Can You Win the Presidency While Losing the Popular Vote?

Last updated 14 hours ago. Our resources are updated regularly but please keep in mind that links, programs, policies, and contact information do change.

On the morning of November 9, 2016, Donald Trump had a problem that was not a problem at all. He had just been elected president, yet nearly 2.9 million more Americans had voted for Hillary Clinton.

Both things were true. Both were legal. And neither canceled the other out.

So here is the direct answer to the question in the title: yes, you can absolutely win the presidency while losing the national popular vote. According to Ballotpedia’s tally of Electoral College splits, it has happened four times since the modern popular vote was first recorded, and each winner was a Republican: Rutherford B. Hayes in 1876, Benjamin Harrison in 1888, George W. Bush in 2000, and Donald Trump in 2016.

The reason is that Americans do not elect presidents with a national headcount. They never have.

The Constitution and federal law hand the job to the Electoral College, a system of 538 electoral votes divided among the states and the District of Columbia, where 270 wins. We walk through the full machinery in our guide to how the Electoral College works. What matters here is the consequence: because the presidency is decided state by state, the distribution of your votes can beat the total number of them.

Why the National Count Doesn’t Decide the Winner

Article II of the Constitution says each state appoints electors “in such Manner as the Legislature thereof may direct,” in a number equal to its House seats plus its two senators.

When you vote in November, you are not voting for a candidate. You are voting for a slate of electors pledged to that candidate. Those electors meet in their state capitals in December and cast the votes that legally count. As the Constitution originally put it in Article II, the electors “shall meet in their respective States, and vote by Ballot for two Persons.” The Twelfth Amendment replaced that two-vote procedure in 1804 with separate ballots for president and vice president, but the core design has never changed: electors, not the national tally, cast the votes that legally decide the presidency.

The national popular vote total is real, but it has no legal role. The Federal Election Commission, which publishes official popular vote numbers, does not even run the Electoral College. State election offices are responsible for certifying results, and the FEC notes that “the results of the elections are certified by the state government”, with the National Archives coordinating certain Electoral College functions between the states and Congress.

So the national tally is a sum of fifty-one separate certifications. It is a scoreboard nobody is playing to.

The full comparison of the two systems, and why they diverge, is the subject of our earlier piece on the Electoral College versus the popular vote.

The Mechanics of a Split: Where Votes Go to Waste

If most states award all their electors to whoever wins the state, then winning a state by one vote and winning it by three million votes buy you the exact same thing.

Only Maine and Nebraska break from this winner-take-all rule, splitting some electors by congressional district. Everywhere else, running up the score is pointless.

Every vote for the losing candidate in a state is wasted, because it earns zero electoral votes. And every vote for the winning candidate beyond the number needed to win is wasted too.

Picture two candidates. One wins ten states by a single point each. The other wins ten states by twenty points each. The second candidate could easily pile up more raw votes nationwide and still lose, because those twenty-point cushions do nothing.

Two structural features sharpen the effect. The first is the small-state bonus, the extra sway smaller states get. Every state gets two electoral votes for its senators no matter how few people live there.

Run the numbers and the imbalance is clear. Take a hypothetical small state of 600,000 people with three electoral votes: each electoral vote carries the weight of about 200,000 residents. A state of six million with ten electoral votes spreads each electoral vote across 600,000 residents. By that math, the small-state voter carries roughly triple the weight.

Applied to real states, the gap is starker still. Using Census apportionment figures, a voter in Wyoming, the least populous state, has close to four times the effective presidential weight of a voter in California, the most populous. The hypothetical triple and the real-world Wyoming-to-California gap describe the same effect at two different scales.

The second feature is the swing state. A handful of closely divided states decide the whole thing, which means a few tens of thousands of votes in the right places can outweigh millions in the wrong ones.

Put all three together, winner-take-all, the small-state bonus, and a few decisive swing states, and you have the recipe for a split. The candidate whose voters are efficiently spread across just enough states can lose the country and win the office.

The Four Elections That Split

The numbers behind each split are worth seeing side by side, because they reveal how modest the popular vote margins usually were.

Sources: National Archives (1876), 1888 results, FEC Federal Elections 2000, and FEC 2016 Presidential Popular Vote Summary. The 1876 electoral totals shown are the final counts certified after the Electoral Commission awarded the twenty disputed votes to Hayes; on election night the count stood at 184 for Tilden to 165 for Hayes, with 20 votes unresolved.

1876: The Split That Needed a Commission to Resolve

The 1876 race between Democrat Samuel J. Tilden and Republican Rutherford B. Hayes is the strangest of the four, because the country genuinely did not know who won for months.

On election night, the Miller Center reports, Tilden had 184 electoral votes to Hayes’s 165, with 20 votes disputed across South Carolina, Florida, Louisiana, and a single Oregon elector. In three of those states, rival slates of electors each claimed victory and each sent their votes to Washington.

