Electoral College Days
With yet another Presidential race yielding a victor who won the Electoral College but lost in the national popular vote, there is again talk about the need to abolish the Electoral College as the system the United States uses to elect its Presidents. There is a contradictory stance I’ve seen: on one hand, people are arguing that we should abolish the Electoral College, because it’s offering a victor different than the popular vote winner, and on the other hand, perhaps we could use the Electoral College itself to overturn last week’s outcome. The positions are not entirely contradictory though, if the idea is that the Electoral College, when it meets, should adjust its votes to reflect the popular vote total, thereby giving Hillary Clinton the victory. Putting aside the question of the desirability of the move in this instance, the more substantive question is whether the Electoral College ought to exist at all if we see it as only legitimate to the degree that it rubber stamps the winner of the popular vote. Though there has been talk that the Electoral College currently is biased in favor of Republicans, this is hardly a necessary feature. In 2004, Bush won Ohio by only a small margin. A small reversal in Ohio, enough to have given Kerry the state, wouldn’t have changed Bush as the popular vote victor, but it could have presented a dramatic reversal of 2000: Bush as the popular vote winner, with Kerry as the Electoral College victor, and therefore, the legal President. There are two separate issues here that need to be disentangled: first, is the Electoral College itself a good idea? Should it be abolished in favor of a direct national popular vote?
On the first, I’m basically a half-hearted supporter of the Electoral College. That is, I’m not strongly committed to it necessarily being the best way to run elections, and I’m open to alternatives if one can be offered that offers some great advantage. However, I’m not aware of any. There are some that may be as good, but given the costs that come with changing, there should be at least some great advantage in doing so, and that’s not what I’m seeing. The main reasons I like the Electoral College are not, I think, especially novel. They have to do with the structure of the United States as a political entity, being organized as a hybrid between a federal and a national system. In a country as large and diverse as the United States, such a system is an attractive means of dividing power and sovereignty. The Constitution establishes the United States divides the powers of government between a national government, given sovereignty over some matters, and state governments, given sovereignty over others. The states are not fully sovereign independent nations, but they do have autonomy and sovereignty over far more than is typical, say, of mere provinces in most other countries. So the United States is neither a fully federal body of sovereign states, like the UN or the EU, but neither is it a purely national entity, like France or Germany. This is reflected in the make-up of Congress as a bicameral body. The House represents the people according to population, and the Senate represents the states as co-equals. In the latter body, we have large states like California represented as equals with small states like Wyoming, just as Liechtenstein and Nauru are equals to India and China in the UN.
I mention Congress here also because this is what representation in the Electoral College is modeled after: every state guaranteed at least two votes for its two senators, and one vote for every House district. This is seen as unfair, in that, proportionially speaking, it makes the votes of people in small states like Wyoming, the District of Columbia and Vermont, carry somewhat more weight than the votes of people in large states like California and Texas. And thus, the logical possibility of, and actual reality of, a mismatch between the winner of the popular vote and the winner of the Electoral College, as we saw in 1876, 2000 and now 2016. However, note that this assumes a standard of legitimacy based on the popular model itself. That is, this is only a problem if one already assumes that legitimacy can only be achieved by securing the popular vote, on the nationalist model of political organization, and that the Electoral College itself is only good to the degree that it secures the same outcome as the popular vote. If one preferred the federal model of political organization, where all states vote as equal sovereignties, where we see nothing wrong with Nauru being as powerful as India, then we can find examples going the other way. I didn’t look at every election, but just going back as far as 1960, I noticed that in 1960 and 1976, a minority of states were able to win in the Electoral College. Specifically, in 1960, 26 states voted for Nixon, but Kennedy was given the victory by the Electoral College with 22 states, and in 1976, 27 states voted for Ford, but the Electoral College gave the victory to Carter with 23 states plus DC. That’s the nature of the Electoral College as a hybrid system: in close elections, more populous states, though fewer in number, can swamp the greater number of less populous states as easily as the reverse.
