The Connecticut effect: the great compromise of 1787 and the history of small state impact on Electoral College outcomes.
But the Three-Fifths Compromise was not the only deal to emerge from the Philadelphia Convention. Even more significant for subsequent presidential elections, since it continues in operation to the present day, was the Connecticut Compromise (or Great Compromise), which resolved a standoff between large and small states over representation in Congress. This compromise, settled at the convention on 16 July 1787, gave all states equal representation in the Senate, thus reassuring the small states that they would not be outvoted in both houses by the large states. The Compromise paired equality of representation in the upper house with proportional representation in the lower. A few weeks later, when the Convention delegates turned to the creation of the presidency, the familiar Connecticut Compromise principle was invoked a second time. The principle of protecting the small states through equality of representation in the Senate was thus extended to the practice of choosing the president in the Electoral College. As one scholar astutely observed, the creation of the Electoral College was "simply a second round of the Connecticut Compromise in settling large state-small state differences." (3) The mechanics of choosing presidents thus benefited the smaller states, guaranteeing them a greater proportion of electoral power than they should have enjoyed on the basis of their population numbers alone.
The fact that representation in the Electoral College incorporated the principle of state equality in the Senate meant that even the smallest states could have no fewer than three electoral votes. (4) In the House of Representatives, Virginia (the most populous state at the time) had ten representatives while Delaware (the least populous) had one representative, but the extension of the Connecticut Compromise principle meant that in choosing a president, Virginia received twelve electoral votes and Delaware three: Delaware and other small states gained a much greater voice in selecting presidents relative to larger states. Sparsely populated states with only one representative received two additional electoral votes, tripling their relative power when it came to choosing presidents. States that had two representatives cast four electoral votes, doubling their significance in the Electoral College. Ever since 1787, this sustained practice of awarding bonus votes (the Connecticut Effect) has artificially enlarged the share of electoral votes for small states well beyond what their share of the United States population would have entitled them to. But what has been the impact on election results of this compromise between large and small states at the Convention? Have the bonus votes made a difference in electoral outcomes? If so, how and when? Has an advantage consistently accrued to one or another political party, or to the winning or losing candidates? What patterns have emerged and how have they changed over time? In short, what has been the Connecticut Effect?
Every US presidential election since 1828 has witnessed a Connecticut Effect. The Effect has benefitted both Democrats and Whigs in the second party system as well as Democrats and Republicans in the modern party system. It has aided both winning candidates and losing candidates with no readily discernible pattern save one: In every single election from 1828 through 2008 the candidate who had the lowest rate of electoral votes per state that he won (calculated simply by dividing a candidate's total electoral votes by the number of states he carried) has been the beneficiary of the Connecticut Effect. This has been true regardless of the candidate's party, whether he won or lost the election, or even whether or not the beneficiary carried more small states than his opponent. The clear consistent link in every election is between the average electoral vote per state and the candidate with the lowest rate benefitting from the Effect, even if it made no difference on the outcome. In almost all cases, removing the bonus votes from the victorious candidates' totals does not change the outcomes, but instead merely alters the margins of victory.
But for every rule there are exceptions. Three times in US political history, in 1876, 1916, and 2000, the winning candidates owed their narrow victories at least in part to the Connecticut Effect. Peculiar combinations of circumstances in each of the three elections tipped the outcome in favor of candidates who, without the bonus votes of the Connecticut Effect, otherwise would have lost the electoral vote and thus the election.
Acknowledgement of the Connecticut Effect complicates the conventional narratives of those three elections by providing a new dimension to add to the familiar accounts. But recognition of the Connecticut Effect also makes it clear that each election was decided as it was because the winning candidates prevailed in enough small, lesser-populated states to carry slim Electoral College majorities thanks to the bonus votes. The Effect has been too little noticed in the scholarship on presidential elections, but its existence and impact demands that we revise and expand our conventional narratives of these contested elections. Thus, while usually ignored albeit omnipresent, the Connecticut Effect deserves to become part of the way we analyze Presidential election returns because it always plays a part, as the following analysis of past elections shows.
While I am not the first scholar to note the effect of the bonus votes on the 1876, 1916, and 2000 elections, below I seek to go well beyond just those three cases. (5) This article offers the first analysis of the impact of the Connecticut Effect in all US presidential elections to discover its role across time. It starts by making simple, rudimentary arithmetical calculations for each election in US history by taking the actual electoral votes earned, subtracting the bonus votes, and assessing the differences between the actual percentage of electoral votes received and the adjusted percentage of electoral votes after subtracting the bonus votes. The Connecticut Effect applies whenever there is a difference between a candidate's actual percentage of electoral votes received and his adjusted percentage. A candidate is judged to be the beneficiary of the Connecticut Effect if his actual percentage of electoral votes is greater than his adjusted percentage of electoral votes, demonstrating a benefit from the bonus votes. Conversely, a candidate has been disadvantaged by the Effect if his actual percentage of electoral votes is less than his adjusted percentage of electoral votes.
The article then will trace this Effect through historical eras demarcated by party systems to determine patterns and trends. It will also analyze in detail the three contested elections in which the Effect was decisive but will consider these elections in the context of those that preceded and followed them, so as to place them in fuller perspective. Lastly, it will establish the precise impact of the Connecticut Effect and highlight that impact as just one of several factors that have shaped Electoral College outcomes over the years. Just as the Three-Fifths Compromise, the district system of voting, and the winner-take-all formats have played roles in determining electoral results (including the 2008 election when Barack Obama picked up a single electoral vote from Nebraska because he won the second congressional district in a state that awards electoral votes by district), the Connecticut Effect, too, has played its part. (6) In short, this article seeks to identify and place the Connecticut Effect in fuller historical context and document, for the first time, its effect on each election since 1828.
Although the bonus vote condition giving the states additional electoral votes has been in place from the outset of the Presidential election system, its effects were mitigated in early elections because the distribution of electoral votes followed no uniform pattern. Beginning in the 1790s, states constructed various systems for allocating votes and there was no single predominant method. As early as 1796, some states chose electors by popular election on a district-by-district basis. Others chose from statewide slates based in a popular election. And some states allowed the state legislature to appoint electors. Since the Electoral College was designed in the absence of political parties, the rapid development of a party system in the 1790s further complicated the various methods of choosing a president. Sometimes methods of selecting electors were changed depending on the changing fortunes of political parties within individual states. (7) Massachusetts, for example, switched away from popular voting for electors to legislative selection in 1800 because the Jeffersonian Republicans were gaining strength in the state but Federalists controlled the legislature. New Hampshire did the same thing. Virginia, the home of Republican strength, changed its rules for party advantage as well, moving away from picking electors on a district-by-district basis to a statewide at-large selection process to prevent Federalists from picking up stray electoral votes. States, in short, used a hodgepodge of methods to select electors, and electoral votes themselves were awarded in a scattered way instead of the winner-take-all method. In the controversial election of 1800, for example, three states (Pennsylvania, Maryland, and North Carolina) divided their electoral votes between Thomas Jefferson and John Adams; Maryland split its vote 5-5, essentially negating its contribution. This practice continued in subsequent elections with states frequently shifting from one method of elector selection and allocation to another.
But political campaigners realized the advantages of moving to a winner-take-all apportionment of electoral votes. It allowed a party candidate to take all of that state's electoral votes after even a narrow win. (8) After 1800, these calculated political considerations decisively led states away from a district selection of electors toward a statewide, winner-take-all allocation based on popular election. By 1828 all states had adopted this practice except South Carolina (which still had its legislature choose electors). (9)
The lack of a uniform standard for allocating electoral votes rendered the Connecticut Effect irrelevant in early elections, even if they were beset by major problems regarding the Electoral College, highlighted by both the 1800 and 1824 elections being decided in the House of Representatives. By 1828 a number of developments had coalesced to give more uniformity to the electoral process. First, the Twelfth Amendment in 1804 clarified the practice of voting separately for president and vice-president; second, the two-party system was by 1828 well-established as the basis for electoral competition and political campaigning; and finally, as noted above, only South Carolina still deviated from the now common practice of awarding electoral votes based on statewide popular ballots and on a winner-take-all principle. Once the electoral practices thus became more uniform, a fuller breakdown of election results categorized by eras of party systems and periods of shifting party dominance further illuminates the operation of the Connecticut Effect.
From Andrew Jackson's election in 1828 to James Buchanan's victory in 1856, Democrats dominated presidential politics, claiming victory in six of the eight elections. (10) During these elections, the presence of the Effect was persistent but slight, never affecting an outcome and not changing the electoral vote percentages by more than a point or two. The most striking feature of the Connecticut Effect during these years is that it tended to benefit the losing side. In 1828, 1832, and 1836, the Effect favored the losing Whig candidates each time. It did not benefit the Democrats until 1840 when they lost. Again in 1844 and 1848 (despite the fact that each party won one and lost one) the Effect again benefitted the losing side. Not until Franklin Pierce's victory in 1852 did the Effect work to benefit the victorious candidate, a result replicated in 1856.
