Study title

Elections for the municipal council were held in 380 Dutch municipalities on March 5th 2014. These are the rows of the data matrix. In 133 of them at lest one Vote Advice Application (VAA) was used. Three different sets of variables appear in the columns of the data-matrix. First, whether a VAA was used, and if so, which one(s) out of the six VAAs. Next, the dataset includes for each of the 964 parties that offered candidates in at least one municipality the number of votes and the number of seats in the municipal council for the municipal elections that were held in 2014 and in 2010 (four variables per party). In addition to basic aggregate features of municipal voting (number of eligible voters, turnout, number of invalid votes) calculated aggregate features such as volatility (net shift of votes / seats in 2014 as compared to 2010) and fragmentation (effective number of different parties) are included. For each of the municipalities basic socio-demographic statistics from the Dutch Central Bureau of Statistics (CBS) are included (population size, crime, dominant age category, average income and ethnic fragmentation (the latter is a composite based on detailed CBS data on the number of inhabitants from different ethnicities)). Furthermore composite muicipal characteristics were added based on a factor analysis of measured indicators (scores on socio-economic rich-poor dimension and on cultural dimension (rural-monocultural versus urban-multicultural). The data set was set up to study whether VAAs are helpful to countervaill the trends toward lower turnout, higher volatility and further fragmentation.