I've seen some of the YouTube and news outlet pieces on the mechanics of how the USA's electoral college voting system works.

But if electors are supposed to follow the popular vote and the popular vote is for candidate A, then how does candidate B amass enough elector votes to win in A's place?

In the case of the 2016 race just over, is this down to electors voting against their state?

Looking at Al Gore's previous defeat by the electors, there was a not insignificant margin of half a million popular votes for him, was this then down to faithless electors too?

EDIT: Follow-up; if the electoral votes can bring a different result than the popular vote then what use are polls and demographics assessments? Lots of media outlets are pointing to the "rise of the uneducated white male" in Trump's win, but I have to assume that electors have a certain standard of education and that some who voted for Trump this year are not all white or male...

  • 16
    In response to the media quote, I would says it's less of the "rise of the uneducated white male" and more of the "rise of the voters that didn't want Clinton". Much of this election came down to the thought process of "I don't want that candidate, so I'll vote for the other one". Commented Nov 10, 2016 at 16:52
  • 13
    To @DavidStarkey 's point, the last poll I heard of which gave people four options (for Clinton, for Trump, against Clinton, against Trump), over 50% of people were against a candidate rather than for one.
    – Cort Ammon
    Commented Nov 10, 2016 at 17:34
  • I think that the answers at this question better answer the question of how someone could win the electoral college without winning the electoral vote.
    – Brythan
    Commented Nov 10, 2016 at 20:04
  • @CortAmmon What a horrible poll. Gary Johnson is rolling over in his grave.
    – coburne
    Commented Nov 10, 2016 at 21:14
  • 1
    @coburne It all depends on what you are trying to get from the poll. I agree with the sentiment about Johnson, but that particular poll did capture information that is typically missed in most poll question choices. You have to give it credit for that.
    – Cort Ammon
    Commented Nov 11, 2016 at 0:11

8 Answers 8


The electors are supposed to follow the popular vote within their state. In other words, whichever party "won" the state gets all of the electoral college votes. They don't follow the national popular vote.

This means that a candidate can win by a slight margin in states which have a disproportionately high number of electoral college votes compared to their population, and those electors will vote for them.

While Trump has a lower national popular vote than Clinton, he has a higher popular vote in a number of states with disproportionately many electoral college votes.

Consider a situation where you have only three states. Two states have 11 people in them and 3 electoral college votes (the minimum). One state has 100,000 people but only 5 electoral college votes. Candidate A can win with 6 votes against 5 in the two smaller states and win the electoral college, even if candidate B gets 100% of the vote in the larger state. In this contrived example, candidate B has practically 100% of the national popular vote, but candidate A wins because the vote result is not set up proportionally.

In response to your follow-up question, polls often try to gather data by region to understand how those areas will vote, and how it will affect the election. The reason they're off isn't that they ignore the way the election works, but because they're not able to obtain complete and accurate data.

  • 17
    Even with a completely proportional system (say, three states with the same number of voters and electoral votes) odd things may happen, with A winning two states by 100 votes each and B winning only one state by 20.000 votes.
    – SJuan76
    Commented Nov 10, 2016 at 10:41
  • 12
    @SJuan76 That's right. Any system that divides voters into blocks, in which each block contributes only one answer, can never be guaranteed to totally and accurately represent each individual vote.
    – Samthere
    Commented Nov 10, 2016 at 10:51
  • 8
    @ChrisCirefice No; the point of the system (in part) was to balance state concerns. The US is a federal republic, not a unitary one. The states are sovereign with their sovereignty partly limited by the Federal govt. The electoral college gives the states some "vote" towards the federal executive
    – eques
    Commented Nov 10, 2016 at 16:19
  • 8
    In other words, whichever party "won" the state gets all of the electoral college votes. - This actually is not generally correct. How the EVs are distributed is state-level policy. In Maine and Nebraska, for example, it's one EV per congressional district. For example, Maine has 4 EVs, this year Clinton got 3 and Trump got 1. In Nebraska Trump still got all 5. Most states do winner-take-all, but it's not a federal-level rule. Each state-sized voter block doesn't have to contribute only one answer. Theoretically, one could fiercely lobby for their home state to use a different system.
    – Jason C
    Commented Nov 10, 2016 at 16:49
  • 5
    @ChrisCirefice The electoral college was developed over 200 years ago when travelling and even communicating across the US took a long time. The nation was very loosely unified. People thought of themselves as a citizen of their state not the US (kind of like the EU today.) There was a fear that each state would just vote for their own candidate. Electors were meant to be political elites that would think nationally and people would vote for them, not the candidates. To eliminate this system, many of the states that benefit from it would need to ratify it so it's unlikely to change.
    – JimmyJames
    Commented Nov 10, 2016 at 20:23

