Tuesday, April 28, 2009

India's Elections

Those who follow international news will already know that India—the most populous democracy in the history of the world—is currently in the middle of its (nearly month-long) federal elections. Poli-Tea has pointed me to a pair of dueling editorials on whether India would be better-served by becoming a two-party country (like the United States) or by fostering as many parties as possible.

First, some background. India uses single-winner plurality districts to elect all (well, all but two...) members of its lower parliamentary house, which if you've been reading the blog you know means it will tend toward two-party dominance. But currently, India has a lot of parties, and many of them actually have representation. Three facts contribute to this apparent contradiction. First, parliamentary governments better resist two-party domination (as compared to presidential governments like the United States has). Second, India has been a Republic for a relatively short time, 60 years, and the pressure to fall into the two-party trap operates over multiple election cycles. Third, and perhaps most important, India is a very regionally-divided country; most of those parties are very regional-focused in nature, and only six meet the rather low requirement to be listed as "national parties", that requirement being simply recognition in just four of India's 28 states.

The two-party side argues that, with so many parties, it was difficult to form a government (under a parliamentary system, a majority of the parliament must come together to form the government—what we would call the administration—and put forward a prime minister), or that government coalitions would collapse–usually over budget issues–necessitating a new election. And there's precedent for that fear, because it has happened at least three times since the year 1989. He goes on to say that, with just two parties, you always know the winner has a true majority (well I guess that's true), that the government will be less extreme (probably not), and corruption will decrease (I doubt it).

The multi-party side argues that two-party domination means many far-flung regions will have their issues ignored and eventually politics will devolve into a one-party dominated system. And there's precedent for that fear since one party, the Indian National Congress, dominated the government until the year 1989. He goes on to argue that regional parties better serve their regions (no argument there), and that perhaps a three-party system would be best (unlikely).

But there's more than these two options. What I think (as a mostly uninformed American) is that India could be well-served some form of proportional representation (PR). They already have a work-around to try to shoehorn a minimal amount of PR in: some seats are reserved for specific castes and tribes. PR has worked well in many other parliamentary systems to preserve regional differences while still leading to effective governments (Ireland's lower house and Australia's upper house come to mind). It might be something worth considering.

Specter Boogaloo

In case, you hadn't heard the news, Senator Arlen Specter of Pennsylvania has switched parties, from Republican to Democrat. There's plenty of analysis out there about what that means for this congress, but I want to use it to make some points about the two-party system.

Pennsylvania is an interesting state. Until today, it had a pro-choice Republican and a pro-life Democrat as its Senators. This, I think, perfectly illustrates one of the major problems: on the issues of the day, there is a strong incentive (mostly money in the form of party campaign contributions, but other incentives as well) for politicians to adopt the full party-line on all issues. A pro-choice Republican is noteworthy because it is such an exception. But what thread runs through a person's views on, for example, abortion, gun control, immigration, and corporate taxation? Is there any reason that a person's view on those four issues should go together with such high regularity? I don't see one. And yet, from the 16 possible combinations of stances on those four issues, typically voters only get to choose between two pre-defined packages.

Let's be clear on something: Specter is leaving the Republican party for one reason and one reason only, and that's because polls show him losing by 20% to his most likely challenger in the Republican primary (Pennsylvania has closed primaries; only registered Republicans get to vote in the Republican primary, and 200,000 moderate Republicans jumped ship to vote in last year's Democratic pressidential primary).

Changing parties is his only hope for survival. But he's always been on the left outskirts of the Republican party, ever since the last time he switched, from Democrat to Republican. He's a man whose views just don't line up perfectly with either of the two major parties. Which, while making him an exception in Congress, makes him part of the American plurality: more Americans identify as neither Republican nor Democrat (38% versus 35% Democrats and 21% Republicans). So if most people are independents, why are barely one-third of one-percent of congress independents? It's not a conspiracy, it's the system. The system we chose for our elections—singe-winner districts chosen by plurality vote—always tends toward two-party domination. And Specter, like most Americans, finds himself stuck in the middle, not liking either option.

