Brendan O'Leary

The US presidential, Senate and House Elections of 2012
Part One: Before The Elections

One of the most able public intellectuals of the last generation of American citizens, John Kenneth Galbraith, maintained that modesty was the most overrated of the virtues (Parker 2005). Yet the task set by the editors of this review obliges genuine modesty. American, or rather US politics, is not among my fields, either as a researcher or teacher. The assignment was to explain the US elections to the intelligent Irish reader, to highlight what may be missed by a European eye, and, last but not least, to account for why contemporary Republicans are such powerful contenders for office, to the surprise of a majority of observing Europeans, including the Irish.

In the first part of this article, published before the elections, the focus will be on the Electoral College and explaining the bases and trajectories of US political parties. In the second, mostly written before the elections but published afterwards, the focus will be on explaining the strength of the Republicans, and evaluating Romney and Obama’s leadership respectively and retrospectively.

Whatever my scholarly disqualifications, composing this letter in Philadelphia is entirely appropriate. The Constitution of the United States was drafted here in 1787, when it was the capital of the federation (Beeman, 2009; 2010). The city soon lost its chance of becoming the permanent capital, though it was the largest urban concentration at the declaration of independence. The loss was mostly because the Virginians, including Thomas Jefferson, James Madison and George Washington, successfully manoeuvred to ensure that what would become Washington DC would be located in the slaveholding South (Wills 2003: chapter 15; for a more traditional account see Cooke 1970). Indeed Philadelphia lost out from a series of exchanges between New York financiers (initially led by Alexander Hamilton, intent on building a powerful federation through assuming all the debts incurred by the states in the revolutionary war of independence), and Southern slaveholders (fearful of an urban, urbane, democratic, cosmopolitan and abolitionist capital). Sordid exchanges between New York-based financiers and southern white racists have been recurrent features of US politics.

A dumb institution made by smart people in the Philadelphia summer

Philadelphia is also an apposite site from which to write because the Electoral College that every four years determines the choice of president and vice-president was invented here. The college’s mechanics always matter, and its intimate sub-components may be decisive in the week and month ahead. Its design was a botched compromise between those who wanted the entire (male) citizen body directly to elect the president, those who wanted the Congress to do so (or the House of Representatives, or the Senate), and those who wanted the states to make the choice (either through their governors or legislatures). The compromise was concluded in the peak heat of a typically steaming summer, at the end of the constitutional convention, long before air-conditioning was invented. Created without the counsel of electoral system specialists it provides sustained fodder for my profession.

Why the Electoral College is Bad for America, by Professor George Edwards, is still probably the best single guide to the relevant anomalies, contradictions and potential for harm (2004). Edwards emphasises that the original design was a disaster waiting to happen, not least because initially the electors voted for two persons for president. As Professor Jack Nagel puts it, “The 1800 election produced an Electoral College tie between Thomas Jefferson and his running mate, Aaron Burr. Under the Constitution’s contingent election procedure, the decision then passed to the House of Representatives, which deadlocked for thirty-five ballots. On the thirty-sixth ballot, Jefferson was elected president, thus pulling the fledgling democracy back from a constitutional crisis and the brink of civil war.” The governors of Pennsylvania and Virginia had prepared their legislatures to mobilise their militias. The system was modified in 1804 by a constitutional amendment (the twelfth), “which substituted a conventional balloting system that gives each elector only one vote per office” (Nagel 2007: 43). The twelfth amendment, however, did nothing to repair the other flaws of the system.

The slavish origins of Electoral College numbers

Electors, not US citizens, are supposed to determine the presidency (and vice-presidency). The electors are, however, chosen by each state’s citizens. Each state is entitled to the total number of electors that derives from adding its number of senators (two) to its number of members in the House of Representatives. The latter number varies every ten years, and is apportioned proportionally to the state’s population (not its number of registered voters) at the last relevant census. Next week California will pick fifty-five (two plus fifty-three) and Texas thirty-eight (two plus thirty-six) electors, whereas numerous states with small populations, including Alaska, the two Dakotas, Delaware and Vermont, choose three (two plus one) each.

Gary Wills (2003: Chapter 3) explains lucidly why the original design of the college deliberately advantaged the slaveholding states. The number of electors, as we have seen, is mostly proportional to each state’s population, not to its citizens. For these purposes, slaves, who had no votes or representatives ‑ these days these facts need to be stated ‑ were counted as three-fifths of a person. The chosen fraction was not an accident: it balanced the northern and southern states’ allocations in the original House of Representatives. It also gave an absurdly distorted advantage to the South in the Electoral College until the civil war (shortly after which a different form of distortion, accomplished through disenfranchising blacks, became the norm until approximately 1965; sustained efforts are still being made to revive this southern tradition).

