Twitter Users Are Nerds

Twitter users are nerds (and that includes me). Twitter has a new feature called Twitter Tracking. You can track search terms, and you'll get tweets from any user whose update matches the terms. I just tried the terms sex, MacBook, and Leopard and the first two results I got were for Leopard - and yes, they were referring to Mac OS X 10.5. The third result matched MacBook. Where are all the dirty Twitter users talking about sex?

Are you voting for MMP?

As part of Ontario's provincial election tomorrow, we are having a referendum on electoral system reform. We must choose between the existing "First Past The Post" (or FPTP) system and a "Mixed Member Proportional" system, MMP. I have been waffling on which I will choose for a few days now. Initially, I jumped on the MMP bandwagon, but after watching a debate about the alternatives on TVO, I've started considering things more carefully.

I do have a few issues with the existing system:

1. The number of seats won by each party does not reflect the popular vote.
   This is exactly the issue that proportional representation systems are designed to address. Although only 39 of the 129 seats in MMP would be elected based on party vote, this is enough in many cases to guarantee that each party receives a number of seats proportional to the percentage of votes they received.
2. Voters feel compelled to vote strategically to prevent certain parties from winning seats.
   Suppose parties A and B mostly share an ideology that I support, but I prefer party B whereas party A tends to be more popular. If a third party C has a completely different set of values, I may feel that I need to vote for party A just to prevent party C from winning the seat. Parties A and B are splitting the vote.
3. Women and minorities are under-represented.
   I'm not fully convinced that this last problem is directly related to the electoral system. Rather, I think it stems from the fact that parties themselves use FPTP in nominating their local candidates. Men and majority groups will usually garner the highest level of support, and so parties tend to run fewer female and minority candidates.

So how well does MMP address each of these?

Clearly, MMP solves or at least lessens the severity of the first problem. One caveat is that MMP cannot take away seats from a party that has won an inordinate number for its level of popular support; it can only address the imbalance by awarding the list seats to the party's opponents. For example, in some elections, I believe parties have managed to win in excess of 90% of the seats with about 40% of the popular vote. 90% of 90 seats is 81, which is almost 63% of the 129 seats in the provincial parliament. For me, this is not a huge concern, as having a majority government does help move legislation along more quickly. In fact, this scenario probably makes MMP seem less scary to a lot of people, since minority and coalition governments can be shaky. I think the idea of parties cooperating is nice, but sometimes this means that a small, relatively unpopular party can hold the balance of power. Granted, this example is extreme, and I think we would likely have mostly minority governments under MMP. Even in the extreme cases, other parties still get a voice in the legislature, even if they ultimately hold no power.

MMP doesn't do much to address issue #2. Advocates of the new system will say that it does, since, in my above example, I could vote for my local representative from party A while voting for B in the party vote. It is still possible, however, that party A will get more seats than it deserves, meaning the 39 list seats will not be sufficient for party B and the other parties to have fully proportional representation.

A better system for addressing issue #2 is the Single Transferable Vote (STV) system, which was proposed in a B.C. referendum in 2005 but did not receive enough votes and was not implemented. In this system, candidates are ranked. If no candidate has majority (50% + 1) support using the voters' first choices, the lowest vote-getter is removed from consideration. Then, the votes of those people who voted for this candidate are transfered to another candidate, their second choice. This is repeated until one candidate has majority support. In my three-party example, I could rank the candidate from party B as my first choice followed by the candidate from party A, any other candidates, and lastly the candidate from party C. That way there is no worry about splitting the vote, and I can vote with my conscience. If it so happens that lots of other people felt the same way I did, the candidate from party B will win. Alternatively, perhaps the candidate will be dropped in the first round, and my support will be transfered to party A.

MMP advocates say that proportional representation (or PR) systems always elect more women and minorities. As I mentioned before, I think this has to do with how parties choose their candidates. I can see how choosing candidates for a list would likely result in a broader range of representation. When forced to choose just one candidate in each riding, a white man is likely to win almost every time. So on the one hand, I can't really argue with the statistics. On the other hand, I don't really think it's fair to blame the system. If we got more women and minorities interested in and involved with politics, especially at the party level, we would nominate more diverse candidates.

My biggest concerns about MMP are related to the party lists. One problem is the granularity of choice. When voting for a party, we get to see the list of list seat candidates in advance. But what happens if we disagree with a party's top choices, even though on the whole we support that party? Assuming parties will use open and democratic processes for compiling the lists, one could join his or her party of choice and vote for better candidates, although that is probably more involvement than most people care to have.

Another problem with the party lists is in deciding which candidates will run locally, and which will be on the general list. I didn't do an exhaustive search, but I was unable to find out if the party list can include individuals who are also running in a riding, so I am assuming that this is not allowed. Consider a small party like the Green Party with low popular vote and geographically dispersed support. It would be in their best interest to run their strongest candidates on the list, and perhaps not even bother to run any local candidates in the ridings. With the Liberals, the situation is very different. They are likely to run their best candidates in local ridings, scraping the bottom of the barrel to fill the party list. What about other parties between these two ends of the spectrum? I think choosing the list members could result in some political in-fighting, and I think that in many cases, having two types of candidates will mean that we don't always elect the best ones.

My ideal system is something of a hybrid system that addresses issue #2 that I have with FPTP and also eliminates the party lists and the concerns I have about them. I like to think that I invented it myself, but I am not well-read in electoral systems, so it's likely that others have thought of this before. In my system, which I'll call Transferable Vote Listless Mixed Member Proportional (TVLMMP), each person still gets 2 votes - one for a local representative and one for a party - but the vote for the local representative is a ranking, as in STV. This eliminates vote splitting and strategic voting. There would still be 90 ridings and 39 list seats, but rather than choosing the list seat representatives from a list, the seats would be awarded to the local representatives who received the highest proportions of votes in their ridings without winning the seat. (I guess that necessitates a new name for "list seats"... perhaps "general seats.") This way, there is basically a single ranking of the candidates that is determined democratically by the public, rather than having a separate ranking that is determined (hopefully) democratically within parties. Without a list, I don't have to worry about supporting an unliked candidate that the party felt deserving of a spot on the list. If I don't support my local candidate, but I do support his or her party, not voting for the candidate will decrease his or her chances of getting one of the general seats, while my vote for the party will help to get the next best candidate, as decided by my peers, elected should any general seats be required.

In the end, I think I will probably support MMP, because I still think it is better than FPTP in many ways, even though it doesn't address all of the problems. Another factor is that I don't have much interest in provincial Ontario politics, and so I don't mind experimenting with it. Finally, since MMP will require 60% support and majority support in 60% of the ridings, I think it is unlikely to pass. And somehow in my twisted logic, that makes it the endearing underdog.