Home | About | Donate

Abolish the Electoral College, Empower the People

Abolish the Electoral College, Empower the People

Robert Alvarez

Senator Elizabeth Warren is hell-bent on dismantling the systems that feed inequality in this country, including the Electoral College.

“Every vote matters,” she said at a recent CNN town hall. That’s why we should “get rid of the Electoral College” and institute “national voting.”

Americans don’t directly elect their president—states do. In most cases, states award all of their “electoral votes” to the candidate who wins the popular vote in those states. Whoever gets 270 electoral votes wins the election.

Abolishing the Electoral College wouldn’t be necessary if we:

"Abolished the Repugnantcant.Party."

Of course you pick the most insincere response to argue against. And follow that with the second most insincere response- the national popular vote compact.

Democrats feel they can win the popular vote so they support the compact.

But the problem with the electoral college is not that the winner of the electoral college may not be the winner of the national popular vote. That is how it is designed to work and part of the reason we have it (no matter what other reasons it was initiated) is to level the playing field for smaller states.

The problem with the electoral college is that states award their electoral votes to the winner of the popular vote in their state. The NPVIC does not solve this problem- it makes it worse.

The NPVIC only transfers the injustice from the state level to the national level. A state could vote 70% for a candidate and under the compact would award their electoral votes to another candidate that won the national popular vote.

While Rich Lowry may not be suggesting all states use proportional distribution of electoral votes, I am.

How about a pact for states to agree to proportional distribution?

Neither Republicans or Democrats want this because it could open the door for third party and independent candidates to obtain electoral votes.

Often it seems when people talk about leveling the playing field they are only talking about leveling the playing field between the Current Major Parties, not leveling the playing field for everyone.

There are good reasons why no state awards their electors proportionally.

Electors are people. They each have one vote. The result would be a very inexact whole number proportional system.

Every voter in every state would not be politically relevant or equal in presidential elections.

It would sharply increases the odds of no candidate getting the majority of electoral votes needed, leading to the selection of the president by the U.S. House of Representatives, regardless of the popular vote anywhere.

It would not accurately reflect the nationwide popular vote;

It would reduce the influence of any state, if not all states adopted.

It would not improve upon the current situation in which four out of five states and four out of five voters in the United States are ignored by presidential campaigns, but instead, would create a very small set of states in which only one electoral vote is in play (while making most states politically irrelevant),

It would not make every vote equal.

It would not guarantee the Presidency to the candidate with the most popular votes in the country.

The National Popular Vote bill is the way to make every person’s vote equal and matter to their candidate because it guarantees the majority of Electoral College votes to the candidate who gets the most votes among all 50 states and DC.

Abolish representative government.

The "iron law of oligarchy " states that all forms of organization , regardless of how democratic they may be at the start, will eventually and inevitably develop oligarchic tendencies, thus making true democracy practically and theoretically impossible, especially in large groups and complex organizations .

If its not direct, it’s not democracy.

You forgot to add to your unsubstantited claims the unsubstantiated claim that proportional distribution would cause the sun to stop shining.

If your going just make stuff up you may as well go for the gold.

So proportional distribution of electoral votes in the electoral college that is designed to sometimes elect the candidate that wins the national popular vote and sometimes elect the candidate that does not win the popular vote fails because it doesn’t always elect the candidate that wins the national popular vote.

Your reasoning for why proportional distribution fails fails.

And it could lead to the election being decided by the House.

And it doesn’t make every vote 100% equal.

I don’t see any of those things as a reason that proportional distribution fails.

The only thing of those that is a problem is making every vote 100% equal.

Proportional distribution comes closer than the current system and the NPVIC.

The current system and NPVIC give many more votes greater weight than other votes.

Under the current system the votes for the winner of the popular vote in that state have their votes counted 100% toward awarding the states electoral votes and voters that do not vote for the winner of the popular vote in that state, perhaps even more than 50% of the voters in that state, have their votes counted 0% toward awarding the states electoral votes.

Under the NPVIC, as stated previously, the winner of a state’s popular vote even with as much as 70% of the state’s vote would award the state’s electoral votes to a different candidate if that other candidate won the national popular vote. That gives those voters and anyone else in the state that did not vote for the national popular vote winner 0% of their vote counted toward awarding the state’s electoral votes and the voters in the other 49 states that voted for the national popular vote winner 100% of their votes counted toward awarding the electoral votes of a state they don’t live in.