But there's good news: we can change the system. We can use different voting methods that don't suffer from two-party domination, such as score- and approval-voting, or we could change from single-winner districts to a system of proportional representation. There will be opposition; afterall, it's not easy to change the law when all the law-makers owe their position to the law you want to change. But today, I hope, we have received a new ally in this fight. Please, contact Senator Specter to bring this issue, to bring the solution to his own troubles, to his attention.

Monday, April 20, 2009

Mistakes Squared: North Carolina

Prompted by a reader comment, I did some research on North Carolina's instant runoff voting pilot program, and I am simply stunned. It's no secret that I'm not an IRV fan; it fails to fix the "spoiler" problem, it fails to help third party candidates, and it fails in a host of other ways. But what they're doing in North Carolina goes beyond even that. They've actually managed to create a system that is even worse than IRV.

Attack of the Clones

One of the few things IRV actually has going in its favor, and one of the few ways it actually improves over plurality, is that it is immune to "cloning". To abuse cloning under plurality, you prop up candidates who are ideologically identical to your opponent; i.e., clones. All else being equal, this will lead to voters splitting their ballots among the clones, hopefully (for you) more or less equally. So, instead of you losing 40% to 60%, if you introduce a clone, you'd win 40% to 35% and 25%. In practice, things are a bit more complicated, and usually results in, for example, Republicans paying for Green Party ads to shave a couple points off the Democrat in a close election (or Democrats paying for Libertarians).

IRV actually fixes the cloning problem, because 100% of the clone-first voters will have your original opponent as their second choice. So, when the clone is eliminated, you're back to losing 40% to 60%, where you should have been all along.

Short Circuit

But North Carolina added two short-cuts to the IRV process. One, they limited a ballot to listing at most three candidates. This re-introduces the threat of cloning, because if there are four ideologically similar candidates on the ballot, a voter can't list them all, so the incentive to list one of the major two parties highly is even stronger than regular IRV. Two, rather than eliminating just one candidate at a time, as is normally done for IRV, North Carolina has decided that they will eliminate, all at once, all except the top-two. This makes the already vanishingly-small odds for a third-party candidate under IRV an order of magnitude worse, as there's only one opportunity for second- and third-choices to cascade together.

Consider an election where two "third" party candidates each have 17% support, and lets say they are effectively clones of each other, with the voters for each listing the other as their second choice (even if all they agree on is "anyone but the Republicrats"). The two major parties evenly split the remaining votes at 33% each. Even though the third-party coalition has a 34% total—more than either major party candidate has—both third party candidates are eliminated immediately. Even though under normal IRV, only one of them would have, and then one of the major party candidates would have been. Every third party effectively becomes a spoiler for every other third party, while the major party candidates are insulated.

Gentleman's Agreement

What this amounts to is a gentleman's agreement between the two major parties whereby they agree not to prop-up third party candidates as spoilers (even if it's cost effective, it just feels like cheating I guess), while simultaneously making it harder for a coalition of third parties to act together against them (which is unlikely under IRV, but still more-likely than it was under plurality). And thanks to IRV's inability to eliminate the spoiler effect, they can continue to have a virtual guarantee that their duopoly will continue.

As much as I rail against it, I want to state that, by the metric of Bayesian regret, IRV is an improvement over plurality (at least, as long as some voters vote honestly rather than strategically.) This is probably also an improvement (I haven't run the simulation) but certainly less of one than normal IRV.

It Can Always Get Worse

Even if this systematically-assured lack of change weren't enough, NC still has to deal with all the normal problems of IRV: difficulty counting, costly machines and software, inability to centrally count, etc.. But there's one thing that's even worse: North Carolina wants to use IRV in multi-seat elections.

First, a brief history lesson: IRV was invented as a single-winner analog to a multi-winner method called "single transferable vote". It's actually been pretty successful and effective. But NC is not using STV, they're using a home-grown and absolutely insane system that is so convoluted I can't even begin to analyze it (mostly because I start crying before I can finish reading the description).

A New Hope

There is hope, though. This is just a pilot. If NC is open to the idea of trying new voting methods, maybe they can be convinced to give score voting a try. Heck, maybe they could be convinced to try actual instant runoff voting and single transferable vote too, instead of these abominations. Almost anything looks good when compared to plurality, but time is running out on this opportunity; lets get some real options out there for North Carolinians to look at, and let them make an informed decision.

