How much could someone win the popular vote by and still lose the US presidency?

Factual Questions
1

Something I’ve been thinking about wrt the 2000 and 2016 Presidential election: how extreme could it get? By how much could a candidate win the popular vote, but still lose the presidency?

The states’ differing electoral systems throws a big wrench in my ability to calculate this. I know there can’t be specific numbers, but I figured there must be a way to get a rough estimate, setting aside a certain percentage for third party candidates.

Anyone know, or can simplify the math better? Or is there no solid theoretical extreme? I could see that as well, but I’m not a math type.

2

Are you assuming equal voter participation across the states? Otherwise, you could win with just a handful of votes in the big states, while the loser gets all the votes from the small states

3

Theoretically, aren’t there all sorts of rare fringe cases to consider, like faithless electors, contingent elections, and similar?

4

Is there a quorum that applies either in your state or Federal elections requiring a minimum turnout to be valid? If West Dakota plunged into apathy could 2 determined Republicans sweep the state over the solitary Democrat?

5

Presumably if everyone in the most populous states voted for one candidate but the other candidate picked up exactly 50%+1 vote in just enough of the least populous states to get to 270 EC votes (or even 269 with a favorable Congress) there could be an enormous disparity. But I’m too lazy to do the math.

ETA: Obviously you’d also need full turnout in the loser’s states and absolute minimum turnout in the winner’s states. It’s crazy, man.

6

There is no quorum requirement for the Electoral College. In theory, one could win the presidency with 11 votes.

I just crunched it on 270towin.com and got 272 electoral votes from these states:

California
Texas
Florida
Georgia
North Carolina
Virginia
Illinois
Michigan
Ohio
Pennsylvania
New York

As odd coalition, indeed. 11 votes to win. That’s all it takes.

7

There’s 11 states whose cumulative total of electoral votes is 270, which is exactly enough to win. They are CA, TX, NY, FL, PA, IL, OH, GA, MI, NC, NJ. If, for some crazy reason, only one person voted in those 11 states, all for candidate A, while every voter in the other 40 states (counts DC, since they get to vote for Pres, too) votes for candidate B, then the vote disparity will be maxed. To find the actual number, you’ll have to find out how many people are old enough to vote in those 40 states and assume they all register, if necessary, and vote. Oh, yeah, subtract off those who are inelligible to vote because they’re in prison or something and add on those who are out of country, but still elligible to vote.

This whole scenario is utterly ridiculous in real world terms, but it’s the theoretical maximum.

8

Mmm, better redo your addition. Those 11 states add up to 269 by my calculation, but if you substitute NJ for VA, you get one more to bring it to 270.

9

And why should something so ridiculous not be enumerated … :slight_smile:

In 2016 Presidential Election there were 226,410,346 eligible votes in the 50 states.
If Candidate A received 62,959,821 spread precisely so as to win by a single vote in 11 states (California, Texas, Florida, New York, Illinois, Pennsylvania, Ohio, Georgia, Michigan, North Carolina & New Jersey) they would pick up the requisite 270 EC votes.
Then Candidate B receives every other eligible vote cast and thereby totals 163,450,525 votes.
Consequently losing the Presidency but winning the popular vote by over 100 million.

10

In the numbers above Candidate A wins with 27.8% of the eligible vote.

Sounds like the US system would be profoundly outlying in this misproportionality and evidence of a singularly broken electoral system, but (at least by this metric) it isn’t.

In Australia, if Party A contested the 76 smallest Federal electorates and won them by a single vote (after preferences) and did not stand candidates in the other 74 they would be able to form a majority government with just 23.8% of the enrolled two party preferred vote.

11

And if you wanted the scenario in post #9 run based on 2016 Presidential votes cast then Candidate A could win 270 EC votes with 37,536,776 votes and Candidate B could get 99,217,160 to lose while and winning the popular vote by 61 million.

12

Indeed, it seems like this aspect of elections is regularly worse than the USA in the UK at least (I haven’t followed politics in Canada and Australia as much: even though they share similarities their votes haven’t had as much impact on the world stage in the past few years.)

Just taking into account the factors of third parties getting more of the vote, and being able to be the dominant member of a coalition even without a majority of parliament, it seems that the percentage necessary to win control of the government would regularly be smaller in the parliamentary first-past-the-post countries than in the USA. And that’s even before taking into account any lopsided versus tiny district outcomes.

13

Are we looking for the maximum margin in the popular vote, or the maximum ratio? If you’re looking for the maximum margin, then there’s no difference between a vote of 1-0 and a vote of 10,000,001-10,000,000, and you’d get the greatest margin by making the small states the close ones and the large states the blowouts.