Beginning with Abraham Lincoln's election in 1860, Republicans dominated presidential politics for the next several generations, winning seven of the nine elections during this era, with only Grover Cleveland's victories in 1884 and 1892 breaking the juggernaut. (11) Meanwhile, the Connecticut Effect broke nearly evenly between the two parties, favoring Democrats in 1860, 1864, 1880, and 1888, and the Republicans during the other five elections. One pattern that persists from the Jacksonian era, however, is that the Effect worked overwhelmingly for the losing candidates, inflating their electoral vote totals in six of the nine elections. Only in three consecutive elections (1868, 1872, 1876) did the Effect benefit Republicans, who won all three elections. Another trend that gained momentum during this era is that the Connecticut Effect's beneficiary carried the most small states (those with three or four electoral votes) in five of the nine elections. This was also the case in the last four elections of the Jacksonian era (1844 through 1856), although that fact did not consistently aid one party over the other, favoring Republicans in 1868, 1872, and 1876 but favoring Democrats in both 1884 and 1892.
It was not until 1876 that the Connecticut Effect made a difference in the outcome of an election. The 1876 election remains infamous for the fraud and vote stealing on both sides and the controversial awarding of twenty disputed electoral votes from Florida, Louisiana, South Carolina, and Oregon to Republican Rutherford B. Hayes of Ohio. They enabled him to prevail over Democrat Samuel J. Tilden of New York by a single vote (185-184), even though Tilden won the popular vote. (12) While the early election returns presaged a huge Tilden victory, Republican operatives sensed an opportunity in the returns of the Southern states that could be contested. Sizing things up astutely, Republican leaders put out the word to Southern partisans to hold their states for Hayes, calculating that if they got all the disputed votes that their man would be president. The controversy dragged on from election day in November almost until inauguration day in March, and involved arguments about the finer points of election and constitutional law as well as rabid political passion. Ultimately, Congress created a special commission consisting of seven Democrats, seven Republicans, and one Independent to decide which sets of returns to count. Voting along party lines, and with the Independent voting Republican, the commission decided in a series of 8-7 votes to award the disputed electoral votes to Hayes. Hayes owed his victory not only to the special commission, however, but also to a Connecticut Effect: It was worth a net eight electoral votes. (13)
Hayes won more states (21-17) but his average electoral vote per state was only 8.8 whereas Tilden averaged 10.8 votes per state. By winning 21 states, Hayes received 42 extra electoral votes while Tilden received 34 extra electoral votes for his 17 states. Subtract the extra votes from both candidates and Tilden would win the adjusted vote by 150 to 143 electors, a cushion of seven electoral votes and an electoral vote percentage (51%) that would exactly match his popular vote percentage nationwide. This result holds even with all the disputed electoral votes still assigned to Hayes. (14)
A closer analysis of the returns suggests that Hayes won the election not so much in the disputed states as in the smaller ones. Of the five states in 1876 that cast the minimum three electoral votes, Hayes carried four (Colorado, Nebraska, Nevada, and Oregon) while Tilden carried only one (Delaware). In effect, Hayes won the election because of the eight bonus electoral votes he got in the smallest states which easily offset the two extra votes Tilden won there. Florida and Rhode Island were the next smallest states, each casting four electoral votes (two bonus votes each) and Hayes carried them both. Five states cast five electoral votes in 1876 and only one of them, West Virginia, voted for Tilden. The others (Kansas, Minnesota, New Hampshire, and Vermont) all went for Hayes. Added together this means that the twelve smallest states in 1876 gave Hayes a 40 to 8 electoral vote margin (a net margin of plus 32) since he carried ten of them. Hayes thus received nearly 22% (40 of 185) of his total electoral vote from the smallest states. He lost the other twenty-six larger states by 145 to 176.
But exactly half of those 40 electoral votes Hayes won in the smallest states were bonus votes. These extra votes are what Electoral College scholars Lawrence D. Longley and Neal R. Peirce call the "constant two" senatorial votes. (15) When these extra votes are eliminated, Hayes's margin over Tilden in the twelve smallest states shrinks from 40 to 8 to only 20 to 4, a massive reduction in his net margin from +32 votes to only +16 votes. Hayes owed his winning margin of one electoral vote, the slimmest in U.S. history, as much if not more to the bonus electoral votes awarded to the smallest states as he did to the commission that ultimately awarded him the disputed electoral votes.
There is even more to the story of the small states in the 1876 election. Charles Stewart and Barry Weingast have argued that the electoral math of the 1876 election was the direct result of shrewd and aggressive strong-arm tactics by Republicans, designed to maximize their political power. (16) Stewart and Weingast contend that Republicans in Congress manipulated the admission of new western states into the union, often ignoring precedents and longstanding conventional practices so as to increase Republican power in the Senate and, hence, the Electoral College. This would further Republican chances of controlling the presidency. Central to this strategy, they suggest, was the timing of the admission of states together with the criteria used for determining admission. By manipulating the politics of statehood admission from the Civil War through 1896, Republicans were usually able to maintain control of two of the three main levers of government (presidency, Senate, House) and thus block the Democrats from undoing their legislative program. (17)
This created the American equivalent of English "rotten boroughs," with states being admitted to the union on questionable grounds to assure Republicans of having additional votes in the Senate and provide Republican Presidential candidates with the additional votes in the Electoral College. Specifically, this practice had a decisive impact on the 1876 vote. The three most recently admitted states (Nevada in 1864, Nebraska in 1867, and Colorado in 1876) were all very small in population but also very strongly Republican. These three newest states gave Hayes nine electoral votes. The most brazen case of all was the admission of Colorado, admitted under an enabling act passed on 3 March 1875, exactly one day before the Democrats gained control of the House of Representatives for the first time in fourteen years. (18) Furthermore, the act allowed Colorado time to complete the necessary ratification referendum approving the state constitution. Once that was done, all that remained was for President Ulysses S. Grant (a Republican) to declare Colorado admitted, just in time to participate in the 1876 presidential election. But leaving nothing to chance, the 1875 enabling act granted the newly chosen Colorado legislature, and not the voters in the state, the right to select the state's three electors.
Without Colorado's three decisive votes, Hayes would have fallen two votes short of Tilden, even taking all the disputed electors into account. Thus, the Republicans' push on the last day of the Forty-Third Congress to get Colorado admitted in time to cast its three electoral votes in 1876 might have been as important as anything the party did on or after Election Day. (19) One might say that the 1876 presidential election was won on 3 March 1875. Of course, such manipulation, most of which was duly legal, characterized the fierce and deeply-ingrained partisanship of politics in the fabled Gilded Age. (20)
The second consecutive era of Republican dominance was ushered in by William McKinley's victory over William Jennings Bryan in 1896, starting a trend of seven Republican victories in nine elections. (21) Nonetheless, the Democrats benefitted most from the Connecticut Effect during this segment: six times in nine elections, including all three of Bryan's losses and both of Woodrow Wilson's victories. In contrast to past eras, the Effect favored winners slightly (five times in nine elections) and there was a tighter link between the Connecticut Effect beneficiaries and those candidates carrying the most small states. Six times the Effect favored the candidate with the greatest number of small states; five of those candidates were also victorious. Overall, the Effect was fairly small during this period with one great exception: Wilson's 1916 re-election victory over Charles Evans Hughes.
Most accounts of the 1916 election feature the razor-thin margin of victory Democratic incumbent Woodrow Wilson won in California, whose thirteen electoral votes spelled the difference between victory and defeat. (22) This result is usually attributed to an alleged snub by Hughes of California governor Hiram Johnson. While Hughes was campaigning in California, he spent part of a day in the same Long Beach hotel where Johnson also happened to be. Primarily because staffers kept each man from knowledge of the other's presence until after the fact, it seems, Hughes and Johnson never met. Johnson, however, believing he had been deliberately snubbed by Hughes, privately stewed and publicly provided only perfunctory campaign help. (23) Had Johnson not sat on his hands that fall, Hughes might easily have made up the miniscule three-tenths of one percent of the vote by which he lost the state.
A less explored yet decisive feature of the 1916 result was that Woodrow Wilson carried 30 states to just 18 states for Hughes, a result that gave Wilson 24 extra electoral votes--exactly one more than his margin of victory over Hughes. Subtract the bonus electoral votes and Wilson's 277 to 254 victory would become a 218 to 217 Hughes victory, even with California going to Wilson. Wilson's victory depended as much on his carrying twelve more states than Hughes as on winning California. As Hayes did in 1876, Wilson piled up votes while carrying many of the smallest and medium-sized states. The more populous states generally went for Hughes, taking him to the brink of electoral victory. The 30 states Wilson carried averaged only 9.2 electoral votes; the 18 states that Hughes won, however, averaged 14.1 electoral votes. Hughes won six of the nine largest states.
