Duverger's law is the basic explanation of the dominant influence of election rules on the number of parties in a democratic polity (Duverger 1972).  This generally hold among many democracies (Cox 1997; Lijphart 1994; Lijphart 1999, Ch. 5).  But William Riker said SMDP (single-member-district plurality) was conducive to reduction in number of parties, but in certain conditions that number is not reduced all the way down to two (Riker 1982).  That condition is geographic concentration, which permits a regional party to dominate SMDP elections only in that locale.  We'll look at three major countries to get at Riker's important qualifier to Duverger's law. 

Two of those countries are Canada and the United Kingdom.  Wilfried Derksen's longtime Electionworld.org site has now moved to Wikipedia with a generic guide to Parties and Elections located here: Electionworld - Wikipedia > Elections and Parties.  Scroll from there to Elections by country for Elections in Canada and Elections in the United Kingdom.

And meanwhile to better understand plurality voting's effect, go to Elections and Parties > First Past the Post electoral system (name of recipient file:  Plurality voting system).  All of these names coalesce around winner-take-all, a single person who gets the most votes and therefore wins the office.  This can happen in a first-past-the-post single round (customary for most American elections, but not for Senator Mary Landrieu in the runoff system of the State of Louisiana), or two rounds with a runoff of two or more leading vote getters (just two in Mary's case, resulting in her getting the majority and another term in the U.S. Senate).  This is our SMDP that Duverger and Riker (with qualifications) say will reduce party number down toward just two (a duopoly).

The chief democratic alternative to SMDP is proportional representation or P.R.  It is widely used among democracies and has strong advocates who invoke its superior fairness and nonbias, but it is barely used at any level in American elections to sorrow of its chief advocates here (FairVote > FairVote - The Proportional Voting Solution).  There's no doubt that national-election P.R. is conducive to having numerous parties represented in the country's national parliament.  Small homogenous Iceland in the frigid north Atlantic is a fine example.  See Elections by country > Elections in Iceland > 2009 Parliamentary election with 7 parties getting votes and 5 of them earning seats in their Althing (arguably the world's oldest popularly elected parliament; see Icelandic parliamentary election, 2009 for details).  Other Nordic countries follow suit, including Elections in Denmark, Norway, and Sweden.  Germany as well typifies the P.R. principle of faithfully converting a share of votes to a closely comparable share of seats (Elections in Germany; and German federal election, 2005 selecting the 16th German Bundestag).

Now the U.K. and former Commonwealth countries, along with the U.S. and many Latin American countries, practice various forms of SMDP.  We leave proportional representation to continental Europe and elsewhere in the democracy world.  But the United Kingdom national election of 2005 shows more than 2 parties, as the third-party Liberal Party got 8.2% of seats.  The district map of Britain's House of Commons also shows that this is an SMDP system, with 1 winner per district just like allocation of America's 435 congressional districts in our lower house.  Maps with tiled space colored for party tell this tale (2005 UK Election Map.svg and mother site United_Kingdom_general_election,_2005; and United States House of Representatives elections, 2008 - Wikipedia).

Canada tells a similar story with a concentration of separate partisan votes from the large French-speaking districts of Quebec Province.  See 39th General Election of Canada, 2006 for indication that Quebec is very distinct.  SMDP district division is essentially the same for Canada as for the U.K and the U.S.  And you can easily find other examples with regionally concentrated third parties winning seats under SMDP rules--but not in the U.S.

So SMDP alone is clearly not sufficient to bring off the strict party duopoly we experience in the U.S.  Why, then, are we different from our friendly brethren and sisters of those countries?

History comes in here.  There is a very robust and longstanding duopoly at all state and national levels in America.  This has basically been true for 150 years since the Civil War, even to the point of having the same two parties--a Grand Old Party (the Republicans, birthed in 1854) and an even older one (the Democrats, dating from the 1820s emergence of Jacksonian politics).

It probably should not be so, in America.  After all, we have few "barriers to entry" in formal constitutional or legal terms.  Those who want to create a party, or another factional interest group, are pretty darned free to do it.  What's more, Americans demonstrate in most things private and public that they like a range of choices and an array of different results.  American parties grew rapidly to a mature system during our first and second "party systems periods"--first in 1793 to 1824, and second from 1828 to 1854, with maturity by 1840 (First Party System - Wikipedia; Second Party System).  This transition produced a radical change in selection of presidential nominees, from control by "King Caucus" within the Congress up to 1824, followed by the 1824 legitimacy crisis and its replacement by the nominating convention system by 1840 in both the Democratic Party and Whig Party.  Those conventions allocated voting power to delegations from every state in rough proportion to their relative sizes (Democratic National Convention; Republican National Convention).

