That other diary is right. The two party system is great for corporate lobbyists, and horrible for ordinary citizens. The lobbyists have two parties vying for their money, and when they manage to buy off both, the citizens have no real choice on the ballot - except an ineffective protest vote for a third party.
So, is the solution to start a third party? Not so fast. History is littered with the (figurative) bodies of those who tried that without reforming the system first.
There are several important areas for reform. We have to rein in corporate cash and spread public election financing - especially after Citizens United. We need to end filibustering, and hopefully even move to filibuster-proof proportional representation such as multimember districts (which are not unconstitutional, but are federally illegal right now). We need to push the media away from its pathetic two-way-horse-race framing of politics, which systematically excludes any view outside the two-party "spectrum". But this diary is about a different reform, one which would help provide an underpinning for all of those: how do we count the votes.
In most places in the US, we use plurality voting, or as some call it, "pick one and shut up." You know the drill: if the best candidate happens to be from a third party, you must choose between an ineffective vote for them and the a bad-tasting vote for the less-evil of the two frontrunners. The result: a duopoly of power. The phenomenon is so inevitable, there's a name for it: Duverger's law.
And that's not the only problem. Since everybody knows that only the frontrunners can win, that status is a self-fulfilling prophecy. That makes early advertising especially important; and the resultant need for money all too often means that corporate PACs can pick the candidates.
It doesn't have to be that way. There's several ways (see below) to reform the balloting process to even the playing field for third parties. And I do really mean even the field; this would not give any unfair advantage to third parties, and it's virtually certain that the established parties would still win most of the time. But these reforms would help keep them honest, because nobody would be forced to vote for their favorite hideous space reptile.
As I said before, voting reform would help support the other necessary systemic reforms. Public financing would be easier, because if it's possible for a lesser-known candidate to win, the desperate need for advertising money would be reduced. It would help end gerrymandering, since third parties would be real factors especially in "safe" districts, and because the exposure to better ballots would help voters be open to proportional representation. And it would push the media towards more issue-based coverage, and the candidates away from purely negative advertising.
Better voting systems: the options
Here's my opinionated summary of some good systems*:
- The simplest is
approval voting - just vote for as many candidates as you wish, and the one with the most votes wins. While no voting system is perfect, this is truly one of the best. Just count all the votes - it's simple to explain, easy to administer, and gives excellent results.
Its only real downside is that it does require voters to choose between voting for just their favorite ("bullet voting") or also approving a "lesser evil"; but because it would never be sensible to vote for a lesser evil INSTEAD OF your favorite, the two-party deadlock would be ended.
-
Majority Choice Approval is a little better, but also a bit more complex. It's similar to approval, but instead of just approving or disapproving each candidate, you can also especially prefer some of them**. Here's how it works: If just one candidate gets a majority (50%+1) of preferences, they win. The same happens with approvals. If two or more candidates get a majority of preferences, they go into a runoff (plurality based, since there would essentially never be 3 such candidates). If several candidates get a majority of approvals, the one with the most preferences wins. And if no candidates get a majority of approvals, the top two on approvals go into a runoff. Runoffs would be an infrequent affair - essentially, they'd only happen when voters can't figure out which two candidates are the frontrunners.
The exact MCA system I describe has never been used in public elections, but it is similar to some systems which have. It is a form of Bucklin voting, which was called "American Preferential Voting" when variants were used in several US cities during the Progressive era. It's also actually somewhat reminiscent of the system which used in the middle ages to elect the Pope; a system which stood up well, considering the inevitable context of extreme intrigue and corruption. Like Approval, it can be easily hand-counted or counted using any existing voting machines. Unlike approval, it doesn't force you to choose rank a "lesser evil" alongside your favorite, since you can simply prefer your favorite and approve the LE. Also unlike approval, it guarantees that the final winner has an actual majority of the votes.
-
Range voting is where you simply rate the candidates on some scale - say, from 0-100 - and the one with the highest sum wins. It's the system that Olympic ice-skating judges use, or any website which asks you to rate things from 1-5 stars. It works extremely well with honest voters, with strategic voters (who rank everything at the top or bottom of the range to increase their voting power), or with anything in between - that is, any randomly-distributed fraction of honest and strategic voters. In simulations, it leads to the greatest "overall happiness" of the simulated voters of any system. The only downside is that when one side is honest and the other side is strategic, the strategic side has an unfair advantage.
- Another good option is
Round Robin Voting systems (RRV, aka Condorcet). This is where you rank the candidates in order of preference, every possible race is tallied separately in a matrix, and the pairwise champion wins. Different RRV systems have different "tiebreakers" for the "circular tie" situation where A beats B, B beats C, but C beats A, but in real conditions (almost always at most three strong candidates) these tiebreakers would give the same results, so the difference is academic**.