While most treatments of the election stress the geographical feature of the returns, noting Wilson's coalition of southern and western states, Wilson's election might just as accurately be described as resulting from a coalition of the small and medium-sized states. Of the five states with the minimum three electoral votes, Wilson won four. Of the five states with just four electoral votes, Wilson again won four. Wilson defeated Hughes by 28 electoral votes to 7 in these 10 states, a margin of 21 votes in an election he won by just 23 votes. More than half of Wilson's votes were simply due to the bonus rule, which accounted for 16 of those 28 votes. If the bonus votes were taken away, Wilson would have netted just 12 electoral votes from winning those eight states and Hughes would have earned just three votes for his two states. Instead of a 28 to 7 advantage (a margin of +21) in the smallest states, Wilson would have held only a 12 to 3 edge (a +9 margin). When the next seven smallest states (those casting five and six electoral votes) are added to the ten smallest, the small-state trend favoring Wilson continues. Wilson won 11 of the 17 smallest states (those with between three and six electoral votes); Hughes won only six. Those 11 states gave Wilson 45 electoral votes while Hughes collected 28 from his six states, a difference of 17 votes, thus comprising the vast majority of Wilson's 23-vote margin of victory. Again, the bonus votes accounted for almost half of Wilson's total. Of course, of the three Connecticut-Effect elections, the 1916 election is the only one where the victorious candidate also won the popular vote. Wilson's winning margin was three percentage points, or over half a million votes out of over 18 million votes cast. (24) In 1916, unlike in 1876 and 2000, the elimination of the bonus votes would have overturned the electoral outcome but would have installed the candidate who lost the popular vote in the White House, an event that has occurred only four times in American political history: 1824 (John Quincy Adams instead of Andrew Jackson), 1876, 1888 (Benjamin Harrison instead of Grover Cleveland), and 2000. The abolition of the Connecticut Effect would have reversed the results of 1876 and 2000, but would not have altered the outcome in 1888 and, as just noted, would have created a new instance of the popular vote loser winning in 1916. (25)
As for the alleged snub, although historians have emphasized this in their treatments, Hughes, for his part, denied that any snub had taken place and that the failure to meet up with Johnson, what he termed "the misadventure in California, was not due to any fault of mine. I did not know that Governor Johnson was at Long Beach. I had been very desirous of meeting him, and had I known that he was at Long Beach when I was there, I should have seized that opportunity to greet him." (26) After Johnson's death in 1945, Hughes declared again that "I did not snub Governor Johnson," and that the latter "certainly knew that I did not snub him." (27)
Democrats dominated the next era (1932-64), winning seven of nine elections. (28) But the Connecticut Effect worked in favor of Republicans in six of those nine elections. Even more noteworthy, the Effect benefitted the losing candidate in every single election except Harry Truman's victory in 1948. A noteworthy point: the 1960 election was the only really close one in this era and the Effect helped make it even closer than it was, inflating Republican loser Richard Nixon's share of the electoral vote by almost three percentage points. Although Kennedy still won the electoral vote handily, his margin would have been even greater without the Effect.
In the most recent era of elections (1968-2008), the Connecticut Effect has worked in favor of the losing candidate in eight of eleven elections and for the Republican candidate in seven of those eleven contests. (29) However, Republicans were aided by the effect in three very close elections they won (1968, 2000, 2004), while Democrats benefitted only when they lost by decisive margins (1972, 1980, 1984, 1988). The Effect has worked to the Republicans' benefit in the last five presidential elections. The only other stretch where the Effect benefitted one party in five consecutive elections was from 1848-64, when Democrats were the beneficiaries. And this most recent era also includes the third and final election that would have been the reverse if not for the Connecticut Effect: George W. Bush's election in 2000.
The conventional wisdom about this very recent election is that ballot irregularities and disputed recounts in Florida were decisive, since that state's 25 electoral votes gave George W. Bush the 271 electoral votes he needed to win. As with the 1876 and 1916 elections, however, this is only a part of the story. Indispensable to Bush's victory was the fact that he carried 30 states to Al Gore's 21 states (including the District of Columbia). Many of these were small states whose electoral votes were disproportionally higher than their share of population. While Bush won more states, they carried a lower per state average electoral vote: 9.0 votes per state compared to Gore's 12.7 votes per state. (30) The power of these small states' electors was crucial to the 2000 outcome, over and above anything that happened in Florida. Bush won 60 bonus votes to 42 for Gore, a net margin of 18 votes in an election he won by just four votes. Without the 18 extra bonus votes Bush received from the Connecticut Effect, he would have lost the election to Gore by 225 to 211. (31)
Of the eight states (including the District of Columbia) that cast the minimum three electoral votes, Bush carried five of them, Gore three. Bush thus won fifteen electoral votes among these states, Gore nine, a margin of six votes. But ten of the fifteen votes Bush won here were bonus votes. Subtracting the bonus Connecticut Effect votes from the equation, Bush's margin in these smallest states would be only five to three, a net margin of just two votes. Since Bush won the election by just four electoral votes, his entire winning margin came from these smallest states. But Bush won by big margins not only in the very smallest states but among all those smaller states casting between three and six electoral votes. Altogether, in the twenty states with between three and six electoral votes, Bush carried thirteen states, Gore just seven. Bush's actual electoral vote margin in these smallest states was 54 to 26 over Gore. Almost exactly half of that total was inflated with the bonus votes. Subtracting Bush's 26 to 14 edge over Gore in bonus votes from his totals leaves Bush with a 28 to 12 margin in adjusted electoral votes among the smallest states. Bush's margin of sixteen votes is still clear-cut; however, the adjusted vote total represents a significant shrinkage of his plus 28 vote margin in the actual, unadjusted totals. Clearly, Bush (who again, won the election by just four electoral votes) piled up most of his winning margin in these smaller states where the Connecticut Effect has always most heavily been felt.
Pushing this analysis of the smaller state vote even further reveals that Bush piled up a huge electoral vote margin in the 29 states that cast between three and eight electoral votes. Bush carried 19 of these states for 101 electoral votes; Gore won in 10 states (including the District of Columbia) for 48 votes. More than a third of Bush's electoral votes (38 of 101 in these smaller states) were bonus votes. Gore's 48 electoral votes included 20 bonus votes, thus giving Bush a margin of 18 electoral votes in the 29 states that cast the fewest electoral votes. Taking away the bonus votes from these states would reduce Bush's margin from 101 to 48 to just 63 to 28, a reduction in his electoral vote margin of 18 votes, down to a net of 35 from a net of 53.
While the most obvious similarity between the 1876 election and the 2000 election was the drawn out dispute over which votes to count and how to count them, the other similarity they have is that the ultimate winner relied on the small states to provide his winning margin.
Retrospectively, we can observe the Connecticut Effect in practice. But did anyone at the time of these elections perceive that electoral bonus votes in the small states cost them or their candidates the elections? It seems not. Samuel Tilden immediately settled his focus on the disputed electoral votes and, specifically, on the charges of fraud. During the election controversy he was quiet and cautious in public, staying informed about everything but keeping a very low profile. While maintaining a quiet (and to some supporters, maddening) passivity in the public eye, behind the scenes he prepared elaborate memoranda for a legal case about election, trusting in the process and also believing that specifics of election and constitutional law would be followed. Tilden ignored the political nature of the dispute, apparently unwilling or unable to realize that the counting of ballots was simply treated by the Republicans as an extension of the campaign, as the necessary final step toward installing Hayes in the presidency. (32) Because so much was being done on his behalf, Hayes could afford to be quiet and passive; Tilden's listlessness, however, left the Democrats leaderless and voiceless in the fray.
Once the commission assigned the disputed votes to Hayes, Democrats turned to fingerpointing. First and foremost they accused the Republicans of stealing the election, of course. Tilden himself frequently noted that "the men who were elected by the people ... were counted out; and the men who were not elected were counted in and seated." (33) Tilden additionally believed that he had been done in by disloyal Democrats in Congress and in his own campaign organization, which had allowed the creation of the Electoral Commission. (34) Meanwhile, some Democrats blamed Tilden himself for his passivity and timidity in failing to claim in the days just after the election the victory that even some Republicans conceded he had won. Even Tilden's biographer concurred with these judgments, noting that the candidate's characteristic penchant for indecision "probably cost him the Presidency." (35)
With all the fingers being pointed, variously, at Republican fraud, Democratic perfidy, or Tilden's own fecklessness, no one at the time seems to have noticed the role of the small-state bonus votes in Hayes's triumph. Even when Colorado's crucial role in the Republican efforts was affirmed in the October 3 election that gave Republicans a 2 to 1 margin in the legislature (and with it, the critical authority to insure that the state's three presidential electors would cast ballots for Hayes), Democrats saw the results as "disappointing hut not disheartening." (36) They were still banking on a solid Southern strategy paired with victories in New York and a combination of other key states such as Indiana, Ohio, New Jersey, Connecticut, or Wisconsin. Despite spreading the blame far and wide, Democrats failed to observe the way that the Connecticut Effect was yet another factor that did them in.
