Majoritarian elections in multi-member districts have not gained much attention in the literature so far. This is partially surprising, given that this has been the most common system in the 19th century and it is still is the most widely used system worldwide because very many private organisations and companies use it for their governing bodies and also many local and sub-national constituencies use this system. We will study the effect of multi-member majoritarian elections on elite and voting behaviour using a new dataset from more than 200 Swiss cantonal government elections (which are the provinces) using this system between 1971 and 2010. Multi-member majoritarian elections confront political elites with a coordination problem, in systems where no party has a clear majority, which is the case in most Swiss canton. If a party does not have a big enough vote share to win seats on their own, they have to find other parties that support their candidates and vice versa. However, on the same time they have to make sure that their own candidates do better than the candidates of the other parties. Further, incumbent candidates have an interest in not getting too much competition from other candidates from their own party or from allied party candidates, because this would put their re-election at risk. We can show, that in multi-member majoritarian elections parties tend to present a number of candidates which is slightly higher than their vote share, however they don’t inflate the number of candidates or present full list. If they do, they are often punished by voters and don’t get their candidates through and/or it increases the likelihood of incumbent defeat. This leads in sum to a fairly proportional outcome of multi-member majoritarian elections, which is counter to what we would expect from such an electoral system.