RRV is arguably the "right" answer; if one candidate could beat all the others one-on-one, shouldn't they win? It does, however, have two downsides. First, it's a bit more complex to vote and count than the reforms mentioned above; some voting machines couldn't handle it without modifications. Second, it is sometimes rational to "betray" your honest favorite candidate and strategically vote a lesser evil. Although rational cases for dishonest strategic voting would happen rarely and be fiendishly difficult to figure out from polling data when they did, some overclever voters might do it too often and throw off the results.
- Finally, I come to perhaps the best-known option,
Instant Runoff Voting (IRV). As with RRV, you rank the options in order. The top-rank votes are counted, and if one candidate has a majority, they win. If not, the weakest candidate is eliminated, and a "virtual runoff" is held, counting all the ballots which voted for the eliminated candidate as votes for their second choice (if any). Eliminations continue until there is a majority candidate; of course, when there are only two candidates left, one of them must have a majority.
While IRV is an improvement over one-round plurality, it is not the best available method. If a moderate "compromise" candidate is eliminated, they can lose - even if they would have beaten either of the "stronger" candidates in one-on-one elections. This pathology showed up when IRV was recently implemented in the three-party town of Burlington, VT - and it led to IRV's repeal. Thus, you can still be forced to dishonestly vote for the "lesser evil" to keep them from being eliminated. (IRV hides this fact a bit, because if your favorite candidate is truly a minor player, you can vote them safely. But precisely when it matters, you're back to the "lesser evil" dilemma.)
Why is IRV better known than these other, better, voting systems? Because there are a few interlocking organizations - notably, FairVote and the Center for Voting and Democracy - which are dedicated to promoting IRV. They made their choice when the problems with IRV were less understood, because IRV is more similare to STV, a form of proportional representation which they also support. And now these organizations, and their leaders, are completely closed to supporting the methods now known to be better. Meanwhile the people who support Approval, MCA, RRV, and other better systems tend to be better mathematicians than activists. (Just contrast FairVote's website with that of the Center for Range Voting, and you'll see what I mean. One is slick, the other is earnest.)
The legal and political situation for voting reform
Voting reform activism, like most activism, is an uphill battle. But it's not a lost cause. Existing third parties are naturally strong allies. And even the major parties are interested; by removing the spoiler problem, voting reform actually helps the stronger of the major parties in a region.
So, who is against reform? Lobbyists. Anything which increases citizen power, decreases their backroom pull. But, shortsightedly, supporters of the minor party in a region sometimes oppose it as well, since their hoping for third-party spoilers on the other side to help their chances. Thus, when there was an initiative for IRV in Alaska in 2000, the League of Women Voters there (ostensibly nonpartisan, but definitely Democratic-leaning) broke with the research of the national LWV and opposed the initiative.
Fairvote/CVD has been successful in getting reform - unfortunately, IRV, the minimal improvement - adopted in over 10 US cities, starting with San Francisco. Unfortunately, the problems with IRV led to its repeal in Burlington, and a trial period being abandoned in Aspen.
There is momentum building for Approval Voting in Colorado. It's strongly supported by third parties there, especially the Libertarians; and this has gotten a cautiously-approving mention from the Secretary of State:
"Buescher would be willing to cautiously experiment with what is called
“approval voting,” which allows voters to choose more than one
candidate in a race. The idea is favored by Libertarians and other
minor parties who often siphon votes from Republicans and Democrats."
The IRV implementations have helped clarify the legal situation for voting reform. The problem is, during the rollback after the progressive era, when party machines were taking power back from the citizens, there were a couple of poorly-reasoned court decisions against voting reforms; notably,
Brown v. Smallwood in Minnesota. However, voting reform has stood up to all recent court challenges, and this has showed the way to avoid the "problems".
For use in presidential elections, reformed voting systems would have to mesh with the Electoral College. Luckily, the National Popular Vote movement has already shown the way to do that: an interstate compact. A state could potentially sign on to both the NPV compact and some improved version which used, say, approval voting. The strongest compact with a majority of electoral votes signed on would be implemented.
Footnotes
*You can experiment for yourself and see how IRV (called "Hare"), RRV (called "Condorcet"), Approval, Plurality, and Borda (a system I don't recommend, because of its vulnerability to strategy), using this impressive flash-based voting simulation by Ka-Ping Yee, a nonpartisan voting security expert. Using Yee's assumptions about the electorate, MCA would function identically to Approval. Especially notable are the IRV pathologies with 3 or more candidates.
** If you care, "Minimax" is the simplest round-robin tiebreaker, while "Schulze" is the theoretical best.