A common claim about the affluent democracies is that protest is trending, becoming more legitimate and widely used by all political contenders. In the new democracies, protest is seen as having contributed to democratization, but growing apathy has led to protest decline while in authoritarian regimes protest may be spurring more democratization. Assessing these ideas requires comparative trend data covering 15 or more years but constructing such data confronts problems. The major problem is that the most available survey item asks “have you ever joined (lawful) demonstrations,” making it difficult to time when this protest behavior occurred. We advance a novel method for timing these “ever” responses by focusing on young adults (aged 18-23 years), who are likely reporting on participation within the past 5 years. Drawing on the Survey Data Recycling harmonized data set, we use a multilevel model including harmonization and survey quality controls to create predicted probabilities for young adult participation (576 surveys, 119 countries, 1966-2010). Aggregating these to create country-year rate estimates, these compare favorably with overlapping estimates from surveys asking about “the past 5 years or so” and event data from the PolDem project. Harmonization and survey quality controls improve these predicted values. These data provide 15+ years trend estimates for 60 countries, which we use to illustrate the possibilities of estimating comparative protest trends.

关于富裕民主国家的一个普遍说法是，抗议正在成为趋势，变得更加合法并被所有政治竞争者广泛使用。在新民主国家，抗议被视为促进了民主化，但越来越冷漠导致抗议减少，而在专制政权中，抗议可能会刺激更多的民主化。评估这些想法需要涵盖 15 年或更长时间的比较趋势数据，但构建此类数据会遇到问题。主要问题是，最可用的调查项目询问“你有没有加入（合法）示威”，这使得很难确定这种抗议行为何时发生。我们通过关注可能在过去 5 年内报告参与情况的年轻人（18-23 岁），提出了一种对这些“永远”响应进行计时的新方法。利用调查数据回收协调数据集，我们使用包括协调和调查质量控制在内的多级模型来创建青年参与的预测概率（576 次调查，119 个国家，1966-2010 年）。将这些汇总以创建国家/地区年率估计值，这些与询问“过去 5 年左右”的调查和来自 PolDem 项目的事件数据的重叠估计值相比具有优势。协调和调查质量控制改进了这些预测值。这些数据提供了 60 个国家超过 15 年的趋势估计，