Can a president win the electoral college and lose the popular vote?

Let me preface this by saying I'm not American (but I'm British so I can't be too high and mighty about voting systems -Brexit).

Last chance I got to look at the popular vote figures, Hillary is amazingly close (<100k votes) behind Trump. But I realise that irrespective of how many votes she gets from here on, Trump has won. So my question is:

Has there ever been an elected president who was not first by measure of popular vote (not needing 50%, just more than anyone else)?

I'm assuming there's not a long since forgotten rule that makes this an important question, but it's interesting nonetheless.

• In addition: Hillary currently has 0.6% more votes than Trump, according to the popular vote. nytimes.com/elections/results/president Nov 9, 2016 at 14:17
• Just for the record, it's happened in the UK that a government has been elected with a plurality of MPs but not of voters: en.wikipedia.org/wiki/United_Kingdom_general_election,_1951 . This possibility is an inevitable consequence of picking a government by a majority of regional contests rather than a single election. Nov 9, 2016 at 16:48
• The electoral college has not yet voted. We are anticipating that they will vote for Trump, but if Trump were to die before then, the electors would be unbounded and could even vote for me - if they felt I was the most qualified candidate to be president. In that case, the winner of the electoral college would not have received a single popular vote. Nov 9, 2016 at 17:03
• Yes. A more in-depth answer can be found at en.wikipedia.org/wiki/… Also, it appears to have just happened again in the 2016 US Presidential election, with Hillary Clinton winning the popular vote and Donald Trump winning the electoral vote.
– WBT
Nov 9, 2016 at 20:10
• You might like this video and this video. CPBGray looks at how to become president by winning a mere 20% of the popular vote. Feb 1, 2019 at 15:20

There sure have been a few instances of this occurring.

2000 Election

• This's the one that many people still remember. Gore won the popular vote by 0.51% while Bush won the electoral college. The tight margin in Florida automatically triggered a recount. Ultimately, a court case stopped the recount and Bush became President.

It has also occurred in the 19th century -- 1824, 1876, and 1888 -- though they aren't that well known.

However, it doesn't actually matter since only the electoral college counts.

With this happening before, it isn't unprecedented if it happens again this year. The NYT does forecast Clinton to win the popular vote by around 1.2%.

• It must be mentioned that, after the court case stopped the recount, one of the big newspaper chains paid for a complete, precinct-by-precinct recount of the entire state of Florida. They were hoping to prove that Gore in fact had won, and the court stole the election. Unfortunately, the final result was not to their liking: it turned out that Bush won Florida, fair and square. Nov 9, 2016 at 16:39
• @JohnR.Strohm "Fair and square" is an interesting assessment, given all the other shady things going on around there. I don't think it's so easy to assume the Bush campaign didn't have a hand in influencing which voters were struck from the register. Nov 10, 2016 at 9:33
• @Samthere: I don't think it's so easy to assume the Bush campaign didn't have a hand in influencing which voters were struck from the register. Any reason to think any more so than the Gore campaign? Nov 10, 2016 at 15:55
• @loneboat Well, definitely more reason than for the Gore campaign - many individuals unable to vote were in demographics more likely to favour Gore, and Florida required those convicted in other states to request the restoration of their rights to vote. This was at the discretion of the Governor, Bush's brother. I don't have all the details but there's absolutely more reason to think one way. The main point is that I don't think absence of a particular kind of underhanded tactic is enough to say that any major politician is acting "fair and square". Nov 10, 2016 at 16:25
• @JohnR.Strohm: Citation please. en.wikipedia.org/wiki/Florida_election_recount claims that if there had been a recount of all ballots statewide, that Gore would have won. It also says that using the methodology that the Democrats were proposing (and implementing) that was blocked by the Supreme Court, that Bush would have won. Nov 10, 2016 at 20:13

I stumbled on this illustration here, and it's captioned:

Note how somebody could win the popular vote, and still lose the electoral vote.