In the presidential election of 1800 the sole reason Thomas Jefferson was even in contention against the incumbent, John Adams, was because disenfranchised slaves were counted as fractions of persons. That is why his Northern critics called Jefferson the “Negro President” (Wills 2003: passim). The Jefferson-Burr ticket, by the way, is an example of a Southern slaveholder and Northern alliance that went awry, but otherwise it set a precedent. At least nine Democratic presidential-vice-presidential tickets have been alliances between New York money and Southerners.

Guaranteed disproportionality

The upper limit on the size of the Electoral College is given by the prescribed limit of 435 members of the House of Representatives, added to one hundred members of the Senate (fifty by two), and three electors from the District of Columbia, that is 538.

The design of the college guarantees some disproportionality, that is, a mismatch between Electoral College outcomes and popular vote outcomes. Small states are overrepresented. Wyoming recently had 0.2 per cent of the US national population, but with three electors it has 0.56 per cent of the total college, that is it has nearly three times as much weight in the college as in the national popular vote. Similarly, three of the smallest states (three by three) may have as many electors as one medium-sized state with nine, such as Colorado, even though it may have a significantly higher population than the three small states combined.

The overrepresentation of small states guarantees disproportionality; yet more important in many respects is the fact that each state is free to choose how its Electoral College votes are allocated. In the US there is no federation-wide uniformity in voting procedure to determine electors, across the states, nor, notoriously, is there uniformity in electoral administration within each state, regarding such consequential matters as voting times, ballot design, absentee ballots, use of computers, counting and appeal rules, ID requirements, and so on. As is often said among officials of the United Nations Electoral Assistance Unit, the US system fails most international best standards for electoral administration.

Most states choose to adopt winner takes all to determine their electors, because they correctly judge that doing so makes their voters more pivotal than they would otherwise be. That means that the presidential and vice-presidential team that wins most of the votes in the relevant state wins all that state’s Electoral College vote (even if in a two-team contest 49.99 per cent of voters voted for the other team). A minority of states have used “proportional” representation. Nebraska says it does so today, but it is not what Europeans would recognise as proportional representation. Instead, votes counted in the US congressional districts inside Nebraska are used to determine who should be the electors according to winner takes all within each district, not the state as a whole, with the proviso that two electoral votes, corresponding to the two senate seats, go the statewide winner. This formula meant that in 2008 the Republican John McCain won four electors in Nebraska whereas Barack Obama won one. The allocation of electors from Nebraska is certainly more proportional than elsewhere; it has, however, the side effect of making the state less likely to be pivotal in presidential campaigns.

Even if all states used the same winner takes all system to choose their electors, significant disproportionality between the national popular vote and the Electoral College outcome would occur regularly. For example, one team may win a large majority in the Electoral College by winning wafer thin majorities of the popular vote in each state in which it is victorious, whereas the major rival team may pile up enormous but ultimately wasted majorities in a losing Electoral College combination of states. Conversely regional strength and narrow victories elsewhere can be the road to success. Abraham Lincoln had a dramatic Electoral College victory (over 59 per cent), with less than 40 per cent of the popular vote, in a field of four. Shortly after the civil war broke out.

The college, it may fairly be said, displays all the pathologies of winner takes all voting systems (surveyed in O’Leary 2010), but then adds some more. This exposition should be sufficient to explain why the system does not ensure that the winning presidential team in the college will have a majority in the national popular vote. For similar reasons, the winner of the college vote need not even be the plurality winner of the popular vote, especially when there are multiple candidates running.

Adding oddities to disproportions

A further bizarre consequence follows from winner takes all. In a reasonable voting system a candidate or a party should benefit if they receive a higher turnout in a particular locality. In 2012, however, the Republican team may win Florida’s twenty-nine electors with a significantly lower turnout than the Democratic team wins exactly the same number of electors from New York state, with the same share of the vote as in Florida. That would make a Floridian’s vote far more salient than a New Yorker’s, even though the college outcomes would cancel each other out. It also follows that when and where a team believes it is going to win a state come what may, it has little incentive to increase the turnout of its voters (unless it believes that will help its party’s candidates for the Senate and the House, which is not always true). So it may therefore withdraw its resources to fight elsewhere, probably joined by the certain loser. In turn, this too may lead to disproportionality between the Electoral College outcomes and shares of the national popular vote. Rational presidential candidates, and those who fund them, therefore focus their attention on states where the margin of difference between the parties in the most recent elections has been roughly ten per cent or less (the so-called “battleground” or “swing” states). The margin falls to five per cent or less as more polls are taken and election day approaches.

After the college is chosen

The patient reader is now almost finished seminar one of Electoral College 101, but wait. We must ask what happens when the Electoral College is chosen, assuming the courts are adjudicating no protracted recounts or disputes. The chosen electors must congregate in their respective state capitals, not in Washington DC, to vote on the second Wednesday in December. They send their certified results to Washington. A point to make at the bar counter, or as part of a quiz, is therefore available: December 12th, 2012 will be the most likely date on which the president will be elected, not November 6th (or 7th). The editors of the Dublin Review of Books would not have been pleased, however, had an email declared that I was legally entitled to another month before handing in my article at the agreed time, namely the week before the election.