Wednesday, April 8, 2009

Let's Talk

First, let me say that I still get absolutely thrilled whenever I get comments; I've only been at this for a few months, and I've been really squeamish about publicizing it. I got one comment though in response to my last post about how IRV fails to fix the spoiler problem or grant success to third parties. And it really got me thinking. So let's talk:

IRV has elected Progressive Bob Kiss as the mayor of Burlington, VT and Ross Mirkarimi, a Green, to the San Francisco Board of Supervisors. While score/range voting is a worthy system, third parties in this country have enough problems with credibility without trying to advocate for a voting system that has never been used anywhere in the world.
--Anonymous

I Owe It All to IRV?

This comment bring up two points, and they're both worth responding to. The first is the assertion that IRV does fix the spoiler problem and allow third parties to succeed, as evidenced by Bob Kiss of the Vermont Progressive Party winning the election. I have three responses to that. One, if that's true, than plurality voting also fixes the spoiler problem, because plurality elected Bernie Sanders mayor of Burlington, VT in 1981 when he was the Progressive candidate. Obviously, that's ludicrous; if plurality was immune to spoilers, none of us would be wasting our time trying to fix the problem. No, the truth is, that Burlington (and San Francisco as well) is something of an exception in American politics, in that it has a very strong "third" party that is already capable of winning elections. But IRV has virtually nothing to do with their success.

My second response to this first point is a bit more flippant, but makes the same point: in Burlington, the Republicans are the third party; something that I went into more detail with in the too-long post I referred to in the previous post. (Mental note: maybe I should try to turn Burlington's Republicans on to score voting. Or would that be overly antagonistic?)

Third, and more substantive, if IRV were actually any help to third parties, then you would expect a country that has used IRV for years, like Australia, to actually have some in their legislature. But of the 150 Australian legislators, currently every last one is a member of one of two parties. (Technically, the Liberal party and the National party are different, but the two have been in a tight coalition since 1922, and are currently working on a merger plan that has already succeeded in the state of Queensland. In the minds of virtually every Australian, they are a single party. Furthermore, the three "independent" representatives were all, at one time, members of the National party, before leaving over intra-party disagreements; a situation very similar to "independent" US Senator Joseph Lieberman's position with the Democratic party.) And let's not forget senator Bernie Sanders; the same Progressive who won a plurality election to become mayor of Burlington also won a plurality election to become a senator from Vermont; by this logic, plurality is better for third parties than IRV!

But It's Doing Nothing Right Now!

The commentator's second point is that, even if IRV isn't helpful at fixing the spoiler/two-party problem, it's still a better idea than anything else out there because it's actually being used already. This is, one, foolish, and two, a bit disingenuous. It's foolish to put all of this effort into making a huge change, knowing that it doesn't actually fix the problem. Besides being a clear waste of effort, it creates a hostile environment for any future real improvements; once an electorate has been bitten by a false "fix-all" solution like IRV, they become much more shy about trying some other solution. Many locales that try IRV find themselves slipping back into the even-worse world of plurality; North Carolina, for instance, has a pilot program that is facing some strong motions this week to go back.

Besides, the truth is score voting has been used for a very long time, and is a system people are already quite familiar with: from Olympic judging, to movie and restaurant reviews, to Hot or Not. We use it every day; we just haven't used it to democratically elect our leaders. At least, not recently. But the two longest-running democratic governments in world history did. For 500 hundred years, Sparta used a (loud and very analog) score voting method for all their elections, and for another 500 years, Venice used a three-valued (+1, -1, 0) score voting method. IRV, on the other hand, was first used in 1893. Call me in three or four centuries.

The Real Winner

But really, it's not important which idea is older, or was used for longer. Even clearly showing that IRV can't completely fix the spoiler problem doesn't show that it's, overall, no good. Taken alone, each of these arguments can be brushed off as simple anecdotes, and we want data (and the plural of anecdote is not data!) Spoilers are bad, because it means a lot of people get something they strongly don't want; two-party domination is bad, because it restricts voter's choices to "the least of two evils" and we often don't want either. But how bad? Which is worse? How often is it a problem? Warren Smith's insight was to try to side-step all the anecdotes, and directly measure "how bad"; that's the idea behind Bayesian regret. And even he was honestly surprised at just how overwhelming the data was.

