate with the most votes wins. However, this in not the case in parliamentary democracies that do not produce usually a majority of legislative seats for a single party (so-called coalitional systems). In these systems, a period of inter-party negotiations follows each election. This negotiation determines who becomes the prime minister and what other parties have a share of executive power. Little is know about ho voters manage the greater complexity of these systems. Do they simply vote for the party that they most prefer, or are their voting strategies more complex? Do they worry about wasting their vote on a party that has little change of getting into the cabinet? Do they even know whether different parties are more or less likely to get into the cabinet? Most previous work ignores these questions and simply assumes that the post-election bargaining over cabinets is simply too complex for voters to understand, and, thus, has little impact on how they vote. In this project, the investigator uses national election surveys to ask voters in coalitional systems questions that reveal what they understand and do not understand about the coalition-formation process and its results. This inquiry is motivated by a specific theoretical model (developed in Stevenson, 1997) that suggests that voters are actually quite capable of making sense of coalitional systems and using this understanding to modify the way that they vote (that is, to use their vote to try to affect who gets into the cabinet). They gain this understanding, not through knowledge of the details of the coalition-formation process, but by relying on informational shortcuts in the form of empirical regularities that tend to characterize the output of this process in their countries. For example, although voters in a country may not know the rules that govern how proposals for different governing coalitions are put forward, they may know that the largest party in the legislature normally holds the position of prime minister. Stevenson's theoretical model simply suggests that this kind of information is useful to voters and (along with some other empirical regularities) allows them to cast rationally votes that are intended to affect the composition of the cabinet. These theoretical claims, of course, depend on the idea that voters actually perceive the empirical regularities that researchers can identify about the system (for example, that the largest party holds the prime ministership, or that the prime minister tends to pick ideologically compatible parties as partners). However, it is not necessarily the case that this is true. Voters may be ignorant of these simple facts. The survey instrument that the investigator uses in this project helps to settle this question by revealing whether voters: 1) perceive the simple empirical regularities that characterize coalition formation in their countries, and 2) whether they use these regularities to structure their expectations about the likely results of the coalition-formation process. If the answer to these questions is positive, then researchers can have more confidence in the specific theoretical model of voting in coalitional systems put forward by the investigator. Moreover, and perhaps more importantly, such a result can revise the common view that, in coalitional systems, the process of coalition formation is too complicated to affect how individuals vote.