February 25, 2004

Third Party Realities

This is an article that was submitted to kuro5hin.org and then rejected due to partisan voting.

Over the last couple of years, I have done quite a bit of research into U.S. political voting systems and various vote-counting methods. I've come up with the following conclusions, some of which I didn't expect:

  1. The presidential race should absolutely not have third-party candidates.
  2. Preference voting is not the answer for presidential voting while the Electoral College exists.
  3. Democrats and Greens need to share fault for 2000, and share the responsibility to join forces in 2004.
  4. Bush technically had more electoral support than Gore even without the Supreme Court's help.

Submitting to Systems; Pragmatic Principles

One of the basic principles that guided my research was the conviction that when one submits to a system, one also has to submit to the pragmatism that that system requires. I can argue all I want about the inequities of our current voting system, but if I'm going to use it (as I should since the political effects will ultimately affect me), I should use it in a way that aligns with my political intent.

For instance, Nader asked us to "vote our principles" in 2000, when many of us had principles that said we liked Nader best, but preferred Gore to Bush. The problem was that our voting system did not have room for these principles. By asking us to ignore our preference of Gore over Bush, those who had that preference but voted for Nader were ultimately casting an unprincipled vote. (Those that believed Nader actually had a shot at winning, or who honestly saw no difference between Bush and Gore, however, were casting principled, if uneducated, votes.) I came to the conclusion that what would have been more principled would be to vote for the candidate closest to my beliefs that had a shot at winning, and then work to implement a voting system that had more room for my principles.

The Electoral College

One system we are forced to submit to is the Electoral College. The first thing to realize about the Electoral College is that it will be very, very, very hard to change it. If 2000 didn't change it, it is hard to imagine what it would require given our current congressional makeup. Replacing it with a nationwide popular vote won't happen as it is too counter to the interests of Congress; it would likely require a different scheme to advocate. At this point, it looks like it would require massive congressional change over a period of years to get enough representatives in there that would vote to overturn it.

Until that happens, it is the only game in town for electing Presidents. And there's one important facet of it to keep in mind: it requires a majority of 270 votes (out of 538) to win.

What this basically means is that for someone to win the Electoral College outright, they have to get more electoral support than all other Presidential Candidates combined. It's not enough to make a good show. You have to decimate the field. Right now when we have two major parties, it's not a big deal, because no other party strongly competes with them.

But imagine what would happen should a Green, Libertarian, or Reform candidate get enough support to actually win a state or two. First, absolutely nothing happens for the third-party candidate. They would have to win at least the eleven most populous states (which includes California, Florida, and Texas; a strange trifecta). But, what would happen is that either the third party would be ideologically similar to one of the two major parties, splitting its support, or it would lead to none of the parties getting 270 votes.

When no candidate gets 270 votes, the election goes to Congress for them to decide among themselves. And unless the third party has strong congressional representation, they are out of luck.

The long and short of it is that if a third party wants to become president, they are either going to have to have a strong enough national party to enable them to win several states outright, or they are going to have to have a strong enough national party that they would have a plurality of congressional representatives in Congress. Either way, it requires a strong national party with significant local and statewide support and a significant number of elected officials. None of our third parties have this level of strong support at this time.

Now, the lazy conclusion would be to think that the third parties should shut up and go away. This is of course wrong. Instead, they should seek to form coalitions.

When more than two parties are represented in parliament or Congress, and when a bill requires a Yes or a No vote, the multiple parties coalesce into two camps. What's important to realize is that coalitions always ensue in democracy. The difference is when they happen.

In parliamentary democracies, the coalitions happen after the representatives are elected. In an effective American democracy, the coalitions can still happen; they just need to happen before the election. This is the advantage the GOP has over the Democrats right now - they have learned this lesson, while the Democrats have not. In 2000, the Greens and the Democrats failed to form a coalition, which is the fault of both parties.

Due to all of this, the reality is that until a third party has a strong enough national party to compete in a presidential election or in Congress, the only way to pragmatically play the game in presidential races is to form coalitions with the party that they are closest to in ideology.

Preference Voting

One way to make it easier for third parties to build nationwide support is to allow preference voting. Preference voting allows people to specify their full preferences and protect themselves from vote-splitting and "lesser of two evils" voting. However, preference voting is not a good idea for presidential elections.

For as long as the electoral college exists, presidential elections will be decided on the state level, state by state. Given the requirements of the electoral college, preference voting only makes it easier for a third party to win a state here and there, throwing the election to the House. This should not be toyed with until a third party honestly has a realistic chance of winning 270 votes worth of states all in one year. For those seeking to implement preference voting, the responsible approach is to seek to implement them for local and statewide races (including national congressional races) in order to build strong national support. But in the meantime, for presidential elections, forming coalitions is the best approach.

