I just want to make four big points, drawn from the five-page statement I sent last week. One is that I find it distressing that so much of the commentary on electoral reform represents the House as if it's a species of the U.S. Congress or some entity that is part of a congressional framework and not the constitutive chamber for the government, which inevitably it is in a parliamentary system. Voters care about this. They care about this as much under PR as they do under majority rule, because PR will almost always, at least the relevant examples, accompany parliamentary systems. Even under PR, considerations about the formation of the government induce strategic voting. You've probably heard about strategic voting induced by thresholds and that sort of thing.
Even apart from thresholds, many voters actually care about the likely composition of the government once the returns are in. There's some evidence to suggest, for example, that voters on the flanks, or voters on the near flanks, vote for more extreme parties than they themselves prefer in order to pull the government toward them, so to speak. The general point is that you can't isolate the voting moment from the government formation moment. I would say there's been a sort of intellectually lazy separation of those two in much of what's been said here.
The second point, and maybe the one I care the most about at the moment, is that I'm struck that so often wholesale packages are put on the table, and then virtues are claimed for them or vices imputed to them. I think it's absolutely important to remember that any electoral system is actually at a minimum—there are other pieces as well, but at a minimum—a compound of three things. The interaction amongst the three is absolutely necessary to get a grip on what is the likely product of a change in that direction.
First of all, you have to have a ballot. The most important though not the only distinction is whether it's a categorical or a preferential ballot.
Second, the whole question of district magnitude, the number of seats per district, is absolutely a critical element. You can't have proportional representation, obviously, if you have single-member districts. If you have multiple-member districts with a plurality formula, you have grossly disproportional results. Nonetheless, the magnitude of the district is important to a lot of things, not least to the proportionality of a proportional framework.
The third question is the formula itself, where the formula is crystal clear in the case of the plurality rule; semi-clear in the case of the majority, which has kind of two operable forms; and then there's the family, among whom the differences are very consequential for proportional representation. In many cases, effects that are attributed, say, to the formula are actually the product of another feature of the system. It may be that it's a feature of the system that is only relevant under a proportional formula, but it's not the formula as such that's critical.
Let me give you an example. It is generally true, as you've heard, that descriptive representation of females is higher on average in PR systems, but we also observe that there's massive variation within the PR family. It's also within the more majoritarian families, but there's particularly massive variation within the PR family. Some of that variation actually reflects the ballot. If you want to maximize the descriptive representation of any group, women or otherwise, in some sort of guaranteed fashion, it helps to have the largest district magnitude. The larger the number of seat per district, the easier it is to mix up demographic characteristics in the menu that a party presents.
Most importantly, you want to have a ballot that has a closed list, that has a categorical vote, one only per voter. Additionally, outside the electoral formula itself, you want to have a centralized placement of names on the ballot. If you yield to the temptation of derogating from any of those things, you will not have as powerful a mechanism for creating the prospect for guaranteed representation of categoric groups. On the other hand, if what you want to do is facilitate the representation by women or by other groups of forces that in some sense are not incorporated into the central conflict among the parties, you actually want the opposite of all of those things. The general point is that the ballot is often as important as the general formula for the achievement of certain political or social goals.
My third point is that most electoral system change, although New Zealand would be an exception, and possibly Germany, because it was kind of tabula rasa at the time and has a history of partisan objective, either to advance the interests of a party or to retard the interests of another. I think we need to be clear about that. To be silent on that I think is...well, it's dishonest, frankly. It is clear, for one thing, that proportional formulas in general empower the left relative to the right, and the opposite is true for majoritarian formulas. It's just the way it works. If you average across all the consolidated polities postwar, it's really a remarkably stark difference.
Second, a question that will always be in play in a place like Canada is which form of formula, majoritarian versus plurality, politicizes ethnic questions the most? It's actually not a simple answer. It is true that under PR you facilitate the coming into existence of micro-parties, which can be ethnic or otherwise but certainly can be ethnic. That's on the one hand. On the other hand, you do not augment the power of groups that are appealed to by those parties. It is easier for parties with a more national appeal to penetrate into those communities and in some sense dissolve the singular claim of a particular party to speak for them.
On the other hand, or actually a variation on the same theme, our system does create certain privileged, geographically related possibilities. The ancient and most important one in Canada is that until 1993, at least, it made Quebec the pivot for government. By augmenting the power of whatever majority was prevailing in the province at the time, and a whole other set of considerations about whether...but it made Quebec the pivot for government. It has since not been the pivot for government in the same way, but in some sense the pivotal task has been handed over to suburban Vancouver and suburban Toronto.
Those are all good-news, bad-news stories, but we should recognize that geographically differentiated groups, and this often includes ethnolinguistic groups, can actually have their power augmented in the formula we have now. That could be either the price of Canadian unity or the price of successful incorporation of groups, but it does produce a situation in which some votes count more than others. Those counts can help ethnic minorities.
My final few thoughts are about transition, if there is to be one. I am neither in favour of nor opposed to a referendum. I don't think they are to be absolutely abhorred or to be required. The issue I would care about, frankly, is the institutional stability of whatever framework you put in place. If you're going to change the electoral system, don't do it in a way that merely invites a change back or a change to something else in short order. I just think that's corrosive to the legitimacy of government and it's corrosive to the operation of political parties. The countries that do that, and there are a few, Italy and France being prime examples, I think have paid for that.
I think stability is key. To that end, the referendum could be a contribution to stability in the sense that it has a kind of morally binding force from the population at large. I'd invite you to consider whether some of the purposes of a referendum could be achieved by other means. Part of why a referendum appeals to people is that in some sense it takes if not the final choice out of your hands, then in some sense it raises the costs to you and limits your freedom of action in making the ultimate choice. That could be done by ways other than a referendum—a citizens' assembly, expert group, or whatever.
I have a couple of things in mind. For example, for the politically very fraught task of closing military bases in the U.S., or passing highly divisive trade agreements, the U.S. Congress has a route in which they basically invite outsiders to make the proposal and Congress decides up or down. In a manner of speaking, that's a bit like changing electoral boundaries in this country.
The general point is that to the extent that all politicians, rightly or wrongly, are perceived as in the business of self-dealing and self-interested action, you might want to think about, if you're really serious about change, executing the change in a way in which there is an independent voice that nonetheless has to come back and talk to you.
Thank you very much.