When they congregate, the electors are constrained in what they may do. They may not vote for two candidates from their own state. This fact led Professor Sanford Levenson to propose and participate in a losing lawsuit that argued that Texas electors could not vote for both George W Bush and Dick Cheney, because both were Texans. (Cheney maintained his official residence in Wyoming, which he had represented as a Congressman, even though he mainly lived in Texas.) It is less clear, however, that the electors are constrained from casting their ballots against the preferences of their states’ citizens. Whether the electors are delegates or representatives remains uncertain. The answer is another disaster waiting to happen, or, to put it in another way, waiting to be adjudicated by the current or a future Supreme Court. Let us restrain ourselves from peering down that abyss, not least because there are some further crevices ahead.

If a majority of all the electors diversely assembled in their states vote for a given presidential and vice-presidential team then that team is elected. If there is not such a majority, however, then an ill-devised route out of the morass is prescribed. The three presidential candidates with the highest number of electors’ votes then become candidates in an election decided by the House of Representatives ‑ which decides according to the formula of “one state, one vote”. Note, first, that it is not exactly clear how an equally divided state, for example one with equal numbers of Republicans and Democrats, should cast its single vote. Lawyers suggest it would have no vote, because in 1800 such states did not cast a vote. The impasse was broken when Federalist Congressmen from two equally divided states (Vermont and Maryland) abstained, creating a majority for Jefferson amid the rest of their delegation (Nagel 2007). Note, second, that an absolute majority of state delegations is required for election by the House, enhancing the prospects of deadlock. Note, third, that this rule makes the House of Representatives resemble the Senate, as the vote is based on the equality of states, but with this proviso, namely, that this second selectorate (the college being the first) has not had its composition for these purposes determined by any direct mandate. Differently put, whereas the electors arguably have a mandate to vote as they do, the House of Representatives as constituted for these purposes does not.

Meanwhile, amid all this confusion, the Senate will separately be electing the vice-president: I assure you, I am not making this up. The Senate’s task is easier. It must choose among the names of the two highest number of electors’ votes. Each senator casts a single ballot. The vice-president elect (and the future president of the Senate) is then available to act as president should the House remain deadlocked. Yes, you have read correctly, and I hope I have communicated correctly.

The immediate implication is that if there is no decisive outcome in the Electoral College, and if current projections for party performances in the Congressional elections are correct, the House, under these bizarre rules, would choose Republican Mitt Romney as president, and the Senate would choose Democrat Joe Biden as vice-president. That is because the Republicans are almost certain to control a majority of state delegations in the House, even if they lose their majority of members, and it is currently likely that the Democrats will retain (and possibly increase) their majority in the Senate.

This would be a power-sharing outcome that no reasonable power-sharing adviser would recommend or seek to design. When, after what I hoped was a careful reading, I understood these implications of the US Constitution I simply could not believe it, so I emailed the Professor Nagel cited above, under the heading “Please tell me that this is not true.” Having affirmed that it was indeed true, Nagel explained that “Amendment XII retained this provision from the original Article II, Section II, even though by 1804 they should have recognized the role of parties and the possibility that different parties would control the two houses.”

When – perhaps that should be “if” – the reader’s head recovers from this spin, please recall that there are 538 Electoral College votes, and that the winning team requires a simple majority of electors (270). We have been discussing what happens when there is no Electoral College winner, which would most likely arise in multi-candidate contests when each candidate has some significant strength (enough to win some electors). The designers, however, managed to create a system in which not only might there be no decisive winner, but one in which a tie is not only logically possible, but not altogether unlikely. A tie happened in 1800, that is in one of the fifty-four contests to date. In 1824 there was no majority winner because four candidates won electors. Moreover, in four elections since World War II, according to Edwards, very small shifts in the popular vote would have produced Electoral College ties, namely in 1948, 1960, 1968, and 1976 (in two of these cases there were significant third party candidates).

Should you wager, and, if so, at what odds?

Currently, the two most scientific websites that integrate polls of polls, namely RealClearPolitics and FiveThirtyEight, the first of which is alleged to lean Republican, and the second Democratic, agree in putting Obama and Romney at a tie in the popular vote. A tie is here defined as one in which the weighted average of all the polls puts the two candidates within 1 per cent of each other. Does the US then face the prospect of a matching tie in the college? There could, in principle, be such an outcome, but it seems unlikely, not least because of all the reasons already given that generate possible discrepancies between popular vote and Electoral College outcomes.

Nate Silver, the Bayesian statistical specialist at the The New York Times, who runs FiveThirtyEight currently rates the prospect of an Electoral College tie at three chances in a thousand (New York Times, October 27th, 2012). Since Silver in his useful and persuasive book on predictions (2012) describes himself as a prudent, empirically cautious fox rather than an attention-seeking hedgehog, his judgment is to be commended. A rational person should not punt on an Electoral College tie, except for sheer enjoyment. If there were to be a tie, however, then Romney would win the presidency because the likely House voting by state delegations, each counting as one, would give the presidency to him. Think about that, however. One of its weird consequences is that Obama needs 270 Electoral College votes to win, whereas Romney needs 269, and that Biden needs one less elector than Obama to retain the vice-presidency!

