The Uncertain Elections: 1876 and 2000
One of the problems with the American presidential electoral system is that it does not handle close contests smoothly. Winner-take-all…
One of the problems with the American presidential electoral system is that it does not handle close contests smoothly. Winner-take-all rules in the Electoral College mean that a narrow result in a single state is enough to cast doubt on the result of the entire presidential election.
The foundational justification for democracy is that it captures the will of the people. From a simple scientific perspective, if our measurement of the will of the people is unclear, we should measure it again. Close elections should be re-run under close supervision, or at a minimum re-counted in a timely and transparent fashion.
Instead of re-measuring the will of the people, however, the two uncertain American presidential elections have been resolved on an ad hoc basis. This places a lot of power over close presidential elections in the hands of a small handful of officials — who can be expected to be biased in their own partisan interests.
The two American elections that were decided in the absence of a clear winner happened in 1876 and 2000. The 1876 election pitted Samuel Tilden (D) against Rutherford B. Hayes (R); the 2000 election pitted Al Gore (D) against George W. Bush (R). The final results were immensely controversial in both cases, for one very good reason:
Nobody truly knew who had received more votes in key states at the time the election was officially decided. The numbers were too close and the methods used to count ballots simply weren’t reliable enough. What is worse is that weren’t clear and rational procedures in place for resolving these types of disputes, so the official results came from negotiation and litigation. Having a firmer grasp on the appropriate levers of existing political power is important in that type of process; how many votes are actually cast in favor of which candidate is not.
The elections in brief
In the case of 1876, all three states held elections under the administration of a Republican governor — three out of the four Republican governors remaining in office in the former Confederacy. The recognized state governments submitted returns favoring Rutherford Hayes, after throwing out results from some counties, parishes, and cities amidst allegations of fraud, voter suppression, and unfairly designed ballots that might trick some Republican voters, particularly illiterate ones, into accidentally voting for the wrong candidate. Democrats cried foul, and two sets of electors emerged from each state.
With no precedents for resolving disputes over which set of electors to recognize, Congress, divided between a Republican Senate and a Democratic House, put together a special Electoral Commission comprised of members of Congress and justices of the Supreme Court.
In the case of the 2000 election, Florida was the unique key state. There were allegations of voter suppression, voting machine malfunctions, and unfairly designed ballots that might trick Democratic voters into voting for the wrong candidate. The Republican administration of Florida, headed up by the brother of George W. Bush, certified Bush as the winner of the state. Efforts to force a recount went up to the Supreme Court and were shot down in a narrow 5–4 decision by the Court.
Commonalities
The two elections have a number of common features. One is that they were exceptionally close elections. Any small structural change is likely to change them. For example, these are two out of the three elections in American history where the two bonus “Senate seat” electoral votes mattered (the third being 1796). Similarly, both elections featured disagreement between the Electoral College and the popular vote, which is associated with close elections.
A few other common features are coincidental.
It is coincidence that the Republican party emerged victorious in both cases. The two major political parties have undergone several major political realignments since that time, and their fortunes have risen and fallen accordingly. In political science parlance, we would be comparing the Third Party System to the Fifth or Sixth Party System. It is worth noting that Tilden, a northern Democrat who had come out of the moderate anti-slavery wing of the party, had support both among white Southerners (who would never vote Republican) and within the greater New York metropolitan area, carrying the tri-state NY-NJ-CT area — a coalition difficult to imagine assembling in 2000.
Also largely coincidentally, Florida played a pivotal role in both elections, although it was not the only state whose results were within the margin of error. In the 1876 election, nobody truly knew who had received more votes in South Carolina or Louisiana either, and it’s not truly certain who received more votes in New Mexico in 2000, although neither campaign contested the official results.
What is not a coincidence is that in each of the four cases where the outcome was contested, the official results accepted favored the presidential candidate from the same party as the governor of that state. The Electoral College system is very sensitive to small shifts in key states, and it is perfectly clear in retrospect that had the state government been in the hands of the other party in those key states, the other party would have won the presidency — both in 1876 and 2000.
Many of the complaints on the ground in both elections are similar, including local authorities issuing deceptively (or simply badly) designed ballots that favored one candidate over another and voter suppression tactics, although those factors in particular favored Tilden (the losing candidate) in 1876 and Bush (the winning candidate) in 2000. The reason for the similar catalogue of complaints is simple and twofold. First, when an election is expected to be close, dirty tricks pay off, and there are only so many different dirty tricks available. Second, when an election is tightly contested, the procedures under which that election was carried out come under close scrutiny, and complaints that might otherwise be ignored are circulated.