Proportional voting awarding electoral votes by percentage is not perfect but is closer to achieving equality than the current system or NPVIC because it achieves better than 0% for many more voters in aligning electoral votes with the votes cast in an election.

It is not too complicated to implement.

Simply divide the states electoral votes into 100 to the nearest whole number.

In a state with 15 electoral votes a candidate would need 7% of thev state’s vote to achieve an electoral vote. Any leftover electoral votes could be awarded to the winner of the state’s popular vote.

This will never lead to a majority of citizens not having their vote counted 0% in awarding electoral votes. It is unlikely that much more than 10% in a state and never more than 20% of citizens nationwide will have their vote counted 0% toward awarding their state’s electoral votes.

And if this leads to the house having to decide the election- so what?

If they can’t do such a basic part of their job then the solution is not to avoid making them have to do their job, the solution is to put people in Congress that can do their job.

oops. that should have been “This will never lead to a majority of citizen not having their votes counted in awarding electoral votes.” or “…a majority of citizens having their vote counted 0% in awarding…”.

Some Republicans support the compact. It was almost the case that Bush lost Ohio in 2004 if Kerry picked up enough votes there which would not have gotten him the popular vote win (but he would have won the EV count). I so wish that would have happened because this argument wouldn’t be used nearly as often as it is.

The bottom line is the EC is stupid because it makes the presidential general campaign (not the primaries of course) all about the swing states. Republicans in CA and TX are just as ignored as Democrats are. Also states that aren’t swing states have no incentive to increase turnout - the result is the same either way in EV count.

A pact for all states to agree to proportional distribution changes swing states to swing states + swing districts. It’s still stupid and now it brings gerrymandering into the issue since how this districts are drawn now comes into the argument.

[EDIT - sorry, I’m mixing two concepts up, I thought you were referring to the Maine and Nebraska methods, which is winner take all district for the EV associated with the district and winner take all for the 2 EV associated with the state - most critics of the NPV propose this as there is heritage). If you are proposing taking all parties into play and multiplying fractions by the EV count of the state and rounding and then adjusting so the sum is the same, that is going to be a lot more confusing and I’m not sure it is on the table in any discussion I’ve heard]

An NPV is the only way to move to Ranked Choice Voting in the General and this could give a better foot in the door for third parties because more people may be willing to vote for Libertarian > Republican > Democrat > Green instead of just Republican and on the flip side Green > Democrat > Libertarian > Republican (some would flip the last two I guess) vs just Democrat. NPV is definitely better for third parties and it is better in any way I can imagine, though sure it reduces the power of a person’s vote in a small state to be the same as everybody else’s power (as it should be) assuming it is a swing state because if it isn’t, it’s EV count is taken for granted anyway.

correction: When I said in my previous comment that more than 10% of citizens in a state would never have their votes counted as 0% in awarding electoral votes, that only applies to states with more than 10 electoral votes. So it is possible that nationally more than 20% could get 0%.

It would take an amendment to give each state at least 10 electoral votes with proportional distribution and eliminate the electors, which is a better solution than both non-amendment options or remaining with the status quo, but proportional is the closest we can get without an amendment.

Not sure what you’re talking about with the multiplying fractions, rounding, adjustment so I don’t think I’m proposing that.

I am saying that the number of electoral votes in a state is divided into 100 (for 100% of the vote).

In a state with 15 electoral votes, 15 into 100 comes out between 6 and 7 in whole numbers. So any candidate gets 1 electoral vote for each 7% of the state’s votes. 14% gets 2 electoral votes, 21% gets 3 electoral votes, etc.

The electoral votes not awarded by this method could go to the winner of the state’s popular vote as previously suggested to keep it simple or could be awarded by which candidate(s) are closest to the 7% in their remaining votes.

For example, one candidate gets 17% of the vote in this state. That candidate gets 2 electoral votes and has 3% leftover votes that are not enough to qualify for an electoral vote.

Another candidate gets 31% of the vote in this state. That qualifies for 4 electoral votes and there is 3% leftover votes that do not qualify for an electoral vote.

