The size of the district or the district magnitude refers to the number of candidates to be elected from that district. Depending on this number of seats, they can either be referred to as single-member or multimember districts. A single member district has a single seat while the multimember districts have several seats with the minimum number being two; a two-member district.
Electoral process formulas; plurality and majority formulas can be applied to both this type of districts while the proportional representation (PR) and the single non-transferable vote (SNTV) are applied to only multimember districts ranging from two-member districts to the whole nation where all the members of parliament are elected from (Lijphart, pg. 150). The district size has a strong effect on proportionality and the number of parties in two respects (Lijphart, pg. 50). Studies carried put by Horwill (1925) and, Taagepera and Shugarts (1989) have strongly supported this. First of all, it affects both plurality and PR systems but in a contradicting way. While increasing the size of the district will increase disproportionality by favoring the larger parties with the application of plurality and majority systems, it is vice versa with the PR systems as it leads to greater proportionally by favoring the smaller parties.
A good example to illustrate this with regard to plurality is a contest between party A and B in an area where A enjoys a greater popularity than B. A is likely to take all the seats if the area is a three-member district which is unlikely to be the case if the area is divided into three single-member districts; B may be able to capture one. In most plurality cases, multimember districts are smaller than single-member districts with rare cases being seen in countries like Mauritius (Lijphart, pg. 151).
It has twenty three-member districts and one two-member districts producing a total of sixty-two legislators. It should be noted that multimember districts have been on the decline due the fact that it increases disproportionality. However, it should be noted that in Mauritius that the three-member districts have improved another kind of proportionality by encouraging parties and party alliances to nominate ethnically and religiously balanced states (Lijphart, pg. 152) which has resulted in a better representation of the minority groups.
Another point to note is that the size of the district varies greatly in the PR systems unlike in plurality and majority systems (Lijphart, pg. 152) making it impact greater on the proportionality being achieved. A party representing a 10 percent minority is likely to be successful in a ten-member district which may not be the case in a five-member district. A nationwide district therefore is optimal for a proportional translation of votes into seats (Lijphart, pg. 152) with Israel and Netherlands being the two examples with such PR systems.
Two levels of districts are usually used by many of the list PR countries so as to enjoy the advantages of a closer voter-representative contact in small districts and the higher proportionality of large nationwide districts (Lijphart, pg. 152). Larger districts improve proportionality in the smaller districts as in the case of mixed member proportional (MMP) systems but they are less pronounced in the smaller multimember list PR districts than in the MMP single-member districts. Denmark, Norway and Sweden are examples of countries applying the two-tiered list PR systems with a nationwide district.