º It's Not Ideology
º It's Our Election Laws
º Works Cited
Americans are fond of diversity in the private market. We like a broad array of choices in ice cream, in radio stations, in ethnic food, in automobiles, and in religion. We also celebrate diversity in the political marketplace. Political parties exist in democracies to put their candidates into seats of power and keep them there, but all of them face serious competition by ambitious rival parties. Perhaps then it ought to surprise you that America has only two major national political parties with any realistic chance to win the presidency or the majority of seats in the two houses of the national Congress. The same is true within states for governors and state legislative seats. We have a thorough political duopoly of Republicans and Democrats dominating our politics. This has been so for a long time (since the 1860s without a break) and is highly likely to remain so long into the future. My job is to explain why.
It's Not Ideology Top; Next Down
Students often want political conflict to disappear. Rodney King's 1992 plea (during L.A. riots) was "why can't we all just get along together"? James Madison answered that two centuries ago. People do fight over basic moral values (fundamentalists v. teaching evolution, stem cell research opposition, pro-choice v. pro-life on abortion). And people also fight over scarcity and shares of material goods (labor v. capital, rich v. poor, debtors v. creditors). Democracy doesn't put an end to that; no, it gives open expression to it. The trick is to channel these conflicts well enough to make the fights into fair ones. That way government assumes legitimacy; those in power won office fair and square, and they're naturally "paying off" those who invested votes and labors and capital in getting them elected. It's a vast sight better than being ruled at point of a gun or by people who proclaim that God, not voters, put them in power and will keep them there.
American politics operates on principles enunciated by Madison in Federalist 10. One is that numerous factions exist in a large diverse society. Each is "united by a common impulse of passion or of interest"; in other words, faction members may share basic values, or basic material interests. But none is a numerical majority of citizens or voters, since factional division lines are so numerous and varied. How then will political leaders form majorities to get themselves elected, pass laws, and make policies? Or short of that, how will they form pluralities large enough to get those things done? I define a political coalition as an alliance of two or more political factions to achieve what no faction can achieve by itself (Wikipedia, Coalition and Political coalition). The standard usage of this term in political science is usually applied to European-style multiparty systems and governance, but it has application to America as well. Coalition in our Madisonian context is a remedy to absence of "natural" (factional) majorities of people. You make your faction as big as possible, but once that limit is reached, you either go forth to find partners or yield power to rivals who do so. The result is a large plurality or small majority under a common roof.
So even though the Founders themselves abhorred the spirit of factionalism that underlay formation of political parties, America in the 1790s saw parties form anyhow--and right there inside the Washington Cabinet, with the forceful Alexander Hamilton leading one group (calling themselves Federalists) and the equally able Thomas Jefferson leading another (the Democratic Republicans) with James Madison as one of his leading collaborators. Over much time they and their issue divisions faded, but new issue divisions and parties and party leaders inevitably arose. You cannot have democracy itself without at least two thriving rival parties! Please forget about Rodney King and all getting along together. Politics is about resolving conflict. It can be done via compromise. It can also involve logrolling or vote trading, as when Hamilton won adoption of his cherished national bank in exchange for abandoning New York City's hope to become the nation's capitol city. But not always; sometimes one side wins and the other loses. Elections are designed that way. In 2008 we'll elect one new president, not two.
Contemporary American political life is dominated by conflicts of modern liberalism and modern conservatism (see also: Conservatism - Dictionary of the History of Ideas and Liberalism - Dictionary of the History of Ideas). That's a convenient way to organize things. There are two dominant clusters of people, and one party housing each group. Conservatives cluster around the Republicans, also called the GOP for "Grand Old Party" denoting its 1854 origin. Liberals cluster around the Democrats, who date even further back in time to the Jacksonian era of the 1830s. Not that Republicans were always conservative and Democrats always liberal; no, each party has survived through periodic wrenching changes of issue conflict and coalitional structure known as political realignments. I leave you to the text to see that rich history. But today, the GOP houses nearly all conservatives, and Democrats house their liberal antagonists.
