Disproportionality in a Two-Party System
Numerous studies of electoral systems have demonstrated that single-member district, plurality elections are prone to high levels of disproportionality between the distribution of votes among the parties and the party allocation of seats in the legislature. This is important not only from the point of view of fairness, but also because disproportionate electoral results discourage party formation, voter turnout and issue-oriented campaigns. The result is an enfeebled political culture characterized by low levels of participation, personality-centered campaigns and under-representation of minority political positions.
1992 Elections for U.S. House of Representatives
In the 1992 elections for the U.S. House of Representatives, there was very little difference between the vote shares of the two major parties (3.4%), yet the difference in seat shares was 19.8%. If the electoral system had allocated the seats in proportion to the vote shares of the parties, the Republicans would have won 36 more seats. The average vote/seat difference of the two largest parties (8.2%) was comparatively high for U.S. House elections: 63% higher than the average for the decade of the eighties.
In the proportional representation (PR) systems of Western Europe and Scandinavia, the corresponding figure is only one-fifth as high. Yet this overall figure obscures even wider disparities in many states which canceled one another out. In half of the states (see table, next page) the average vote/seat difference was greater than 10%, and in eight states it was greater than 25%. In ten states the votes of one party were completely wasted, winning no seats at all. The Democrats won a seat percentage in excess of their percentage of the vote in two-thirds of the states. Four states even saw manufactured majorities; e.g., the electoral system awarded the majority of seats to a party which won only a minority of the vote.
Effects of Disproportionality
The effects of electoral disproportionality on political participation are clear in terms of both party formation and voter turnout. The number of electoral parties is depressed by the recognition that votes for smaller parties are wasted. According to the standard method for calculating the effective number of legislative parties (which adjusts the number according to the relative equality of seat shares), the 1992 election allowed the representation of only 1.98 parties. The level of effective party formation is two-to-four times higher in the PR countries.
Furthermore, because coming in second is unrewarded in single-member district elections, the persuasive activities of the two major parties are geared toward winning the allegiance of the middle of the political spectrum. Having reduced their ideological differences, the parties compete by attacking each other's candidates in terms of their personal characteristics. The appeals of smaller parties are largely ignored since attitudes toward them are nearly inconsequential for rational voting.
Faced with such limited choices and distasteful campaigns, the electorate is not inclined to participate in the election. Even though the turnout (55%) for the 1992 election was the highest since 1972, it was 30 percent lower than the average for proportional representation elections in Europe.
Of course, diminished party formation and voter turnout mean under-representation. The under-represented parties are those who would seriously consider basic changes in the social system, and research on the socio-economic characteristics of non-voters suggests that progressive labor and minority-oriented parties would be the main beneficiaries of increased voter turnout. There could be some truth in Noam Chomsky's assertion that Americans are actually more dissident today than in the late 1960's, but our institutions won't allow us to express it freely.
Stephen Winn is a professor of sociology at Marshall University. His recent work includes a sociological analysis of the transfer of ballots in preference voting elections in the United States.
* Calculated from Congressional Quarterly
11/7/92, "Election '92 Results"
Table of Contents