The States Choose the President

The companion to this essay argued that America is a republic and not a pure democracy, that a written law sits above the majority, and that the structures people call undemocratic are usually the republic doing its work. Beyond the Reach of the Vote made the case in general. Nowhere is it clearer, or more furiously denied, than in the way we choose a president. So take the most attacked institution in American government and look at what it actually is.

Each state chooses a number of electors equal to its whole representation in Congress: one for every member of the House, which tracks population, plus one for each of its two senators, which every state holds in equal measure. The District of Columbia was granted three by amendment. Add them up, reach for a majority, and you have chosen a president. The design is not subtle once you see it. The presidency is elected on the same blended principle as the Congress that makes the laws, population in the one part and equal statehood in the other, the House and the Senate fused into a single count. It is bicameralism carried into the executive. I have argued elsewhere, in The States’ House, that a bicameral legislature needs two different principals to be worth having. The Electoral College gives the presidency the same two principals at once.

The framers had three doors and walked through the third. They could have let Congress pick the president, and rejected it because an executive who owed his chair to the legislature would never be independent of it, and the separation of powers would have collapsed on the first morning. They could have held a single national popular election, and rejected that as well, for reasons partly of their moment and partly permanent. The permanent reason is the one that still holds. This was not a nation of individuals who happened to share a landmass. It was a union of states, real political communities that existed before the federal government and agreed to enter it on terms. A president chosen by the states, as states, was the executive that fit the thing they were actually assembling. So that is the one they assembled.

This is the heart of the whole defense, and everything after it is detail. We are the United States, and the states are not administrative provinces drawn for the convenience of a central office. They are the parties to the constitutional bargain, the units the union is actually made from. When a union of that kind elects a chief executive, the honest question is not which candidate has the most individual admirers scattered across the continent. It is which candidate has built the broadest support across the federated communities the union is composed of. Those are different questions, and the Electoral College is built to ask the second one on purpose. A candidate cannot win by running up enormous margins in a handful of dense metropolitan areas and writing off the rest. He has to assemble a coalition that reaches across regions and states, cities and countryside, one end of the union to the other. The system rewards breadth of support over sheer depth in a few places, and it does so because breadth is what holds a continental union of very different states together.

The standard objection is a moral one, and it deserves a straight answer: why should a vote cast in Wyoming carry more weight toward the presidency than a vote cast in California? Put that way it sounds indefensible, and if the presidency were meant to be a single national headcount, it would be. But it is not, and the mismatch is not a bug that slipped in unnoticed. It is the federal principle showing through, the very same principle that gives tiny Wyoming and enormous California two senators apiece. No one who understands the Senate calls it a scandal that a Wyoming senator answers to fewer people. That overweighting is the deal, the standing promise that the small states will not simply be ruled by the large ones with no voice of their own. The Electoral College carries the identical promise into the choice of president. You may argue the promise was a mistake. You cannot treat it as an accident, because it is the most deliberate feature in the document.

Then there is the objection that actually moves people, the one that arrives after an election in which the candidate with the most votes nationwide loses the presidency. It has happened, in 2000 and in 2016 and three times in the eighteen hundreds, and each time it feels like a wound to something sacred. I understand the feeling, and I think it rests on a mistake. There is no national popular vote in the American system, which means there is nothing to win or lose at that level. The national tally is a figure somebody adds up afterward, not the contest anyone entered. Candidates campaign to win states under rules known to everyone in advance. If the rule were instead a single national vote, both campaigns would run completely differently, chasing raw turnout in every crowded place they now ignore and neglecting the states they now fight over, and the national totals we point to would not survive the change that produced them. To say the popular-vote winner “really” won is to score the game by a rule neither side was playing under. And beneath that lies the deeper point the companion essay made: raw national majority was never meant to be the one thing that confers the presidency. A republic of states does not award its executive to whoever gathers the largest single pile of individual votes, any more than it awards the Senate to the party with the most voters counted nationwide.

Now let me do the thing the loudest defenders of the College refuse to do, which is admit where it genuinely fails, because the principle is strong enough that it needs no one to pretend the machinery is flawless. It is not flawless.

