In defense of the Electoral College followed by a plea for real reform

The objections to the EC are merited. But, I argue here, there are also some good reasons for some of the seemingly bizarre stipulations of the EC. I am not, however, defending all aspects of it. Indeed, it is most certainly not the system I would support if we were designing a system from scratch today, although there are some properties I would try to keep. Let me also note that I do consider the current voting system massively unfair, but not because of the EC.

Objections to the EC

1. Violates "one person, one vote" principle. If the popular vote and the EC outcome do not match, then clearly not all votes are equal. A voter in Florida has had more influence on the outcome than me, a voter in California. Indeed, the candidates know this which is why they put much more effort into campaigning for votes in Florida than they did in California. Those Florida votes (as ones in Ohio, Pennsylvania, Michigan, etc) were simply more valuable then votes in California, New York, the Carolina's, etc. This is a direct consequence of the "winner-take-state" property of the EC. It doesn't matter the margin by which you win a state. Getting a simple plurality of votes in the state is worth as much as getting 90% of the vote in the state. Likewise, losing by 1% is no worse than losing by 80%.
2. Indirectness of voting Technically, Americans vote not for presidential candidates but for "Electors committed to a particular candidate". There is nothing that actually binds these electors to their commitment. Originally, the idea was that the electors wouldn't even be committed, but that voters would really be selecting individuals who would then select the president. While there may have been some virtue in that 200 years ago with the idea that voters may trust the judgment of local officials that they know, who in turn knew more about national politics, there is clearly no justification for that kind of indirect election today. Although, one advantage is that it provides a mechanism for selecting a president if the EC favored candidate dies between the time that voters' ballots are finalized and the time that the Electors select the president. Some procedure like that would save the court battles we are going to see regarding the Missouri election for US Senate. But there are ways to get this benefit without having indirect election.
3. Clumsiness of procedure The actual formal procedures reflect concerns and technologies 200 years out of date. They became out of date with the invention of the telegraph.

I will not be defending either indirectness or the clumsiness of the EC. I suggest that those two problems alone don't justify amending the Constitution. In practice the Indirectness is not really an a problem, and the Clumsiness we can just view as some quaint ritual, and even enjoy if we are so inclined. My focus is on Winner-Take-State.

The People or the States?

The clearest place that this happens is with the make up of the Senate. A voter in Wyoming has a far greater say in the composition of the US Senate then me, a voter in California. And the purpose is clear, the Senate is where the states have their voice, while the House of Representatives is where the population as a whole have their representatives.

Now should be President be selected by the states (as the Senate is) with each state having an equal voice (but each voter having an unequal voice), or should the President be selected like the House where each voter has an equal voice, but states are very unequal?

The "winner take state" system is a compromise between the two. It makes each states voice roughly proportional to its population, but it also means that big states can't take advantage of their large populations beyond that. The EC prevents a candidate from trying to win very big in a few big states (or population centers) while ignoring the rest of the country. While with the EC an election may depend on a few crucial states, you can never win by only winning in a few states. You can only with the EC by winning in a substantial number of states.

While there are alternatives to the EC which would do the same much more cleanly (such as using national popular vote, but adding on "bonus" votes for each state won were the bonus votes are on a "winner take state" basis), the EC is a remarkably clever compromise.

As the US becomes more homogeneous by region and state the the importance of state voices declines. But I do not think that that has occurred to the point where it would be wise to ditch the compromise the EC achieves.

Other voting system reform

In particular I support the idea of "Preference Voting" in which individual votes don't just mark there first choice, but actually rank the candidates on a ballot from first choice to last choice. If the system is designed properly it eliminates the situation where one might vote for the "lesser of two evils". (If it is designed improperly, like most instant run off systems, then it doesn't actually improve things). Let me just say that I support a tallying system for Preference Voting called Condorcet.

I will write more on this, probably in a second document with links to good resources, at some later date, but let me provide one source of technical information about voting methods.

And a note to Nader voters

I think you were wrong to agree with Nader. I consider him very wrong on many extremely important things, but I passionately dislike the electoral system that tries to make people choose the lesser of two evils instead of voting for whom they support. People shouldn't have to hold their noses when the vote. And don't believe that it has to be that way. The Condorcet voting system would fix that.

So, Nader voters, although I disagree with your candidate and political outlook and may be very unhappy with the consequences of your votes, I support your decision to vote for a thrid party candidate in the face of an unfair voting system. You refused to let your vote be corrupted by an unfair voting system. Let's support electoral reform so that nobody has to let their vote be corrupted.

Most of this applies not just to Nader voters, but to all who voted for third parties. The only thing that makes Nader voters special is that they may have swung the election, and so are being presented as "spoilers". Yet, all they did was vote honestly. Is a system where voting honestly makes you a spoiler democratic? A democratic election should come to a collective decision based on the preferences of the electorate. Surely the electorate should state their preferences honestly.