• In realistic terms, the two on the right would be states like California and New York Nov 10, 2016 at 16:26
• This graphic can be seen as misleading in regards to the US because it shows every state getting the same amount of electoral votes per population; while in the US, less populous states get a higher electoral vote to population ratio. Dec 27, 2019 at 6:50
• The deficit which you describe is even worse when it comes to the Senate. The ten largest states, have 53% of the population but only 20% of the senators. (The smallest ten also have 20% of the seats but only 2.8% of the population). Were senate seats awarded "democratically" by population, Trump would have been impeached and given his marching orders long ago.
– WS2
Dec 28, 2019 at 11:28
• In early nineteenth-century Britain, some parliamentary constituencies were so small, their constituents could be bought (pocket boroughs) - some had even ceased to exist, having fallen into the sea (rotten boroughs) - but continued to send members to parliament. Large burgeoning industrial cities like Manchester, Birmingham, Leeds etc were largely unrepresented. But that was all sorted out in Grey's Reform Act of 1832. All 650 constituencies today are roughly of the same population size. However the position is not perfect, and Conservatives plan to do further equalisation.
– WS2
Dec 28, 2019 at 11:36

Grover Cleveland won the popular vote three elections in a row but was only president twice. That was 1884, 1888, and 1892. He is the only president to have served two non-consecutive terms. Every other two term president served consecutive terms. Cleveland also won the most consecutive popular votes to that point. Franklin Delano Roosevelt would surpass him with four wins.

This has the most effect on the way that the campaigns compete. The concept of a swing state matters a lot to the electoral college vote but not so much to the popular vote.

There have been proposals to make the popular vote the important vote. The National Popular Vote project is currently trying to convince states to bind their electors to the winner of the national popular vote rather than to the votes of their states. It currently has commitments from states representing 165 electoral votes. It only needs 105 more to take effect.

Yes. This happened yesterday. Here's why:

This is a snippet of a much larger map and explanation located here: http://www.slate.com/articles/news_and_politics/map_of_the_week/2012/11/presidential_election_a_map_showing_the_vote_power_of_all_50_states.html

The issue is that not every vote for president counts the same. Electoral votes are based on population--but every state is given a minimum of 3 electoral votes. So if you live in a low population state, your vote actually counts for more than if you live in a high-population state.

Because, in general, low population states tend to be rural, and rural leans republican, the electoral college for the past few elections, and likely going forward, is going to favor the republicans. And that's how Trump can win the electoral college, but not the popular vote.

• Electoral vote weight is not the only mechanism for this. The phenomenon is also about close or large margins. Even if every state had an equal population and an equal number of electoral votes, the popular vote-receiver can lose if states won are won big, and states lost are lost small. E.g., given three states A, B, C; 99:1 for A, and 49:51 for B and C, the first candidate won the popular vote by a huge margin but lost the electoral vote, as B and C were narrowly won by the second candidate. Nov 10, 2016 at 1:41
• @ErikE that's certainly mathematically correct. But not as likely to happen.
– user1530
Nov 10, 2016 at 2:19
• It should also be noted for the benefit of the OP, who states that he's British, that the reason that each state has a "minimum of 3 electoral votes" is that the number of electoral votes is tied to representation in Congress--two Senators for each state, as well as a minimum of one member in the House of Representatives. Nov 10, 2016 at 15:46

It's very important to note that the people that go out to the polls and place their votes are not actually voting for the president. They are telling their representatives who they want them to vote for.

So it is totally possible and legal for voters to vote for candidate A but the representatives cast their vote for candidate B. In a way this happens all the time.

48 out of the 50 states have a rule that says if the majority of districts want to vote one way, then the entire state will vote that way. So it happens all the time that District 1 votes for John, and District 2 votes for Jack, District 3 votes for Jack and thus All three districts vote for Jack.

But in another way, it is legal and possible for people in a district to vote for Jack, and that district just decides instead to vote for John. This would not sit well, and those representatives would have a lot to answer for, but it would be legal. See https://www.archives.gov/federal-register/electoral-college/electors.html#restrictions

Lastly there is the "tradition" that a president should be elected by everyone. While not law, the general tradition is that the nation should "come together" to elect a president. So what generally happens is once a winner is clear, the loser backs out of the election. The general feeling is that the "great debate" is over and it's time to come together and move on.

