From Kilometres to Alienation: Examining electoral accountability in amalgamated municipalities

Dataset comes from an exit poll survey from the 2022 local elections in the Czech Republic. The data from voters were obtained after they left the polling stations. The survey was conducted in 53 non-randomly selected municipalities. The selection process aimed to maximize variance. Voters that took part in the survey were selected randomly. Data are subset limited to amalgamated municipalities. Data collection was done in centres and peripheries of such municipalities. The respondents living in the central part of the municipality were coded in the analysis as 1, while respondents residing in the peripheral parts were coded as 0. Furthermore, the data include information about the distance between the respondent’s unit of residence and the centre. The distance is measured as the shortest distance on a public road between the intravillan of the centre and the intravillan of the unit since such a measure is the closest to the assumed mechanism of the physical barrier.

All the voters that participated in the survey were asked various questions regarding their electoral behaviour, their state of knowledge about local politics, their evaluations of the performance of the board of councillors or municipal council, and various sociodemographic questions. Voting for the incumbent is coded 1 when a respondent chose a political subject that was on the board of councillors in the past electoral term, or when they selected a candidate from a political subject.
Additionally, dataset covers several control variables - perceived performance of incumbents related to the voter’s most salient issue, mostly the management of the municipality. Original 5 point scale is recoded into a dichotomous variable, where 1 indicated a positive voter assessment (original values 1 and 2) and 0 a not positive assessment (3, 4, 5).
Secondly, there is support for the political party in the 2021 parliamentary elections, the most important elections in the Czech Republic. controlling whether respondents who voted for an incumbent voted for the same party in the last parliamentary elections. Third, location of the candidates for local government office is included to capture whether they were from the same part of the municipality (centre or periphery) as the respondent. Respondents’ level of trust in the board of councillors (incumbents) is also included in the data as well as sociodemographic controls (age, education, gender) .

本数据集源自捷克共和国2022年地方选举的出口民调（Exit Poll）。选民数据采集于选民离开投票站之后。本次调研覆盖53个非随机选取的市镇，选取原则为最大化样本方差。参与调研的选民为随机选取，且数据子集仅涵盖合并后的市镇。数据采集覆盖上述市镇的中心城区与外围区域。分析中，居住于市镇中心城区的受访者被编码为1，居住于外围区域的受访者编码为0。此外，数据集包含受访者居住单元与市镇中心的距离信息。该距离以公共道路上，市镇中心镇内区域与受访者居住单元镇内区域之间的最短路径进行测算，这一测算方式最贴合物理屏障的假设作用机制。所有参与调研的选民均需回答多类问题，涵盖其选举行为、对地方政治的了解程度、对市议会（Board of Councillors）的履职评价，以及各类社会人口学问题。若受访者投票支持的政治主体为上一选举任期内的市议会成员，或其选择了来自该政治主体的候选人，则将“为现任参选方投票”编码为1。此外，数据集包含多项控制变量：一是受访者对现任参选方在其最关注议题（主要为市镇治理）上的表现评价。原始5级量表被重新编码为二分变量：1代表受访者给出正向评价（对应原始分值1与2），0代表非正向评价（对应原始分值3、4、5）。二是受访者在2021年捷克共和国最重要的议会选举中的政党支持倾向，用于控制“为现任参选方投票的选民是否在上次议会选举中支持同一政党”这一变量。三是地方政府公职候选人的所在地信息，用于判断候选人与受访者是否来自市镇的同一区域（中心城区或外围区域）。数据集同时包含受访者对市议会（现任参选方）的信任程度，以及年龄、学历、性别等社会人口学控制变量。