The most serious practical complaint is simply true. In practice a small number of closely divided states draw nearly all the attention, while the rest, safely one color or the other, are taken for granted, so a system meant to force breadth ends up concentrating the entire contest in a few battlegrounds. But notice where that defect comes from. It comes from winner take all, the rule that hands every one of a state’s electors to whoever carries the state by even a single vote, and winner take all is nowhere in the Constitution. It is a choice each state makes for itself, and two of them, Maine and Nebraska, already choose otherwise and split their electors. The swing-state distortion is a flaw in how the states have elected to allocate, not in the federal principle of state allocation, and the states could repair it tomorrow without touching the College at all.

The genuinely dangerous flaw is the one almost no one argues about. If no candidate wins a majority of electors, the choice falls to the House of Representatives, where it is decided not by the members but by state delegations, one vote per state, so that the delegation of a state of forty million and the delegation of a state of six hundred thousand count exactly alike. That mechanism chose a president once, in 1824, and produced a result half the country regarded as a corrupt bargain. In a fractured election with a strong third candidate it could do the same again, and it is a far weaker link than anything the College’s critics usually name. A serious defender of the Electoral College should be the first to say that this contingency needs rebuilding.

Which points at the honest position, and it is not the reflexive one on either side. The principle the Electoral College embodies, that a union of states chooses its executive through a mechanism weighted by statehood and not by national headcount alone, is not an antique to be scrapped. It is required by what this country actually is, a federal republic rather than a unitary mass democracy. But the specific eighteenth-century machinery that carries the principle is improvable, and pretending otherwise surrenders ground that need not be surrendered. Fix winner take all. Rebuild the House contingency. Bind the electors, which the Court has now confirmed the states may do. Do all of it. What you may not honestly do is abolish the principle by stealth. There is a live effort to do exactly that, an interstate compact whose member states would pledge their electors to the national vote winner and so neutralize the College without ever amending the Constitution, the moment enough states join to control a majority of the electors. Set aside whether you like the goal. The method is the same drift this country keeps mistaking for reform, a foundational structure quietly rewritten by a workaround instead of by the amendment process the Constitution provides for this very purpose. If the people want to trade the federal republic for a national democracy, that is their right, and the door is Article V. Walk through it in the open. Do not pick the lock.

Two objections need meeting head on before I finish. The first is that the Electoral College is a relic of slavery, tainted at the root and disqualified by its origin. The historical claim is not baseless and I will not wave it off. The College interacted with the three-fifths clause, which inflated the electoral weight of the slave states by counting people who were given no vote, and the founding was shot through with concessions to that evil. But an institution is not refuted by the worst company it once kept, or half the Constitution would fall beside it. The federal principle underneath the College, that the states are the electoral units of the union, stands entirely apart from slavery. It is the same principle that seats two senators from every state, which no one calls a slaveholders’ relic. Judge the principle on its own terms, and on its own terms it is about federated consent, not human bondage, and it would do its work identically in a country that had never held a slave.

The second objection is aimed the other way, at the people who defend this institution as a wall the countryside throws up against the cities, a way to keep the clever few in the metros from ruling the virtuous many outside them. That is not the argument, and its resentful, anti-urban version is worth disowning as plainly as the abuse on the other side. The Electoral College is not a verdict that a rural vote is worthier than an urban one. It is a structure for federated consent among communities, and the communities it protects are not ranked by virtue or by soil. The coalitions it forces a candidate to build will flip over time, as they always have. The party that curses the College in one decade blesses it in the next. Defend the principle because it is right for a union of states, not because it happens, this cycle, to favor your side. A principle you love only when it wins is not a principle. It is a preference wearing a principle’s coat.

So here is the whole of it. The Electoral College says out loud what the Senate says, and what the amendment rule says, and what the entire compound structure says: that this is a union of states and a republic of laws, not one undivided national democracy taking a headcount. That is not a defect the founders failed to catch. It is the country they were building, drawn into the way it chooses its head of state. If you want a national plebiscite for president, what you want is not a better version of this republic. It is a different country, unitary where this one is federal, and you are owed the honesty of saying so and the discipline of amending the document rather than hacking it. Defend the principle, which is sound. Fix the machinery, which is not sacred. And keep the two apart, because almost no one in this argument manages to.