So with two dominant ideologies, it's tempting to say "aha! That's why we have just two parties! One for the right wing, another for the left! That's pretty simple." Yeah, but it's entirely too simple. Here's why. See Patterson Chapter 6, pp. 177-79 on ideology. That entry affirms that most Americans aren't ideological at all, and that most are not consistently liberal or conservative. Political leaders and the "political class" often are; but most citizens aren't. So why should they confront only two choices at the voting booth? Why not a moderate party in between these two? Here's where political space matters; if I'm a moderate voter seeing a conservative 9 steps to my right, and a liberal 8 steps to my left, well, I suppose I'll hold my nose and vote for the "lesser of two evils" 8 steps leftward. But in idle time I might wonder why there's not a moderate party just 2 or 3 steps away from me.
What's more, liberalism and conservatism are not truly as cohesive or natural as underlying belief systems as their advocates would have you conclude. A former text of mine (Fiorina et al. 2005b, p. 138) said "In our view, ideologies are social constructions. ... can anyone explain why a person who prefers lower taxes and a strong military necessarily should oppose abortion but not the death penalty and should support strip-club regulations but not gun control?" By the way, all six of those issue positions are now "conservative," but what is there about any two or three of these that compel you to adopt the remainder? The parties are coalitions, not factions.
Nowadays we have a broad liberal v. conservative heritage on economic policy derived from the 1930s New Deal and 1960s Great Society liberal eras. Liberals loved those policies (creation of national minimum wage law, Social Security, medicare, assistance for labor against capital, affirmative action for racial minorities), conservatives loathed them. Add to that the Cold War of 1947-to-1990 and now the War on Terror of 2001-2008; conservatives want military strength above all while liberals call for more diplomacy and international cooperation. And since the 1970s, don't forget religion v. secular conflicts over moral values (abortion, gay rights, teaching evolution, stem cell research). That divides ideologues as well. We have economic liberals and also cultural liberals, economic conservatives and also cultural conservatives. The first two cluster in a coalition called the Democrats, the latter two in its rival called the Republicans.
But most Americans are moderate, not out on the left (liberal) or the right (conservative) spatial pole. And most Americans don't devote nearly the time and attention to politics to even become "ideological" in the first place. Morris Fiorina in the influential book entitled Culture War? The Myth of a Polarized America (2005a) shows that with policy questions listed on p. 17 (Table 2-1). Only Fiorina's "political class" is truly ideological, and is polarized around the liberal v. conservative poles (2005a, p. 14). The everyday self-identified Republican and Democratic citizenry are not nearly so polarized as media leaders, talking heads on television, or convention delegates would have you believe.
Please don't tell me that somehow a natural alignment of conservatives on all three of these will align in one party, v. liberals on all three in the other. It'd be just as easy to envision a Populist Party (liberal on the New Deal's economic regulation, conservative on church and religion matters) alongside a Libertarian Party (conservative on New Deal economics, secular and permissive on personal behavior and sexuality). That's what the Patterson text has (2008, p. 178). Add those two to the Republicans (conservative on both) and Democrats (liberal on both). And hey, then we voters would have four choices instead of two. Surely that's better than facing just two, with one 9 steps too right-wing and the other a mere 8 too left-wing.
So yes, ideology is a social construction, and we had but two choices in elections 2004, 2006 and 2008: largely conservative, or largely liberal. Which potion do you want? For most Americans, neither is ideal. But something about American political design keeps the choice at two, not four.
It's Our Election Laws Top; Next Down
All democratic countries (called polities here) have at least two parties which gain a lot of votes and obtain seats in power (in a parliament or legislature) at the nationwide level. Let's call all of these "national assemblies" for simplicity. But why do a few democratic polities like the U.S. tend to have only two parties there, while most have three or more? The answer is structural. It was first expressed by Michel Duverger as Duverger's Law (Duverger 1972): the election rule used to convert popular votes into electoral seats in power is the primary determinant of how many parties will exist.