To get to specifics in the 2016 race, at the time that I’m writing this (late Sunday night, November 13th), I’m seeing that the current numbers being reported for the national popular vote are Trump at 60,350,241, at 47.30%, and Clinton at 60,981,118, at 47.79%, roughly a difference of 600,000 in favor of Clinton. If one only looks at this, it must seem puzzling that this results in Trump getting 306 Electoral votes, and Clinton only 232. But again, remembering the federal system of co-equal states we considered as an alternative, this difference seems much less mysterious if we consider that this outcome is generated by 30 states, plus a Congressional district in Maine, for Trump, as against only 20 states plus the District of Columbia for Clinton. For Clinton to win as the favorite of 20 states over 30 states would also seem strange if the states are supposed to be equals among each other. We would not consider the United Nations or the European Union to be democratic if 2/5 were able to overrule the desires of 3/5 of the member states. Indeed, smaller nations would be unlikely to even join such a union if the rules functioned this way. Perhaps such bodies would have larger nations like China and India, but even the United States would be unlikely to join if China, with its 1.3 billion citizens, had the ability to overwhelm the United States with its 300,000,000.
This is what I’m getting at with the notion of the Electoral College as a hybrid. Most of the time, historically, winners of elections by that system are going to coincide both in terms of the national popular vote and the majority of states. It’s only in close elections where occasionally there is a divergence in favor of either number of states (1960, 1976) or popular vote total (1876, 2000, 2016).
One might argue at this point, well, so much the worse for the federal co-equal state model. The United States may have once been a hybrid federal/national polity, but no longer. Before the Civil War, it was common to refer to the United States in the plural, as a union of states, and for citizens to think of themselves as Virginians, New Yorkers and so forth before being Americans, but the Civil War settled this issue once and for all: the United States is a single unitary nation-state. The Senate was once a body of members appointed directly by state legislatures, but they are now directly elected. So states, and state governments in particular, are in a clearly subordinate position institutionally speaking. So here, I would simply protest that the genius of the American system is its clearly divided powers and divided sovereignty. One need not be a fan of the retrograde “states’ rights” ideology to see that the great benefit of such a system is that it makes power difficult to centralize into a single pole, despite enormous political pressure to do so. The Electoral College is one part of this system, by having the President selected by a system which, in effect, splits the difference between the national and federal models of union. Abolishing the Electoral College would further diminish the importance of states as separate sovereign entities, leaving only the Senate itself as the last remaining fig leaf on a purely unitary nation-state model of popular sovereignty.
It might still seem like a matter of unfairness if numerically more people vote for Trump than Clinton. Let’s say that there is a normative pull toward legitimacy here being a matter of popular votes. Even granting that, however, we should not infer that the fact that the numbers favored Clinton in this election shows that the Electoral College is giving us an outcome at odds with that necessarily being true here. Consider an example: you and I play a game of chess. Suppose that we’re tightly matched, so you decide to make a few sacrifices to win. You give up more pieces, even your queen, but the end result is that you get my king into checkmate, and therefore win. Now, one way I could respond is like this: “Well, I ended the game with more pieces still on the board than you, and I still had my queen, so if you think about it, it’s just a technical formality that you won. I should be regarded as the ‘real’ winner, and it’s a sham game if you’re the winner after I still have more pieces in play.” But this would be churlish and more than a bit unfair. After all, the whole reason I played the way I did was because the rules define victory by the achievement of checkmate over the other player’s king. The number of pieces on the board, even who still retains a queen, doesn’t matter; checkmate does. If the rules had been that different, I wouldn’t have made those sacrifices, and I would have played very differently.
