- Debate—Proportional RepresentationThe Problem with PR
Arend Lijphart's article on "Constitutional Choices for New Democracies" [Journal of Democracy 2 (Winter 1991): 72-84] attempts to provide scientific evidence for the superiority of proportional representation (PR) to the system of plurality elections. The author presents a comparative analysis designed to show that regimes based on plurality elections do not measure up to parliamentary-PR regimes in terms of "democratic performance."
Lijphart considers the effects of electoral systems on eight variables, which we will consider successively. The first correlation suggests that PR favors the representation of "minorities" and pressure groups. As clearly shown by the statistics, women legislators are more numerous in Nordic countries, which also tend to spend more money on family policies. Although tradition plays a role, the phenomenon is made possible by PR: candidate slates are chosen by party leaders, who are more easily influenced by strong women's movements.
The relationship between PR and voter participation is not as clear. If it were calculated on the European basis (as a proportion of registered voters), U.S. voter turnout would be similar to that of Western Europe. Moreover, Lijphart's figures would look quite different if he had not made some questionable decisions in categorizing countries. France, for instance, might be counted as a presidential-plurality democracy alongside the United States. Germany (whose mixed electoral system has majoritarian effects) belongs among the parliamentary-plurality regimes, while Spain and Portugal (which Lijphart ignores) should be included among the third group, the parliamentary-PR democracies. [End Page 30]
Lijphart's next set of figures indicates that northern European countries have a more equal distribution of income, which is not surprising. If there is a link between the electoral system and the greater degree of economic equality in these countries, it may not have much to do with democracy. When conservatives win elections in such countries as Sweden, Denmark, Norway, and Finland, they must form coalitions with other parties, which makes it hard for them to pursue their democratically mandated program of reducing the welfare state.
Lijphart's use of Robert Dahl's system for rating "democratic quality" raises the question of what criteria best measure democratic performance. If the index includes variables such as turnout (as measured in the U.S.), the number of parties, and the strength of interest groups, it introduces a strong bias in favor of PR with this sample of countries.
Finally, the correlations with inflation, economic growth, and unemployment (underestimated in Nordic countries because of highly protected jobs) are difficult to exploit. These indicators are much more powerfully influenced by many other important factors.
Consequences of Electoral Systems
Lijphart accepts from the beginning a fundamental hypothesis—namely, that the electoral system largely determines the party system and through it the structure of the government. Thus, countries where PR is the rule end up with multipartism and coalition governments, while plurality elections favor the two-party system and single-party governments. But as Lijphart notes, opinions diverge on how the party system affects the exercise of democratic governance. It is precisely on this point that it would have been fruitful for Lijphart to test the hypothesis against empirical data.
Such an analysis would have clearly shown that bipartism favors governmental stability and decision-making capacity as well as periodic alternations in power. Multipartism, on the other hand, is positively correlated with ephemeral governments, periods when the chief executive office goes unfilled, repeated elections, and long tenures in office for fixed groups of key politicians. The more parties a country has, moreover, the greater is the incidence of these phenomena.
When the government rests on a homogeneous majority, it remains in power for the duration of its mandated term (stability); can apply its program (efficiency); and is likely, should it falter, to lose power to a strong and united opposition (alternation). By contrast, the coalition governments so common in PR systems often cannot survive serious disagreement over particular measures (instability); need inordinate amounts of time to build new coalitions (executive vacancy); and when they fall, call new elections that generally return the same people (nonalternation). [End Page 31]
The contention that PR favors...