Why Do We Still Have the Electoral College?

Historian Alexander Keyssar on why the unpopular institution has prevailed 

A protester holding a sign that reads "ELECTORAL COLLEGE = VOTER SUPPRESSION" on November 13, 2016.
Protestors demonstrate against president-elect Donald Trump on November 13, 2016, in Philadelphia. Photograph by Mark Makela/Getty Images

The title of Alexander Keyssar’s new book—Why Do We Still Have the Electoral College?—is also, he says, the question Americans ask themselves every four years. The Stirling professor of history and social policy at the Harvard Kennedy School recently spoke with Harvard Magazine about the book and about why the Electoral College, an institution that’s been unpopular almost from the moment it was founded, has prevailed to this day.  


Harvard Magazine: What’s the origin story for your new book? How and why did this project start? 

Alexander Keyssar: The book really began, I think, after the 2000 election, when the winner of the electoral vote received only a minority of the popular vote. I began to wonder why we still have the Electoral College, what had prevented its reform or abolition. After doing a bit of reading and research, it seemed that the most standard answer to that question—that small states prevented reform—simply was not accurate. As I read about past efforts at reform, I also came to realize that those efforts were very extensive and had been going on for a very long time, which made the question all the more compelling. The question that is the title of the book is both a scholarly, analytic question, and a question about public understandings. It’s a question that several hundred million Americans ask themselves every four years. 

There is no right to vote for president guaranteed in the United States Constitution.

What do you see as the connection between this book and your earlier work, much of it focused on voting rights? 

Interestingly, when I wrote the first edition of The Right to Vote, I said very little about the Electoral College, which then troubled me because that book was published in 2000—just before the election, in which the Electoral College played a significant role. But the Electoral College just wasn't on my mind. 

But the 2000 election made me think about what voting rights meant, given the peculiarity of the presidential election system, of the Electoral College. There was also another very specific link, which was something that arose during the dispute over the 2000 election: it became clear that individual state legislatures retained the right to choose electors by themselves, without holding popular elections. One of the Supreme Court justices—I think it was Justice Scalia—in the Supreme Court case in 2000 pointed out there is no right to vote for president guaranteed in the United States Constitution. So that is the intellectual link between the two books.

More than 1,000 constitutional amendments have been introduced to change the Electoral College. What are the defects in the system that have caused people to want to reform it for almost its entire existence? 

The Electoral College began to be unpopular within a few years after it was adopted. There were a number of features of [its] design that were regarded as problematic, and the foci of concern changed over time. In the early years, the central concern was the fact that each state could decide how to choose electors, which meant that it could choose all its electors by “winner-take-all,” it could have an election that allocated electors by [congressional] district, or it could have no election at all, and the state legislature would just choose electors. All of those things happened. Moreover, by the 1790s, different states started changing how they were going to choose electors from one election to the next, to game the system in a partisan fashion. By the time you get into the 1810s-1820s, you have a presidential electoral system in which there was no uniformity from state to state, and each state’s processes could (and did) change from election to election. That was regarded by large numbers of people, including members of Congress, even James Madison, as no way to run a railroad. (Admittedly, there were no railroads at the time!) What resulted was a desire to create uniformity and stability. Most leading politicians also thought that there should be district elections for electors, and the Senate approved constitutional amendments for district elections four times between 1813 and 1826. 

One set of problems, thus, had to do with the variability in how each state chose electors. That variability comes to an end by the 1830s, when just about all states adopted a winner-take-all system. Since then, there have been repeated and insistent objections to the use of winner-take-all. People just don’t think that it’s appropriate or fair to have a system in which a candidate who gets 52 percent of the popular vote in a state receives all of the electoral votes. That feature is thought of as undemocratic; and it deforms election campaigns. Other critics believed that the presence of intermediaries, meaning the electors themselves, was profoundly undemocratic and hazardous, because the will of the people could be ignored or overridden. As American political values on the whole became more democratic, the presence of any kind of intermediaries and any weighting of votes so that not all votes counted equally began to seem more and more objectionable. 

What would be the difference between district elections, which you mentioned a moment ago, and what we have now? Would electors be assigned on a district level rather than the state level? 

Exactly. Electors would be assigned according to who won the popular vote in each district, which meant that individual states would likely not wield large blocks of electoral votes. One worry about district elections, in the nineteenth century and now, is that they could readily import the problem of gerrymandering into presidential elections. 

Let me go back to your larger question of what were the things people objected to in the early decades. There was also considerable concern, as there has been only occasionally since then, about a feature of the system that’s called the contingent election system. The Constitution requires that the president receive a majority of electoral votes. If no candidate gets a majority, then the election gets turned over to the House of Representatives, using an odd formula in which every state delegation gets one vote. That part of the system greatly advantages small states because every state, no matter its size, had the same weight. This system was used in 1800, it was used in 1824, and a lot of people believed that it would be continue to be used very frequently. So that was another source of concern and desire for reform, to get rid of the contingent election system and if possible, to figure out some way to not have Congress involved at all in choosing a president. 

