Sunday, November 30, 2008

A nice piece from the Fair Vote UBC blogsite

There’s another blog on electoral reform that was started by some UBC students in which they wrote an article that kind of follows along one of the themes in my "Why Change to BC-STV" post. In it, they discuss how our First-Past-The-Post system encourages campaign strategies that concentrate on character assassination, and demoralization of the voters of other parties as opposed to trying to appeal to the entire electorate.

Negative Campaigns: Is it in their nature or is the system to blame?

October 21, 2008

Further thoughts on the recent Federal election. One of the reasons people are so apathetic and upset about elections and politicians is the negative tone that is so prevalent. It seems that one of the primary strategies for winning an election is character assassination. If one of your opponents has a perceived weakness then attack that weakness and forget about talking about policy or issues. A good example this past election was Liberal leader Stephane Dion. He had a perceived leadership weakness, which even he now admits. It doesn’t matter if it was true or not, the appearance was there that he was soft, elite, not decisive, not a great communicator. The Conservatives and the NDP went after that. They talked about policy as well, but you could argue they got a lot of mileage out of tearing down Dion. Just look at how much time Jack Layton spent attacking Dion in the debates. Whether Mr. Harper or Mr. Layton were doing it, it hurt Dion and in the end helped the Conservatives form another government.

Is there any way out of this? Isn’t this just because politicians are bad people who have no goodness left in them? Just empty husks of human beings who have sold their souls for power?

Call me an idealist, but I don’t think so. I wouldn’t remove personal responsibility from any choice a political leader makes, but the current electoral system we have provides certain incentives. Politicians are competitive, goal oriented people who optimize their behavior based on the system of incentives that is in place. The fact of the matter is that first-past-the-post voting lets you win government with much less than a majority (in this election, just 38% was needed). So there is no need, no incentive to appeal widely to everyone. The optimal strategy is clearly:
  • mobilize your base
  • try to cause infighting amongst your opponents to encourage vote splitting
  • be vague enough not to scare off all the undecided voters

No one party is to blame here more than another, the system rewards you for negative campaigns that sow doubt based on character. It rewards a strategy of dividing your opponents against themselves. The essence of this strategy is that by dividing your opponents and unifying your side you ensure that many more votes on the other side won’t count. Thats because in our system the winner wins and the losing votes are thrown away. So, under this strategy, the Conservatives don’t need care if the Liberal voter they are convincing votes NDP or Green or Conservative. Regardless who they vote for, their vote will likely not contribute to the make up of parliament or its for you, either way you win, as long as it isn’t a vote for your closest opponent.

So, if we just set the question of the existence of good politicians aside for a moment, we’ll never get a better, more respectful campaign where these strategies aren’t optimal until the system changes.

In a proportional system, even if you attack one leader and shift their votes away, those voters will still be heard. Voters who are torn between two alternatives on the left, for example, will still contribute to the makeup of parliament if there are enough to pass the threshold. And in Canada the Greens and NDP regularly pass this threshold. What’s more, once people know this is how it works they will start new parties or vote for other alternative voices. The only optimal strategy then will be to appeal to the widest population of voters. Furthermore, you couldn’t burn all your bridges with character assassination of your close opponents, because if you want power, you may need to work with them to form government.

So next time you hear someone complaining about negative campaigning and blaming politicians think about how the system motivates their actions. One day we’ll have a better system (like May 12, 2009) Then we can see if politicians rise to the challenge of putting issues and ideas before character and strategy.

Friday, November 21, 2008

Interesting Andrew Coyne Article

Here's an interesting article about how our first-past-the-post electoral system is really only adequate to deal with two political parties at a time, that it is being used to handle five political parties, and failing. And because of it's failures to do so, it creates a situation that encourages the type of behaviour, that Canadians profess to find so distasteful, in politicians and political parties in their desperate attempts to manipulate the electorate into inadvertently voting in a majority. Where, instead of building themselves up, they have to resort to tearing each other down.

What if they gave an election and nobody won?