Congress had two sets of returns and no rule for choosing between them. So on January 29, 1877, it passed the Electoral Commission Act, creating a temporary body consisting of five senators, five representatives, and five Supreme Court justices.

The commission’s job was to decide “the true and lawful electoral vote of such State,” and its rulings would stand unless both houses of Congress overrode them.

It voted 8 to 7, along strict party lines, to award all 20 disputed votes to Hayes. That gave Hayes exactly 185 electoral votes to Tilden’s 184, one more than he needed, as the National Archives records.

The cruelest detail: Tilden had won an outright majority of the popular vote, not just a plurality. He remains the only popular-vote loser to have done so, drawing roughly 4.28 million votes to Hayes’s 4.03 million out of an electorate near 8.3 million.

The resolution came wrapped in a larger bargain, the informal “Compromise of 1877,” in which Southern Democrats accepted Hayes in exchange for the withdrawal of federal troops and the end of Reconstruction. The mechanics of the split and the fate of Black voters in the South were tangled together from the start.

1888: The Cleanest Example of All

If 1876 was chaos, 1888 was almost clinical. President Grover Cleveland won more votes than Benjamin Harrison and lost anyway, with no disputed slates and no commission.

Cleveland took about 5.54 million votes to Harrison’s 5.44 million, a margin near 90,000, or roughly 0.8 percent. Harrison won the Electoral College 233 to 168.

The whole thing turned on New York. Harrison carried Cleveland’s home state by 1.09 percent, and that narrow win delivered a decisive bloc of electors. Flip New York and Cleveland wins 204 to 197.

Cleveland’s national vote lead depended partly on Black citizens in the South being blocked from voting, most of whom favored Harrison, with Republicans of the era pointing to the suppression of hundreds of thousands of Black voters. The popular vote reflected only the ballots officials permitted, not what all eligible citizens wanted.

2000: 537 Votes in Florida

Most Americans alive today first learned that splits were possible on a November night in 2000, when the race between George W. Bush and Al Gore came down to a single state and would not resolve for five weeks.

Gore led the national popular vote by roughly 540,000, about half a point, with just under 51 million votes to Bush’s 50.5 million, per the FEC’s Federal Elections 2000 report. But the Electoral College stood at 271 to 266, and everything hinged on Florida.

Bush’s certified Florida margin was 537 votes out of about six million cast. Those 25 electoral votes were the presidency.

The recount fight climbed to the Supreme Court, which decided Bush v. Gore on December 12, 2000. The Court held that Florida’s recount, with different counties applying different standards for reading a ballot, violated the Equal Protection Clause, the constitutional promise of equal treatment, and that no compliant recount could be finished in time.

The right to vote rests on giving each ballot equal weight and each voter equal dignity, according to the opinion as published by Democracy Docket, which closed the door on the recount: “Because it is evident that any recount seeking to meet the December 12 date will be unconstitutional for the reasons we have discussed, we reverse the judgment of the Supreme Court of Florida ordering a recount to proceed.”

The Court was careful to limit how far its ruling could apply, stating in the opinion as published by Democracy Docket that its consideration was limited to the present circumstances, since equal protection in elections generally raises far more complicated questions. That single line has fueled two decades of argument about what the case stands for.

2016: The Widest Gap Yet

The 2016 split was the largest in raw terms. Clinton won the national popular vote by nearly 2.87 million, about 2.1 percent, with roughly 65.85 million votes to Trump’s 62.98 million, per the FEC’s 2016 Presidential Popular Vote Summary. No losing candidate had ever led by so many votes.

The geography tells the story. Clinton won California by more than four million votes. Trump, meanwhile, took Michigan, Pennsylvania, and Wisconsin by a combined margin under 80,000 votes, three states that had leaned Democratic and that together delivered 46 electoral votes.

The official count was Trump 304, Clinton 227, after a scattering of faithless electors in Hawaii, Texas, and Washington nudged the totals off the expected 306 to 232, the National Archives recorded.

The reactions split along predictable lines. Senator Barbara Boxer of California introduced federal legislation to abolish the Electoral College the week after the vote, as CNN reported. On Fox News, former House Speaker Newt Gingrich defended the outcome with a football analogy, arguing that Clinton’s popular-vote lead was like a team gaining more yards but losing, because the points that count are electoral votes.

The Near Misses, and Why Landslides Almost Never Split

Four splits in roughly two centuries makes them rare. But the system has come close to producing more, which tells you the four were not flukes.

Consider 2004. Bush beat John Kerry in the popular vote by about three million, a comfortable margin. Yet Bush won Ohio by only about 118,000 votes out of 5.6 million, and Ohio’s 20 electoral votes were the whole ballgame, per the National Archives.

A swing of roughly 60,000 votes in Ohio would have handed Kerry the presidency while Bush still led the national vote. That would have been a fifth split, running in the opposite partisan direction from all four real ones.

