The recent election has understandably drawn attention to the Electoral College’s role in selecting American Presidents. One argument presented in its favor deserves particular attention: that it rewards candidates for having geographically broad support.

This argument, it turns out, is not simply wrong but directly opposite the truth: the Electoral College grants an advantage to candidates with geographically concentrated support. This is clear from both a simple toy model and from the actual history of American elections

Theory

The current practice in the Electoral College is found nowhere in the Constitution; instead, it is the result of early 19th century practice. While the Constitution makes no demands on how electors are chosen and arguably forbids them from being bound by state law to support one candidate or another, in practice almost all electors are awarded to the winner of a plurality of the vote in their state. This creates obvious benefits to concentration.

Suppose a toy electoral college of three states, each containing 10 voters and each casting 1 electoral vote for the winner of the popular vote. In one state, candidate A wins 8 votes; in the other two, Candidate B wins with 6 votes in each. Candidate B wins the electoral college 2 votes to 1, in spite of losing the popular vote 16-14. But which candidate has the more geographically dispersed support?

While this is ultimately a subjective determination, I claim that candidate A’s support is more widely dispersed. First, his floor was higher, winning a minimum of 40% compared to B’s minimum of 20%. Second, fully half of his voters live in states won by candidate B, whereas only 14% of B’s voters live in states won by candidate A. Indeed, in this toy example, a popular vote-electoral vote inversion must rest on exactly this sort of distribution, with a candidate winning small majorities in two states while being crushed in the third.

The opposing argument is that victory in two states rather than one demonstrates dispersion, while winning a state 8-2 demonstrates excessive concentration. Considering the argument’s inverse, however, demonstrates its absurdity: this argument claims that candidate B’s support is shown to be dispersed by virtue of his weakness in the state he lost.

The actual system used in American elections is more complex than this toy in a number of ways, but an examination of actual Electoral College results shows exactly this pattern: the Electoral College boosts regional candidates and harms candidates with broadly dispersed support.

Case 1: The 1860 Election

The 1860 election is the most dramatic and straightforward case of geographic concentration in the Electoral College. Abraham Lincoln, running on a free soil platform, was unable to even get on the ballot in 9 states (all of which would go on to secede) and won single-digit percentages of the popular vote in three more; he won the election on the strength of 15 outright majorities, pluralities in California and Oregon, and a close fought race in New Jersey that split its electoral votes 4-3 in his favor. Lincoln won no electoral votes in a slave state, while taking 180 out of 183 in free states.

Lincoln’s opposition was fragmented between three major candidates: two Democrats, Stephen Douglas and John Breckinridge, and Constitutional Union nominee John Bell. Of these three, Breckinridge finished second in the Electoral College by winning outright majorities in 6 states and pluralities in 4 more. Bell finished third, with pluralities in three states.

Stephen Douglas, however, beat out both of these candidates in the popular vote but paid the price for dispersed support in the Electoral College: he only won electoral votes in Missouri, where he pulled a plurality of just over 35%, beating Bell by fewer than 500 votes, and New Jersey, where he won the 3 electors who did not go to Lincoln. He was the only candidate to win electoral votes from both free and slave states, and he finished with at least 10% of the popular vote in 19 states, only 1 shy of Lincoln’s mark in the same measure and ahead of Breckinridge (17) and Bell (15).

While Lincoln benefited from his opposition fragmenting between three candidates, this was not the primary cause of his electoral strength: he won outright majorities in states worth 164 electoral votes, well beyond the 150 needed to win the election. While some of these were close enough to plausibly be endangered had the Democrats been able to field a single candidate against him, the Republican base in the North would have made him competitive, at the very least, with a single national opponent.

Case 2: Third Party Candidates

In the postwar period, most third party runs have been minor sideshows; on two occasions, however, they have won electoral votes. Both of these candidates exploited the growing regional rift within the Democratic Party to win a few states, while mostly or entirely neglecting the rest of the country.

In 1948, Strom Thurmond, then Governor of South Carolina, walked out of the Democratic Convention due to a platform fight over civil rights. With the support of a number of Southern Democrats, he ran for President as the candidate of a splinter faction in the party. Thurmond was the archetype of a regional candidate: he was on the ballot in only 13 states, all former slave states, and he won only 2.4% of the total popular vote. However, as the official Democratic candidate in four states, Thurmond was able to win 39 electoral votes.

Twenty years later, another Southern candidate, George Wallace, followed Thurmond’s path to winning votes in the Electoral College. Wallace’s support was much more widespread than Thurmond’s: he was on the ballot in all 50 states, and he won more than 13% of the popular vote. However, in the Electoral College, he demonstrated somewhat less success than Thurmond, winning only 5 of the 6 states carried by Thurmond in 1948 (South Carolina went for Nixon) and 46 electoral votes. Wallace’s support, though broader than Thurmond’s, was still very heavily concentrated in the Deep South: his victory in Alabama with nearly 66% of the vote was a higher share of the popular vote than either major party candidate won in a single state (though Washington DC went for Humphrey with over 80% of the vote), but he failed to reach double digits in 27 states and DC.

Thurmond and Wallace’s success in the Electoral College stand in stark contrast to the most successful modern third party run in the popular vote, Ross Perot’s 1992 campaign. Perot led national polls at one point and finished with nearly 19% of the popular vote in spite of taking a hiatus from campaigning for a significant part of the summer; however, lacking the strong regional base of the Southern candidates, Perot did not manage a single electoral vote. In stark contrast to Thurmond and Wallace, Perot managed at least 10% of the vote in 49 states; however, his best state, Maine, only gave him a touch over 30%, leaving him 8 points behind Bill Clinton.
Case 3: The Solid South
The Republican Party was born a regional party of the North and West, and while it expanded into the South during Reconstruction, winning a number of southern states in the 1868, 1872, and 1876 elections, that same experience ensured that the white elites that reasserted control after the end of Reconstruction would overwhelmingly support the Democratic Party.

Yet the GOP’s lack of competitiveness in a significant region did not prevent the party from winning Presidential elections; beginning with the election of 1876, the first election after a significant number of states had begun rolling back black enfranchisement, and ending with the election of 1944, before Strom Thurmond’s Dixiecrat revolt marked the beginning of the end for the Democratic South, Republicans won 10 Presidential elections to the Democrats’ 8. Though the Republicans were able to make headway in the South by the 1920s, with Herbert Hoover winning 5 former Confederate states in his 1928 landslide, they never penetrated a core from Arkansas and Louisiana in the west to South Carolina in the east. However, with few votes wasted in safe Democratic states, they consistently ran ahead of their popular vote totals in the Electoral College, including winning the 1876 and 1888 elections in spite of losing the popular vote.

Conclusion

The Electoral College may have virtues; however, by granting a state’s full electoral heft to the winner, even if by a bare plurality, it favors regional candidates over national. This is ordinarily not a major issue: no election has been close enough to be thrown to the House by third party candidates, and even during the era of the Solid South Republicans contested enough of the country to win elections and claim broad support. Nevertheless, anyone concerned about rewarding candidates for geographically concentrated support should oppose rather than support the College.

Feature Image by mccun934