The United States does not have a national election. Nor is there an official, legal 'national vote' total. Instead, the US has 50 state elections for electors who vote for President. People sum up the legal totals in the 50 states, but the result does not mean as much as some appear to think, as it mixes together contested and uncontested state totals (see below).

In all but a few states, the popular vote winner gets all of the electors. At least 2 states distribute their electors more or less proportionately, according to fixed rules.

This year, at least 35, perhaps 40 states were pretty much uncontested by the two major candidates. I believe both focused their resources on about 10 states where they thought they could get the most benefit in terms of electoral votes.

To answer the followup question, national polls and unofficial national totals are only roughly correlated with electoral college results. To predict the latter, it might be better to focus polling on the 10 to 15 contested states.

A national election, with an official national total that determined the winner, would be a different contest. Campaigns would be quite different, with the major candidates competing more nationally instead of in a small subset of the states. An in-between proposal that might have a similar effect would be one elector elected in each House district and two for each state (which are the Senate 'districts').

  • 2
    To your last paragraph: Also, if we had any serious presidential election reform, we'd be likely to also have primary election reform. We could be talking about President-Elect Rubio or Sanders today.
    – dan04
    Commented Nov 11, 2016 at 15:11
  • 2
    "2 states distribute their electors more or less proportionately", well no. They're just winner-take-all at a district level instead of state level.
    – Ben Voigt
    Commented Nov 22, 2016 at 21:25
  • 1
    Which is still in the spirit of representative government, which is the point of the electoral college (and of the House of Representatives), since all politics is (and should be) local. Commented Dec 28, 2016 at 16:18

There are three effects: First, some states have more electors compared to the population than others, especially small states. Let's say 10 states with 3 electors and one million population, one state with 25 electors and 25 million population. 10 million votes in the 10 small states beat 25 million votes in the big state. In reality, it's not quite that extreme, but it can make a difference.

Second, in many states you need just the majority. Whether you have 50.1% of the votes or 99.9%, doesn't make a difference. So one candidate could win all their states with 50.1% majority, and the other one wins theirs with 99.9% majority. The second candidate can have almost three times as many votes and still lose the electors.

Third, it doesn't matter how many people vote. One candidate could win states where 99.9% of the voters actually vote, and another one wins states where only one percent of the voters actually vote. The second candidate could get more electors with fewer votes, by winning states with only 100 voters advantage, while the other wins fewer states, with 100,000 advantage in every state.

It won't be as extreme as I said, but enough to make a difference.

I think a few years ago Florida was won with 500 or so votes. So just 500 votes made a difference of 25 electors. These 500 votes wouldn't have changed the country wide popular vote, but they changed who became president.


Think of the way Congress is set up, with a Senate with two Senators per State and a House of Representatives with Representatives by population. The Founders wanted to make sure that factions had to work out their differences by negotiation in the political process and that very populous areas would not control the process by sheer size of population. Many people do not know that we are a Republic of States and that States are Sovereign per the Constitution.

  • 5
    I downvoted this for two reasons. First, it doesn't answer the question. Second, the states are not sovereign (either by law or defacto). Commented Nov 10, 2016 at 17:19

But if electors are supposed to follow the popular vote and the popular vote is for candidate A, then how does candidate B amass enough elector votes to win in A's place?

Electors are bound to follow the popular vote in their state (but not always), not the US as a whole. On a national level, the electoral vote differs from the popular vote for two reasons:

(1) The winner-take-all system

In all but two small states (Maine and Nebraska, which allocate electors by Congressional district), a candidate need only win a bare plurality of the popular vote to win the state's entire slate of electors. Getting 50.001% of the vote (or even less, with third parties on the ballot) is just as good as 100%.