By contrast, Silver rates the prospects of Obama winning the popular vote but losing the Electoral College at fewer than two chances in a hundred, or just over 50/1 against for those who like to wager. The possibility, however, that Romney might win the popular vote but lose in the Electoral College is rated by Silver as much higher: 5.3 chances in a hundred. A flutter on that outcome is therefore worth it provided the bookmakers will offer you better than 19/1 against. Be prepared, however, to await the outcome of lawsuits.

The Electoral College is, in short, a standing indictment of America’s reputation for innovation. It is clear evidence that the United States has become a global superpower despite rather than because of its institutions. The smart people who made it cannot have expected it to survive: it was a short-run compromise, made easier because all assumed that Washington would be the first president. The American political class, however, shows little shame over failing to agree on a rational alternative for two hundred years. Their conduct, in turn, is an indictment of the US Constitution, which makes it so difficult to abolish or reform such a key institution (amendments require a qualified majority of two-thirds in each chamber of Congress and of three-quarters of the state legislatures), unless the Supreme Court conveniently rereads the Constitution (though it has scant interpretive room here).

Why should anyone be concerned?

Does the Electoral College really matter? Is raising these questions typical of political scientists who focus too much on counterfactuals rather than facts, or give too much attention to low probability events?

The college does matter: the college, not the voters, decides who holds the most powerful executive position in the most powerful country in the democratic world. Abraham Lincoln could not have won in 1860 without the college’s design. After the civil war the college has also mattered decisively. In 1876 another Republican, Rutherford Hayes, “won” the presidency in a hotly disputed contest. Even when we allow for large-scale fraud, on both sides, Hayes definitely lost the popular vote to Samuel Tilden of the Democratic party (which was then, among other things, the party of Southern racists). The Republicans claimed, however, that Hayes had won the college by one vote. In a violent atmosphere, several states disputed their college results, and sent rival certifications to Washington.

As the historian of “Reconstruction”, Eric Foner, puts it: “As on so many other questions, the Constitution was maddeningly ambiguous as to how the validity of disputed returns should be decided.” (2005: 576). Hayes “won” only after a special electoral commission, containing Congressmen and Supreme Court justices, awarded him twenty disputed Electoral College votes, after a series of eight-seven votes within the commission. The Democrats had in fact intimidated and ballot-stuffed on a massive scale in the South, so there may have been some narrow justice in the formal outcome. But, Hayes’s victory had a price. In return for eventually agreeing to it, the Democratic party succeeded in reversing the reconstruction of the South, and the Republicans ceded to the southern states full control of their institutions and franchise arrangements. The military occupation of the South ended, and Southern white racists, then in control of the local Democratic party, were able largely to exclude local blacks from the franchise for eighty years. Radical republicanism, the project to remake an egalitarian America, was extinguished for Hayes’s victory. So, yes, the Electoral College matters, as do the predictable generation of crises and disputes over its procedures. The world still wonders what would have happened had Al Gore won the presidency in 2000.

Nowhere else in the democratic world has directly imitated the Electoral College, and for very good reasons. The distributive requirements imposed on Nigerian and Indonesian presidential candidates, which oblige them to show a certain minimum of the vote across the numerous states of their federations, come closest. They also have similar in-built pathologies, and are disasters waiting to happen, but at least they have spared themselves an Electoral College.

What is to be done?

Eliminating the executive presidency and vice-presidency is beyond the American imagination ‑ trust me. Yet the outright abolition of the college has often been proposed. One proposal would let US citizens vote the way that Irish citizens vote for Ireland’s president, namely through the majority-preferential vote (primaries and party conventions could be kept). Another proposal would imitate the French (perish the thought). That is, if no candidate wins a popular majority in the first round, the two top-placed candidates go into a final run-off, which guarantees a majority-winner.

“Walk-around-the-obstacle” reforms, which would not require constitutional amendments, have also been proposed. One idea is that all the states would agree among themselves to use proportional representation to choose their electors. Unfortunately, this agreement would still be vulnerable to any state tempted to switch back to winner takes all to enhance its chances of being pivotal, and it would not solve the overrepresentation of small states. Another idea would persuade enough states to amend their statutes to award all their electors to the winner of the national popular vote (provided enough others do so), which would, in effect, render the college redundant, and would presumably have to pass muster before the Supreme Court at some juncture. This, the “National Popular Vote” plan, has actually made some progress: its story can be found on its website of the same name.

Such ideas will have a prospect of success only when both main parties feel equally liable to suffer badly from the college’s mechanics. How would a professional probability specialist calculate the chances of that? Presumably the number lies somewhere between Silver’s estimates of the chance that there will be a college tie this time (three in a hundred) and his estimates of the chances that Obama will win the college but lose in the popular vote (less than six in a hundred).