Another candidate gets 45% of the vote in this state. That qualifies for 6 electoral votes with 4% leftover that does not qualify for an electoral vote.

Unless one other candidate gets all of the remaining 7% of the votes which could only happen if the other candidates got exactly those percentages of the vote with no fractions that leaves three electoral votes. But it could end up with only one or two leftover electoral votes depending on how the percentages turn out and how many candidates are in the race and getting enough votes to reach a whole number in the percentage of the vote in that state.

If there is one leftover electoral vote it would go the the candidate with the highest amount of leftover votes. If two, the two highest candidates, etc.

This way even a candidate that got 6% of the vote could get an electoral vote if the other candidates had a lower leftover percentage.

This would also in a state with three electoral votes allow a candidate that got 20-32% of the vote in that state to get one of the electoral votes to bring the awarding of electoral votes in that state more in line with the votes cast in that state.

While this is not easy it is also not too hard. And sometimes being as fair and equal as possible is hard- but that is not an excuse to not try or not do it.

As for the NPV being better for third parties I just don’t see it. That just means a third party candidate has to win the national popular vote to get electoral votes from participating states instead of just having to win the state under the status quo or getting electoral votes just by getting the percentage to earn an electoral vote in a state with proportional distribution.

Rank Choice Voting is also not the answer. In most cases it allows the voters to rank their choices when they vote, but is in reality Instant Run-off Voting when it comes to counting the votes.

The problem with instant run-off voting is that it eliminates candidates as each round progresses. This may force citizens to validate with their vote candidates they do not find suitable for office as the candidates they do find suitable are eliminated in order to have their votes counted in all the rounds.

They are forced to choose lesser evil voting or not having their vote counted in the total to achieve a majority in any round where all the candidates they voted for have been eliminated.

This means the majority is not achieved from all the votes cast in the general election, it is achieved from only those voters that voted for the remaining candidates in any round. It is clearly not a majority of all the voters.

A candidate could be the second choice for many voters that have another candidate as their first choice (even a majority of all voters when added to the voters having that candidate as a first choice), but if that candidate is eliminated in the first round then those citizens don’t get their second choice when their first choice is eliminated.

Only rank choice voting that does not eliminate candidates in the counting is true rank choice voting.

If no candidate gets a majority in the general election, all candidates move to the second round.

The first round votes are added to the second choice of each voter that makes a second choice. If one candidate achieves a majority they are elected.

It is possible that more than one candidate could achieve a majority in this or subsequent rounds and the one with the higher majority would be elected.

If no candidate gets a second round majority it continues with the same principle for each subsequent round.

If no candidate reaches a majority in any round and there are no more round to go through then the candidate with the highest plurality gets elected.

I agree your scheme for describing what to do in the 15 EV example is definitive. The ‘error’ in the final EV being related to the proportional vote is worse in small EV states - maybe you want to just multiply all EV values by 100 in all states and then repeat your algorithm.

I’m only claiming NPV is better for third parties with RCV, with plurality it doesn’t matter - they are going nowhere.

I agree that IRV is the most common way to count RCV ballots (and these days I use the terms interchangeably because I don’t believe we will ever see schemes like Condorcet counting used with RCV ballots in the US).

Only rank choice voting that does not eliminate candidates in the counting is true rank choice voting.

I read a lot on this topic, and I have never come across this type of claim - all schemes have their upsides and downsides, nothing about IRV is automatically worse than any other counting scheme.

All known RCV counting schemes have strategy voting problems, and in fact all possible schemes have these problem (as I presume you know), but the problem you state here does not make sense to me. If all voters rank all candidates (and I suggest they do, or at least make damn sure they rank the two or three most likely to make to final rounds), then everybody’s vote is taken into consideration - of course in the final round you could be on the losing side, there is nothing that can be done about that (besides proportional representation, but we are talking about single winner elections here).

I am not aware of your scheme of adding in the literature. It certainly isn’t obvious to me that is better than IRV or any of the Condorcet schemes. But if you are citing a well known counting scheme and have a reference to go along with it, I will take a look. I am very interested in RCV counting schemes - I’m not as interested in score voting or approval voting.

Another poster who posts on RCV often is @webwalk, perhaps he/she has heard of your scheme.

