Replication Data for: Who Defects? Unpacking a Defection Cascade from Russia’s Dominant Party 2008–12

Under what conditions do individuals withdraw support from dominant parties in nondemocratic regimes? Employing an original panel survey, we measure the same individuals’ support for Russia’s dominant party first at the peak of its dominance in 2008 and again shortly after it suffered a cascading defection of regime supporters in 2011–2. This allows us uniquely to explore the microfoundations of theories of regime defection cascades, generally supporting the argument that they involve complex “informational” as well as “reputational” processes. Accordingly, we find that early and eager movers in such a cascade tend to come from less socially vulnerable segments of the population, to have less need to rely on other people for interpreting events, to believe the regime has lower levels of popular support, and to come from more heterogeneous communities. We find little role for mass media (including social media) or democratizing zeal in driving Russia’s 2011–2 regime defection cascade.<br /><br />The u variables are from the 2007-2008 survey and the v variables from the 2001-2012 survey. The file I’m sending you is an additive merger, i.e., it includes all cases and all variables from the two variables, merged by case (respondent). The panel is embedded in the larger data set, i.e., respondents who were interviewed both times. Other respondents were either interviewed in the first survey and not re-interviewed in the second, or alternatively not interviewed the first time but interviewed in the second survey.

在何种条件下，个体将撤回对非民主政权（nondemocratic regimes）下主导政党（dominant parties）的支持？
本研究依托一项原创面板调查（panel survey），对同一批受访者的支持态度进行了两次测量：一次是2008年俄罗斯主导政党处于影响力顶峰之时，另一次则是2011至2012年该政权支持者出现连锁叛逃之后不久。这一研究设计使我们得以独家探究政权叛逃连锁反应理论的微观基础，研究结果总体支持这类连锁反应涉及复杂的“信息性”与“声誉性”过程的核心论断。据此我们发现，此类连锁叛逃中的早期响应者与积极叛逃者，往往来自社会脆弱性更低的群体，更少依赖他人解读公共事件，认为该政权的民众支持度更低，且来自构成更为多元化的社区。我们未发现大众媒体（含社交媒体）或民主化热情在推动俄罗斯2011至2012年的政权叛逃连锁反应中发挥显著作用。

u变量取自2007-2008年的调查，v变量取自2001-2012年的调查。本次发送的文件为按个案（受访者）合并的加法式合并数据集，即包含两次调查的所有个案与全部变量，合并依据为受访者身份。该面板数据集嵌入于更大的整体数据集之中，仅包含两次调查均完成受访的受访者。其余受访者要么仅参与首轮调查、未在第二轮受访，要么未参与首轮调查但在第二轮受访。