It happens that Trump won several important “swing states” with thin margins:

  • Florida (29 EV), 49.1% to 47.8%
  • Pennsylvania (20 EV), 48.8% to 47.6%
  • Michigan (18 EV), 47.6% to 47.3% (preliminary — not yet “called” by most media outlets)
  • North Carolina (15 EV), 50.5% to 46.7%
  • Arizona (11 EV), 49.5% to 45.4%
  • Wisconsin (10 EV), 47.9% to 46.9%

This translates to an Electoral College landslide even though the popular vote is close.

(2) Electoral College malapportionment

The number of electors a state has is equal to its number of seats in the House of Representatives (apportioned based on population) plus its number of Senators (always 2). These two “Senatoral“ votes skew the votes/population ratio in favor of small states.

(Note that since the ratio of Representatives to Senators is not fixed by the Constitution, this effect depends arbitrarily on the size of the House.)

California, the most populous state (37 341 989 according to the 2010 Census on which the current apportionment is based) has 55 electoral votes, or one electoral vote per 678 945 residents.

Wyoming, the least populous state (568 300) has the Constitutional minimum of 3 electors (for 1 Representative + 2 Senators), or one for every 189 433 residents.

In this sense, a Wyoming voter is “worth” 3.6 times as much as a California voter.

Trump appealed more to rural voters (and thus, small-population states), so benefited more from these “Senatorial” votes.

  • 1
    The electors in many States actually are not bound by law to vote in line with the popular vote in their Stare, although it is rare that they don't. Commented Nov 11, 2016 at 8:41
  • 1
    Quick query on the notion of votes being "worth" more or less: Is this how it is generally seen by Joe Voter in the US? To my mind if a vote is said to be worth more in this way, then by definition it is in a state with only a few electors. With fewer electors and electors of most states effectively having to vote as a single block, this means the state has less say in the total EV count and thus, comparing 55 to 3 EV's with the 1:3.6 ratio (55/(3*3.6)), the WY PV votes are actually worth 5.6 times less than the CA ones in who gets elected...
    – Toby
    Commented Nov 11, 2016 at 10:03
  • Dang mobile phone keyboard and tiny font, I obviously meant State. Commented Nov 11, 2016 at 14:30
  • 3
    @Toby: If you look at the Banzhaf power index, voters in larger states have more power.
    – dan04
    Commented Nov 11, 2016 at 14:58

Part of this is that Hillary Clinton ran up the vote in California, where

  1. It was a foregone conclusion that she was going to win the state, so no reason for Republicans to turn out to vote for president.
  2. There was no Republican candidate for Senate, so no reason for Republican to turn out to vote there.
  3. Only a fraction of its fifty-three Representatives are in competitive districts (perhaps a third if we count all the Republican districts; ten if we count only districts with a Partisan Voting Index of five or lower), so no reason for many Republicans to vote for Congress at all.

With no reason to vote except to have done it, many Republicans stayed home. Result? Hillary Clinton won the popular vote by about three million votes, both in California and the United States as a whole. There are other states with similar situations; e.g. New York, Massachusetts, and Connecticut also have many lopsided districts and were not competitive for president or Senate. But it's only in California that the state margin is about the same as the federal margin.

We can't know how Republicans would have voted if the popular vote was actually binding and their vote would have counted. Or Democrats for that matter. But we do know that turnout tends to be higher in more competitive states. And Democrats have more population in lopsided states.


All answers are great, but I think miss two very important aspects of the voting system in USA.


USA is not a country. It's a federation of independent states... Or better yet: sovereign states. It's more de jure than de facto (since that fascist Lincoln, even less Republican than Trump today), but the fact remains. Yes, I mean that Arizona is like France. That is why even if Congress ratifies, for example, Kyoto accords (on CO2 etc), every state must have do it as well, in order for it to become law in that state. Federal law cannot trump state law, otherwise it won't be a federation, but one state with administrative borders and peculiarities. Yes, there are federal laws enforced on a state level, but always they are first incorporated into that state's laws.