The 1916 loser, Charles Evans Hughes, later pinpointed the regions of the country where he lost the election but did not specifically cite the effects of the bonus votes in his defeat. (37) In his autobiographical notes, Hughes again dismissed the idea that the alleged snub of Hiram Johnson did him in. "I still should have been elected," he wrote, "had it not been for the effectiveness, particularly in the middle west, of the Democratic slogan, 'He kept us out of war."' (38) Hughes also observed that the Republican National Committee "was strangely lacking in appreciation of the effect of this appeal." (39) Had he been apprised of this situation, Hughes noted, he could have again visited that part of the country before the campaign ended. In fact, one ally in Michigan "did sense the situation ... and asked me to come there once more, which I did, with the result that I carried [Michigan]." (40)
Other Republicans blamed Hughes's defeat on the still simmering tensions of the 1912 split between Republican regulars and Progressives. Since Progressive sentiment was strongest in the western states, Republicans needed to do considerable work there to conciliate the Progressives and bring them back into the fold. But, as Hughes himself suggested, this never happened, prompting one of his backers to note that the Republican defeat "was wholly unnecessary [for, if] courageous action been taken in every State and had both Mr. Hughes and the Eastern Committeemen understood the Progressive spirit of the West, we would have won an overwhelming victory." (41) Neither this supporter nor Hughes himself noted the vital electoral bonus-vote power of those Mid-West and Western states and the role they played in sealing Hughes's loss in 1916. (42)
Similarly, Al Gore and his campaign staff largely focused on the disputed vote counts in Florida in the 36-day election standoff of 2000, not realizing the considerable impact of the bonus votes. In a style highly reminiscent of Tilden in 1876, Gore adopted a post-election posture of being above the fray, trusting that the Florida recount and all subsequent court verdicts were part of a rational, legal process instead of, as the Republicans quickly grasped, an extension of the political campaign. (43) Like Tilden, Gore showed indecision, lassitude, and a lack of leadership despite his clear popular-vote victory. After the election was finally settled, Democratic post-mortems focused almost entirely on the Florida imbroglio and on various technical aspects of the campaign, such as how Gore's staff was unable to reposition him to overcome the burden of "Clinton fatigue" stemming from the scandals plaguing Clinton's Presidency (whose Vice-President Gore was), or how the campaign might have run a better set of advertisements. As one senior advisor said about the Democratic campaign:
The bottom line on our shortcomings? We didn't convince enough voters that Al Gore was his own man not only to carry the popular vote and carry Florida relatively narrowly but to provide a cushion against the Florida factor and the Supreme Court ploy. (44)
Evidently, neither candidate nor campaign observed the huge boost Bush got from the bonus votes of the small states. (45)
The simplest explanation is sometimes the best, however. In all three elections it may be that no campaign challenged the results on the grounds of the Connecticut Effect (if they even noticed it) because they all accepted the ground rules stipulated by the Constitution going in and thus, thought they could not challenge the results except where they believed laws had been violated.
While all of the small states gain a huge benefit from Electoral College bonus votes, it is a particular concentration of small states whose bloc-like voting behavior has contributed greatly to the results of the 1876, 1916, and 2000 elections: the Mountain West and Plains states. These states have long been able to exert great power in the Senate, to the consternation of some observers. (46) This analysis has been more concerned with their ability, by voting nearly in unanimity as a bloc of states (however random or unintentional that may have been), to deliver the 1876, 1916, and 2000 elections to the Electoral College victors. Some of these states were admitted to the Union in a haphazard manner, driven by raw partisan political calculations and manipulations. (47) The lingering influence of the overrepresentation they enjoy in voting has continued to skew election results, since they also have tended to vote over time as a bloc of states, serving to maximize their collective overrepresentation in the Electoral College.
The political allegiances of these states have shifted over time. (48) For example, the Plains and Mountain states were reliably Republican for the 1870s and 1880s. In the 1890s many were carried first by the Populist candidate James B. Weaver, and then by Democrat William Jennings Bryan. After a shift back to the Republicans, most of them went twice for Wilson, returned to the Republican fold in the 1920s, and then switched again to the Democrats in the 1930s. In the 1950s these states began a swing back to the Republicans, for whom they have voted almost without exception to the present.
That regional uniformity is illuminated by the three Connecticut-Effect elections and provides the particular illustration of this general trend. In 1876, Nebraska and Kansas from the Plains joined Colorado and Nevada from the Mountain West to cast a total of 14 electoral votes for Hayes. Tilden carried none. In 1916, of the states in these regions, only South Dakota voted for Hughes, while the other ten states in the two regions went for Wilson. And in 2000 Bush swept the Plains and Mountain states from the Dakotas down through Oklahoma and west to the coastal states, while Gore, like Tilden before him, was shut out completely. While these small Mountain West and Plains states have not been permanently entrenched with either party and while individual states have occasionally broken away (like South Dakota in 1916), the states in these regions have frequently voted as one bloc. The result has been a huge cache of electoral votes for whichever party that has swept the region, and three times in history this collective Connecticut Effect has provided the winning margin to candidates because of the bonus votes. That this has been an electoral accident rather than a targeted strategy of campaigns is suggested by the way these states are routinely ignored by both major political parties during elections. (49)
Years of evolving electoral customs and practices have produced the effect detailed in these pages. We can discern clearly, in the application of the Connecticut Compromise principle to the creation of the Electoral College, the effects that it has had on American presidential election history. In fact, we should add mention of this to our standard narratives of these controversial elections as it provides a new wrinkle. It is also clear that the Connecticut Effect does not play consistent partisan favorites. The Effect is unpredictable and does not inherently favor one party or another. It does, however, favor the small states and the candidates who carry them, and sometimes that advantage has proven decisive.
It should be emphasized that the Connecticut Effect is largely silent and invisible, as is the influence of the small states. After all, no presidential candidate runs on small-state issues. Instead, most presidential candidates ignore the small states, not bothering to campaign or even run television advertisements in them. And small states, and voters living in them, do not intentionally vote as a bloc. Obviously, there is nothing unconstitutional about the Electoral College and its effects, but that is precisely the point. The Connecticut Effect has been a largely unseen and accidental process, operating under the radar so as to be barely detectable, even to those candidates who gained and lost from its workings.
Ultimately, the Connecticut Effect bonus to the small states is only one of many ways in which the awarding of electoral votes is and has been skewed through history. The systematic disfranchisement of black voters in the Southern states after Reconstruction and up through the Voting Rights legislation of the 1960s made the Southern states into rotten boroughs as well, reflecting only the choices of small numbers of their citizens and leaving large numbers of blacks and some poor whites unrepresented. The winner-take-all format means that, as in the case of Al Gore in Florida in 2000, a candidate who lost by only 537 votes out of several million cast got no electoral votes; his opponent, who won by only 537 votes, got all 25 electoral votes. On the other hand, George W. Bush won 42% of the vote in California but received none of its 54 electoral votes. A proportional system (which has been tried by some states in the past and has been considered by others) would correct this particular skew, but only if adopted nationwide. (50) So, too, would awarding electoral votes according to congressional districts carried, as Maine and Nebraska have done. These and various other alternative models would have their own effect on historical election results, which might be traced through time. (51) Indeed, this article serves as a call to stimulate further research into the variety of ways in which the Electoral-College results have been skewed through the years due to a host of different factors, and might have been skewed in other ways had attempts at some other reforms been successful. (52)
All of the various methods for awarding electoral votes deserve the scrutiny of scholars as well as citizens. Recent debates (which flared up again after the 2000 election) over whether or not to abolish the Electoral College might be influenced by the scholarly study of the ways that proportional voting, district by district voting, and the winner-take-all system along with the Connecticut Effect have all shaped electoral outcomes down to the present day, just as the Three-Fifths Compromise once influenced earlier U.S. elections. (53) The reality may be that there is little chance of Electoral College reform since three-fourths of the states would need to approve a Constitutional amendment to do so, and there is no compelling reason for the small states to change a system that benefits them. (54) Nevertheless, as legislators, scholars, and citizens alike reflect on their methods of choosing presidents and reevaluate the venerable institution that is the Electoral College, attention deserves to be paid to the historical impact of the Connecticut Effect over time. Further examination of this effect and others that have or might have shaped election outcomes in the past will create a context for better understanding the impact of the Connecticut Effect by showing how its effects compare to those of other scenarios for allocating electoral votes. A detailed examination of these other means is beyond the scope of this article, but if other studies join this one, we will have a fuller context for understanding the ways the variety of rules that have governed the allocation of electoral votes has shaped, and continues to shape, the workings of an electoral practice created over two hundred years ago in Philadelphia.
