Proportionalitet i svenska Riksdagsval: En analys av den reformerade vallagen med Dirichlet-fördelade simuleringar

Sweden has a proportional electoral system where the seats in parliament are distributed according to the modified St. Laguë method. In 2015, the Swedish parliament approved amendments to the electoral law in order to improve proportionality in the national elections. The amendments came about after a failure to produce a proportional distribution of parliamentary seats after the 2010 national election. The amendments included changes of the first divisor from 1.4 to 1.2, and a new law for returning excess seats of overrepresented parties. The amendments were first implemented in the 2018 parliamentary election. This thesis aims to investigate the probability of achieving proportionality in parliament without the occurrence of excess seats with the use of Dirichlet-simulations for a number of different scenarios. My results show that the risk of excess seats is small if the current voting pattern persists. However, the electoral system has difficulties in handling many small parties. If additional parties gain representation in parliament, the probability of excess seats will increase significantly, causing the proportionality within constituencies to deteriorate