Countries with more than two parties holding seats in their national assembly usually employ proportional representation (PR). So if party A harvests 20 percent of the vote, it will receive a proportional 20 percent of the seats. For voters, the emphasis is all about party label, and not so much on who the particular candidate(s) may be.
PR is popular among dozens of democracies, but it's not universal. A number of English speaking and Latin American democracies including the U.S. do not use PR. We use single-member, simple plurality (SMSP). Plurality rule means the candidate who wins the most votes in a defined geographic territory wins the office. Districts have single members; so for example, Representative Jo Ann Emerson is the Congresswoman from Missouri's Eighth Congressional District covering 27 full counties and parts of two more. A look at national Electoral College or congressional district maps reveals a tiled space with every residence in one district and state, and no one in two. SMSP is also called "first past the post" or "winner take all" or (for states and D.C. in the Electoral College) the "unit rule." Duverger tells us that this rule tends to reduce party number toward just two major political parties. That's right.
The job of political parties is to put slates of candidates into public office. The crucial first job is to pick those candidates (nomination) and the second is to enable them to win office (election). Nomination is done either by party leaderships in caucuses, or by voters in the reformist-inspired party primaries. The former prevailed for selecting presidential candidates up to 1968 in the U.S. Since 1968 each American major party has depended on party primaries to select its presidential nominee. That produces the distinctive American candidate-centered politics where calendar 2007 saw numerous "self starting" candidacies from both parties seeking their 2008 nominations in the busy presidential primary season in calendar 2008. Each party has but one candidate for one post in the general election.
And each office has but one occupant from a congressional district or a state or the nation. With SMSP, if you (and your party) win the most recorded votes in a given election, you win that post outright. The 50 state governorships, 100 U.S. Senate seats from the 450 states, 435 U.S. House seats from that many districts, and 51 Electoral College "units" (50 states plus D.C.) with 538 presidential votes are all decided on this SMSP basis. It's not really majority rule; it's plurality rule.
Now here's the rub with plurality voting: there is no formal or practical limit on permissible number of parties and candidates on the ballot except that it's one per party per office in the general elections held in our even-year Novembers. Most (not quite all) of these American general elections require only a plurality (the most votes of any candidate) rather than an outright majority (more than all others combined) to win that office. If elections were to require majority winners, then unresolved elections like Florida in 2000 would quickly become very common.
Duverger says SMSP is a profound disincentive to third parties (RangeVoting.org - Duverger's law, two-party domination). They can be and are born, but they die out fast. That's so even though third parties in America have historically been very important. They arise to contest elections; they often act as spoilers causing an established party's candidate to lose; they introduce new issues and arguments into campaigns; they boost voter turnout by offering an interesting third way to voters. But it asks too much for one to arise from the grassroots and actually win on its first or second try; and they rarely get more serious tries than that since our fixed election calendar typically puts 2 or 4 years between elections.
What happens if a third party does arise across a large nation and fare well enough to capture 20 percent of the voters' allegiance? In PR countries, they'll be in business with approximately 20 percent of the seats in the national assembly. In SMSP countries, they're likely to lose everywhere and get zeroed out.
