Postelectoral Data Analysis

From the provisional electoral results available by minimum voting units it is possible to estimate voting movements or transfer matrices between two elections (Autonomous 2015 vs 2019) taking into account that the vote is secret - that is why we speak only of simulations - and that in all units there are discharges (young people who vote for the first time and new registered in the census section) and low (deaths and voters who change residence with respect to the previous election).

There are different methodologies to calculate these transfers and, consequently, different estimated results that do not have to coincide. In our case, we resort to methodologies based on ecological inference.

To apply these methodologies, by the hand of JM Pavia -Dtor. of the Research Group on Electoral Processes and Public Opinion at the University of Valencia- we assume a certain homogeneity in the electoral behavior among voters of tables that are closer geographically to each other, of similar census size, of demographic and socioeconomic characteristics, predominances of the same political type , etc ... .. these groups will allow us to exploit the variations that occur in registered vote distributions. In this way, we will obtain congruent estimates of "electoral mobility" (origin-destination) for each voting unit, adding the result by municipality, district or province and Autonomous Community.

Next, we publish the matrices of those Communities (and their municipalities) that make available the autonomic electoral results at the table level and authorize MINSAIT to disseminate the transfer simulations.