Sounds like @D.Harris is referring to a form of Borda count? Borda is a ranked choice ballot that does not eliminate candidates through successive rounds, but assigns points to every candidate from every ballot based on their ranking on that ballot. There are different ways of assigning relative points, as you’ll see if you read the Wikipedia article. Otherwise i’m not sure exactly what system they are referring to, with multiple rounds but no elimination, so if i’m missing it perhaps they will clarify.

As you point out, every voting system has pluses and minuses, all systems are subject to some form or another of strategic voting, and there is no “perfect” voting system. It gets more complicated because in addition to the mathematical “perfection” of each system or the relative likelihood of “wrong” outcomes that are possible under each system, we also face the problem of popular understanding of the system. A system that is “better” in many ways than another system, may not be as easily understood or preferred by voters who are asked to adopt a new system. So simplicity and understandability become as important as the mathematical probability of “good” outcomes with each system. This might be a problem with the system D.Harris is proposing: People have to want to adopt the new system in the first place, and so it helps if they can easily understand it. This is one of the reasons the Approval Voting advocates like their system, since it is simple to change the ballots (no change needed, just allow voters to check as many candidates as they like), simple to understand and simple to count (just add up all the votes for each candidate and the one with the most wins), and is a lot better than the stupid plurality system currently in widespread use. i prefer Score Voting because i get to indicate more nuance than simply “yes” or “no” for each candidate (ranking each candidate not against each other but against a scale [0 - 3 or 0 - 9 for example, and Approval Voting is just a scale of 0 - 1]), with the same simplicity of just adding up all the scores with the highest total score being the winner. Approval, Score, and RCV/IRV are all vastly better than Plurality Voting. i might get sold on STAR Voting (Score Then Automatic Runoff) which attempts and claims to get the benefits of both scoring and ranking systems with less likelihood of “bad” outcomes, but i’ve only been studying voting systems for a few years and it’s complex!

1 Like

Interesting. I see the analogy to Borda count (or one of its variants). But if on the first round there happens to be a majority of first place votes for a given candidate, the proposed scheme picks that candidate but with points accounted for in other places the Borda count winner could be different. Then if there is a second round the first and second choice votes are treated equally in point value. Makes me wonder how many RCV schemes are out there.

I think I’ve said this before. On approval and score, I’m a bit averse to the strategic voter problem. Approval voting is easy to understand - easy enough that a bunch of people choose to bullet vote their top candidate and we don’t get to know what their preference on other candidates are. On score/star, I’m worried it just isn’t intuitive how you should score. I intuitively know if you tell me two people are running which one if prefer, and when more than two are running, my pairwise decision will not be circular and will be deducible from an RCV ballot (I’m well away of the circular problem on the aggregate of ballots). But how much better do I like Tulsi Gabbard than Bernie Sanders? I don’t want the think that hard (I’m having a hard enough time just choosing which one to rank first - it would be great if they were in the same debate night).

1 Like

Thanks. Maybe you’ve thought about this too:
In RCV, there can be different requirements or opportunities for how many you rank. A system could require you to rank all candidates, or allow you to rank as many as you like including allowing bullet voting for only one candidate. In Maine i believe you only get to rank three, and i’m not sure if you’re required to rank three or could decide to rank just one or two.

Do you have a sense for which form of RCV works better from your perspective?

i think one of the main reasons for limiting it to three, is the complexity of the count and of reporting the count, which is one of the “issues” some advocates have with RCV: You don’t just take the ballots and count them, you go through a series of counts, and the process goes back and forth between the precincts and the central tallying place, which in a nine-candidate race (think the upcoming Democratic Party presidential nomination process) would get very complicated not just to do, but to report on.

Looking at http://www.lwvme.org/RCV.html, it does seem it is top 3. That seems pretty limiting - it would have been OK in the 2016 general for me (I’d choose Stein, Clinton, Johnson, and I don’t need to rank Trump), but I’d want to be able to rank a lot more than 3 or even 5 of the Democratic primary candidates given the number there are (and we aren’t even done yet).

I don’t want a requirement to rank all candidates, you should always be able to choose only 1 or 2 if you prefer - the last thing I want to do is frustrate an older voter who is used to doing things a certain way.