That particular problematic feature of the Electoral College is one that we do not pay much attention to now, because it hasn’t been deployed in a long time. But there was a lot of attention focused on it as recently as 1992, when Ross Perot ran for president. It looked for a while as though he would get enough electoral votes to toss the election into the House.

How did we get the present-day winner-take-all system? Isn’t that at odds with the reason the Electoral College was invented in the first place?

Winner-take-all was not really envisioned by the framers. When the Constitution was written, there were no political parties. It was written in the belief that there really would not be political parties and that that there should not be. If you think about the design of the Electoral College and the need to have a majority to win an election, a lot of people thought that what would happen would be that states would cast their electoral votes for four or five or six different people. Nobody would get a majority, and then it would go to the House of Representatives and they would sort it out. 

Winner-take-all emerges out of the dynamics of partisan competition. There was a famous incident in 1800, following in the wake of the previous election in 1796, when Thomas Jefferson had lost the presidency to John Adams very narrowly. Virginia’s electors were chosen through a district system, and there had been one or two Virginia electors who voted for Adams. The votes that Adams received from Virginia helped him to win the election. When the 1800 election was approaching—a replay between the same two candidates—Virginia changed its system to winner-take-all, to ensure that Jefferson would get all of [its] electoral votes. Massachusetts retaliated by doing something similar, and that same partisan logic begins to take over, extending to all states by the 1830s. The dominant political party in each state preferred having winner-take-all. Everybody at the time said that they would be happy to go to a district system if everybody else would. But they weren’t going to do it by themselves. That stance has reappeared often in more recent history.

Can you think of any society whose election system works in a way remotely comparable to ours?

I can’t say that there isn’t one or that there hasn’t ever been. But our electoral system is just absurdly complex and distanced from its original design and the political world that it was designed for. I don’t know of any system now that is even remotely as complicated. The complexity of the system, we don’t often think about—we often tend to think that the problematic features of it are winner-take-all, and the fact that the small states get a slight advantage in the number of electoral votes. In fact, there are other problematic features as well, such as the ability of the state legislatures to change the process and the contingent election system. To some extent these different features were designed to offset or counterbalance one another. The very complexity of the system has meant that it’s extremely difficult to change any one piece of it without changing something else. 

You’ve talked and written elsewhere about the possibility of enacting a federal law or even a constitutional amendment enshrining the right to vote. If we had such a law, could that then be the basis for challenges to the Electoral College in some way?

A federal law would not have any impact on a constitutional provision. A constitutional amendment that guarantees every American citizen of voting age the right to vote for president would in effect create a conflict between two features of the Constitution. It would create a conflict between that amendment and the current design of the Electoral College, as outlined in Article II and the Twelfth Amendment. So a national right-to-vote amendment would—at the least—compel modification of at least one piece of the Electoral College, which is the piece that says that each state legislature can decide by itself the manner in which electors will be chosen.

The number of electoral votes a state gets is not entirely in proportion to its population relative to other states. Is that a modern problem? 

That’s in the Constitution. The number of electoral votes a state gets is proportional to its number of representatives and senators. At the Constitutional Convention in 1787, they had a lot of trouble figuring out how to choose the president. They didn’t have any models. They really just could not agree. The default position was that Congress would choose the president. And in a series of straw votes in the course of the summer, that was the position that had the most backing. But then within a day or two or three of these straw votes, they realized that that was actually not a good idea. Because then you don’t have separation of powers. And then the president would be beholden to Congress and that arrangement would set up all sorts of corruption.  

The framers floated a whole lot of other ideas, including having the governors choose or having a national popular vote, which James Madison supported. But they couldn’t reach agreement. It was a long hot summer in Philadelphia, everybody was tired, and at the end of the summer the convention went on vacation for a week. They left a small set of knotty, unresolved issues to a committee, the Committee on Unfinished Parts. It’s that committee which comes up with the design of the Electoral College, which, with modifications, is then adopted by the entire convention. One way to understand what the Electoral College is, by the way, is that it’s a replica of Congress in its composition. The same number of representatives and senators from each state. It’s a replica of Congress, but it does not legislate. It only meets once and only has one order of business to do, and thus the problems of corruption and lack of separation of powers are solved. What that institutional design did from the outset was give a slight advantage to small states. 

The perversity of the design of the Electoral College is that a state gets the same number of electoral votes regardless of turnout in elections and regardless of how much voter suppression it engages in.

You mentioned just now the idea of a national popular vote. Your book discusses how, before the Civil War, a national popular vote was a nonstarter for slaveholding states because, obviously, slaves wouldn’t be allowed to vote, and the voting power of southern states relative to other states would be diminished. What about after the Civil War? Why and how did race continue to play a role in opposition to a national popular vote?