You can see this quite readily via Dave Leip's Atlas of U.S. Presidential Elections illustration for 1992, when Ross Perot's lavishly funded third party presidential bid harvested 18.9% of the national popular vote but did not win any of the 50 states plus D.C. with their 538 Electoral College votes (1992 Presidential Election Results). This isn't a new story. In 1912 Theodore Roosevelt was spurned in seeking to oust incumbent Republican William Howard Taft from the Republican presidential nomination; so he turned to a Progressive or "Bull Moose" third party and actually outpolled Taft nationally in popular vote and Electoral College. But with the once-majority Republicans cleaved like this, the smaller party Democrats' Woodrow Wilson easily won vote pluralities in state after state to cruise to a decisive 435 to 88 Electoral College victory (1912 Presidential Election Results). T.R. had had enough after that; by 1916 he retired from this enterprise and returned to the GOP. This time they ran one candidate, Charles Evan Hughes, and almost defeated Wilson despite the latter having a spectacularly successful first term as president (1916 Presidential Election Results). Later, in 1924, the Progressive Party tried again, with their champion Robert LaFollette of Wisconsin, but his 16.6% of the nation's vote won only his home state of Wisconsin with 13 votes (1924 Presidential Election Results). The 1928 election saw no repeat of this. But 20 years later in 1948, southern defenders of segregation were angry enough with the Democratic Party's endorsement of integration to run Strom Thurmond on a third party. It worked to win four of the deepest southern segregationist states but failed to deny general election victory to Democrat Harry S. Truman (1948 Presidential Election Results); in 1952, the Democrats heeded the warning and avoided talking about integration during that campaign. Another 20 years after 1948, southern segregationist George Wallace of Alabama went third-party and helped defeat the national Democrat; but the Republican, not Wallace, won the election (1968 Presidential Election Results). And in 1972 when he was shot and nearly killed during his springtime presidential campaign, Wallace was contesting the Democratic Party's Maryland primary election. He had abandoned his own party after it served his onetime purpose. Now Ross Perot in 1996 was different; he'd won 18.9% in 1992 and had visions of grandeur and success four years later. But despite keeping its candidate and its funds aboard, this third party did not encourage voters to believe it had a chance. So in 1996 its vote harvest declined by more than half to just 8.40% (1996 Presidential Election Results).
That's the SMSP story. American third parties matter a lot, but they do not win and do not survive to hold seats in power. Our electoral game is rigged for just two players to stay in business. American separation of powers contributes a great deal to this. Look at the dominance of the two parties in holding 20th century congressional seats (Renka, Presidents and Congresses). The "other" category for third party seats is virtually empty. It's a lot to ask a third party not only to run a national presidential campaign, but also to produce a viable slate of congressional candidates in 50 states and hundreds of congressional districts.
But if you study regionally based presidential elections, you'll quickly notice that parties have often enjoyed a concentrated geographic power base. This is evident in my Presidential Election Maps, by County. The 1968 third-party George Wallace campaign won 4 southern states and a great many counties (colored in green). Surely that party might have won the senatorial and congressional seats from there, too, if they'd run a full slate of candidates for various offices (which they didn't). So geographically concentrated third parties can win something on a regional (not national) basis.
Thus even use of the SMSP election rule will not kill off a regionally dominant third party that runs a slate of legislative candidates. Canada uses SMSP to choose its House of Commons parliament from single districts (called "ridings" there) within provinces, just as America does for the House of Representatives with congressional districts within its states. Canada's large French-speaking province of Quebec houses the semi-separatist Bloc Québécois party. Bloc Québécois fares very well in French-speaking districts but poorly to nonexistent elsewhere. Election results are at The Atlas of Canada - The 39th Federal Election, 2006 and also Canada Election Maps 2005. We see that more than 2 parties have significant seat holdings in parliament; but SMSP within ridings produces largely two-way contests determined by plurality winner. This pattern crosses elections rather than being strictly temporary. The 2006 result is essentially similar to 2004, per The Atlas of Canada - The 38th Federal Election, 2004 and The 37th Federal Election, 2000. So its career politicians win office, win reelection, and build careers as third-party members. The "third-party" Bloc Québécois has its Quebec-based national parliamentary seats, a stake in power, and no reason whatsoever to fade out of existence after an election.
Canada differs from America in one respect: being a parliamentary system, parliament chooses its Prime Minister. America is not parliamentary. Our Electoral College operates almost like direct national popular election of a president. So the two American parties are organized to exist in all states, to participate in national presidential nominations, and to contest for the presidency. Both parties are truly national, not regional or local. There is no American parliament to negotiate the choice of a Canada-styled Prime Minister out of some coalition among two or three parties; there is only the nationwide Electoral College with electors obediently casting votes in deference to the state's popular vote plurality of November 2-8.
The result is that our once-separate Southern Democrats largely stayed in the national party (the Democrats) rather than becoming a separate geographically dominant Dixiecrat third party akin to a Bloc Québécois. Look back through Presidential Election Maps, by County in the southland to witness decades of one-party southern dominance by Democrats before the 1960s civil rights era brought that to a close. They could have separated but never did.
