Unrepresentative Democracy, Part 1: Instant-Runoff Voting

Canadian subtitle: Harper’s fake majority

American subtitle: How’s that two-party system working?

Canadian billionaire’s response: Hey, we’re accomplishing some real damage here!

American billionaire’s response: Great 🙂

I’m so excited for this. I finally get to write about electoral systems! I really like math, democracy, and fairness, so this is one of my favorite topics.

Here’s the thing: The United States’ and Canada’s electoral systems are among the worst (least democratic, logical, and fair) in the “developed” world.

Let’s start South. As we know, the US political system is dominated by two parties that do a pretty bad job of representing the political positions of a large part of the citizenry. But no challenge to their dominance is possible when every federal and state level officeholder is elected by the first-past-the-post (FPTP) system, in which the candidate with the most votes wins, even if they don’t receive a majority. In this system, voting for a third party is useless in terms of actually winning a seat. Almost never will a sufficient quantity of voters cast a vote for a third party, knowing that doing so eliminates their ability to exercise a preference between the main party candidates. This results in the “spoiler effect” in which the presence of a third-party candidate is understood to have siphoned votes from the more similar of the two major party candidates. Remember everyone blaming the Nader (Green Party) voters for Bush’s “victory” over Gore in Florida in 2000

It seems clear we need a system where people can vote for the candidate they truly prefer without helping the worst candidate win.

The most common solution to this, globally, is to have a two round runoff election. If no candidate receives 50%+1 of the vote in the first round, the top two vote-getters compete in a second round at a later date. This partially eliminates the problem of spoilers, in that you can express your preference for a third candidate in the first round and still participate in the choice between the top two in the second round. However, spoilers are still possible when a potential winner loses enough votes to minor candidates in the first round that they don’t even reach the second round. This happened in the French presidential election in 2002, when the left vote was so fragmented/diverse in the first round that the two candidates who advanced were the center-right and far right candidates, even though the Socialist candidate likely would have won the election if he had reached the second round.

The two round system also has the significant problems of frequent lower turnout in the second round (in the mayoral race in our town of Athens, GA in 2010, turnout in the runoff was less than half of the first round) and increased cost vis-a-vis single day voting systems.

Two round systems are pretty clearly better than first-past-the-post, in our view. But is there a better system?

Instant-runoff voting (IRV)! (also known as Ranked-Choice Voting, the Alternative Vote, or the Preferential Vote) When you vote, you rank the candidates in the order of your preference. If no candidate receives 50%+1 (a majority) of the first place votes, the candidate with the fewest first place votes is eliminated and those voters’ votes are transferred to their second choice. This process (eliminating the last place candidate and transferring their votes) is repeated until one candidate receives a majority. That candidate wins, obviously.

Image

What would this do in practice? IRV means you can vote for the candidate you want and still express a meaningful preference among the other options. If Florida had used IRV in 2000, most of Nader’s votes (upon his elimination) would have been transferred to Gore (as the second preference of those voters) and Gore would have won, reflecting the actual preference of a majority of Florida voters for Gore over Bush. In France in 2002, the smaller leftist parties’ votes would have been transferred to the Socialist as they were eliminated until he eventually achieved a majority, reflecting the actual preference of a majority of French voters for the Socialist over the center-right candidate. In these cases and many others, you would end up with a winning candidate a majority of voters actually prefer. And, you can vote honestly, rather than strategically.

Canada isn’t trapped between a Rock Party and a Hard Place Party. There are three nationally competitive parties on the federal level and a fourth in some places. But because Canada also uses FPTP, this means that it’s exceedingly common for candidates to win ridings (electoral districts) with less than 50% of the vote, and sometimes a lot less. In the 2011 general election, 53% of seats in Parliament were won by candidates who received less than 50% of the vote (in 2008, it was 61% of seats). This is a big part of the reason that the Conservative Party was able to win 54% of the seats in Parliament with only 39.6% of the vote, giving them essentially unchecked power in government even though 3 out of 5 Canadians voted against them. Even a quite conservative analysis of the direction and quantity of ranked preferences strongly suggests that the results would have been significantly less disproportionate (the Conservatives wouldn’t have won a majority) if IRV were in place. In other incredibly disproportionate results, the 1984 election saw the Conservatives win 211/281 seats (75%) with only 50.03% of the vote. In 2001 in British Columbia, the Liberals won 77/79 seats (97%) with 57.6% of the vote.

So let’s see how IRV could make for fairer results by looking at the Vancouver South riding. Rounded to the nearest 50 votes, the results were [Conservative = 19,500, Liberal = 15,600, NDP = 8,550, Green = 1,150, Marxist-Leninist = 200]. So under the current first-past-the-post system, the Conservative won, and he (Wai Young) is the current MP, despite only receiving 43% of the vote.

Let’s imagine how this would have played out with IRV, making educated, simplified guesses at people’s preferences. The last place Marxist-Leninist would be eliminated first, and we’ll give their second preferences to the Green candidate, bringing their total to 1,350. The Green Party candidate is next eliminated, and we’ll assume that their voters preferred the NDP over the Liberals or Conservatives. These additional votes bring the NDP candidate to 9,900. Finally, the now-last-place NDPer’s votes are transferred to the Liberal, giving them [9,900+15,600=25,500]. And we have a new winner! (Clearly, in real life not all M-L, Green, and NDP voters would prefer the Liberal over the Conservative, but due to the margin of victory, the result would very likely be the same.) The Liberal candidate, who 57% of voters prefer over the conservative (and even stronger majorities prefer over the others), would represent the district as the Conservative no longer benefits from the splitting of votes between the Liberal and NDP candidates.

Even better, in practice this riding, and others, would see even more different results as voters no longer need to strategically vote for the “lesser-of-two-evils”. How many Vancouver South voters voted Liberal despite actually preferring the NDP or the Green Party because they felt that the Liberal had the best chance of defeating the Conservative? That kind of strategic thinking doesn’t take place in IRV, where you can express your true preference, while still expressing a meaningful preference for one major candidate over another.

So, Canada and the USA should follow the example of Ireland, which uses IRV to elect its president, Australia, which uses it for its House of Representatives, the Academy Awards (oh yeah), and several cities in the US (San Francisco, Berkeley, Oakland, Minneapolis, Portland of Maine) that use it for local elections. It’s simply the most democratic way to elect representatives in single-member districts.

But there’s more to fair, representative elections than just having a good method for single-member districts. So Part 2 of our Electoral Reform series will focus on the case for (and different types of) proportional representation.

Yay!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

– Adam
from a chair near Josie the cat

Image

 

Advertisements

One thought on “Unrepresentative Democracy, Part 1: Instant-Runoff Voting

  1. Pingback: Unrepresentative Democracy, Part 2: Proportional Representation | sowmanyreasons

What do you think?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s