Now, maybe the simulation is flawed. But maybe it's not. But it's certainly compelling enough to suggest that score voting would be something worth trying. Meanwhile, everything CRV has said would happen with IRV, has happened, including spoilers and continual failure for third party candidates. So again, I would urge you: if you're supporting IRV because of spoilers, or because of third parties, you should STOP supporting IRV. Other options do a better job of fixing the issues you care about.

But if you're still not convinced, I'm game to continue the debate, so keep the comments coming!

Monday, April 6, 2009

FairVote's Flash

As I mentioned, off handily, in an earlier (and far-too-long) post, FairVote has a not-too-shabby video showing off how instant runoff voting "fixes" the spoiler problem.

At the start of the video, we're presented with two candidates, A and B, and shown the results of the election between them: seven for A, eight for B; a narrow win for B.

Then, the plot thickens, and a new candidate, C, who agrees on most issues with B, is introduced. Three of B's voters, and one of A's, decide they prefer C, and we are shown the TERRIBLE TRAGEDY of how C's presence caused A to win instead of B, a clear SUBVERSION OF THE DEMOCRATIC ORDER. And I agree; that's terrible. Adding a candidate who is not a winner themself should never change who the winner of the election is. Academically, this is called independence of irrelevant alternatives, or colloquially, the spoiler effect.

"Luckily" (according to the video), instant runoff voting saves the day, because it eliminates C, and redistributes their votes back to A and B, and the victory returns to it's proper place, with B.

But I have a question for FairVote.

What if C was just a bit more popular, and managed to convince just one more of B's supporter's to vote for them?

Now, B, having the fewest first-place votes, is the candidate who is eliminated. And, depending on B's voter's second-choices (which we don't ever see in video) either C or A may win. If two or more of the four prefer A over C, then A wins. Which is the same SUBVERSION OF THE DEMOCRATIC ORDER which instant runoff voting was supposed to guarantee never happened again; an "irrelevant alternative" changed the outcome of the election. That should never happen!

IRV's advocates are trying to popularize a false notion (or maybe they're just trying to deal with their cognitive dissonance) that there's some magic cutoff for "significance", beyond which a candidate doesn't count as a spoiler. But if it quacks like a duck, then it's a duck. I believe they may be confused because they (and all of us) are most familiar with plurality voting, and under plurality third parties virtually never get more than about 15% of the vote. There's a good reason for that value; it's that the difference between the Republican candidate and the Democratic candidate in any election is virtually never more than about 15% of the vote. Third parties tend to grow, until they get big enough that they spoil an election, and then they're destroyed in a terrible backlash (Nader has yet to recover from Gore's loss; come to think of it, neither has Ross Perot from Bush I's). The only thing IRV changes is what that cutoff-limit is; it goes from about 15% to closer to 30%. But don't let them trick you: it's still there, waiting for a fatted third-party to grow large enough to sacrifice itself upon its altar, and SUBVERT THE DEMOCRATIC ORDER by spoiling an election.

If you're an IRV advocate because you want to eliminate the spoiler effect, or because you want third-parties to have an actual chance of wining elections, then you should STOP being an IRV advocate. There are other voting methods that better achieve your ends. Score voting actually does completely eliminate the spoiler effect, which means third-parties have a real chance to win.

Wednesday, April 1, 2009

Look at Other Options, Too, LA!

The LATimes' blog is reporting that L.A. County is going to look into Instant Runoff Voting. Currently, they use a plurality election, with a top-two runoff election if no one gets a majority, but the added cost of a second election and the lower turnout for it, have them looking elsewhere.

Please, L.A. County, take a look at some other alternative election methods, like score voting. If money is the worry, it's a winner:

And that's ignoring the utilitarian benefits: score voting reduces two-party domination, which means better options for voters, which means better government.

EDIT: Oh, and if turnout is the worry, remember that IRV also suffers from potential no-show paradoxes!