You will also find that all those southern states did have state Republican parties. Why would southern Republican parties even bother to exist in the 11 old confederacy states below the Mason-Dixon line? They obviously could not win elections in this one-party terrain filled with white voters who were taught from childhood about the evils of Republican-run occupation of the southland. More generally, why have two parties in a state when one cannot win there? Democrats in Utah win practically nothing statewide. Is it worth having a Democratic Party in such Republican-dominated states?
The answer comes this way: Democrats in Idaho and Utah do exist, do participate in presidential primaries, do go to the convention to cast votes for the party's nominee. After that, the gig is up as the Republican wins both states in November. But it's still worth having a party and showing the flag there. If the Democrat is elected President, then ambitious Idaho and Utah Democratic leaders are in line for coveted executive branch and national party organization jobs. Besides, for ambitious political persons, these parties offer far fewer internal obstacles on the climb toward prominence than the state's dominant party does. Parties thrive only to the extent they work as vehicles for the ambitious to advance.
This same behavior existed in the segregationist south for a century. During the long decades of Democratic dominance in the south, Democratic state parties were led by luminaries whose grip on power and highest office was virtually unbreakable. Imagine being a young George H.W. Bush in the 1940s, journeying to Texas to make your political fortunes, and discovering that within its state Democratic Party resided Senator Lyndon Johnson (aged 40 in 1948) AND Speaker of the House Sam Rayburn (known universally as "Mr. Sam")! The Texas GOP was moribund, but it did exist, did send delegates to presidential conventions, did contribute to national decisions affecting the presidency, and did offer some hope that if ever Texas turned away from one party, another was already on hand to provide an alternative. And that is precisely what happened in the long upward path for Daddy Bush in the 1960s, 1970s, and 1980s.
In other words, America not only discourages third parties via use of SMSP. It also discourages them through existence of decentralized, state-based parties that participate everywhere in choosing national presidential nominees. That's a double deterrent to any sustained existence of American third parties. A regional third party like George Wallace's American Independent Party of 1968 offers the chance to change racial politics in a conservative direction. But it cannot hope to actually win national office, and its supporters forfeit the chance to participate within the emerging southern Republican parties. That's an historical wrong path for them. The white South nowadays has gone to the GOP, and the nonwhite south along with a handful of culturally liberal bastions (Austin, Athens, Raleigh) have gone to the Democrats. What other realistic choice do they have? Two parties hold all the real power in American politics.
Duverger, Maurice. 1972. "Factors in a Two-Party and Multiparty System." in Party Politics and Pressure Groups. New York: Thomas Y. Crowell, 1972, pp. 23-32. URL: janda.org/c24/Readings/Duverger/Duverger.htm.
Fiorina, Morris P., Paul E. Peterson, Bertram Johnson, and D. Stephen Voss. 2005. The New American Democracy, 4th ed. New York: Pearson Longman.
Fiorina, Morris P., with Samuel J. Abrams and Jeremy C. Pope. 2005. Culture War? The Myth of a Polarized America, 2nd ed. New York: Pearson Education, Inc.
Leip, Dave. Dave Leip's Atlas of U.S. Presidential Elections. URL: www.uselectionatlas.org/.
Patterson, Thomas E. 2008. The American Democracy, 8th ed. New York: McGraw Hill.
RangeVoting.org - Duverger's law, two-party domination. URL: rangevoting.org/Duverger.html. Parent site: RangeVoting.org - Center for Range Voting - front page at rangevoting.org/RangeVoting.html.
Renka, Russell D. 2008. Presidents and Congresses. URL: cstl-cla.semo.edu/renka/ui320-75/presandcongress.asp.
Renka, Russell D. 2006. Presidential Election Maps, by County. URL: cstl-cla.semo.edu/renka/ui320-75/renka_papers/party_system_maps.asp.
Copyright©2009, Russell D. Renka