For six decades the civil society and democracy thesis has generated great interest. How dependably can the relationship of associational membership to democracy be demonstrated empirically? Using a data set derived from 37 national surveys in eight countries, this article finds that the questions used commonly to measure civil society are unreliable or simply measure different things, thereby imperiling the assumed universal and robust relationship between civil society membership and democracy. By emphasizing problems related to question wording, this article intends to prompt the profession to improve and standardize measurement of membership in civil society organizations to determine whether this tantalizing hypothesis has the powerful predictive value ascribed to it by the field of comparative politics.

六十年来，公民社会和民主论据引起了极大的兴趣。如何凭经验证明协会会员制与民主的关系？本文使用从八个国家进行的37次国家调查得出的数据集，发现通常用来衡量公民社会的问题是不可靠的，或者仅仅是衡量不同的事物，从而损害了公民社会成员与民主之间假定的普遍而牢固的关系。通过强调与问题措辞有关的问题，本文旨在促使该行业改进和规范民间社会组织中的成员资格衡量，以确定这种诱人的假设是否具有比较政治领域赋予的强大预测价值。