What is Duverger’s Law?
Duverger’s Law – The First Past the Post Voting System Creates a Two-Party System
In the mid-twentieth century, French sociologist Maurice Duverger noticed a peculiar tendency: employment of the first past the post system in a single-member constituency system leads to a two-party system equilibrium. This principle has come to be known as Duverger’s Law.
Both theoretical modeling and historical example support Duverger’s Law. Under first past the post, all constituencies have an incentive to ally with one of the two leading parties. A small faction that is unable to secure one of the top two positions will be completely shut out of the government – unless the faction allies itself with one of the two leading parties. Compare this to the runoff system, under which first round losers can enjoy leverage as “kingmakers” during the runoff round. Or compare it to the incentives for a small constituency under the proportional system. Under a proportional system, single-party legislative majorities can be difficult to achieve. When the party with the most votes still lacks enough to enjoy an outright majority, minor parties become a necessary component of any coalition government, giving substantial leverage to such parties.
A telling historical example is found in the United States’ electoral college, which, while not a legislature per se, features the same electoral attributes. In 1992, independent Ross Perot secured about 19% of the popular vote. Nonetheless, he won a grand total of zero electoral votes. The same fate would likely befall any other nationwide group that sought to promote a presidential candidate. This is a powerful incentive for any such group to join with one or the other of the two major parties in order to promote its agenda or interests.
An important exception, though, is regional parties. While Perot’s support varied from state to state, it was legitimately nationwide. Contrast this with George Wallace’s 1968 bid, which did secure electoral votes despite winning a smaller percentage of the popular vote than Perot would win 24 years later. The difference? Wallace’s support was concentrated in the South, and he was able to win southern electoral votes – albeit still not enough to deny the winning candidate from between the two major parties an electoral college majority.
Does Duverger’s Law Still Hold? Considering Recent Anomalies.
Recent history has challenged Duverger’s Law. One challenging case is Canada. Governed by a parliament with first past the post, single-member districts, Canada enjoyed a stable two party system (at least at the federal level) prior to 1993. Yet in that year’s election, the separatist Bloc Quebecois won a dominant position in Quebec districts. This in itself is largely consistent with Duverger’s Law, as the Bloc is a regional party; the Bloc’s separatist leanings suggest further that maximizing power within the Canadian government is less important to the party than it is to most parties.
But more unusually, the conservative vote split between the venerable incumbent Progressive Conservatives, and the upstart, Western-based Reform Party. Once again, this is not unheard of – in Britain during the early-to-mid 20th Century, the Liberal party was replaced by Labour as the major left-leaning party.
Still, the ensuing – and ongoing – political era confounded the expectations set by Duverger’s Law. The Liberal Party dominated national elections as the divided right-leaning parties struggled to (re)unite. Finally, just as the conservatives cohered, the Liberals lost their dominant position on the center-left, challenged by the historically-marginalized New Democratic Party.
In all likelihood, equilibrium will return, with a single united center-left party facing a single united center-right party, with the Bloc Quebecois either marginalized or else watching as a regional bystander. But the current “detour from equilibrium” has now gone on for almost twenty years. This is not what Duverger’s Law would have predicted. (It is interesting to speculate whether the persistent presence of the Bloc Quebecois has somehow interfered with the operation of the Duverger principal.)
And, on top of the Canadian example, Britain has confounded Duverger’s Law, with the longtime third party Liberal Democrats achieving an electoral breakthrough in the most recent election sufficient to force the leading Tories into a coalition government. Time will tell if they can sustain their recent success. But the very existence of such success, on the heels of the recent Canadian experience, is fascinatingly inconsistent with Duverger’s Law.
In time, the Law will likely continue to enjoy overwhelming explanatory power – its theoretical modeling is simply too strong. But the recent failure of first-past-the-post, single-member-district electoral systems to come to rest at the expected two party equilibrium presents an interesting challenge to the model and will no doubt be a topic of study for political scientists for years to come.