The same is true for the Electoral College. If the victor were a matter of achieving more popular votes, to use the 2016 election as an example, we would have seen the candidates adopt very different strategies. Trump, for example, would have campaigned heavily in conservative areas of California, downstate Illinois and upstate New York, to run up his numbers there, where it wouldn’t have made sense to do that in the status quo. Clinton, for her part, would have spent her campaign mostly in coastal urban areas, particularly in New York City, Los Angeles, Chicago, and other major urban areas across the country. Deeply red and deeply blue states alike would have both had much more robust GOTV efforts. It is difficult to predict how the national popular vote would have turned out had they done so, since the issues they campaigned on, and the number and type of voter who turned out, would have all likely been different.
One could also point to peculiarities of 2016 that would distort the overall total. In California, for example, Republicans likely had a very low turnout, not only because California was not in serious contention for its Electoral College votes, but also because in the only statewide race, for the US Senate, the two candidates were a Democrat against a Democrat. Had there been a national popular vote system, Trump very likely would have gotten hundreds of thousands, if not millions, of more votes out of California alone.
I could see the reply made here that this only proves the need to abolish the Electoral College. People in these areas are not inclined toward active voting in Presidential elections because they know that their votes don’t matter, so let’s make them matter by having a national popular vote election. That’s a fine argument as far as it goes; I pass no judgment on it here, though I have my doubts. But let’s say that we accept it; this would only provide a normative reason to adopt a new system. Under the status quo system, it would only tacitly accept the claim I’ve made, that one cannot look at the existing popular vote totals to assess whether the victor of the Electoral College is truly legitimate on the basis of popular sovereignty, because the existing popular vote numbers aren’t a reliable indicator of how people would have voted if victory was defined by the national popular vote. As with the chess game, where the moves made by the players were governed by the rules they were operating within, it would be illicit to deem one player as the “true” winner by an alternate set of rules they weren’t playing under.
There is a separate argument about political relevance: the Electoral College only makes the “battleground” states important in Presidential campaigns, making states like California, New York, Illinois and Texas mere bit players and onlookers. But certainly, smaller rural states would fade in importance in a national popular vote system, and it begs the question to say that the former is a better outcome than the latter. This may only be a matter of happenstance, but I note the “battleground” states in my adult lifetime have generally been quite representative of the whole. Major urban areas like Philadelphia, Pittsburgh, the Northern Virginia suburbs of DC, Milwaukee, Detroit, Las Vegas, Cincinnati, Miami and Denver are all included in battleground states. The Rust Belt, the Southwest, the Rocky Mountains, the Midwest, Mid-Atlantic, New England, even parts of the South are all included as well as deeply rural states like Indiana, Iowa and Wisconsin. These areas are also racially and culturally diverse as well, with the Hispanic and African American vote playing pivotal roles.
Second, an argument is made about the recount in Florida in 2000; for the election to be decided by a few hundred votes there, over hanging chads, also seems strange if it’s the case that the popular vote totals between Gore and Bush were strong enough to decide the election on Gore’s behalf. In reply, I’d note that the general effect of the Electoral College is to amplify a majority, and make it decisive. The Electoral College itself never needs a recount, but Florida in 2000 shows that the winner of a single, very close state could place the overall outcome in doubt. The argument was made in 2000, however, that the closeness of that race was relative to the total number of votes cast in Florida. Let us consider again the national popular vote in 2016 as a matter of percentages (again, with the caveat that these percentages are incidental, given the rules that governed campaign strategy and voter turnout): 47.79% vs. 47.3%. In any election where the margins are this close, we’d very likely have recounts nationwide. The chaos of the situation that emerged in Florida would instead be a nationwide affair in this scenario, as lost and absentee ballots were sought out, and hanging chads scrutinized in hundreds of counties. One benefit of the Electoral College was that the chaos was confined to Florida. Indeed, a similarly close election in New Mexico that same year was also subject to a recount, but conducted in a relatively orderly manner, as New Mexico’s contribution to the Electoral College was not going to be decisive relative to Florida’s. It seems that the Electoral College, if anything, makes the resolution of close elections a far simpler matter.