So to answer your question it is possible that the popular vote goes for one candidate, while the electoral collage votes a different way. Generally this is thought of as a good thing as it avoids situations like the Civil War where really populated places had more political control than less populated places.

• This is sort of incorrect. First, it is not the congressional districts that vote, it is the state as a whole in a winner-take-all-the-EVs fashion, except in Maine and Nebraska, and even there the districts have only one EV each, while the winner of the state overall gets the other two EVs.
– user15103
Oct 18, 2018 at 1:01
• See paragraph 3 Oct 18, 2018 at 7:30
• it's not if "a majority of districts vote" though, it's if a plurality of citizens vote. Votes are not winner-take-all at the district level except in Maine and Nebraska.
– user15103
Oct 18, 2018 at 15:14
• No, it is exactly, district level, except states, have laws or arrangements that modify that. The constitution and the EC don't give a whit. Oct 18, 2018 at 18:40
• That is absolutely not true. The number of electoral votes is based on the number of Congressional districts plus two for each state, but absolutely nothing in the Constitution nor in actual practice makes the EVs hinge on the outcomes at the district level, except in those two states.
– user15103
Oct 18, 2018 at 19:17

FWIW, in response to @JohnRStrom, that's not quite accurate.

Many groups used a variety of standards to determine who "really" won in Florida in 2000. Some counts using some standards showed Bush, but other counts using other standards showed Gore had, in fact, won. Sometimes the margins either way were in the hundreds or low thousands. In one case, it was Gore by 3. Not three thousand or three hundred. Three votes.

Considering that there is no such thing as a national standard, but an accumulation of votes as regulated by smaller groups of decision makers, it seems to me that we should count by whatever the rules are for the area that is being counted.

In that particular case, a review of all ballots statewide according to the standard as set by each county canvassing board during their survey: Gore by 171

Factoring in the Butterfly Ballot errors*, and the overvotes (that were tossed from the count) where people voted for Gore twice because the ballot said "Write in vote" as opposed to "Write-in vote" **, and many other factors - including the count above - it's easy to see why many people would dispute your simplified assertion that Bush won Florida.

*: Poor ballot layout design in a hugely Democratic voting area caused confusion and led some people that thought they were voting for Gore to instead vote for third-party candidate Pat Buchanan in numbers large enough to eclipse any victory margin of either Bush or Gore in any recounts. Pat Buchanan picked up 3,704 votes in Palm Beach County (where the Butterfly Ballot was used), which was roughly 2700 more than he picked up anywhere else in the state, including where his headquarters was located. Given Mr. Buchanan's positions and the demography of people in that area, some were horrified to learn that they might have voted for him accidentally. Even Buchanan himself stated openly that he didn't earn that many votes in that area.

**: Some people misunderstood that that was a space provided to add a vote for an unlisted candidate and instead thought it was an explicit instruction to also write in the candidate's name. In those cases, many voters both selected Gore on the ballot and also wrote in his name below, resulting in an "overvote", rather than counting as a vote for the candidate that they clearly chose.

• Responses to comments should not be posted as answers. Jul 20, 2021 at 0:48

Just as a pure hypothetical:

If you start with the states in increasing order of population, and assume half of each small population (+ 1) votes for one party, those votes ensure that all the state's electoral college votes go to that party.

Continue this into larger and larger states, until you reach the magic number of 270 electoral college votes committed to that party. Let all the voters in all the remaining states go to the other party.

You'll find that the winner of the electoral college become President with around 22% of the popular vote.

All highly artificial and never to be seen. But it shows how the Electoral College skews the popular vote...

• Or as pure nonhypothetical consider a candidate who wins only the top 20% most populated of the counties. She could win a plurality the popular vote while losing most of the country. Showing how the popular vote does not actually require a candidate to have broad support and how plurality voting fails to demonstrate anything. Also, this seems to answer another question rather than this one. Dec 15, 2016 at 14:22
• @Brythan Plurality does in fact show something, namely that no other candidate has more support. Of course it's possible to win without broad support. It just take 50%+1 to win. You seem to be implying that someone who wins more counties is somehow more deserving of the presidency than someone who wins more votes, which is absurd. Jul 20, 2021 at 0:51

Yes there was. In 2000, Al Gore received the most votes but George Bush became president.