Table A1. Democratic Dominance, 1828-1856 Candidate States EV per EV Bonus Adjusted state EV 1828 Jackson (D) 15 11.8 178 30 148 Adams (W) 9 9.2 83 18 65 1832 Jackson (D) 16 13.7 219 32 187 Clay (W) 6 8.2 49 12 37 1836 Van Buren (D) 15 11.3 170 30 140 Harrison (W) 7 10.4 73 14 59 1840 Harrison (W) 19 12.3 234 38 196 Van Buren (D) 7 8.5 60 14 46 1844 Polk (D) 15 11.3 170 30 140 Clay (W) 11 9.5 105 22 83 1848 Taylor (W) 15 10.9 163 30 133 Cass (D) 15 8.4 127 30 97 1852 Pierce (D) 27 9.4 254 54 200 Scott (W) 4 10.5 42 8 34 1856 Buchanan (D) 19 9.1 174 38 136 Fremont (R) 11 10.3 114 22 92 Candidate Actual Adjusted % % 1828 Jackson (D) 68.2 69.5 Adams (W) 31.8 * 30.5 1832 Jackson (D) 76.5 78.6 Clay (W) 17.1 * 15.5 1836 Van Buren (D) 57.8 57.8 Harrison (W) 24.8 * 24.4 1840 Harrison (W) 79.6 80.9 Van Buren (D) 20.4 * 19.0 1844 Polk (D) 61.8 62.8 Clay (W) 38.2 * 37.2 1848 Taylor (W) 56.2 57.8 Cass (D) 43.8 * 42.2 1852 Pierce (D) 85.8 * 85.4 Scott (W) 14.2 14.5 1856 Buchanan (D) 58.8 * 58.1 Fremont (R) 38.5 39.3 * indicates beneficiary of Connecticut Effect. Explanation of abbreviations: states = states carried; EV = electoral votes; Bonus = number of bonus votes won by candidate; Adjusted EV = electoral votes after subtracting the bonus votes; Actual % = percentage of total electoral votes won by candidate; Adjusted % = percentage of adjusted electoral votes won by candidate. Sources: Compiled from data in Presidential Elections, 1789-1996 (Congressional Quarterly, Inc., Washington, D.C. 1997). Table A2. Republican Dominance, 1860-1892 Candidate States EV per EV Bonus Adjusted state EV 1860 Lincoln (R) 17 10.6 180 34 146 Breckenridge (D) 11 6.5 72 22 50 Bell (I) 3 13.0 39 6 33 Douglas (D) 2 6.0 12 4 8 1864 Lincoln (R) 22 9.6 212 44 168 McClellan (D) 3 7.0 21 6 15 1868 Grant (R) 26 8.2 214 52 162 Seymour (D) 8 10.0 80 16 64 1872 Grant (R) 29 9.8 286 58 228 Greeley (D) 6 10.5 63 12 51 1876 Hayes (R) 21 8.8 185 42 143 Tilden (D) 17 10.8 184 34 150 1880 Garfield (R) 19 11.2 214 38 176 Hancock (D) 19 8.1 155 38 117 1884 Cleveland (D) 20 10.9 219 40 179 Blaine (R) 18 10.1 182 36 146 1888 Harrison (R) 20 11.6 233 40 193 Cleveland (D) 18 9.3 168 36 132 1892 Cleveland (D) 23 12.0 277 46 231 Harrison (R) 16 9.0 145 32 113 Weaver (Pop) 5 4.4 22 10 12 Candidate Actual Adjusted % % 1860 Lincoln (R) 59.4 61.6 Breckenridge (D) 23.7 * 21.1 Bell (I) 12.9 13.9 Douglas (D) 3.9 * 3.4 1864 Lincoln (R) 90.9 91.8 McClellan (D) 9.0 * 8.2 1868 Grant (R) 72.7 * 71.7 Seymour (D) 27.2 28.3 1872 Grant (R) 81.9 * 81.7 Greeley (D) 18.0 18.3 1876 Hayes (R) 50.1 * 48.8 Tilden (D) 49.9 51.2 1880 Garfield (R) 57.9 60.0 Hancock (D) 42.0 * 39.9 1884 Cleveland (D) 54.6 55.0 Blaine (R) 45.4 * 44.9 1888 Harrison (R) 58.1 59.4 Cleveland (D) 41.9 * 40.6 1892 Cleveland (D) 62.4 64.9 Harrison (R) 32.6 * 31.7 Weaver (Pop) 4.9 * 3.4 * indicates beneficiary of Connecticut Effect. Explanation of abbreviations: states = states carried; EV = electoral votes; Bonus = number of bonus votes won by candidate; Adjusted EV = electoral votes after subtracting the bonus votes; Actual % = percentage of total electoral votes won by candidate; Adjusted % = percentage of adjusted electoral votes won by candidate. Sources: Compiled from data in Presidential Elections, 1789-1996 (Congressional Quarterly, Inc., Washington, D.C. 1997). Table A3. Republican Dominance, 1896-1928 Candidate States EV per EV Bonus Adjusted state EV 1896 McKinley (R) 23 11.8 271 46 225 Bryan (D) 22 8.8 176 44 132 1900 McKinley (R) 28 10.4 292 56 236 Bryan (D) 17 9.1 155 34 121 1904 Roosevelt (R) 33 10.2 336 66 270 Parker (D) 12 11.6 140 24 116 1908 Taft (R) 30 10.7 321 60 261 Bryan (D) 16 10.1 162 32 130 1912 Wilson (D) 40 10.8 435 80 355 Roosevelt (I) 6 14.6 88 12 76 Taft (R) 2 4.0 8 4 4 1916 Wilson (D) 30 9.2 277 60 217 Hughes (R) 18 14.1 254 36 218 1920 Harding (R) 37 10.9 404 74 330 Cox (D) 11 11.5 127 22 105 1924 Coolidge (R) 35 10.9 382 70 312 Davis (D) 12 11.3 136 24 112 1928 Hoover (R) 40 11.1 444 80 364 Smith (D) 8 10.8 87 16 71 Candidate Actual Adjusted % % 1896 McKinley (R) 60.6 63.0 Bryan (D) 39.4 * 36.9 1900 McKinley (R) 65.3 66.1 Bryan (D) 34.7 * 33.9 1904 Roosevelt (R) 70.6 * 69.9 Parker (D) 29.4 30.0 1908 Taft (R) 66.4 66.7 Bryan (D) 33.5 * 33.2 1912 Wilson (D) 81.9 * 81.6 Roosevelt (I) 16.6 17.5 Taft (R) 1.5 * 1.0 1916 Wilson (D) 52.2 * 49.9 Hughes (R) 47.8 50.1 1920 Harding (R) 76.1 * 75.8 Cox (D) 23.9 24.1 1924 Coolidge (R) 71.9 71.7 Davis (D) 25.6 25.7 1928 Hoover (R) 83.6 83.6 Smith (D) 16.4 * 16.3 * indicates beneficiary of Connecticut Effect. Explanation of abbreviations: states = states carried; EV = electoral votes; Bonus = number of bonus votes won by candidate; Adjusted EV = electoral votes after subtracting the bonus votes; Actual % = percentage of total electoral votes won by candidate; Adjusted % = percentage of adjusted electoral votes won by candidate. Sources: Compiled from data in Presidential Elections, 1789-1996 (Congressional Quarterly, Inc., Washington, D.C. 1997). Table A4. Democratic Dominance, 1932-1964 Candidate States EV per EV Bonus Adjusted state EV 1932 Roosevelt (D) 42 11.2 472 84 388 Hoover (R) 6 9.8 59 12 47 1936 Roosevelt (D) 46 11.4 523 92 431 Landon (R) 2 4.0 8 4 4 1940 Roosevelt (D) 38 11.8 449 76 373 Willkie (R) 10 8.2 82 20 62 1944 Roosevelt (D) 36 12.0 432 72 360 Dewey (R) 12 8.2 99 24 75 1948 Truman (D) 28 10.8 303 56 247 Dewey (R) 16 11.8 189 32 157 Thurmond (I) 4 9.8 39 8 31 1952 Eisenhower (R) 39 11.3 442 78 364 Stevenson (D) 9 9.9 89 18 71 1956 Eisenhower (R) 41 11.1 457 82 375 Stevenson (R) 7 10.4 73 14 59 960 Kennedy (D) 23 13.2 303 46 257 Nixon (R) 26 8.4 219 52 167 Byrd (I) 1 15.0 15 2 13 1964 Johnson (D) 45 10.8 486 90 396 Goldwater (R) 6 8.6 52 12 40 Candidate Actual Adjusted % % 1932 Roosevelt (D) 88.8 89.2 Hoover (R) 11.1 * 10.8 1936 Roosevelt (D) 98.5 99.1 Landon (R) 1.5 * 0.9 1940 Roosevelt (D) 84.5 85.7 Willkie (R) 15.4 * 14.2 1944 Roosevelt (D) 81.3 82.7 Dewey (R) 18.6 * 17.2 1948 Truman (D) 57.0 * 56.7 Dewey (R) 35.6 36.1 Thurmond (I) 7.3 * 7.1 1952 Eisenhower (R) 83.2 83.6 Stevenson (D) 16.7 * 16.3 1956 Eisenhower (R) 86.0 86.2 Stevenson (R) 13.7 * 13.5 960 Kennedy (D) 56.4 58.8 Nixon (R) 40.8 * 38.2 Byrd (I) 2.8 2.9 1964 Johnson (D) 90.3 90.8 Goldwater (R) 9.6 * 9.1 * indicates beneficiary of Connecticut Effect. Explanation of abbreviations: states = states carried; EV = electoral votes; Bonus = number of bonus votes won by candidate; Adjusted EV = electoral votes after subtracting the bonus votes; Actual % = percentage of total electoral votes won by candidate; Adjusted % = percentage of adjusted electoral votes won by candidate. Sources: Compiled from data in Presidential Elections, 1789-1996 (Congressional Quarterly, Inc., Washington, D.C. 1997). Table A5. Republican Dominance, 1968-2008 Candidate States EV per EV Bonus Adjusted state EV 1968 Nixon (R) 32 9.4 301 64 237 Humphrey (D) 14 13.6 191 28 163 Wallace (I) 5 9.2 46 10 36 1972 Nixon (R) 49 10.6 521 98 423 McGovern (D) 2 8.5 17 4 13 1976 Carter (D) 24 12.4 297 48 249 Ford (R) 27 8.9 241 54 187 1980 Reagan (R) 44 11.0 489 88 401 Carter (D) 7 7.0 49 14 35 1984 Reagan (R) 49 10.7 525 98 427 Mondale (D) 2 6.5 13 4 9 1988 Bush (R) 40 10.6 426 80 346 Dukakis (D) 11 10.2 112 22 90 1992 Clinton(D) 33 11.2 370 66 304 Bush (R) 18 9.3 168 36 132 1996 Clinton (D) 32 11.8 379 64 315 Dole (R) 19 8.3 159 38 121 2000 Bush (R) 30 9.0 271 60 211 Gore (D) 21 12.7 267 42 225 2004 Bush (R) 31 9.2 286 62 224 Kerry (D) 20 12.6 252 40 212 2008 Obama (D) 29 12.6 365 58 307 McCain (R) 22 7.8 173 44 129 Candidate Actual Adjusted % % 1968 Nixon (R) 55.9 * 54.3 Humphrey (D) 35.5 37.4 Wallace (I) 8.5 * 8.2 1972 Nixon (R) 96.8 97.0 McGovern (D) 3.1 * 3.0 1976 Carter (D) 55.2 57.1 Ford (R) 44.8 42.9 1980 Reagan (R) 90.9 91.9 Carter (D) 9.1 * 8.0 1984 Reagan (R) 97.6 97.9 Mondale (D) 2.4 * 2.1 1988 Bush (R) 79.2 79.3 Dukakis (D) 20.8 * 20.6 1992 Clinton(D) 68.7 69.7 Bush (R) 31.2 * 30.3 1996 Clinton (D) 70.4 72.2 Dole (R) 29.5 * 27.7 2000 Bush (R) 50.4 * 48.4 Gore (D) 49.6 51.6 2004 Bush (R) 53.1 * 51.3 Kerry (D) 46.8 48.6 2008 Obama (D) 67.8 70.4 McCain (R) 32.1 * 29.6 * indicates beneficiary of Connecticut Effect. Explanation of abbreviations: states = states carried; EV = electoral votes; Bonus = number of bonus votes won by candidate; Adjusted EV = electoral votes after subtracting the bonus votes; Actual % = percentage of total electoral votes won by candidate; Adjusted % = percentage of adjusted electoral votes won by candidate. Sources: Compiled from data in Presidential Elections, 1789-1996 (Congressional Quarterly, Inc., Washington, D.C. 1997) and Dave Leip's Atlas of U.S. Presidential Elections at http://uselectionatlas.org/RESULTS/ (accessed on 28 September 2010).
