In the year 2000, after 36 days of recounting Florida was handed to George W Bush setting the scene for his next 8 years in Washington D.C. The decisive margin, a supreme court amendment that forced the votes to no longer be recounted and for the presidency to be secured by a single man. Almost twenty years later, the same question is being asked: what would have happened if not for the supreme court ruling, and who would ultimately have won. The 2000 election is one of the 5 elections in the US, in which the candidate which won the popular vote did not end up winning the election, and of the last three presidents, two have lost the popular vote. This highlights the obvious issue with the electoral college system: someone can have more support and yet still lose the presidency to an illegitimate winner; legitimacy is conferred through popular support.
But the issues with the electoral college system run far deeper than the overall popular vote. Even inside states, the disparity in where the votes come from has significant effects on who ultimately wins elections. In California in 2016, despite Trump winning 25 counties and Hilary Clinton only winning 61.5% of the popular vote, she won all 55 delegates. Similarly, in Illinois, despite Clinton getting 3.1 million votes to Trumps 2.1 million, she won all the delegates in the state. Since 1824, 48 of the 50 US states use a winner-takes-all method of assigning delegates, making there be a huge issue in electoral politics, as millions of votes get disregarded across the US, since they are a numerically smaller number inside the state.
It is not enough to say in the abstract that the system should be changed, and it is instead necessary to see what other system would be better and fairer. As per Arrow’s impossibility theorem, it is impossible to design a voting system by which non dictatorship, pareto efficiency, independence of irrelevant alternatives, unrestricted domain and social ordering all exist. These tenants all seem very intuitively necessary, as non-dictatorship means no single voter should be able to decide who wins the election; pareto efficiency means that if every voter prefers one candidate to another, then they should always beat the other candidate; independence of irrelevant alternatives means that if you remove an irrelevant candidate then there should be no effect on ordering of the other candidates; unrestricted domain means that voting has to account for everyone’s preferences and social ordering means that every individual should be able to order their choices in anyway the feel necessary. This means that there are very few viable voting mechanisms. For example, a simple plurality vote means that whoever won the popular vote would win the election, but it has several drawbacks: it only allows you to vote for a single candidate, harming those who have similar policies, as they take votes away from each other, and also forces people to not vote for the favourite candidate, because they feel like they have no chance of winning. Similarly, other voting systems require a complete overhaul of the system and of federalism, by which states individually get delegates, much of which underpins US politics. No matter the system, it means that devolved government to the states would be on its way to removal, at the point at which the presidential election is no longer a state by state election.
What would be better however is rather than tearing down the system, simply rebuilding it. A simple plurality voting system within states evidently does not work, but there are many other options. For example, the ranked-choice voting system adopted in Maine in 2018 for their state-wide elections could be used for presidential elections state-wide, and then delegates could be assigned based on each state, as that would create a far fairer system. This would mean that scenarios like the election of Paul LaPage, who won the mayoral election with less than 40% and 50% of the vote in 2010 and 2014 would not have won, assuming that there was a ranked choice system. This would have allowed there to be a significant shift in policy, such as the adopting and expansion of Medicaid, which has been largely held in its tracks by him. This huge impact on public policy highlight how a ranked choice system could help solve several of the issues remaining. A system like the Borda count, where the candidates are ranked in order and get smaller scores the lower they place, could perhaps work as it rules out the main argument against ranked choice, which is how its hard logistically, since this relies on a single election rather than a series of run-offs. But even if there are a series of run-offs, is this really too big a price to pay for there to be truly democratic elections. Either of these methods could work state-wide to determine how many delegates get assigned to each party, and together with some reshuffling of how many delegates each state gets assigned, as the fact that a voter is Wisconsin gets three times the vote of an exactly similar voter in California seem ludicrous. The delegates should also be shared more equally throughout each state, in proportion to vote share, with perhaps a small bonus for winning, in a shared system nearer to Maine’s or Nebraska’s than the rest of the USA’s winner-takes-all system. The Borda Count, used in Slovenian presidential elections, could perfect fit into solving the electoral college system. It would mean that the nation would be able to better aggregate preferences, since at the point at which if 50% and a single voter love candidate A while everyone else hates him, while candidate B is the first choice of the remainder of the electorate and the second choice of everyone who wanted candidate A, then it seems reasonable to elect candidate B over candidate A. Only with a system as nuanced as the Borda count can appropriately deal with these preferences.
It is often said that the system should be one man one vote. But when one man one vote relies on breaking down the system and rebuilding a new one, destroying federalism in the process (the senate, where 2 senators come from each state and the house are both not elected 100% proportionally, and therefore could be considered unfair as one voter gets more of a say than another voter), which is even then unlikely to be the best system at aggregating preferences, then clearly this is not the best option. Instead, through working within the system, smaller scale reform of the electoral college system would work far better, allowing more equal representation and a fairer distribution of delegates, which gets shared by the winners rather than directly to one lucky individual who won by razor thin margins.