Blowing smoke

• Myth: Don't mess with the consitution
  Answer: Unless you want to reform the electoral college, fixing the voting system does not require any change to the US constitution. And it can proceed on a state by state basis. Indeed, if you consider that the current system really denies an honest vote for supporters of third parties, then the current system is in violation of the spirit ofthe 14th amendment (one person one vote). So reforming the voting system as suggested here would be making practice more constitutional.
• Myth:Proportional Representation (as in Isreal) leads to unstable coalitions with extremists.
  Answer:That can be true, but the proposed system is not Proportional Representation Indeed, it tends to have the opposite effect proportional representation. This is because Condorcet will reject candidates who are widely disliked, but still allow votes for minority positions to count.
• Myth:It's too complicated
  Answer 1:So are the details of the electoral college, but nobody fusses about that, do they?
  Answer 2:What voters have to do is simple. All voters need to do is mark their preferences, 1st, 2nd, 3rd, etc. Even if the tallying method is complicated (which it really isn't, see next answer), voters don't have to perform the count. All they need to know is that in this system they really are better off voting their sincere preferences.
  Answer 3:For the most part, Condorcet is simple. Each voter marks their preferences, lets say among three candidate, Alice, Bob, and Charlie. Finding the Condorcet winner is simply finding the candidate which defeats all of the others in head to head races. If more people rank B over A than A over B that means that B defeats A in head to head races. If B also defeats C in head to head races that makes B is the winner as Bob defeats all opponents in head to head races. The only complication is that sometimes there may not be a "Condorcet winner". That is, sometimes, A may defeat B and B may defeat C and C may defeat A, so no candidate defeats all others in head to head races. The Condorcet method has various ways of resolving those situations, and those are complicated. But those situations are very unlikely to come up.
• Myth: Well, what about Borda, Hare, Single Transferable Vote, Alternative Vote, etc?
  Answer:There are lots of confusing alternatives to the Condorcet method. And this is probably where most smoke will be blown, simply because there are well-meaning reform minded advocates of those methods. The problem with those listed (which fall into the broad category of "Instant Run-off Voting" (IRV) is that they don't actually eliminate the problem of voting for the lesser of two evils. There are a few reasons why these terrible methods are even considered.
  1. Historically, people ran run-off elections in an attempt to solve the problems of the current system (it doesn't work though). And then later people figured that if they could get voters to list preferences they could avoid having to have multiple ballotings for run-offs and do instant run-offs. All of these methods evolved from tinkering with run-offs. But run-offs never solved the problem in the first place.
  2. The earliest forms of these systems were designed for hand counting of paper ballots. Ballots were sorted into piles and then the ballots from the shortest piles got redistributed to other piles. Condorcet on the other hand is not as well suited to counting that way. When ballots are calculated by computer, however, Condorcet is easier to count (but who cares if the computer needs to do a bit more work).
  3. Lots of academic and professional societies (like the American Psychological Association, etc) started using varients of these methods, so lots of professionals have a passing familiarity with them.
  4. It is possible that they are used deliberately to blow smoke. Recently the British government has introduced such methods for elections to the relative powerless Scottish Parliament, Welsh Assembly and used proportional representation for the even less powerful European Parliament. While they have gone back on any hint of previous pledges to look at voting reform for elections that matter. One might suspect that they are deliberately trying to discredit Preference Voting.
  (And what about the "approval" method, which is not an IRV? I'll write more about that later).
• Myth:If it ain't broke, don't fix it
  Answer:It is broke. If in a democracy you have to hold your nose when you cast your ballot, then that democracy is broken. If in a democracy you have to lie about your preference on your ballot, then that democracy is broken. If in a democracy a candidate wins not because they are the most preferred, but because the others split the vote, then that democracy is broken. It's broken. Let's fix it.
• Myth:Don't just extremists want this?
  Answer:It is certainly true that third parties have the most to gain by getting rid of the unfairness of the current voting system. But Condorcet actually promotes the selection of compromise candidates with nobody having to compromise their ballots. (See the Proportional Representation Myth) For the record, I tend to genuinely prefer the candidates from the two main parties over those from third parties (although in my youth I actively supported left-wing third parties). I may disagree strongly with fellow reformers on the right and the left (and the libertarians), but I want more mainstream views to win on their merits, not by a rigged system.

Over the next few days [From Noveber 8, 2000 when this was written], I will put in links to resources about these issues, but I am a bit tired from a sleepness night and have other things to do as well. [In June 2001, I still haven't done that. Maybe I still will.]

Version: $Revision: 1.12 $
Last Modified: $Date: 2003/01/15 06:29:56 $ GMT
First established November 8, 2000
Author: Jeffrey Goldberg