(1.) For a recent examination of the effects of the three-fifths clause on the election of 1800 see Garry Wills, "Negro President": Jefferson and the Slave Power, Boston, MA: Houghton Mifflin Harcourt, 2003; see also G. Wills, "The Negro President," The New York Review of Books, 6 November 2003, 45-51. But note the corrective raised by Lance Banning about the problematic nature of trying to recalculate the impact of subtracting the slave electors without also adjusting for the extra electoral votes given to the smaller Northern states (Lance Banning, "Three-Fifths Historian," Claremont Review of Books 4, 2004, 54-6).
(2.) For additional discussions of the electoral power of the Southern slave states and the effect of that added influence on the 1800 presidential election see Leonard Richards, The Slave Power, Baton Rouge: Louisiana State UP, 2000, 40-2 and William W. Freehling, The Road to Disunion: Secessionists at Bay, 1776-1854, New York: Oxford UP, 1990, 146-8. For further discussion of the impact of the slave electors on the 1800 election, see Andrew Robertson, "The Election of 1800," in William G. Shade and Ballard C. Campbell, eds., American Presidential Campaigns and Elections: A Reference Guide, Armonk, NY: M. E. Sharpe, 2003, vol. 1, 113-27: 124-5. For a related account about a different controversy concerning irregularities in the Georgia ballot in the 1800 election, see Bruce Ackerman and David Fontana, "Thomas Jefferson Counts Himself Into the Presidency," Virginia Law Review 2, 2004, 551-643.
(3.) Shlomo Slonim, "The Electoral CoLlege at Philadelphia: The Evolution of an Ad Hoc Congress for the Selection of a President," Journal of American History 1, 1986, 35-58: 55.
(4.) The best treatment of the creation of the Electoral College at the Philadelphia convention is Slonim, "Electoral College." See also the discussions in Tadahisa Kuroda, The Origins of the Twelfth Amendment: The Electoral College in the Early Republic, 1787-1804, Westport, CT: Greenwood, 1994, 7-15; Neal R. Pierce and Lawrence D. Longley, The People's President: The Electoral College in American History and the Direct Vote Alternative, rev. ed., New Haven, CT: Yale UP, 1981, 10-30; George C. Edwards, Why the Electoral College is Bad for America, New Haven, CT: Yale UP, 2004, 78-91; Carol Berkin, A Brilliant Solution: Inventing the American Constitution, New York: Harcourt, 2002, 116-148. For an excellent treatment of the entire process, see Jack N. Rakove, Original Meanings: Politics and Ideas in the Making of the Constitution, New York: Knopf, 1996 (for the two crucial compromises at Philadelphia, see ibid., 57-93).
(5.) See Randall E. Adkins and Kent A. Kirwan, "What Role Does the 'Federalism Bonus' Play in Presidential Selection ?," Publius: The Journal of Federalism 4, 2002, 71-90; Glenn W. Rainey, Jr., and Jane G. Rainey, "The Electoral College: Political Advantage, the Small States, and Implications for Reform," in Robert P. Watson, ed., Counting Votes: Lessons from the 2000 Presidential Election in Florida, Gainesville, FL: U. of Florida P., 2004, 170-191. Both articles focus primarily on the 1876, 1916, and 2000 elections. Meanwhile, the authors of both articles use their findings chiefly to serve other ends. Adkins and Kirwan embed their discussion of these elections in a defense of the present system of bonus votes (what they call the "federalism bonus") and an argument against Electoral College reform. Rainey and Rainey use the three elections as a springboard for the discussion of a variety of alternatives to the present Electoral College system, how those variations might be implemented, and what they might mean.
(6.) A scholarly analysis of any Presidential election results should also account for the context of recent voting behavior and trends in local, state, and national contests, and consider candidates' over- or underperformance in specific elections according to normal vote percentages (see Lee Benson, "Research Problems in American Political Historiography," in Mirra Komarovsky, ed., Common Frontiers of the Social Sciences, Glencoe, IL: The Free Press, 1957, 113-183; Albert C. E. Parker, "Beating the Spread: Analyzing American Election Outcomes," Journal of American History 1, 1980, 61-87). I am indebted to Fred I. Greenstein for alerting me to the existence of Benson's chapter.
(7.) See for an excellent short discussion of the means of choosing presidential electors, "The Electoral College," in Presidential Elections 1789-1996, Washington, DC: Congressional Quarterly, 1997, 7-10. The Connecticut Effect may have been at work in the 1796 and 1800 elections. Certainly, the application of some models of calculating electoral votes and subtracting bonus votes alters the outcomes, or expands or contracts the margins of victory. The absence of winner-take-all rules, however, and the prevalence instead of split electoral votes in several states (not to mention the absence of direct popular voting) defeat any fair, simple, or straightforward way of reallocating bonus votes. One encounters too many other variables and uncertainties in the allocation of electoral votes in those early Presidential contests to confidently consider either 1796 or 1800 as Connecticut Effect elections.
(8.) See for the fullest treatment of early presidential elections, Tadahisa Kuroda, Origins.
(9.) An excellent short discussion is found in Jack N. Rakove, "The E-College in the E-Age," in J. N. Rakove, ed., The Unfinished Election of 2000, New York: Basic Books, 2001, 201-34 (see especially ibid., 215-220). Rakove elaborates on the politicizing of the presidency and the evolving functioning of the Electoral College in J. N. Rakove, "The Political Presidency," in James Horn, Jan Ellen Lewis, and Peter S. Onuf, eds., The Revolution of 1800: Democracy, Race, and the New Republic, Charlottesville, VA: U. of Virginia P., 2002, 30-58.
(10.) See Table A1 in Appendix.
(11.) See Table A2 in Appendix.
(12.) See Keith I. Polakoff, The Politics of Inertia: The Election of 1876 and the End of Reconstruction, Baton Rouge, LA: Louisiana State UP, 1973; Sidney I. Pomerantz, "Election of 1876," in Arthur Schlesinger, Jr., and Fred Israel, eds., History of American Presidential Elections, New York: Chelsea House, 1971, vol. 2, 1379-435; K. I. Polakoff, "The Election of 1876," in Shade and Campbell, American Presidential Campaigns, vol. 2,465-76; Eugene Roseboom, A History of Presidential Elections, New York: Macmillan, 1957, 236-52; Ari Hoogenboom, Rutherford B. Hayes: Warrior and President, Lawrence, KS: U of Kansas P, 1995, 256-94.
(13.) For perspectives that consider this election in the broader context of Gilded-Age politics and electioneering, see Mark Wahlgren Summers, The Era of Good Stealings, New York: Oxford UP 1993, 274-99; M. W. Summers, Rum, Romanism, and Rebellion: The Making of a President 1884, Chapel Hill, NC: U. of North Carolina E, 2000, 27-40; M. W. Summers, The Press Gang: Newspaper and Politics 1865-1878, Chapel Hill, NC: U. of North Carolina P., 1994, 279-313. And see also Gore Vidal's fascinating historical novel 1876, which is set against the backdrop of this disputed election, and features Samuel J. Tilden as a background character (G. Vidal, 1876, New York: Vintage, 1976).
(14.) Of course, a key factor in the voting of the Southern states was the systematic racist violence designed to physically intimidate blacks and thus diminish strongly Republican support (see for example Paul A. Gilje, Rioting in America, Bloomington, IN: Indiana UP, 1996, 94-108).