After the Civil War, after a period of contestation and Reconstruction, white supremacist governments returned to power in the South, and they disenfranchised African Americans who had been voting and wielding some political power. Thereafter the South, in terms of representation, benefits not from the [Constitution’s] three-fifths clause, but effectively from a five-fifths clause. In other words, southern states get electoral votes in proportion to their entire black populations, but that population is not allowed to vote. Given that situation, the white South is fiercely resistant to a national popular vote, because it would do one of two things or perhaps both. It would put pressure on a state to try to maintain its political weight by enfranchising African Americans, which they did not want to do, or those states would lose influence in presidential elections, significant influence, because they would have many fewer voters. Under a national popular vote, the influence of a state is determined by the number of voters who cast ballots; under the Electoral College, it is determined by the state’s population. 

The perversity of the design of the Electoral College is that a state gets the same number of electoral votes regardless of turnout in elections and regardless of how much voter suppression it engages in. There was only one southerner from 1890 to the 1960s who introduced an amendment calling for a national popular vote, and he ends up being pilloried for it. By the late 1940s, it’s become a common view that the Electoral College is key to protecting the “southern way of life.” In the end, at the moment at which the United States comes closest to adopting a national popular vote, which is in 1969-70, a constitutional amendment passes the House by an 82 percent vote, but when it gets to the Senate it is defeated by a filibuster by southern segregationist senators. The amendment’s sponsors needed a two-thirds vote to end the filibuster, which meant that all it took was 34 senators to stop the national popular vote amendment.  

I don’t think most people know that we almost got rid of the Electoral College in the late ’60s. 

I agree.

Do you have a point of view about whether and how we ought to reform the Electoral College?  

My view is that our electoral system should embody certain key values. One is that all votes should count equally. No matter where you live, no matter who you are, all votes should count the same. The system should also be stable in the sense of not changing, not varying from one election to the next. And it should be transparent. The principles on which it’s constructed should be transparent and widely regarded as fair by the public. Based on those values, I certainly think that the ideal system would be to have a national popular vote.  

Until the 2000 election, it was widely believed that if there was a “wrong winner” election, we’d get rid of the Electoral College.

Do you think that if we continue seeing “wrong winner” elections, it will eventually force a full-blown legitimacy crisis? How long can that keep happening without something breaking? 

Very well said. Until the 2000 election, it was widely believed that if there was a “wrong winner” election, we’d get rid of the Electoral College. That’s what everybody was saying in the 1970s, 1980s, 1990s. Well, we didn’t get rid of it after 2000, and then, of course, we got 2016—when the gap between the popular and electoral votes was much larger than in 2000. I think that with the election of President Trump, and in part because of the nature of his presidency, we already have the beginnings of a legitimacy crisis. I think it would be intensified if it were to happen again. In the last few pages of the book I ruminate about this a bit. In an era of very sharp political polarization, you want to have an electoral system that is understood to be principled and fair. A lot of people doubt that now. 

To bring current events back full circle to some of the earlier history from your book, I was struck by some rhetoric of segregationist southern politicians about what would happen if there were a national popular vote. They’d say that their votes were going to be “debauched” by the votes of undesirables (i.e., black voters). It’s impossible not to connect that to present-day rhetoric about voter fraud. 

Yes, I think there is a direct line that goes from those kinds of sentiments to some of the support for Trump. 

Anything else you’d like to add?

I want to emphasize that one of the discoveries that propelled me forward in [writing] the book, and that I think is very important, is that—contrary to conventional wisdom—the primary obstacle to reforming the Electoral College has not been the small states. The “small state” explanation has been conventional wisdom for about 40 years, and it was widely believed in the 1940s and 1950s as well. This view that the small states have prevented reform because of the additional electoral vote weight that they get is simply not true. It’s not true in terms of who the leaders in the movements for change were, and it’s not true in terms of the roll call votes when proposals for a national popular vote came to a vote in Congress.

Read more articles by Marina N. Bolotnikova

You might also like

Former Women’s Hockey Coach Sues Harvard

Katey Stone alleges gender bias in handling of abuse allegations that led to her retirement.

Remembering Helen Keller and Anne Sullivan

On a Radcliffe-Harvard memorial to remarkable figures

Harvard Confers 11 Undergraduate Degrees

Protestors now found in “good standing.”

Most popular

Former Women’s Hockey Coach Sues Harvard

Katey Stone alleges gender bias in handling of abuse allegations that led to her retirement.

Mechanical Intelligence and Counterfeit Humanity

Reflections on six decades of relations with computers

Remembering Helen Keller and Anne Sullivan

On a Radcliffe-Harvard memorial to remarkable figures

More to explore

Broadway Director from Harvard Adapting Disney

Broadway music director Madeline Benson on art and collaboration

How Political Tension on Campus Creates Risk Aversion

How overheated political attention warps campus life

Harvard Professor on Social Psychology for Understanding War

Two scholars’ extracurricular efforts in the Middle East