We now know one thing: this electoral system is broken

ANDREW COYNE | October 16, 2008 |
On Friday, Aug. 29, as the Conservatives were putting the finishing touches on election planning, the S&P/TSX composite index closed at 13,771 — down from its record earlier in the summer, but still at historically high levels, and rising. A week later, just before the election call, it was down 955 points. Four days later, another 670.

Still, for all the turmoil on the markets, the gathering crisis in global finance remained for many Canadians a distant thunder — troubling, but not yet occasion for a cold-sweat panic. For all the stock market's gyrations over the first weeks of the campaign, the trend was unclear, and Conservative support held relatively firm. Through the end of the campaign's third week, the Tories maintained a healthy 10-point lead over the second-place Liberals.

And then everything went to hell.

The economy was not just the most important issue of this election. It was the only issue. It wasn't even an issue, as such: nobody really had much to offer in the way of significant policy differences. It was more like a natural disaster. It was like holding an election in the middle of a hurricane. Nobody much wants to know what your policy is on force nine gales. They just want to know how well you'll stand up in it. They just want to know if you can get them in to port.

All elections are about leadership, to a greater or lesser extent. This one became, as few elections are, a test of leadership under fire, played out in real time. Surprisingly, that did not automatically redound to Stephen Harper's benefit.

In the early part of the campaign, the Conservatives had some reason to hope that a diffuse unease over the economy would work to their advantage: that steady-as-she-goes would be an appealing message, that Harper's image as a "strong leader" would attract undecided voters to his side. But when the storm clouds turned into a deluge, it became clear that the "strong leader" image was built on sand. Harper had had great sport beating up on St├ęphane Dion, he'd shown a tactical mastery of the House of Commons, he'd bullied and bruised virtually anyone he'd come into contact with. Politics engages the primitive part of our brain, and Harper's appeal was that of the lead wolf, "red in tooth and claw." But what did it amount to in the end? All those broken promises, all those abandoned convictions, all those jaw-dropping about-faces — they'd won him the element of tactical surprise over his opponents, but at the cost of any relationship of trust with the broader public.

It turns out that matters. Way back at the start of the campaign, it was commonly framed as a contest between strength (Harper's preferred "ballot question") and trust (Dion's presumed comparative advantage). Apparently strength depends on trust. When the crisis broke, Harper was at first unable to call upon those reservoirs of trust a leader needs if he is to, well, lead. A leader whose appeal was based on always being in control was peculiarly vulnerable when it became clear he wasn't in control — not of the markets, not of the televised debates, timed with cruel precision for the very week of maximum turmoil, where a blinking Harper was subjected to a non-stop, four-party barrage of abuse: you don't care (not true), you're not aware (surely not), you have no platform (you got that right).

It must have been bewildering to Harper. In policy terms, he was absolutely right: the Canadian economy was hardly in as bad repair as the American, nor were the sorts of remedies being contemplated there in order here. We were bound to be affected by the crisis in American finance and there were sure to be hard times ahead, but in the short term there was little that any Canadian government could do about it, and even less that any party leader was actually proposing. Yet for a week or 10 days after the markets collapsed, no one was listening. Or not enough people were listening. Or not the right people: the undecided, the swing voters, the voters in parts of the country Harper needed to reach if he was to achieve the majority he sought.

What happened, in that first flush of public panic, was that everybody returned to their corners. Polling data from Harris/Decima tells the story. Of those who said the economy was the issue that would decide their vote — and there were many more of those midway through the campaign than at the start — a disproportionate share broke Tory, from Manitoba west, while from Ontario east, they tended disproportionately to vote Liberal. The regional and partisan split correlated closely with people's views of the nature of the threat to the economy. Those who saw the threat in more general terms were inclined to look for a steady hand at the tiller, Harper-style — and more of them were to be found in the prosperous West. Those who saw a threat in more personal terms — my job, my future — looked for someone who cared about them. And so was born the brief Dion boomlet, in those parts of the country where people are more inclined to see the government as their protector. He may have been slow to recognize how completely the economy had come to dominate public concerns — at the expense, say, of climate change — but he was quicker than the others.