(15.) See Lawrence D. Longley and Neal R. Pierce, The Electoral College Primer 2000, New Haven, CT: Yale UP, 1999, 137-143; Rainey and Rainey, "The Electoral College." Adkins and Kirwan (see note 5) use instead the phrase "federalism bonus," while other terms are sometimes used by scholars, but my preferred term is "bonus votes," which I use throughout this article.
(16.) See Charles Stewart and Barry R. Weingast, "Stacking the Senate, Changing the Nation: Republican Rotten Boroughs, Statehood Politics, and American Political Development," Studies in American Political Development 2, 1992, 223-71.
(17.) See also Michael E Holt, By One Vote: The Disputed Presidential Election of 1876, Lawrence, KS: U. of Kansas P., 2008, 28-32. For a valuable reminder of the role of apportionment on representation in state legislatures during the era, see Peter Argersinger, "The Value of the Vote: Political Representation in the Gilded Age," Journal of American History 1, 1989, 59-90. More broadly, see Rosemarie Zagarri, The Politics of Size: Representation in the United States, 1776-18S0, Ithaca, NY: Cornell UP, 1987.
(18.) On the politics of statehood admission, specifically that of Nevada and Colorado, see Stewart and Weingast, "Stacking the Senate," 230-42, as well as the works cited in note 47 below. In 1876, Nebraska voted for Hayes by 65% to 35%; Nevada favored Hayes by 53% to 47%; brand new Colorado cast no popular votes that year.
(19.) For a fascinating overview of the politics of this entire era, see M. W. Summers, Party Games: Getting, Keeping, and Using Power in Gilded Age Politics, Chapel Hill, NC: U. of North Carolina P., 2004.
(20.) For assessments of election fraud during the era see Peter H. Argersinger, "New Perspectives on Election Fraud in the Gilded Age," in Political Science Quarterly 100 1985-86, 669-687; and also Summers, "Party Games: The Art of Stealing Elections in the Late Nineteenth-Century United States," in Journal of American History 88 2001, 424435. More broadly, see Tracy Campbell, Deliver the Vote: A History of Election Fraud, An American Political Tradition--1742-2004 (New York: Carroll and Graf, 2005).
(21.) See Table A3 in Appendix.
(22.) Wilson won California with 46.6% of the vote to 46.3% for Hughes, a plurality of just 3,420 votes out of 974,250 votes cast in the state. Wilson won the election with 277 electoral votes to 254 for Hughes. Had Hughes managed to win California's thirteen electoral votes, Hughes would have triumphed over Wilson, 267 to 264.
(23.) Standard accounts of the 1916 election may be found in Arthur S. Link and William Leary, "Election of 1816," in Schlesinger and Israel, eds., American Presidential Elections, vol. 3, 2245-70; Maureen A. Flanagan, "The Election of 1916," in Shade and Campbell, American Presidential Campaigns, vol. 2, 646-55; Roseboom, History, 379-388; A. S. Link, Woodrow Wilson and the Progressive Era, 1910-1917, New York: Harper and Row, 1954, 223-51; Dexter Perkins, Charles Evans Hughes and American Democratic Statesmanship, Boston, MA: Little, Brown, 1956, 50-70. An excellent book-length treatment is S. D. Lovell, The Presidential Election of 1916, Carbondale, IL: Southern Illinois UP, 1980, 135-47.
(24.) Hughes's winning margin would actually have been 219-216 (or his losing margin 255-276) but one West Virginia elector voted for Wilson. Hughes carried West Virginia narrowly, winning 49.4% to 48.5% for Wilson, and thus won the state's two bonus votes (which would be deducted from his total under this revised scenario), but only seven of the state's eight electoral votes.
(25.) It is germane, too, to note that the women's vote aided Wilson. Women could vote for President in twelve states in 1916; Wilson won in ten of them, in part because Democrats had recognized the significance of this voting bloc in 1912 and made deliberate appeals to women in 1916. By contrast, the Hughes campaign was strongly criticized by women's leaders for not taking these votes seriously. The women's vote helped Wilson achieve his very slim victory margin in California. Six other states that Wilson carried where women could vote included Connecticut-Effect states that had three or four electoral votes and whose bonus votes gave Wilson his margin: Arizona, Idaho, Montana, Nevada, Utah, and Wyoming. For further discussion see Flanagan, "The Election of 1916"; Molly M. Wood, "Mapping a National Campaign Strategy: Partisan Women in the Presidential Election of 1916," in Melanie Gustafson et al., eds., We Have Come to Stay: American Women and Political Parties, Albuquerque, NM: U. of New Mexico P., 1999, 77-86.
(26.) See The Autobiographical Notes of Charles Evans Hughes, eds. David J. Danelski and Joseph S. Tulchin, Cambridge, MA: Harvard UP, 1973, 182, 184.
(27.) Ibid., 185.
(28.) See Table A4 in Appendix.
(29.) See Table A5 in Appendix.
(30.) This 12.7 figure represents Gore's average electoral vote per state, counting the District of Columbia as a state. If one subtracts D.C.'s 3 electoral votes from Gore's total and calculates only states, his per state average rises to 13.3 (264 votes on 20 states).
(31.) The 2000 election is still too recent to have much historical literature yet, but see the essays in Rakove, ed., Unfinished Election; Melanie Shell Weiss, "The Election of 2000," in Shade and Campbell, American Presidential Campaigns, vol. 3, 1032-1046; James W. Caeser and Andrew E. Busch, The Perfect Tie: The True Story of the 2000 Presidential Election, Lanham, MD: Rowman and Littlefield, 2001; Gerald M. Pomper, ed., The Election of 2000: Reports and Interpretations, New York: Chatham House, 2001.
(32.) In many ways, the 1876 campaign was only getting started, rather than concluding, with election day. For excellent full treatments of the efforts by and on behalf of both Hayes and Tilden from election day to inauguration day, see especially Polakoff, Politics, 201-324; L. Hoogenboom, The Presidency of Rutherford B. Hayes, Lawrence, KS: U. of Kansas P., 1988, 25-50; Harry Barnard, Rutherford B. Hayes and His America, New York: American Political Biography Press, 1954, 314-96; Alexander Flick, Samuel J. Tilden: A Study in Political Sagacity, Port Washington, NY: Greenwood, 1939, 322-416. For a shorter summary, see Pomerantz, "Election of 1876."
(33.) As quoted in Flick, Samuel Tilden, 526.
(34.) Although Tilden grew up steeped in the Jacksonian political tradition and has been seen mainly as an inheritor of Jackson's political philosophy, he reacted very differently to his own "counting out" than Old Hickory did in 1824 when, despite winning the popular and electoral vote, Jackson lost the election in the House of Representatives. Jackson and his supporters immediately denounced what they termed a "corrupt bargain," spent the next four years running against Adams and reminding voters of this "stolen" election, and organizing for a landslide victory in 1828. Tilden, despite similarities in the circumstance (and with arguably much greater evidence of a corrupt bargain) did none of those things.
(35.) Flick, Samuel Tilden, 526.
(36.) See for a discussion of the Colorado results and the Democratic reaction to them, Polakoff, Politics, 149-50. Of those battleground states, Tilden carried all except Ohio and Wisconsin, both of which he lost narrowly.
(37.) Interestingly, although the point is only incidental to his larger purposes, Alan Natapoff, in a mathematical analysis of the relative voting power of large states and small states, noticed the effect of the bonus votes (what he dubbed "Senatorial votes") in the 1916 election; he wrote how Wilson's margin "was provided entirely by the Senatorial votes," and argued that "[a]lthough Senatorial votes are often criticized as undemocratic, they saved the popular-vote winner's victory in 1916," but then incorrectly called the Wilson-Hughes contest "the only election that has ever turned on them [bonus votes]" (see A. Natapoff, "A Mathematical One-Man One-Vote Rationale for Madisonian Presidential Voting Based on Maximum Individual Voting Power," Public Choice 3-4, 1996, 259-273: 266).
(38.) Autobiographical Notes of Charles Evans Hughes, 184.
(39.) Ibid., 184-5
(40.) Ibid., 185.
(41.) James R. Garfield quoted in Lovell, Presidential Election of 1916, 177. Lovell, the leading student of the 1916 campaign, states unequivocally that "[i]t was the West that produced the victory for Wilson; all except Iowa, South Dakota, Minnesota, and Oregon west of the Mississippi gave their electoral support to the president ... [and olf 200 counties carried by a Democratic candidate for the presidency for the first time, most were in the states west of the Mississippi (and in the northern half of that section)" (ibid., 172). Lovell notes that these gains, when added to the solid Southern votes, were enough to counteract Republican dominance in the Northeast and the other highly populous states that Hughes carried (see for a fuller analysis of the returns ibid., 170-82).
(42.) For the Democratic campaign in the west, see Leonard Bates, "Mr. Wilson's Campaign: Winning the West with Wilson in 1916," Journal of the West 1, 1995, 16-23. For a fuller analysis of the western campaign and election results, see David Sarasohn, "The Election of 1916: Realigning the Rockies," Western Historical Quarterly 3, 1980, 285-305.
(43.) For an insightful analysis of the 36-day Florida standoff that emphasizes Gore's defeatism and his focus on a legalistic, as opposed to political, approach to the vote counts (both qualities which made his response very similar to Samuel Tilden's in 1876) see Jeffrey Toobin, Too Close to Call: The Thirty-Six Day Battle to Decide the 2000 Election, New York: Random House, 2002, especially 7-8.