Perhaps only a truly violently contested outcome will trigger reform – though it did not do so in 1876 (the “reform” agreed then led to the constitutional striking down of the Civil Rights Act of 1875). Violence may be as American as apple pie, but it has never yet been significantly co-ordinated against the college, or for its reform, only for its manipulation.

Next week the outcome will be known, or not, as you now know.

Viewing US differences from a potentially swinging state

Philadelphia and Pennsylvania provide good vantage points for observing US elections for other reasons. The best ethnography of lively multiracial and multiethnic Philadelphia may be found in the work of the African American sociologist Elijah Anderson (2011) who powerfully conveys its hard-won though remarkable civility. It is the largest city in a battleground state that has an important harvest of twenty electors for the college. In 2004 John Kerry defeated George Bush in Pennsylvania by 50.9 per cent to 48.4 per cent. In 2008 roughly nine in ten citizens of Philadelphia who voted cast their ballots for Obama, and that enabled him to carry the state against McCain by 54.5 per cent to 44.2 per cent. (Such numbers never sum to one hundred because of rounding, invalid ballots and third party candidates.) Anecdotally, but at no risk of revealing the nature of my networks, the strength of Obama’s support in Philadelphia was such that it would be October 2010 before I met a city resident who claimed to have voted for McCain. It was not for want of asking on my part, and the person concerned was not sober.

Once outside of Philadelphia, however, it takes the law-abiding driver who is headed west at least five and a half hours before another clearly Democratic-majority enclave comes into sight, namely the former steel and now college city of Pittsburgh. Psephologists therefore call Pennsylvania “Pennsyltucky”, hinting that it is pregnant with Republican Kentucky. James Carville, the “ragin’ Cajun”, one of Bill Clinton’s campaign managers in 1992, by contrast, has described Pennsylvania as “two cities with Alabama in between”. He meant rural and white Alabama. In Philadelphia and Pittsburgh old churches in the core of the old cities host rock concerts, are given over to condominium conversions, and often simply rot. In between the two cities, the churches are more recently painted, and increasingly “mega”. Filled with believers on Sundays, black faces are rare. All the frequently filmed clichés of town and country, professional and redneck, urban homeboy and God-fearing farmer, college girl and volunteer soldierboy, have solid sociological and geographical foundations. These cleavages split the parties left and right, as they do in other democracies. A swing state has the heterogeneous demography or sociology of the United States as a whole ‑ this has been true of Ohio for some time. Typically it has urban Democratic enclaves surrounded by suburbs and the more sparsely populated countryside which vote Republican.

When maps of such states are shown at county level and are not weighted by population the effect is one of deep blue dots inside swathes of red. To win requires a party to mobilise its base and do slightly better than the other at breaking into its rival’s space. What makes the US visually different at election time from much of western Europe, notably Great Britain, of course, is that the colour red does not signal socialism or labour politics. Rather, red celebrates the conservative or libertarian (not the same thing) as the hue of the Republican Party. Blue does not signal conservatism, as it does in the UK; rather it denotes the social liberalism of the Democratic Party.

These parties are nevertheless strange entities to the European eye, and not just because their banners seem the wrong colour. The Republicans are an alliance of the super-rich, the “managers” of finance capitalism, and of degree-less whites. They are the “whitebread” party, topped with an upper-crust of hypercapitalists, a misalliance cemented by pan-Protestant evangelism. They control “the heartland”, or what their coastal critics call “flyover land”. The Democratic party, by contrast, is a multi-ethnic and urban coalition, an alliance of teachers, professors, medium and upper-income graduates in the public sector, and unionised workers; up close it resembles the British, French and Irish Labour and socialist parties. It controls the western and eastern coasts, and the cities near the Great Lakes. At their conventions, Republican women look feminine, and are less evident as candidates, speakers or activists than men. Democratic party women look professional, speak a lot, and are in abundance as candidates. There has, not surprisingly, been a “gender gap” for several decades: a majority of women vote Democratic, a majority of men vote Republican, “angry white men”, so the well-founded cliché goes. To which we should add, straight “angry white men”. The Democratic party welcomes gays, lesbians and the transgendered, and its nominee for president now supports gay marriage, after a period of ambivalence. The Republicans are Old Testament (or the Hebrew Bible or the Book of Mormon); the Democrats favour loving thy neighbour as thou choosest.

Americans will solemnly tell you that they do not have racial or ethnic parties, but roughly nine in ten Republican voters will be white on November 6th, and more than nine in ten black voters will vote for President Obama and the Democratic party.