America had some regional parties during these early decades.  But they did not thrive in the long run.  The Federalist Party by 1814 became a sectional New England party, and death followed as their chances to win the nationwide presidency faded to almost nothing.  A sectional party forfeits any real chance to win the presidency unless it could build a broad political coalition in many states.  Now other antebellum parties after 1814 also became sectional, under the forceful divisive influence of the slavery issue.  Death came to the Whigs this way, around 1852.  The Democrats survived after a sort with adherents both north and south, but through some good fortune for the American future, they too were deeply divided in the 1860 presidential election.  That allowed the new free-soil northern-based Republican Party to get Lincoln elected despite the inherent absence of that party in the slave-state south.  The Republicans swept every northeastern and midwestern state (Leip, 1860 Presidential General Election Results).

By conclusion of the Civil War, our familiar D v R duopoly had emerged in mature fashion as a so-called Third Party System (1860-1894; and see Third Party System - Wikipedia).  Both parties continued to rely on national nominating conventions to pick their presidential standard-bearers.  National party affiliations became increasingly "sectional" or geographic, with Republicans dominant in much of the north and Democrats controlling the south and many "peripheral" hinterlands (Bensel 1984; Martis 1989; Martis and Elmes 1993).  You can see both the sectionalism and the duopoly's stranglehold on all presidential elections from 1860 onward here:  Presidential Election Maps, by County.

Parties dominated American political life.  The average citizen got information via party or not at all.  Voting turnout rose, often with help from corrupt operatives and absence of the Australian or secret ballot (not to mention voter registration, a 20th century innovation).  Local and state politics typically saw one party far dominant over the other.  After Reconstruction ended in 1877, the old confederacy South became a one-party stronghold of Democrats.   Democrats comparably had no chance to win anything much within Republican-run New England states like Vermont and New Hampshire.  So why did either party remain alive and vital in such places?  Because southern Republicans and New England Democrats both participated in choosing the national nominee at the GOP convention.

Rise of such dominant parties in each state also was crucial to creating SMDP rules in the first place.  Any dominant party in a given state will insist that statewide election be winner-take-all (for governor, two Senate seats, and allocation of its share of the presidential Electoral College).  The states one by one adopted winner-take-all standards for counting Electoral College votes from those states.  And state presidential electors were no longer free agents (as the founders assumed they would be).  Now they just vote the way the state voting plurality instructs.

Any American sectional third party forfeits two crucial things:  1) the ability to directly win the presidency, barring extraordinary division and luck circa 1860; and 2) any influence on selection of presidential nominees via the national nominating conventions.  That's an awful lot to lose.  American politicians have largely chosen to work inside one or both the big-house coalitional tents known as Democrats and Republicans.  The one-party segregationist South did this for almost a century, from the 1870s all the way to the national civil rights era of the 1960s.  They never formed a separate formally organized "Dixiecrat Party" in all those years despite having distinct views (on race) that warranted such distinction.  To understand why, consider the price of a century-long forfeiture of influence over the election of presidents!  Regional "parties" have in fact existed for long periods in America since 1860, but rarely with true separate labels, organizations, and candidate slates of their own.  Instead they persist as longstanding large factions within one major party (or both) in the peculiarly American party duopoly.

Top

References

Amy, Douglas.  Proportional Representation Library.  URL:  www.mtholyoke.edu/acad/polit/damy/prlib.htm.

Bensel, Richard Franklin.  1984.  Sectionalism and American Political Development, 1880-1980.  Madison:  University of Wisconsin Press.

Cox, Gary W.  1997.  Making Votes Count:  Strategic Coordination in the World's Electoral Systems.  Cambridge:  Cambridge University Press.

Duverger, Maurice.  1972.  Factors in a Two-Party and Multiparty System - The Technical Factor:  The Electoral System.  URL:  janda.org/c24/Readings/Duverger/Duverger.htm.

FairVote.  FairVote - Home.  URL:  www.fairvote.org.

Lijphart, Arend.  1994.  Electoral Systems and Party Systems:  A Study of Twenty-Seven Democracies, 1945-1990.  Oxford:  Oxford University Press.

Lijphart, Arend.  1999.  Patterns of Democracy:  Government Forms and Performance in Thirty-Six Countries.  New Haven:  Yale University Press.  (See Chapter 5, Party Systems:  Two-Party and Multiparty Systems, pp. 62-99.)

Martis, Kenneth C. with Ruth Anderson Rowles and Gyula Pauer.  1989.  The Historical Atlas of Political Parties in the United States Congress: 1789-1989.  New York:  Macmillan.

Martis, Kenneth C. and Gregory Elmes.  1993.  The Historical Atlas of State Power in Congress:  1790-1990.  Washington, D.C.:  Congressional Quarterly Press.

Riker, William H.  1982.  The Two-party System and Duverger's Law: An Essay on the History of Political Science (via JSTOR).  American Political Science Review 76:4 (December, 1982), pp. 753–766.
Top

Copyright ©2009, Russell D. Renka
Monday, August 31, 2009 10:48:10 AM
Disclaimer