(44.) Gore advisor Robert Shrum quoted in Kathleen Hall Jamieson and Paul Waldman, eds., Electing the President 2000: The Insiders' View, Philadelphia, PA: U. of Pennsylvania P., 2001, 74. For a fuller discussion see Shrum's own memoir, R. Shrum, No Excuses: Concessions of a Serial Campaigner, New York: Simon & Schuster, 2007, 370-82.
(45.) But if the strategists inside the campaign failed to notice, others outside the campaigns did. In addition to the sources cited in note 1 above, the existence of the revised electoral count minus the bonus votes in 2000 was noted by a few observers. Within a month of the election an online discussion in "Media Backtalk" hosted by Howard Kurtz at washingtonpost.com referred to the 225-211 numbers (see "Media Backtalk," 4 December 2000, available at http://www.washingtonpost.com/wp-srv/liveonline/00/politics/ mediabacktalk120400.htm. [accessed 28 September 2010]). Jack Rakove did not report the revised electoral vote count but did note that "Bush's margin of victory in the electoral college thus rested on the eighteen additional electoral votes he gained simply for carrying more states" (Jack Rakove, "E-College," 205). Neither source, however, linked the 2000 results to the earlier Connecticut-Effect elections of 1876 and 1916.
(46.) For pointed criticisms of the undemocratic nature of the United States Senate and the great overrepresentation of the smaller Western states see Tom Geoghegan, "The Infernal Senate," The New Republic, 21 November 1994, 17-23, and Richard N. Rosenfeld, "What Democracy?: The Case for Abolishing the United States Senate," Harper's Magazine, May 2004, 35-44. See as well Sam Dillon, "In Some Schools, It's One Teacher, One Student," The New York Times, 19 January 2004, A1 and, A10, and Daniel H. Pink, "Givers and Takers," New York Times, 30 January 2004, available at: http://www.nytimes.com/2004/01/30/opinion/givers-and-takers.html, accessed 15 December 2010.
(47.) On the politics behind statehood creation see Stewart and Weingast, "Stacking the Senate," and Summers, Party Games. I am indebted to Mark Wahlgren Summers for calling my attention to the rich body of literature on the territories, territorial politics, and statehood admission in the works I cite here and in note 50 below. On the political and other considerations behind the drawing of territorial and state boundaries, see Glen M. Leonard, "Southwestern Boundaries and the Principles of Statemaking," Western Historical Quarterly 1 1977, 39-53; Merle Wells, "Idaho's Season of Political Distress: An Unusual Path to Statehood," Montana: The Magazine of Western History 4, 1987, 58-67. For the structure of politics in western territories and their evolution into statehood see Kenneth N. Owens, "Pattern and Structure in Western Territorial Politics," Western Historical Quarterly 4, 1970, 373-392. On the experiences of particular states see Howard R. Lamar, Dakota Territory, 1861-1889, New Haven, CT: Yale UP, 1956; Richard G. Lillard, Desert Challenge: An Interpretation of Nevada, New York: Greenwood, 1949; revised edition 1979; Ronald H. Limbaugh, Rocky Mountain Carpetbaggers: Idaho's Territorial Governors 1863-1890, Moscow, ID: Idaho Research Foundation, 1982.
(48.) See Paul Kleppner, "Voters and Parties in the Western States, 1876-1900," Western Historical Quarterly 1, 1983, 49-68; Jeffrey Ostler, "Why the Populist Party Was Strong in Kansas and Nebraska but Weak in Iowa," Western Historical Quarterly 4, 1992, 451-474.
(49.) For a powerful demonstration of the states in which presidential candidates in 1976 and in 2000 campaigned and made media buys, and those they ignored see Edwards, Why the Electoral College, 92-121. Small states, unsurprisingly, were largely bypassed in favor of states with larger electoral vote payoffs. On the targeting of some states (and ignoring of others) in recent election cycles, see Daron R. Shaw, The Race to 270: The Electoral College and the Campaign Strategies of 2000 and 2004, Chicago, IL: U. of Chicago P., 2006; D. R. Shaw, "The Methods Behind the Madness: Presidential Electoral College Strategies, 1988-1996," The Journal of Politics 4, 1999, 893-913; D. R. Shaw, "The Effect of TV Ads and Candidate Appearances on Statewide Presidential Votes, 1988-96," American Political Science Review 2, 1999, 345-361; Larry M. Bartels, "Resource Allocation in a Presidential Campaign," The Journal of Politics 3, 1985, 928-936.
(50.) Calls for electoral reform along these lines are nothing new. For Gilded Age-era discussions see F. A. P. Barnard, William Purcell, H. L. Dawes, Roger A. Pryor, Z. B. Vance, "How Shall the President Be Elected?," North American Review 339, 1885, 97-129. For articles advocating distribution of electoral votes by districts see "Presidential Election by Districts," Harper's Weekly, 7 February 1880, 82; S. M. Merrill, "Our Electoral System," North American Review 479, 1896, 402-16. For an alternative method of allocation, see Benjamin F. Meyers, "The Single Vote in Congressional Elections," North American Review 403, 1890, 782-5.
(51.) Proportional allocation of electoral votes was enacted in Michigan for the 1892 Presidential election, with the result that the state's 14 electoral votes were divided, nine for Benjamin Harrison (who won 48% of the vote) and five for Grover Cleveland (who took 43%). Four years later Michigan returned to winner-take-all and William McKinley got all 14 votes. Colorado voters in 2004 rejected a proposal to divide the state's nine electoral votes proportionally. In 2008 California and North Carolina voters faced the option of dividing their states' votes by congressional districts carried, as Maine and Nebraska have done (see Hendrik Hertzberg, "Votescam," The New Yorker, 6 August 2007, available at: http://www.newyorker.com/talk/comment/2007/08/06/070806taco_talk_hertzberg, accessed 15 December 2010).
(52.) Some scholars have already made assessments of how presidential election returns might be affected (and have been skewed in the past by winner-take-all) by analyzing elections in terms of increases in the size of the House, proportional allocation, and congressional district allocation (see for example Michael G. Neubauer and Joel Zeitlin, "Outcomes of Presidential Elections and the House Size," in PS: Political Science and Politics 4, 2003, 721-5). For ballot reform laws and the impact on Gilded Age and early twentieth-century elections, see Summers, Party Games, 133-8; John F. Reynolds and Richard L. McCormick, "Outlawing 'Treachery': Split Tickets and Ballot Laws in New York and New Jersey, 1880-1910," Journal of American History 4, 1986, 835-58. Also, the true winner of the popular vote in the 1960 Kennedy-Nixon election has recently been disputed (not because of longstanding allegations [never definitively proven] of vote fraud in Illinois and Texas), but due to the apparently mistaken allotment of votes for presidential electors in Alabama (see Brian J. Gaines, "Popular Myths About Popular Vote-Electoral College Splits," PS: Political Science and Politics 1, 2001, 70-75; Edwards, Why the Electoral, 48-51).
(53.) For the most recent and cogently argued statement in favor of abolishing the Electoral College, see Edwards, Why The Electoral. The editorial page of The New York Times in 2004 also came out in favor of eliminating this apparatus. See "Making Votes Count; Abolish the Electoral College," 29 August 2004, available at: http://www.nytimes.com/2004/08/29/opinion/making-votes-count-abolish-the- electoral-college.html, accessed 15 December 2010). Other books critical of the Electoral College include Longley and Peirce, Electoral College Primer; Longley and Peirce, People's President; Robert Dahl, How Democratic is the American Constitution?, New Haven, CT: Yale UP, 2001. See also the essays in William N. Eskridge, Jr., and Sanford Levinson, eds., Constitutional Stupidities, Constitutional Tragedies, New York: NYU Press, 1998.
(54.) Arguments in favor of retaining the institution are found in Judith A. Best, The Case Against Direct Election of the President: A Defense of the Electoral College, Ithaca, NY: Cornell UP, 1975; Robert M. Hardaway, The Electoral College and the Constitution: The Case for Preserving Federalism, Westport, CT: Praeger Publishers, 1994; and Gary L. Gregg, ed., Securing Democracy: Why We Have an Electoral College, Wilmington, DE: Intercollegiate Studies Institute, 2001. For a recent exchange of views see Jack N. Rakove, "Presidential Selection: Electoral Fallacies," Political Science Quarterly 1, 2004, 21-37; Judith A. Best, "Presidential Selection: Complex Problems and Simple Solutions," Political Science Quarterly 1, 2004, 39-59.
Todd Estes is Associate Professor of History at Oakland University in Rochester, Michigan. He is the author of The Jay Treaty Debate, Public Opinion, and the Evolution of Early American Political Culture (2006) and is at work on a book manuscript about the ratification of the Constitution. He wishes to thank Fred I. Greenstein, Mark Wahlgren Summers, and Bruce Zellers for their very helpful comments on an earlier version of this article as well as the editor and the anonymous readers for The Historian.
|Printer friendly Cite/link Email Feedback|
|Date:||Jun 22, 2011|
|Previous Article:||A difference of complexion: George Fitzhugh and the birth of the Republican Party.|
|Next Article:||The historiography of Bohai in Russia.|