Americans may less solemnly tell you that they do not have religious parties. They know, however, that neither of the two main parties has ever had an open atheist or agnostic on its presidential ticket. It is simply obligatory for candidates to believe ‑ or to pretend to believe ‑ in God. The Democratic party ran William Jennings Bryan for president at the end of the nineteenth century, a full-throated denier of Darwin. This time around the Republican party is running Willard “Mitt” Romney, a former Mormon bishop, whose formal beliefs are incompatible with the second grade science curriculum. One of the fastest growing categories of Americans, people with degrees who do not believe in God, is not directly represented in party politics. They bite their tongues within or near the Democratic party; whereas those who speak in tongues, forked or otherwise, are very content in the Republican party.

The more religious someone describes themselves as being, whether that be Protestant, Mormon, Catholic or Jew, then the more likely they are to vote Republican. Outside the big cities, especially in the South, Americans will ask you courteously which church you attend on Sundays in the way that a city-dweller may ask you where you shop for groceries.

Catholics are more likely to vote Democratic because they include Latinos, who are likely to break over two to one for the Democratic party, and because historically the “white ethnic” working class Catholics (the Irish, Italians and Poles) voted Democratic, and the socialisation effects still linger. Though “Reagan Democrats” (working class white Catholics who may vote Republican) exist in abundance, they are not a majority of their kind. Arab American Christians (Lebanese Maronites and Palestinians) usually count as Catholics and as Democrats.

For political purposes Mormons now count as Protestants. Mormonism is arguably the only truly native American religion, besides those of the true Native Americans. Questions about the “Americanism” of Mormons and the “Christological” soundness of The Book of Mormon have not arisen in the presidential campaign, though they mattered in the Republican primaries. Perhaps that is because Romney’s private adult life has mostly involved the very typical American religion of making money. Perhaps he has shown such flexibility on ethical questions that many voters who care deeply about religion (for or against) find it hard to believe that Mormonism drives any of his calculations.

The two parties now look more like the right and left parties of Europe than they once did, certainly in their policy platforms on economic management and taxation. Or, is it the other way around: do European parties increasingly look more like US parties? Marx warned the readers of Das Kapital, “De te fabula narratur”! (roughly “this story is about you”). He meant that the most industrialised country foretold the future of others, but no one has ever quite believed that of US political parties.

Their histories are very different from their closest current equivalents in Europe, and not just because both need to appear to be religious. Each retains different ethnic and religious bases. They have also changed places on key questions in the last century which have no parallel in Europe (for example on the emancipation of slaves, the emancipation of blacks, the emancipation of women, and on the merits of science). The Republican party did not begin life as the party of God, king and aristocracy, unlike many European conservative parties. It began as a meritocratic and enlightened party of free enterprise (albeit tempered by “high tariffs” on foreigners’ goods and services). The Democratic party did not begin life as a socialist, social democratic or overtly working class party, unlike most European left and centre-left parties. Instead it began as a lower case ‘r’ republican party (indeed the party descends from the original “Republicans” of Jefferson, the party of independent small farmers and craftsmen that opposed a centralised federation).

The fact that the modal Republican white voter is now poorer than the modal Democratic white voter is instructive. It shows that the Democratic party’s whites are more urban and more educated. That the Republican platform is officially opposed to abortion, period, whereas the Democratic party’s platform supports the right to choose, eloquently advertises the changed alliances with patriarchy and liberal feminism respectively. It is difficult to recall that the Republicans were ahead of Democrats on the female franchise.

Compare, by going online, the Electoral College maps of the presidential contests of 1912 and 1916, when the Southerner Woodrow Wilson won two presidential terms for the Democratic party, with the maps in 2000 and 2004 when George W Bush won two terms for the Republican party. It is easy to see that the overall patterns are remarkably similar. The difference is that the winning colours are different across the ninety year time span. Usually where the blues were once strong the reds are now, and vice versa. This stunning contrast is not a statistical fluke. Rather, it strongly suggests two possibilities, namely, that the parties have, over the course of a century, switched their core, or base geographic support, after multiple strategic shifts in their search for votes, or, that within these decades, the underlying demography and sociology of the relevant states have changed drastically, almost reversing previous patterns of party support. A good case can be made for both explanations.

The Republicans are now the dominant party in most of the South, especially the further South one goes, exactly the position that the Democratic party held from 1877 until the late 1960s. The explanation is well known: the decision of the Texan Democratic president Lyndon Johnson, along with Northern liberals in the Democratic party, and Republican liberals, to support the civil rights movement and impose reform on the racist states of the old confederacy. This brave and enlightened (if belated) decision had consequences that were fully felt in the 1980s. Within two decades a majority of Southern whites had shifted from their Civil War allegiances and now voted for the Republicans, who now eagerly sought them out and remoulded themselves. The Democratic party, in contrast, won the allegiance of most southern blacks, who could now vote effectively, but at the cost of its previous regional dominance, a loss magnified because many blacks migrated out of the South for a better life. Instead of the Democratic party having a strong base in the South with which to bid for Electoral College victory, as in Woodrow Wilson’s time, the Republican party now has that advantage. These facts may explain why the Democratic party has several times chosen white Southerners as its presidential or vice-presidential candidates (Jimmy Carter, Bill Clinton, Al Gore, John Edwards), to break the perceived Republican advantage.

