Benjamin D. Seal and Dr. Ray Christensen, Political Science
Political scientists continue to debate the effect of electoral rules on electoral outcomes. One area of intense study is proportional representation versus single-member districts, and single member districts versus multi-member districts. This research tests the effects of electoral rules on election results for the Democratic Party of Japan in the 1996 House of Representatives election and in the 1998 House of Councillors election. Analysis focuses specifically on how the Democratic Party, a minor party during the two elections covered, benefits from proportional representation (PR) and suffers in single-member districts (SMD).
The debate over PR versus SMD centers on the work of Maurice Duverger, and in particular his work known as Duverger’s Law and Duverger’s Hypothesis, which state that “the simple majority [plurality vote] single-ballot system favors the two-party system” (Duverger 1954, 217) and “proportional representation favors a multi-party system”, respectively (Duverger, 239). Duverger’s Law and Hypothesis rest on a mechanical factor and a psychological factor. The mechanical factor refers to how “all but the two strongest parties are severely underrepresented because [smaller parties] tend to lose in each district” (Lijphart 1999, 165). The psychological factor adds to the third party demise as voters realize that they are wasting votes and adjust their choice strategically (Duverger 1954, 226). The psychological factor extends to politicians “whose natural tendency is not to waste their energy by running as third-party candidates but instead to join one of the large parties” (Lijphart 1999, 165). Critics claim that Duverger simply mixed up the causal direction, i.e., that the number of parties determines the electoral system. Other critics claim that the number of parties in a system is determined by the number and type of cleavages in society. Sociologists reject Duverger’s “institutional determinism” which allows only a residual rather than a systematic role for social cleavages. Reaching agreement between the two opposing schools is difficult because of the level of incommensurability between proponents of Duverger who attempt a more scientific approach to studying politics, and detractors who find the scientific approach more problematic or even unwanted (Cox 1997, 15).
Political scientist Gary Cox makes an effort to reconcile the two approaches. He points out that beginning with the assumption that social cleavages matter does not have to mean that institutions are meaningless, and vice versa (1997, 19). Cox suggests that one way to test Duverger’s Law and Hypothesis while controlling for social cleavages is to collect data from countries with upper and lower houses with different electoral rules (ibid. 20-25). In these states, social cleavages remain the same for upper and lower house elections, while different electoral rules for each house allow for testing Duverger’s Law.
Applying Cox’s logic, Japan is an ideal country to use in testing the effects of electoral rules because completely different electoral systems operate within the same election for the same legislative body.
Japan’s unique electoral system, instituted in 1994, includes both proportional representation and plurality vote principles, making it a fascinating case-study. Within the five hundred seat lower house (House of Representatives), three hundred members are elected in single-member districts, and the remaining two hundred are elected by proportional representation from eleven electoral blocks. Thus, the lower house is three-fifths single-member districts and two-fifths proportional representation; voters at the polls cast two votes, one for the candidate in their district, and one for a party in the proportional representation vote.
Members of the 252-seat upper house (House of Councillors) serve six-year terms, with half of the members’ terms expiring every three years. The House of Councillors also mixes proportional representation and plurality principles. Of the 252 seats, one-hundred are elected through proportional representation, while the remaining 152 are elected in multi-member districts. As in the lower house, voters cast two ballots, one for the candidate in the district race and one for the party in the PR vote. Each of Japan’s 47 prefectures is a district in the upper house, and population determines the number of seats in the district. The district with the highest population is Tokyo with a district magnitude of four; there are four districts with a magnitude of three, eighteen districts with a magnitude of two, and twenty-four districts with a magnitude of one.
In order to test the effects of single-member districts in the lower house, I compared the nationwide results for the Democratic Party in the candidate vote with the party results in the PR vote. Because the candidate and PR votes occur on the same day with identical electorates, time is not an influencing factor. Candidate quality is not an important intervening variable because of the limited role it plays in PR votes. In 1996, the Democratic Party won 16.1% of the PR vote nation-wide, but only 10.62% of the vote in the single-member districts (all election data, unless otherwise noted, is from Miyagawa 1999). After breaking the data down by district, I computed the difference between the PR vote and SMD vote in each of the lower House’s three hundred districts. More votes were cast for the Democratic Party in the PR vote than for its candidates in 241 out of 300 districts. Out of three hundred districts, the Democratic candidate outperformed the party PR vote in only fifty-nine districts, and in only thirty-four of the districts did the candidate outperform the party by more than five percent.
Similar results are found when comparing PR and single-member district results for the Democrats in the House of Councillors. As noted earlier, of the forty-seven prefectures, twenty-four of them have a district magnitude of one and therefore are in practice single-member districts. Using the same test, Democratic candidate vote minus Democratic PR vote, the party won more votes than the candidate in eighteen of twenty-four districts. The Democratic Party clearly performed better in the PR vote than in the single-member districts.
Analysis clearly shows that electoral rules substantially influence outcome. The Democratic Party of Japan, a small, incipient party in 1996, won a larger percent of the vote and more seats in the proportional representation vote than in the single-member district vote of the House of Representatives and in the House of Councillors.
References
- Cox, Gary W. 1997. Making Votes Count: Strategic Coordination in the World’s Electoral Systems. Cambridge, England: Cambridge University Press.
- Duverger, Maurice. 1954. Political Parties. New York: Wiley.
- Lijphart, Arend. 1999. Patterns of Democracy. New Haven: Yale University Press.
- Miyagawa, Takayoshi. 1999. Seiji Handobukku (Political Handbook). Tokyo: Seiji Koho Sentaa.