It didn't last. In the end, Harper was able to pull out a surprisingly strong win — at least compared to where he had been with a week to go — maintaining his party's numbers in Atlantic Canada and Quebec while gaining seats in Ontario and the West. Across the country, the Conservative vote was about four percentage points higher than the polls had predicted. It may have been superior Conservative organization, a more motivated base, a late blast of good news from the markets. But it seemed also that Harper found his feet in the last week.

Oddly, things had to get worse in the financial crisis before they could get better for Harper. Measures that he would have been justified in rejecting earlier in the crisis, such as the $25-billion government airlift of bad mortgages off the bank's books, by the last week of the campaign had become entirely justified, given the alternatives, even to the most doctrinaire free marketeer. Just as the prospect of a global financial implosion galvanized world leaders to action, so it freed Harper to break out of a passivity that, whatever its merits as policy, was political poison. By the last weekend of the campaign, Harper was promising to "protect" the economy with something approaching passion, and something very near conviction.

But let us not lose sight of big picture. The Conservatives entered this campaign with a real shot at a majority — perhaps their best shot, perhaps their last shot. This is not a victory for the Tories, except in the most literal sense. It may not have turned out the catastrophe it looked to be at one point. But the Conservatives can hardly be congratulating themselves. In particular, the utter failure of their Quebec campaign — they finished third, behind the Liberals — must be dismaying to a party that had convinced itself, and a good share of the commentariat, that it was the natural inheritor of the federalist vote in Quebec, that it might even knock off the Bloc.

This is an indictment, not just of the particular tactics of this campaign, but of the whole strategic vision of the party's "pragmatists." They have led the polls since they were elected, yet they have been chasing all the way — chasing the middle, chasing Quebec — only to see their quarries recede ever further from their grasp. All that tacking about, all their attempts to denude themselves of anything resembling an ideology, has not produced a more conservative public: it has never been more liberal. The effect of Tory efforts to woo Quebec nationalists has not been to bring Quebec into the Conservative fold, still less to make them more Canadian: it has only persuaded them to withdraw still further from national life, to consider Canada as little more than a ready source of cash and favours. Think of all that the Conservatives have thrown at Quebec. Billions of dollars in the name of the fictional "fiscal imbalance." The status of nation. A growing role in foreign affairs. And it all falls to pieces over a few paltry cuts in arts funding?

But then, it's hard to see the result as a victory for anyone. The Liberals, at less than 27 per cent, have limped home with the worst popular-vote showing in their history, giving up one-quarter of their seats. If Liberals think this is merely a problem of leadership, a simple matter of giving Dion the old heave-ho and running off with that dreamy Michael Ignatieff — or is it Bob Rae? — they should think again. This is the third election in a row that the Liberals have seen their popular vote drop. Indeed, they have been operating with a narrower and narrower electoral base, not just of late, but for the past five decades. While pundits fretted about friendly dictatorships and "gritlock," the Big Red Machine has been dropping one wheel after another. They lost the West in the Diefenbaker sweep, and have never recovered. (Across the West, the Liberals won just seven seats this election. Seven seats.) They lost Quebec in the Mulroney sweep, and have never recovered. And now they have lost Ontario. They have become, for all intents and purposes, the Montreal-Toronto party, with pockets in Atlantic Canada.

But why pick on the Liberals? The NDP, for all the impression of momentum it gave off during the campaign, finished it no higher in the popular vote than it was last time, and no closer to its professed goal of knocking off the Liberals as the government-in-waiting. The Bloc held on to most of its seats, but only with a large assist from Conservative mistakes — and its popular-vote share declined. Even the Greens, the one party to significantly increase its vote, fell far short of expectations — and elected no one. Is it possible for everyone to lose an election?

And the biggest losers? Try the public. Five weeks of campaigning and $300 million in public funds later, the parties finished within a percentage point or two of where they were at the start. It's almost as if the election never happened — and might as well not have, for all the public cared. All those polls, all those ads, all that breathless coverage, and the turnout in this election, it appears, will be the lowest ever: just 59 per cent of registered voters. At some point it will occur to someone: we have a democratic crisis on our hands — a crisis of legitimacy, a crisis of efficacy. We are stuck, spinning our wheels, unable to find a sense of direction. The prospect is for more hung Parliaments, more bootless elections, more stall and drift, and less and less public interest.

If this election proves anything, it is that the process by which we elect our governments is broken. We are trying to run five-party politics through a system that was designed for two parties. The Conservatives look at their steady, incremental progress, slowly spreading eastward, election after election, from their Alberta-British Columbia base, through Saskatchewan, Manitoba, and now Ontario, and see a majority in the making. But the other way to look at it is that this is as close as they're likely to get.

It has become almost structurally impossible to form a majority government in this country. If you start each election, as it appears we are condemned to do, with 50 seats off the table — the Bloc's gift to Canadian democracy — then it is not 50 per cent of the seats you need to win a majority, it is 60 per cent: 155 of 258. Add to that the growing, institutionalized fragmentation on the left, and the mathematics become almost insurmountable. Eight years ago, the NDP and the Greens took less than 10 per cent of the vote between them. Today, it is 25 per cent. Throw in the Bloc, and the two parties with any chance of forming a government, the Liberals and Conservatives, are working with just two-thirds of the vote between them.

In a two-party system, majorities can be won with very little margin between first and second: in the theoretical limit, just one vote. A generation ago, when the two main parties were more dominant than today, you could win a majority with a margin of as little as nine points, say 44 to 35. But the more parties there are, and the more the vote is dissipated among them, the more the leading party must rely on the accidents of split votes to engineer a majority — meaning the larger the gap it must open up between itself and the second-place party. This explains some of the Tories' heavy reliance on negative ads. It wasn't enough for them to raise their own vote. They had to suppress the Liberals' vote, to somewhere close to the NDP's, to have any chance of a majority. As it was, they wound up with an 11-point gap, and still fell short.

THERE IS nothing wrong with minority governments, per se. It depends what kind of minority. Do we want the kinds of minority Parliaments we have had in recent years, a clutch of hobbled regional or quasi-regional parties, fingers perpetually on the button, endlessly threatening to pitch us all into another pointless election in the vain hope that, if the swing voters can be distracted in their direction, if the splits go their way, if they can demean and belittle their opponents enough, if they can depress turnout even further than before, they might just fluke their way into a majority? Or will we accept that, whatever the ancient glories of the two-party system, it no longer exists?

If we must have five-party politics, let them at least be parties with real differences, and national appeal. Away with the system that guarantees the Bloc two-thirds of the seats in Quebec on the strength of little more than one-third of the vote. Away with the ghettos of Conservative Alberta, or Liberal Toronto, where it is scarcely worth campaigning, so predictable are the results. Away with "strategic voting," and other attempts to tell people they may not vote for the party they support, but must vote against the party they fear. Away with the disgraceful situation of a party winning almost a million votes, as the Greens did this time out, and getting zero seats.

Indeed, when you think about it, many of the problems identified in this piece have their origins in the perverse incentives of our highly leveraged, winner-take-all electoral system. Why have the Tories degenerated into mush? Because they face no competition on the right, Reform-style uprisings being more or less outlawed for fear of "splitting the vote." Why did the Liberals ignore their growing weakness all these years? Because they could still count on the bizarre distortions of first-past-the-post to reap a bushel of seats from one region or another. Why has the Bloc become an immovable blot on the national scene, long after its original purpose was exhausted? Ditto. Why have majority governments become next to impossible? Why has politics degenerated into such vicious, empty partisanship? Why do so many people no longer bother to vote? Because the system is broken, and if this election won't persuade us to change it, nothing will.

Friday, November 14, 2008

Criticism #3: BC-STV will mean less women getting elected.

This is an interesting criticism that I’ve seen or heard put forward about a handful of times, the allegation being that if we switch to the single transferable vote (BC-STV), this will lead to a decrease in women being elected to the BC legislature.

On his website, Bill Tieleman even presents some hard numbers to back up his claim:
"In Malta women make up just 9.2% of the country’s legislators, with only 6 women elected out of 65 representatives. In Ireland just 13.3% of elected officials are women.

By comparison, in British Columbia under our First Past The Post system, women make up 22.8% of our MLAs, 18 out of 79. While it isn’t representative of our society, it is significantly better than under either STV system.

And in Canada women represent 21.1% of all elected Members of Parliament, with 65 women out of 308 seats."

The implied conclusion is fairly apparent: somehow, STV is a deterrent to getting women elected.

So the question that then begs to be asked is: by what mechanism is STV able to influence an election outcome? Maybe the STV system can, in some way, recognize a listed candidate as being female and then siphon votes away from her? What if the candidate’s name is a little sexually ambiguous? Like Chris Scully? Or Pat Maclellan? Jiang Yuyuan?

This is, of course, absurd. The STV system can no more prevent women from getting elected than the First Past The Post (FPTP) system can aid them. The fact is that the degree to which women are elected to representative government is dependant on many factors that are independent of the type of electoral system utilized. Such factors could include women’s access to education, the level of a society’s acceptance of women in positions of leadership or the percentage of women that actually run for office in the first place.

But let’s give Mr. Tieleman the benefit of the doubt. In another article written in opposition to STV (Feb 10, 2005, Georgia Straight), he does actually cite Christy Clark who at least tries to give some rational as to why woman may be elected in smaller numbers under STV: geography.
"the fact that STV performs poorly in ensuring that women are represented in legislatures is never mentioned. Liberal MLA Christy Clark rightly points out that STV, where every constituency is much bigger and elects between two and seven MLAs, makes it harder for a woman MLA to meet the challenges of personal life while serving a very large riding."

Interesting theory, although it ignores the fact that the ratio of Members of Legislative Assembly (MLA) per 10,000 people (or whatever the number is) will still be the same under STV as it is now. But at any rate, if this holds true, then expanding the size of government ridings for a given geographical area should lead to a significant decrease in female representation in elected officials. Gee, if only there was some way of testing this hypothesis out…. I wonder if we had any countries or territories where we can compare female representation under different sized ridings but the same electoral system…

Actually, we can and we can turn to Mr. Tieleman’s example as inspiration. As he has stated, in BC women make up 22.8% of our MLA’s. Now jump up a level to our federal government, where the size of the federal ridings in BC are more than twice as large as our provincial ridings. Do we see a significant decrease in the percentage of female Members of Parliament (MP) compared to MLAs? No, we actually see an increase, to 27.7% (as of 2007).

Furthermore, you can expand this test to the whole of Canada, where our 308 federal ridings are, on average, almost 3 times larger than the provincial ridings which, when combined, total 846. Here, we find female representation, provincially, does not increase significantly as Mr. Tieleman and Ms. Clark would suggest but actually decreases to 19% (as of 2007) from Mr. Tieleman’s number of 21.1%. Geography is simply not relevant.

So with all due respect to Ms. Clark, I believe she is confusing issues as her line of reasoning on why BC-STV should not be accepted (that it would be too much work for women trying to juggle career and family) can be applied to a whole host of situations outside of electoral reform.

For example, being a cabinet minister or even a premier would also make it difficult to balance the conflicting obligations of personal life and political careers because of the increased responsibilities that they hold. In fact, by her logic, any job that is in any way challenging would be discriminatory to women. Unfortunately, she is mixing the issues of electoral reform with the issues of women in the work place and how we support them as they try to balance the competing responsibilities of personal and professional life.

But what I find most incredible about this whole line of argumentation is just how badly Mr. Tieleman, in his blind zeal to latch on to any and every argument against BC-STV, no matter how spurious, contradicts himself from article to article.

In the Georgia Straight article, he tries to argue that BC-STV is bad because the larger ridings will lead to fewer women getting elected, but then on his website he cites Ireland as an example with their lower percentages of females in elected positions (compared to BC).

Has he seen the size of the ridings in Ireland?

Here is a picture of the Republic of Ireland’s electoral map wih it’s 43 constituencies:

Here is one for BC and our 79 ridings:

And here is a map of the world, just to get a little perspective (click the pic to zoom in).

World_pol_map cropped2

BC, at about 925,000 km2, is more than 10 times the size of Ireland at 70,000 km2. This works out to an average size of roughly 12,00 km2 per riding for BC, more than 7 times larger than Ireland’s constituencies (which are around 1,600 km2 in size). Yet despite the huge increase in riding size BC has way more women elected to representative government, exactly the opposite of the point Tieleman is trying to make in his Georgia Straight article.

It should be obvious that Bill Tieleman is not being very honest in presenting his point of view. It seems evident to me that he has not made his decision about BC-STV based on actual merit but has simply decided from the outset to take a stance against it, and then has just thrown any and all arguments out there with the goal of confusing or deceiving his readers. That should call into serious question his credibility as a commentator on this subject.

Back to Criticisms Mainpage.

Saturday, November 8, 2008

Criticism #2: BC-STV will lead to more minority governments.

This is another common argument presented that switching to BC-STV will lead to more minority governments, with critics trying to paint this apocalyptical picture of never-ending minorities mired in legislative stand-offs with political life being relegated to one long campaign as endless elections, being called every 6-12 months, usurp political discussion.

First off, I’ll start off by saying that, yes, there will be more minorities. And that’s how it should be; it’s pretty rare for a majority of the population to vote for a single party. So if a party only wins minority support in the ballot box, then it should likewise only win a corresponding minority of seats in the House or Legislature. Conversely, a party should be awarded a majority government only if the electorate wants one i.e. a majority of the electorate votes for that party. So when you think about it, the critics’ argument that FPTP is better because it produces more majorities is poor because it has to do so by misrepresenting the votes of the people.

But that’s not the only flaw in their argument. The critics’ assertion that switching to BC-STV will mean more elections because of more minority governments is based on the assumption that minorities under BC-STV will be as unstable as under FPTP. This isn’t true. In the Republic of Ireland, the average life of a minority in the last sixty years has been 3.7 years. Over the same time period Canada and BC, governments have lasted 3.15 and 3.55 years respectively (info courtesy of Antony Hodgson). And don't forget that we have had more majority governments. There is actually a page on the Elections Ireland website that lists the election years along the top. If you take a look, you see that elections actually happen relatively infrequently, with some minority governments even lasting as long as 5 years.

Why is this? Because there is a flaw, inherent to FPTP, that actually provides incentives for politicians to bring down minority governments sooner rather than later. Since STV addresses this flaw, the incentive is removed, leading to longer lasting minority governments.

This flaw, again, has to do with our FPTP system’s failure to produce proportional results; it simply is not cut out to deal with elections where there are more than 2 candidates. Because it is not able to account for vote splitting, margins of victory etc., you have these elections where small changes in the polls can lead to wildly skewed election results.

Let’s take our 2008 federal election as an example.

Stephen Harper and the Conservatives ended up winning 143 seats, a 15% increase from 2006, despite the fact that their share of the popular vote only increased by 1.3% (from 36.3% to 37.6%). On the flip side, Stephan Dion and the Liberals lost 27 seats, a 26% fall from 2006, even though they only dropped 4% in the polls. (wikipedia)

And this is why politicians call elections so frequently under minority governments. Because they know that small changes in public popularity can lead to huge swings in election results. So they wait until the political winds shift ever so slightly in their favour, and then bring down the government, hoping to capitalize on the windfall.

Mr. Harper called an early election for this very reason, banking that the polls would shift just enough to give him a majority. If the polls went in the other direction, you can bet it would be Mr. Dion working to bring down the government.

Now let’s take this scenario and apply it to the context of STV. Under STV, because it is proportional, a small increase in the polls would lead to.... only a small increase in seats won and visa versa. So the Conservatives 1.3% boon would only translate to about an extra 3-4 seats in the House; calling an election wouldn’t have really changed the look of parliament. The incentive has now been removed because public opinion usually doesn’t vary that much so politicians don’t have a reason to call an election since it’s not going to lead to a dramatic shift in power.

Back to Criticisms Mainpage.