These changes predicted “The Emerging Republican Majority,” to use the words of a famous book by Kevin P Phillips (1970) that I was assigned as an undergraduate a decade after its publication. The transformation of Southern politics, and the end of the misalliance between Northern liberals and Southern white racists was underpinned by other sociological and demographic shifts. The long entrenchment of Southerners in the Senate (seniority led to key positions on committees, and seniority is associated with low electoral turnover or one-party dominance) had led to major federal government spending programmes in the South. These invigorated its economy, enriched it, and led to net (white) immigration and population growth in the South. The depth of ingratitude to federal expenditures among many Southern whites is well known; their hostility toward welfare expenditures (which might benefit blacks and other minorities) is equally famous; less well known is how many of them have and continue to dine on government (as well as pulled) pork. Think military and defence expenditures. Simultaneously, the weakening of traditional and unionised industries in the northeast weakened organised labour, at the expense of the Democratic party. The major movements of southern blacks to Northern cities did not wholly compensate the Democratic party, partly because the “blackening” of Northern cities was matched by “white flight” to suburbia. Moreover, while many mobile white working class Democrats became suburban Republicans, many marginalised urban blacks did not vote at all.

The electoral weakness of overt socialism in US history, especially by comparison with Canada and western Europe, was once much more vigorously discussed (Lipset and Marks 2000 provide a recent overview). To the German intellectual Werner Sombart’s book-length question, Why Is there No Socialism in the United States?, the libertarian HL Mencken had famously answered that the “ship of socialism foundered long ago on the shoals of roast beef and apple pie”. The material wealth and opportunities that nineteenth century America offered its white working class of largely British and Scots Irish Protestant extraction, and later made available to newer immigrant flows of white Europeans, such as the Catholic Irish, Italians, Germans and Poles, must be part of the explanation of why the US never developed a significant socialist party, but not the whole story. Leaving aside the strategic errors of US socialists and how they related to trade unions, all of which mattered, most agree that the values of an individualist Protestant culture in which white men had achieved the franchise before large-scale industrialisation significantly reduced the appeals of statist or collectivist socialism. Systematic evidence of the long-lasting impact of the political cultural values of different types of British Protestant, and their lingering regional aftermaths, has been painstakingly put together in the brilliant work of the historian David Hackett Fischer (1989). Of these, the Scots-Irish, the Protestants of the Anglo-Scots “borderlands” and Ulster and their progeny, have been the least inclined to socialism. They mostly went South.

Two other features of American institutions and their sociology stand out to the political scientist. One is the way in which US political institutions encourage the formation, and reformation, of big yet usually loose two-party coalitions to win power. Winner takes all voting at the district level, combined with the necessity to coalesce to achieve power across the separation of powers, and across a fifty-state federation, generates powerful incentives that work against new parties. Another is the sustained impact of racism and ethnic and religious cleavages on the prospects for working class or indeed middle class solidarity: treat claims of post-racism and indeed most post- phenomena with scepticism.

After the civil war the Republican party was led by big business and finance. Intermittently genuinely in favour of equality of opportunity (though usually racially blinkered) it was a free market party from its inception, and had little place for democratic radicals (let alone socialists) after it abandoned the Radical Republicans and blacks to the restoration of white supremacy in the South. The Democratic party, by contrast, became a curious cross-regional coalition for eighty years. In the South it ran a one-party regime, in an overtly racist and exclusionary cross-class coalition of poor whites, small townspeople and landlords. In the North and West, however, it became the party of the cities that were increasingly dominated by the new immigrant (now the newly named ethnic minority). It was not entirely closed to blacks. Overwhelmingly Protestant in the South, it was increasingly Catholic elsewhere.

The Republican Party, excluded and self-excluded by its own decisions from the South, eventually faced the prospect of becoming the minority party nationally ‑ increasingly white, Protestant and nativist in the North and West it lost its post-Lincoln potential new base among free and post-emancipation blacks. That only happened, however, when the Republicans were deeply damaged by the Great Depression presided over by Herbert Hoover, and Roosevelt eventually put together the New Deal coalition that had been anticipated in Al Smith’s run for the presidency in 1928, and which made the Democratic party dominant until 1968.

Where are we now? Some say we are at the end of the era of the natural Republican majority, and at the start of a natural era of Democratic domination. As early as 2002 John Judis and Ruy Teixeira proclaimed The Emerging Democratic Majority. This claim was, minimally, premature. Their argument, crudely summarised, was that the sociological changes that had benefited the Republicans under Nixon and Reagan were exhausted. The Republicans had mined the very last available ardently Judaeo-Christian vote; they were tapped out in the South. White America was everywhere in decline, at least in its share of the still rising US population. The diverse new immigrant stock issued in by the post-1965 reforms meant that the historically dominant white Anglo-Saxon Protestant core faced minority status. The new immigrants physically and culturally resembled the world (albeit with Masters and PhDs) much more than they did WASPs or the Scots Irish. Many other trends, they claimed, spelled good news for the party of Franklin Roosevelt: the increased share of registered Latino voters, the solidification and mobilisation of the Democratic black base, and, perhaps above all, the consequences of rising numbers of college-educated citizens less likely to endorse traditional or fundamentalist Christian values, deny global warming, or resolve to just say “No” to drugs or premarital sex of whatever kind. It was even hinted that the Democrats would no longer need to be led by Southerners to have a chance of winning the White House.

These arguments looked foolish, of course, when George W Bush won a second term running against John Kerry, on an electoral strategy of appealing to the Republican base rather than the median voter. Yet they enjoyed a revival, and looked far more plausible, when Barack Obama of Kansas, Kenya, Hawaii, Indonesia and Chicago won big in 2008.

Before 2012 few thought, however, that Obama could repeat the scale of his victory in 2008, when he was elected amid the opening of the great financial crash, and war hero John McCain blustered and sleepwalked his way through an unfolding crisis he barely understood. Obama had by contrast looked wholly competent. Until a month ago it was credible to believe that Obama would win big again. That was until his dreadful first debate performance against Romney allowed the race to tighten. Whatever my professional social science colleagues may say about the past repercussions of debates that first debate loss may yet cost Obama his second term (even though he clearly outpointed Romney on the other two occasions).

Later this week will be the time to explain why Republicans are so strong in the US political system, a fact which surprises so many Europeans. It will also be an appropriate time to explain why Romney has been, contrary to many portraits, such an effective candidate. It will also be the right occasion to evaluate Obama’s leadership. The chances are seven out of ten that his first term will be the subject evaluated; the chances that a one-term presidency will be the subject of the conclusion are three out of ten.


Amar, Akhil Reed. 2005. America’s Constitution: A Biography. New York: Random House.
Anderson, Elijah. 2011. Cosmopolitan Canopy. Race and Civility in Everyday Life. New York: W.W. Norton.
Beeman, Richard R. 2010. The Penguin Guide to the United States Constitution: A Fully Annotated Declaration of Independence, U.S. Constitution and Amendments, and Selections from the Federalist Papers. New York: Penguin.
———. 2009. Plain, Honest Men: The Making of the American Constitution. New York: Random House.
Cooke, Jacob E. 1970. “The Compromise of 1790,” The William and Mary Quarterly, 27, 4, 523-545.
Dahl, Robert A. 2001. How Democratic is the American Constitution? New Haven, CT: Yale University Press.
Edwards, George C III. 2004. Why the Electoral College is Bad for America. New Haven, CT: Yale University Press.
Foner, Eric. 2005. Reconstruction: America’s Unfinished Revolution, 1863-1877, with a new introduction by the author. 2nd ed. New York: Francis Parkman Prize Edition History Book Club.
Fischer, David Hackett. 1989. Albion’s Seed: Four British Folkways in America. Oxford: Oxford University Press.
Judis, John B, and Ruy A Teixeira. 2002. The Emerging Democratic Majority. New York: Scribner.
Levenson, Sanford. 2006. Our Undemocratic Constitution: Where the Constitution Goes Wrong (and How We Can Fix It). New York: Oxford University Press.
Lipset, Seymour Martin, and Gary W. Marks. 2000. It Didn’t Happen Here: Why Socialism Failed in the United States. New York: WW Norton & Co.
Nagel, Jack H. 2007. “The Burr Dilemma in Approval Voting.” Journal of Politics 69 (1):43-58.
O’Leary, Brendan. 2010. “Electoral Systems and the Lund Recommendations.” In Political Participation of Minorities: A Commentary on International Standards and Practice, edited by Marc Weller and Katharine Nobbs, 363-400. Oxford: Oxford University Press.
Parker, Richard. 2005. John Kenneth Galbraith: His Life, His Politics, His Economics. New York: Farrar, Straus, and Giroux.
Phillips, Kevin. 1970. The Emerging Republican Majority. New Rochelle, NY: Arlington House.
Silver, Nate. 2012. The Signal and the Noise: Why so Many Predictions Fail – but Some Don’t. New York: Penguin Press.
Wills, Gary. 2003. “Negro President.” Jefferson and the Slave Power. Boston: Houghton Mifflin.

Brendan O’Leary is Lauder Professor of Political Science at the University of Pennsylvania and recently served as the United Nations Senior Adviser on Power-Sharing for the Standby Team of the Mediation Support Unit of the United Nations. In 2013 his book with Chris McCrudden, Consociation and the Courts: Power-sharing versus Human Rights, will be published by Oxford University Press, and the University of Pennsylvania Press will publish Power-Sharing in Deeply Divided Places, co-edited with Joanne McEvoy, and Divided Nations and European Integration, co-edited with Tristan Mabry, John McGarry and Margaret Moore.