I am taking it for granted that we can with open source computer programs deal with a large number of candidates and ranking all. I am aware that makes it impossible to ‘bin’ to ballots (this is an advantage I liked about Condorcet or any scheme that uses as a sufficient statistic the pairwise counts of every possible 1 on 1 match. Then you just have a set of N*(N-1) numbers [N*(N-1)/2 races, 2 numbers for the race being the two totals]. But as you know, when there isn’t a Condorcet winner, the methods to choose a winner are not something I want to explain to the average voter and I agree with you that this a significant aspect to take into account for each scheme. IRV is easy to explain and computers can handle it (I think anyway, I haven’t actually looked at the details even though I work with computer algorithms all the time). Now I’m for paper ballots so you could do a recount, but the actual counting and processing does have to be done by computer and people need to have confidence in this. Open source software and strict standards on computer equipment I think can address the issue, but the devil is in the details I realize. Of course I think we have to move off the system we have which is a huge mess.

I have been very interested to see some of the Adam Eichen (I think that’s the right name) stories on RCV in the primary because it isn’t a single winner scheme - just a scheme to funnel down to a set of candidates with the lowest having at least 15%. Seems to make sense, and it would be super exciting to see a state try something like that by 2020.

So RCV with proportional representation? Or no, RCV as a primary to narrow the field for the general election, with all candidates who meet the threshold in the primary making it into the general election? i’ll do a search for Eichen.

Oh look, two recent articles posted at Common Dreams by Eichen about using RCV in the 2020 primaries. Thanks for the tip!

1 Like

Neither - it was just a scheme to deal with existing Democratic Primary rules that allocate a given state’s delegates proportionally but only if you reach 15%. If in the 50 state total of delegates, no one gets a majority, there is still the mess of super-delegates at the convention and no RCV there to get to the final nominee.

Still, it was an interesting possible application I hadn’t thought of.


Sorry I can’t refer you to “literature” about my approach (scheme sounds devious). That is just my opinion on RCV and me thinking on how to do it better.

While it may not be the standard approach for many people that read what other people write and repeat it, I prefer reading what other people write and thinking for myself.

This explains why you have never come across the claim that only RCV that does not eliminate candidates is true RCV.

The problem you fail to see is in your reply “if all voters rank all candidates”.

In elimination schemes (that word does apply here) citizens have to rank all candidates to have their vote counted in the total to achieve a majority in all rounds.

If there are 5 candidates and the three that a voter ranks are eliminated in the first three rounds their vote is no longer counted in the total to achieve a majority. Only the votes for the remaining candidates are used so that one will get 50% plus one.

So it is possible that if a majority of citizens that voted do not choose one of the two remaining candidates in the final round then less than 50% of the people that voted in the general election will have selected the winner because the less than 50% would be split between two candidates. Yet they still claim that they are achieving majority support for the winner.

Even with just 20% of general election voters not having their vote counted in achieving the majority, one of the two remaining candidates would have get almost every vote to have majority support from the total of general election voters.

Picking a candidate to rank that is likely to be in the last round is fine if you find that candidate suitable for office.

But if you don’t find the two candidates likely to not be eliminated suitable for office, then it becomes mandatory lesser evil voting because it forces citizens to validate with their vote one of two unsuitable candidates in order to have their vote counted in the total to achieve a majority.

For me what makes a candidate unsuitable is if they are a big money candidate. And all Democrats and Republicans are big money candidates- but are likely to be be in the final two.

Why should I have to vote for one of them in order to have my vote counted in the total to achieve a majority in the final round?

If you are only counting people that find the final two candidates suitable to validate with a vote to achieve a majority and not counting the votes of people that do not want to validate one of those candidates with a vote then it is not a majority of all voters, it is just a majority of people that find one of those candidates suitable for office.

But let’s explore what should be pretty universal disqualifications.

The Democrat likely to be one of the final two is a child molester and the Republican likely to be one of the final two is is a racist that wants to deport anyone that isn’t white.

Which one would you want in office?

And remember if you don’t pick one your vote does not count and one will be elected whether you vote for one of them or not.

All that being said, I am only discussing a way that RCV could be made to work properly if it is going to be done. I am perfectly happy with plurality voting and don’t care if RCV is never implemented.

But if it is because that’s what people want, then it should be done right.