For the largest ratio (equivalently, the greatest percentage), you’d want all of the close states to be 1-0. If electoral votes were proportional to population, then you’d want the close states to be the big ones, but with small states being overrepresented in the electoral college, it might be better to have the small states be the close ones.

14

Indeed. If we allow faithless electors, then, theoretically, you don’t need any votes in the popular election to win the Electoral College vote - the popular vote could be unanimously in favour of candidate A, but faithless electors could still elect candidate B in the college.

Whether that would be a legally valid election has not been clarified. So far, faithless electors have been tolerated (in the sense that there is no case law that declares the vote cast by a faithless elector invalid), but that is only because they have never made a difference as to the ultimate outcome. There is case law that holds that electors can be required to make a formal pledge in advance that they would, in the Electoral College, support their party’s candidate (Ray v. Blair, 1952); but there is no case law that determined whether a vote cast in violation of such a pledge is valid.

15

This question is unanswerable as asked, because it is based on a faulty presumption. There is no “popular vote” as separate and distinct from the “electoral system” vote.

Regarding the election of the President of the United States, there is only the vote of the electoral college. Individual citizens vote to inform their electoral college who the individual citizens want to be President. The electoral colleges then vote per their states’ laws to determine the President.

Therefore, there is no “popular vote.” President Trump did not lose a “popular vote” but win the Presidency in spite of this. He got more votes from the electoral colleges, and therefore became President.

Both President Trump and Mrs. Clinton knew how the system worked, and campaigned accordingly. They visited states and held rallies, not to attempt to get the most individual votes, but to get the most electoral college votes. They spent money and ran ads based on trying to get the most electoral votes. They spent there advertising money where, when and for what based on trying to get the most electoral college votes, not the most individual votes. If they were merely trying to get the most individual votes each would have campaigned very differently.

Similarly, the people voted based on knowing that they were participating in an electoral college system, and not a “popular vote” system (at least, they should have. Anyone who doesn’t understand this by now has no business voting).

Many people vote differently because they know how the system works. Many people in California didn’t bother to vote because it was a foregone conclusion that Mrs. Clinton would win the California electoral votes. If this were not the case, if they were voting to see who got the most individual votes, many people who did not vote might have voted, and for both candidates. Some people vote for third-party candidates to “make a statement” because they are safe in knowing that their preferred candidate will win their state’s electoral college votes regardless.

Similarly in states where the Republican candidate getting the electoral votes was guaranteed.

There was no “popular vote,” and there is no way to know how the election would have turned out had there been. Both candidates would have campaigned to get the most individual votes, people would have voted with the knowledge that their vote would count no matter what state they lived in, and the individual vote count would have been completely different. Which way it would have turned out, we can never know.

16

It’s all true that, constitutionally, the Electoral College vote is what matters, but concluding from it that the popular vote doesn’t even exist goes too far. Of course it exists; individual voters cast ballots marked with the names of presidentiaql candidates, these ballots are counted and the results published. That is the popular vote, and it surely exists.

17

Nonsense. There was, of course, a popular vote. It was held on Election Day.

It just didn’t determine the outcome. The real vote was held on December 19, when the electors voted.

18

Eleven votes may be the theoretical minimum number to win. But I would have approached it from the opposite direction.

Your calculations are based on eleven people voting in eleven populous states. With all of the voters in the other thirty-nine states voting for the other candidate. These states have a combined population of around 141,000,000. Let’s assume 75% vote (a high percentage but we’re assuming voter enthusiasm) and say 105,750,000 people voted for the losing candidate.

Instead, let’s take eleven populous states (CA, TX, FL, NY, PA, IL, OH, GA, NC, MI, and SC) and assume all of their voters vote for one candidate. That’s a total of around 183,000,000 people so using the 75% figure, call it 137,250,000 votes. To the 39 votes cast by one voter each in the other thirty-nine states (which have a combined total of 270 Electoral College votes).

So in the second scenario, the winning candidate got more votes (39 instead of 11) but the losing candidate got a much higher total (137,250,000 instead of 105,750,000). I’d say this qualifies as a wider margin in votes.

19

Exactly. This is how we tell our electoral colleges who we want them to vote for.

Again, we are not casting individual votes that will determine who the President will be; we are filling out ballot forms to inform the electoral college of our preference. We know that when we fill out the ballot forms.

What some call the “popular vote” is meaningless numbers; it is raw data that is a by-product of the electoral college system.

Again, there is no meaningful “popular vote” that somehow gets transformed and changed into an electoral college vote. Candidates campaign for, and people cast their ballots based on, the electoral college votes.

20

If you’re going to postulate only 1 voter in each of the states the electoral count winner, you might as well postulate that all eligible voters in the other states vote.