And a word about preference voting: it is well-established by now in voting theory circles that IRV is flawed and Condorcet - where the winner beats every other candidate in a head-to-head matchup - is superior. There is also fondness for Approval voting, but there is one problem with this - many state constitutions already make allowances for ranked voting, but not for Approval voting. There is more of an obstacle against it. Approval voting also has a psychological barrier in that people tend to think in choices, and user testing has shown that voters tend to resent giving an equally-weighted vote to their second choice, especially when it results in their second choice winning the election. This has resulted in Approval voting being approved and then thrown out in at least one locality that I've heard of, although I can't find the reference.

National Electoral Support

One of the irrelevant statistics in judging presidential races is the national poll. This is because our presidency is not decided by a national popular vote. Instead, our less populous states are given extra weighting compared to our more populous states, to recognize state sovereignty. This is consistent with how Congress is set up - more populous states get more Representatives, but every state gets two Senators. For presidential voting, every state gets an electoral vote for each Rep and Senator.

A good way to calculate actual nationwide support is to weigh the nationwide popular vote by each state's electoral power. While the implementation of our electoral college is antiquated, this is a way to calculate its actual intent. I calculated this for the 2000 results. I took each state's electoral votes, and split them up proportionally to each presidential candidate, according to how many votes that candidate got in that state. Using this formula, we find for instance that Gore got 12.16 electoral votes in Texas out of 32.

After totalling up all the voting power for each of the candidates, I calculated the "electorally-weighted popular vote", and came to an interesting result. Bush actually outscored Gore. There was no winner take all factor - a swing of a few hundred votes in Florida made no difference. But by applying the ratios to the popular vote, I had a weighted popular vote, with Bush getting 48.17% of the vote, and Gore getting 48.03% of the vote. Bush actually had more electoral support nationwide than Gore did.

Accepting for the moment that Gore actually won Florida, then why did Gore win? There are many inequities in the electoral college implementation, chiefly among them being the winner-take-all nature of each state. And I found that of the seven closest states, Gore got five of them. The other two were New Hampshire, and Florida. The inequities actually worked in Gore's favor, overall.

Now, it is important to remember that these are counting the votes as they were recorded by the various Secretaries of State. And given what we know about Florida disenfranchising tens of thousands of ex-felons (and non-felons), we know that the intent of the voters were more for Gore than the recorded tally shows. I personally believe that the GOP has a fraud advantage of 0.5 - 1%. However, there is not good reason to believe that the fraud advantage the GOP had in 2000 has lessened since then, especially with the onset of blackbox voting.

The basic thing to realize here is that if Democrats want to win in 2004, it is not enough to wish to duplicate 2000 simply with a fairer count. First, there isn't a good reason to expect a much fairer count, and second, the Democrats actually had less electoral support in 2000, only "winning" due to electoral college inequities working in their favor, which is hard to control in future elections. To win, Democrats really do need to build more support.


Overall, this suggests a clear strategy for those interested in supporting Democrats and Greens, improving democracy, opposing the current administration, and having our votes better represent our intent:

  • Recognize that 2004 is about building new support for Democrats and constructing arguments to appeal to ambivalent and swing voters.
  • To support third parties, immediately seek to always run a candidate against an otherwise unopposed incumbent congressional member. Several House races every year run unopposed, and there would be no vote-splitting cost to run a third party candidate against them.
  • Research state constitutions and secretary-of-state policies about preference voting, and lobby your state legislature to allow preference voting for state issues. Start on the local level to familiarize voters. Advocate vote-counting systems where the winner is always from the Schwartz set of Condorcet voting, because vote theorists agree that this is always the most fair result if the intent is to represent consensus public support. Advocate preference voting for local, state, and national congressional elections.
  • For presidential races, convince your third party to become active in lobbying the major party that is closest to its ideology, to extract concessions for your party's support.
  • Oppose third party presidential races for as long as the Electoral College exists, until a third party has significant nationwide support.
  • Oppose opportunity for vote fraud by supporting congressional bills that require paper trails (which are NOT the same thing as receipts that voters take with them, and are therefore not open to the vote-trading flaw), such as Congressman Holt's H.R. 2239 and its companion Senate bill.

There are of course other theoretical and idealistic ways to advocate better U.S. democracy, but at this point they all require either a belief that everything will suddenly change, or a long view of it taking decades to implement. If you subscribe to the long view, work for what you believe in, but consider supporting these shorter-term pragmatic efforts as well.

Posted by Curt at February 25, 2004 08:02 PM


Excellent post, I commented a bit on it here.

Posted by: Michael Williams at March 1, 2004 12:39 PM
Post a comment

Remember personal info?