Splits cluster tightly around close elections. All four real ones happened with popular-vote margins between about 0.8 and 2.1 percent in the modern cases, never in a landslide.

The reason is mathematical. For a candidate winning 60 percent of the national vote to lose the Electoral College, their surplus votes would have to be almost impossibly concentrated in a few states while the opponent squeaked out narrow wins nearly everywhere else. Real electoral maps do not bend that far. A candidate who wins big nationally tends to win a lot of states across a lot of regions.

So the practical takeaway is simple. Watch the close ones. When national polls show a race within a couple of points, a split is live. When one side is up by ten, the Electoral College will almost certainly agree with the country.

Does the System Favor One Party? It Has Cut Both Ways

All four splits favored Republicans, and the past three decades have leaned the same direction, which has hardened a widespread belief that the Electoral College is a permanent Republican thumb on the scale. The honest answer is more interesting than that.

The recent tilt is real. Analyzing the 2020 result, the University of Virginia’s Center for Politics found a built-in edge of about 3.8 points for Republicans: Democrats won the popular vote by 4.5 points but carried the tipping-point state, the one that pushed the winner past 270, by only about half a point, the Crystal Ball analysis calculated. A 2020 study in the Proceedings of the National Academy of Sciences by Robert Erikson, Karl Sigman, and Linan Yao reached a similar conclusion, placing the Electoral College’s balance point at roughly 51 percent Democratic rather than a true 50-50, and dating the clear tilt only to the last decade.

That Republican edge is not unprecedented, though. In 1948 the map pointed the same way: Democrats won the popular vote by about 4.5 points yet carried the tipping-point state by only half a point, a roughly 4-point pro-Republican bias that mirrors 2020 almost exactly. The difference is that the modern tilt has proven far more persistent.

Run the clock back far enough, however, and the advantage genuinely flips. An NBER study of electoral inversions found that in the post-Reconstruction era it was Democrats who held the Electoral College advantage. Southern states were given electors based on populations that included disenfranchised Black residents, while the actual voters were overwhelmingly white Democrats.

The system itself has no favorite, even when the results seem to. In 1964, Barry Goldwater ran up enormous margins in a few Deep South states, wasting Republican votes in a way that effectively handed Democrats a bias that year. The party that suffers is whichever one clusters its voters too tightly.

Stanford’s Jonathan Rodden traces why the map now hurts Democrats: the link between how crowded an area is and how it votes Democratic was weak in 1916, moderate by 1960, and steep by 2016, his research shows. Democrats now pile up wasted votes in dense cities. That is a feature of current geography, not of the Constitution, and geography moves.

The Fight Over Whether to Change the Rules

If the rules produce outcomes that unsettle a lot of voters, why not change them? That fight is live right now, and it is more constitutionally tangled than either side’s slogans suggest.

The leading workaround is the National Popular Vote Interstate Compact, an agreement among states to award their electors to the national popular vote winner. It takes effect only once member states control 270 electoral votes. As of 2026, eighteen states and the District of Columbia have joined, including California, Illinois, New York, and Colorado, together holding 222 electoral votes, still short of the 270 the compact needs to activate.

That partisan clustering is exactly what skeptics point to. Derek T. Muller, a law professor at the University of Notre Dame, argues the compact is a “political compact” that requires congressional consent under the Compact Clause, a constitutional rule requiring Congress to sign off on agreements between states, because it boosts member states’ power at the expense of everyone else.

The politics are shaky too. Nevada’s Democratic Governor Steve Sisolak vetoed his own party’s bill to join in 2019, warning, as CBS News reported, that it could diminish the role of smaller states like Nevada in national elections. And in 2025, Colorado lawmakers introduced HB25-1102 to repeal the state’s participation, though the measure died in committee within days. Even members are reconsidering.

Defenders of the Electoral College are not merely defending Republican wins; they are defending a theory of the presidency. Tara Ross, an attorney and author who champions the system, argues it creates 51 separate elections that keep fraud and recount disputes contained within individual states rather than triggering a single nationwide recount. Hans von Spakovsky of the Heritage Foundation has written that a national popular vote would “diminish the influence of smaller states and rural areas of the country.”

Gary L. Gregg II, who holds the McConnell Chair in Leadership at the University of Louisville, framed the stakes in vivid terms. In a Politico essay, he argued that Barack Obama’s 2012 national vote margin was roughly the size of his combined margins in just four cities: Chicago, Philadelphia, New York, and Los Angeles. A national vote, he contended, would let dense metros set the agenda.

That is the unresolved core of the debate, and it is not going away.

Which is why the real test is not academic. It is arithmetic. As suburban voters shift party loyalties, the map of where votes count most is already shifting. The next split, if it comes, may not favor the party the last four did, and that possibility alone is quietly reshaping how each side talks about a system it once took for granted.

Our articles make government information more accessible. Please consult a qualified professional for financial, legal, or health advice specific to your circumstances.