FORTY-THREE STATES DO NOT VOTE FOR PRESIDENT. THEY WATCH.

FACT: Twice in living memory, and five times in American history, the candidate with fewer votes became president. In 2016 the margin was roughly 2.9 million votes in the other direction. That is the objection everybody knows, and it is the weaker one.

The stronger objection

Here is the stronger objection. Because 48 states award their electors winner-take-all, a vote only matters if it might change which pile a state lands in. In practice that is about seven states. Campaigns are not being cynical when they ignore the rest; they are being rational. A dollar or an hour spent in a state whose outcome is already known buys nothing.

So a Republican in California, a Democrat in Utah, anybody at all in New York or Oklahoma or Alabama, is — for presidential purposes — a spectator holding a commemorative ticket. Their vote is counted. It cannot alter anything. Turnout in those states could double or halve and the electoral result would not move by one vote. Forty-three states are visited for money and left.

Two mechanisms that deserve more alarm

The system does not produce geographic breadth. It produces seven states and forty-three flyovers.

The usual defence is federalism: without it, candidates would campaign only in cities and ignore rural America. Look at what actually happens. Under the current system candidates ignore rural America and urban America, in forty-three states, and concentrate on a handful of suburban counties in Pennsylvania. Whatever the Electoral College is protecting, it is not geographic breadth — a national popular vote would at least make a vote in rural Idaho worth exactly as much as one in Philadelphia, which is more than it is worth now.

Two further mechanisms deserve more alarm than they get. First, the Constitution contains no affirmative right to vote for president — Article II assigns the choice of electors to state legislatures, and the popular vote is a custom those legislatures grant. Second, if no candidate reaches 270, the Twelfth Amendment throws the election to the House voting by state delegation — where Wyoming and California each get exactly one vote. A contingent election is one of the most dangerous devices in the document, and it is one close three-way race away from being used.

OPINION: Every other elected office in the United States — governor, senator, mayor, dog-catcher — goes to whoever gets the most votes. Only the presidency uses a mechanism the Framers designed for a country of four million people without national parties, national media, or a national electorate, and which they expected would usually fail and throw the choice to Congress.

The strongest case for the Electoral College, taken seriously

The federalism argument is real: the United States is a union of states, presidential elections have always been state-administered, and a national popular vote would require either federal control of election administration or trust in fifty different systems to produce one comparable number. There is a practical point too — a close national election means a national recount, whereas the current system contains disputes inside one or two states. And small-state protection is not nothing in a country where geography and interest genuinely diverge.

The trouble is that the system does not deliver what its defenders claim. It does not produce geographic breadth; it produces seven states. It does not contain disputes; it concentrates the entire national outcome in whichever state happens to be closest, which is a recount risk multiplied, not contained.

What would fix it, and who is stopping it

Abolition requires an amendment, which is functionally unavailable. But two changes are not blocked. States controlling a majority of electors can join the National Popular Vote compact, which takes effect at 270 and requires no amendment. And uncapping the House would shrink the small-state bonus, since electors equal House seats plus two — meaning the 1929 statute quietly distorts presidential elections as well as legislative ones.

The honest summary is not that the Electoral College is unfair to one party. Its bias has shifted between parties across two centuries and will shift again. It is that a system in which most citizens are structurally irrelevant, by design, before any campaign begins, cannot keep asking those citizens to believe their participation matters. Every vote counting equally is not a radical proposal. It is the proposal.