Unfortunately, in both elections, the decision was ultimately taken out of the hands of the voters. The uncertain results were not resolved by a recount or by holding the election over again. While recounts and re-run elections are expensive and inconvenient, they are far less expensive than the results of a botched election. Fortunately, neither election resulted in a civil war, in spite of a significant amount of pre-election violence in 1876.
Resolutions of the elections
In the case of the 1876 election, the specially-appointed Electoral Commission voted 8–7 along party lines to affirm the Republican electors for all three disputed states. (The commission also resolved in Hayes’ favor a dispute over one Oregon elector’s qualifications.) Supporters of Tilden cried foul, but Rutherford Hayes became the next president.
The Republicans held onto the White House, but paid a heavy political price for it. The Democrats took over the Senate in the midterm elections; President Hayes withdrew the Republican state governments from the last of the occupied states and let the Southern Democrats run the South as they saw fit with little to no federal interference.
In the case of the 2000 election, a 5–4 Supreme Court decision affirmed the Florida government’s certified Bush victory and halted the recount; however, the basis of the decision was that Florida’s state government had dug in its heels for long enough that it would be impossible to complete a full recount of the state in a timely fashion. The Supreme Court justices did not know whether Gore or Bush had won more votes in the state of Florida, but thought there was no practical way to resolve this question fairly in time for the Electoral College to convene by recounting all of Florida’s ballots.
A subsequent unofficial independent recount suggested that if all “undervote” ballots in the state of Florida had been counted consistently and accurately in the first place, Gore would have won, albeit by a narrow margin. (In practice, different counties had used slightly different standards for counting votes.) It is likely that a re-run of the election would have been both easier to administer than a thorough election and given a more decisive result.
The Electoral College system makes provisions for what happens when no single candidate achieves a majority within the Electoral College itself: The House of Representatives decides among the top candidates, using an unusual “one state, one vote” rule. This rule is a holdover from the days of the Continental Congress, but it would be a mistake to think of it as just a technicality — the Founding Fathers expected that the House of Representatives would end up deciding most presidential elections. (This has only happened in 1800 and 1824.)
For better or for worse?
Whether or not it was a good idea in principle to leave these elections up to political negotiation or litigation, a question you might ask is this: Did it work out anyway pragmatically? Did we get the better president?
This is not an easy question. Most of the Republicans in the audience are probably nodding sagely that even if Bush II had not been quite a perfect president, he was better than Gore would have been; while the Democrats in the audience are probably nodding sagely, thinking that obviously Gore would have been much better.
It is too soon for many of us Americans to put aside partisanship and judge the second president Bush.
Most presidential historians would say that between Lincoln and McKinley, the US had a string of mediocre presidents in the late 19th century, with the possible exception of Grover Cleveland. It is hard to make the case that Hayes, who ended Reconstruction and made very limited progress with civil service reforms, was a better president than Cleveland. Tilden was a politician cut from a similar mold: Both were reformist Democrats from New York with a record of opposition to Tammany Hall.
There is little reason to think that ad hoc resolutions of the electoral process, like the 5–4 Bush v. Gore decision or the 8–7 Electoral Commission decisions, will generally give us better presidents than a recount or a re-run of the election.
Problems and solutions
The Founding Fathers did not anticipate disputes over the legitimacy of presidential electors. While the Electoral College system gives state legislatures the power to determine the method by which presidential electors are selected, it doesn’t set clear standards for how to enforce the integrity of that method. The result is that every issue must be resolved reactively, on an ad hoc basis.
The problem with this is that most of the time, the people who determine the nature of the ad hoc solution will themselves have a partisan interest. Unfortunately, even the Supreme Court is not free from allegations of partisanship. The only impartial solutions are those determined in advance, and we need to be prepared for this.
Counting can be hard. It may seem simple, but when you’re administering an election and counting a large number of ballots, things go wrong. Sometimes intentionally, sometimes accidentally. Accidental counting error rates on the order of 1% are not unusual, and these errors tend to be clustered. As a general mathematical rule, it is very unwise to place confidence in a difference in two counts when the difference is smaller than the square root of the number of things being counted.
Internationally, disputes over election results are not unusual, particularly but not only in developing democracies. The option that builds the most credibility in the legitimacy of the democratic process is to hold the election over again.
Another option is to perform a careful recount of ballots; this is less effective at building faith in the process, because any non-random problems might simply be replicated in a re-count. As we saw with Florida in 2000, recounts can be very long and expensive compared to a re-run of the election, particularly if local and state authorities are not inclined to be cooperative.
The third option is to affirm the legitimacy of the results until either objections die down or a civil war breaks out. This is what has been done in America in the past, and it’s risky. If we are not sure that the will of the people has been measured accurately enough, we should re-measure it.
We can, and should, have procedures in place to allow for a swift, efficient, and credible re-count of an election — or better yet, to plan for re-runs of elections that are too close to be sure.