What the Founding Fathers devised and made into law was Republic, not a democracy. That is why only until the WW2 term "american democracy" was virtually unused in public and had strong negative meaning. So it's kinda obvious that the very democratic tool will not be used to elect people to the seats of - in their eyes - near absolute power.

To answer your questions:

  1. Electoral college in each state votes according to that state's laws - and most commonly this is according to popular vote in that state. Trump had more votes in more states than "Mad Hillary", which is what counts.

  2. Electoral college isn't bound to the outcome to the state's vote result. So they can vote the way they want, but - obviously - there are political consequences of that. So they usually cast the way the voters did. There is a book by Mark Goodwin - American Reset - in which there's a description of presidential election going awry at the electoral college level (or maybe Congressional [???], can't remember right now), with very interesting consequences. Worth a look.

  3. Polling is still not very accurate. +/- 3% in any election is, as can be seen, quite a lot. And even then there's the very high susceptibility to the "quality of the question". Polling is useful done properly, but since they are a very important weapon in any election they will, of course, be politicized to a large degree.

  4. This is an answer to unasked question. Yes, this is a feature, not a bug. Electoral system was set up by Founding Fathers for precisely that: NOT to have general election based on popular vote.This was done based on the various events in Europe, like the "Cromwell Rebellion", like the way the Polish-Lithuanian Commonwealth elected their kings (that is: who got more money to pay off voters won), like many more more-or-less contemporary situations. Remember that in the first US election voted less than 45k people... A lot of people could not vote then: poor, women, slaves, Catholics etc., so it wasn't impossible to influence election directly, by buying votes (literally).
  • The founders knew that there has always been a drive to concentrate the most power in the fewest hands, and also a drive on the part of much of the population to hand that power over willingly. Many people seem to instinctively want a king--somebody else to take care of their problems for them. Rome started out with kings, experienced how badly that was going, kicked them out and founded a Republic, enjoyed a golden heyday, concentrated all power back into one popular individual called the Caeser, and now Rome is something from antiquity that you study in history class. Commented Nov 11, 2016 at 16:35
  • @Craig - Which is, come to think of it, not a bad thing. There can be only one Captain on any boat. That's why any successful entity (be it a corp, state or a boat) usually has one CEO. This is the rational thing to do. But you're right - it then always leads to problems. But very rarely it's being done in an instant - usually the erosion is gradual, takes time. In case of mentioned Roman kings it still took around 150 years (IIRC)? In my opinion all depends on the quality of the voter... USA did well until everyone could vote...
    – user10424
    Commented Nov 11, 2016 at 17:15
  • That would be an apt comparison if a nation of 320,000,000 people was a boat. But it isn't. The city of Des Moines might be a boat. Maybe even New York City is a boat. But not the whole country. Healthy politics is and should be local. The quality of the voter argument is valid, and also an extremely hot potato. The intent of the founders was that the States have more power than the federal government. This has been turned somewhat on its head, but the structure they put in place is still preserving liberty. We should be very careful about discarding it without thinking long and hard. Commented Nov 11, 2016 at 17:20
  • Thinking long and hard, though, is something that I don't think is done as much as it should be, and that goes to the quality of the voter argument. People should learn how to think critically and independently, but I think that is an art that is in jeopardy. Commented Nov 11, 2016 at 17:21
  • 1
    I think Heinlein had the right idea in a book I value quite a lot - Starship Troopers. In it only people with prior service in the military could vote. And the service was set up so any serviceman would be either in active, combat units or otherwise risking their lives... "Nothing of value is free". And people do not care about free stuff... quality of the voter...
    – user10424
    Commented Nov 12, 2016 at 0:38

Because the Electoral College was created to ensure that the largest populated states, or more specifically, cities do not get to pick the leader for the rest of the country. Looking at the amount of counties won by Trump it stands to reason that the Electoral College, especially in 2016, did its job. Take NY. There are 62 counties in NY and 45 went to Trump but because Manhattan is deep blue, the state went to Hillary Clinton. Illinois has 102 counties. Trump won 90 and Hillary won 12 but Hillary carried the state thanks to Chicago. If states conducted elections with an electoral college, the only places a Democrat would win with consistency would be California, Vermont, and Taxechusetts.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .