Myth #1: proportional representation leads to minority governments

The UK electoral reform blog Make My Vote Count has an interesting, semi-regular series on myths about proportional representation, and in light of some misleading comments that have been made on various blogs recently, it's starting to look like it's necessary to do a similar series for Canada. So let's start with the one that's been rattling through the blogosphere lately as the Conservatives have gotten frustrated with running a minority govermnent: if Canada had proportional representation, we would have the current sort of unstable minority government forever and ever amen.

The myth

One example of this myth can be found in the comments to this post over at Accidental Deliberations:

And this is what Proportional Representation would look like, nothing getting done. Not too amusing when the taxpayer is footing the bill for NOTHING.

and a bit more thoughtfully in this post from the Prairie Wranglers:

Under a PR system, it would be very unlikely, at any point, for any Canadian political party to achieve majority status, as few politicians have been able to achieve over 50% of the popular vote since Canadian first realigned into a 3-party system, sometime between 1921-35 (Diefenbaker did in 1957, and Mulroney did in 1984, but only by 0.03% margin). Many suggest this is an inherent benefit of the system, as it forces political parties to work together in order to achieve a consensus which better represents the nuances of political opinion throughout the country.

How does this jive, however, with our current predicament. Witness the recent pitched battle over who in fact is responsible for preventing parliament from moving legislation through the House; the Conservatives blame the opposition, the NDP and the Liberals blame the government, while the Bloc sits on the sidelines making outrageous demands in return for its support. This lends credence to the criticisms of those against PR that a perpetual minority government in Canada is an unworkable and unstable solution, as it will simply provide a forum for political grandstanding and posturing while obstructing concrete action.

When put like that, it sounds pretty convincing, right? Proportional representation prevents majority governments, right? When Canada has minority governments, things are unstable, and there's more political grandstanding than there is concrete action, right? Therefore, if Canada switches to a PR-based electoral system, we will have a perpetual minority government--which would mean yearly elections, nothing but hot air from our politicians, and never getting anything done. Right?

It would be a lot more convincing if it were true. In fact, minority governments are much rarer under proportional representation than they have been recently in Canada. What proportional representation systems actually tend to produce are majority coalition governments, which, as I mentioned previously in my proportional representation FAQ, are about as similar to minority governments as apples are to kumquats. Majority coalitions aren't minority governments, but real majority governments of more than one party, formed to create the kind of long-term stability that simply cannot be found in a minority government.

How coalitions work

Part of the problem for Canadians is that the word 'coalition' tends to be misused in this country to mean any informal alliance between parties. But a true government coalition bears almost no resemblance to the informal agreements such as the one we saw between the federal Liberals and the NDP in 2005, or the one Jack Layton is trying to build between all parties in the House right now over the Clean Air Act. These kinds of informal alliances are created on an impromptu basis when one party or a group of parties exerts pressure on the party running the minority government, and as we have seen, this pressure usually consists of threatening to bring down the government if they don't jump. Government coalitions, on the other hand, are formed immediately after an election between the winning party and its chosen coalition partner(s), with the goal of forming a majority government (i.e., adding up to at least 50% of the seats between them). There are no threats and no media grandstanding, just quiet talks between the first-place party and the party or parties they feel they can work with on a long-term basis. The point of these talks is not to concoct a plan that will hold a fragile minority together for a couple of months, but to come up with compromises between the parties' platforms that will withstand the test of time.

During these talks, the parties not only agree to a common policy framework, but they also negotiate about who is going to get various cabinet positions and take charge of the corresponding ministries. This means that the leader of the winning party might say something like this to the potential coalition partner: "Okay, you guys hold ten percent of the seats, so it makes sense for you to have around ten percent of the cabinet positions. But it's important for our party to be in charge of all of the main ministries like finance and foreign affairs. We're willing to give you things like immigration and labour--how about that?" The leader of the potential coalition partner can then come back with things like: "Well, if you're not going to let us have any of the main ministries, we want more than ten percent of the cabinet positions. How does twenty percent sound?" Or perhaps: "No, we're not willing to settle for only minor ministries--and specifically, we want the environment ministry because it's been a huge part of our platform and that's something we won't back down on. But we're willing to take fewer than ten percent of the ministries in order to make sure it can happen that way."

Because everyone agrees to these compromises during the formation of the government, the government doesn't get formed until the disagreements are worked out. This can take up to several weeks in situations where there are a lot of conflicts--primarily in cases when the electoral outcome was different from expectations and the winning party's preferred coalition partner didn't get enough seats to add up to a majority--but the end result is a situation in which all members of the coalition have a stake in ensuring the continued persistence of the government. The smaller coalition partner doesn't feel moved to threaten to bring down the government because they are part of the government. The parties then work together to make and implement the common policies that they have agreed on.

The stability of coalitions

There is a further widespread misconception that regardless of the care taken in their formation, majority government coalitions tend to function in practice in the same way minority governments do. This misconception showed up most recently in Olaf's responses to my objections to the Prairie Wranglers' post:

It is my understanding that coalition governments, as opposed to minority governments, are not as fundamentally different as you would suggest, and would appear quite similar for quite some time in our Westminster system of government, in particular. I mean, the Liberals and the NDP were, in my opinion, textbook examples of natural allies in many respects, and we all know how that worked out.

While it's certainly the case that there have been majority coalition governments that have been as battle-torn and unstable as our minority governments (more on that in a moment), it's incorrect to imply that this is the norm, or even that it is common. There are many forms of proportional representation found in the world today, but only two have been advanced as possible models for reform in Canada: Mixed-Member Proportional like they have in Germany, and Single Transferable Vote like they have in Ireland. And when you compare the frequency of elections in these two countries to similar numbers for Canada, the results are striking:

Ireland:
16 elections since 1948, 1 election every 3.63 years
Germany:
16 elections since 1949, 1 election every 3.56 years
Canada:
18 elections since 1945, 1 election every 3.39 years

In these two countries, majority coalition governments are the norm: Ireland hasn't had a single-party government since the 1980s, and Germany has only had one single-party government in its entire post-war history. And yet they have still been incredibly stable. In Ireland, the coalition of Fianna Fáil and the Progressive Democrats has persisted throughout all but five of the last seventeen years (Fianna Fáil chose the Irish Labour Party as a coalition partner from 1992 to 1994, and three other parties briefly formed a different coalition from 1994 to 1997). The recent German coalition between the Social Democrats and the Greens was stable enough not only to last out its term, but to be reelected in 2002, and extremely stable coalition governments of the conservative Christian Democrats and the Free Democrats have also existed under chancellors Adenauer, Erhard, and the seemingly undefeatable Helmut Kohl, who was reelected a total of five times.

These two countries are typical for the two kinds of proportional representation that have been proposed for Canada. But even if we summarize trends from proportional representation systems across the board, the stability picture doesn't look all that objectionable. The average length of government in countries that use Canada's First Past the Post system is three and half years. In countries with various proportional representation systems, the average length of a government is two and half years. This suggests a somewhat more frequent turnover, but not enormously so. Furthermore, the actual effects of this kind of "instability" are greatly lessened in practice, since many of these "changes in government" include restructurings where the main coalition partner stays the same but only the smaller one is different (making them less like Canadian changes in government and more like Canadian cabinet shuffles). And this average includes many countries that Canadian electoral reformers have no desire to emulate, and many forms of proportional representation (such as pure party list proportional representation) that no Canadians want.

Now, in conversations about proportional representation, detractors will inevitably bring up two countries where coalition governments have meant frequent changes of government and/or a great deal of conflict between coalition partners: Israel and Italy. But there is no reason to believe that Israel and Italy behave typically for countries with proportional electoral systems, and there are many reasons to believe that they are in fact completely atypical. In Italy, weakened party discipline can be attributed largely to the practice of secret balloting by MPs in the House, and in Israel, living under a constant threat of war seems likely to have influenced their governments in ways that most European parliamentary democracies have not had to deal with. And finally, both Italy and Israel have historically used versions of pure party list proportional representation (Italy recently switched to a system more similar to Germany's) that Canadian electoral reformers are not interested in introducing in Canada anyway. And both Mixed-Member Proportional and Single Transferable Vote systems do away with the worst of the potential problems of pure list PR by having threshholds for small parties' inclusion in parliament.

Coalitions: not for Canada?

Finally, an idea has been batted around lately that majority coalition governments simply wouldn't work in Canada. Scott Tribe put this viewpoint forward in his comments on my interview with Stéphane Dion, arguing that "multi-party coalitions just don't happen in Canada," and that "the ideological diffrences in the multi-party system in a such a diverse country as Canada are deep and bitter, and I highly doubt that parties would set aside their partisan differences." Olaf from the Prairie Wranglers agreed in his responses to my arguments in his post, saying that "our system of government is reliant on conflict, and has been since day one," and that he "doesn't think that Canada's political culture has progressed to the point where collaboration and conciliation could be the rule rather than the exception."

I find this line of reasoning puzzling. It's certainly true that coalition governments have been far rarer in Canada than they have been in most other parliamentary democracies: since the 1800s, Canada has only had one coalition government (during World War I). But attributing this rarity to something inherent in Canadian culture seems farfetched at best. There is nothing inherently more squabble-prone about Canadians than there is about Europeans--on the contrary, Canadian culture is well-known for its propensity for compromise. Instead, it seems far more realistic to attribute this distinction to Canada's electoral system. In fact, if we compare the few parliamentary democracies with single-party government traditions to those that instead tend to form government coalitions, a noticeable pattern emerges. The only countries in which the winning party insists on governing alone are those that use First Past the Post--countries like Canada, the UK, the Bahamas, and Jamaica. All of the countries that have some sort of proportional electoral system tend to govern by majority coalition.

Part of the reason for this, of course, is the way First Past the Post tends to award a majority of the seats in parliament to parties that do not manage to get a majority of the votes. After all, there is no need for more than one party to cooperate in a coalition when one party alone always gets the majority of seats. But this also means that under First Past the Post, all parties that want to govern will tend to organize their political strategies in ways that aim for single-party majority governments. We see evidence of that in Canada right now--the Tories are so devoted to finding a way to govern alone that they seem to be forgetting that they have a minority government, and the Liberals, too, tend to behave as if they just need to win the next election and they'll be able to govern with a single-party majority. A lack of compromise is built into the system.

As for the argument that switching to proportional representation wouldn't make our politicians any more willing to work together to form coalition governments, evidence from New Zealand would seem to counter that. New Zealand switched from First Past the Post to a Mixed-Member Proportional system by referendum in 1993, and they have had government coalitions ever since. The early coalitions did tend to be wobbly and fractious, since their parties had to completely adjust their strategies and their politicians had to learn how to do their jobs in ways that promoted cooperation over conflict. But they seem to be moving out of the adjustment period now, and more recent years have shown evidence of the kind of long-term stability found in Germany or Ireland.

It's silly to talk about coalition governments as if they were some weird phenomenon that only exists in the Third World when they're the status quo in most of the world's democracies. Belgium and the Netherlands manage to form government coalitions in countries with large numbers of immigrants. Switzerland manages it in a country that has four national languages and four distinctly different cultural groups. Germany manages it in a country that used to be two different countries right up until 1990. There is absolutely no reasonable argument for the notion that Canada is inherently different from the rest of the world on this front. It is only our electoral system that promotes the kind of conflict our political culture is rife with--all the parties are so busy going for broke that they don't realize they're going for broken.

But most importantly, the arguments about whether or not coalitions are possible for Canada make it sound like we have a choice in the matter. We don't. The main argument for First Past the Post is that it produces a clear winner that can then go on to form a strong single-party majority government, and it works this way in countries that have a strong two-party system, like the United States. In Canada, however, that very same electoral system doesn't work the way it is supposed to work anymore. Why? Because the NDP and the Bloc Québécois have become significant enough forces that the advantage granted to large parties under First Past the Post is no longer enough to manufacture a majority government for either the Tories or the Liberals. And while the two large parties like to pretend that we'll soon be returning to a strong two-party system, pretending doesn't make it so. There is absolutely no reason to believe that the NDP and the Bloc are going anywhere anytime soon. On the contrary, the rise in popularity of the Greens suggests that there may soon be yet another significant force on the electoral scene, one that moves Canada even more irrevocably toward being a true multiparty system.

This means that the political strategies of both the Conservatives and the Liberals--strategies that require single-party majority governments in order to function properly--are entirely out of synch with the realities of the Canadian political situation. Even the NDP, with their recent talk of forming a government on the federal level, seems to be more interested in governing alone than in being part of a real, lasting government coalition. But these parties are labouring under a collective delusion, and it's a delusion that's harming Canada. The only realistic, long-term solution to the current instability is for our political parties to learn to work together--and not in an ad-hoc, temporary sense like what Jack Layton is promoting right now, but in true government coalitions. And while proportional representation isn't a prerequisite for coalition governments (Canada's World War I coalition attests to that) it may well be the only way to force our parties to stop denying the need for majority coalitions and start finding a permanent solution to our current predicament.

Ironically, the best argument for proportional representation in Canada right now may not be the run-of-the-mill ones such as more fairness, increased voter participation, or an increase in the number of women parliamentarians, but more stability.

Further reading

True electoral system geeks (as well as anyone who wants to check the facts in this post) are going to want to read Arend Lijphart's Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries. A summary of this book (in .pdf form) can be downloaded from the Fair Vote Canada website here.