BC’s great debate on proportional representation continues, and opponents are ramping up their attacks as the vote draws closer. In the latest fearsome salvo from the No side, John Hansen warns that if we adopt PR, there is a very real danger that we may end up like… Denmark. Yes, Denmark.

Please remain calm.

Denmark. The very name of the place makes the blood run cold, doesn’t it?

Yes, Denmark may be one of the happiest countries in the world, score higher than Canada on the UN’s Human Development Index, take the third spot in the World Bank’s index on the ease of doing business (New Zealand, which also uses proportional representation, takes number one), have the lowest poverty rate in the OECD and unemployment below the OECD average, win top ranking in the Economist’s Democracy Index, be the least corrupt country in the world according to Transparency International (tied with New Zealand), and possess (according to the Heritage Foundation) “one of the world’s most attractive business environments … characterized by political, economic, and regulatory soundness.”

But on the other hand, Denmark does not allow parties that win only a minority of votes to be rewarded with a majority of seats.

I know this is hard to hear. But it’s important. Please, stop screaming just for a moment?

Thank you.

Governments in Denmark are often formed by coalitions of smaller parties. Unlike many other jurisdictions that use PR, even these coalitions are usually minority governments, forcing the parties in government to pay attention to the views of the opposition parties as well.

Deep breaths, my friend.

The worst part? Danish voters have several parties they can choose between with the expectation that their vote will actually end up mattering, as if there aren’t just two points of view worth considering on every issue. I know, I know—I haven’t heard anything so ridiculous since I met that lunatic who claimed that there are colours besides black and white.

The lesson is clear. Denmark may have a lot going for it, but the price… Recognizing diversity? Fostering freedom of choice? Encouraging collaboration? Ending minority rule?

Some prices are just too high for some of us to pay.

In Ethics: Inventing Right and Wrong, J. L. Mackie argues against the existence of objective moral facts on the grounds that they would be “queer”—that is, the existence of moral facts would involve entities “of a very strange sort, utterly different from anything else in the universe”. But on the next page, he acknowledges the importance of the realist objection that moral facts have “companions in guilt”. According to this objection, a wide range of other putative kinds of facts exhibit the same qualities that make moral facts seem to Mackie to be objectionably odd. Mackie’s response is to say that he thinks “most” of these other kinds of fact will turn out to be explainable through empirical means. Regarding the others, he thinks we should adopt the same error theoretic approach he recommends for moral facts:

from Ethics: Inventing Right and Wrong, page 39

This seems to me like a very damaging concession. Mackie’s original problem with moral facts was that they would be “utterly different from anything else in the universe”. But now he admits that they would actually be like some things (facts about identity, maybe, or modality), and unlike others. Everything in the universe is like some things and unlike others! So if moral facts have companions in guilt, then none of these kinds of facts would be “utterly different from anything else in the universe”, so none of them would be queer in Mackie’s sense. This means that the key premise in the argument from queerness depends on the assumption that the error theory is true with respect to various other kinds of facts. And Mackie cannot defend this assumption without the assumption that the error theory is true with respect to moral facts. So the argument from queerness is circular.

In a recent column, former BC Liberal staffer Keivan Hirji issues three complaints against the process to be used in the upcoming referendum on electoral reform. In this post I will explain these charges, and show why all three should be rejected.

Hirji’s first complaint concerns an alleged false dichotomy in the referendum question. A false dichotomy occurs when two alternatives are presented as exhausting all possible options, but other options are in fact available. The false dichotomy in this case, Hirji claims, is between support for the status quo and support for all three forms of proportional representation on the second part of the ballot. One might prefer MMP to FPTP, for example, but prefer FPTP to DMP, and the structure of the ballot does not allow for the expression of this kind of preference.

But to say that the first part of the referendum ballot presents two alternatives as exhausting all possible options simply because there are only two options on that part of the ballot is like saying that an ordinary election ballot presents three or four alternatives as exhausting all possible options simply because there are only three or four names on the ballot. Unless they include a space for write-in candidates, all ballots exclude possible alternatives. No fallacy is committed because the options on the ballot do not purport to represent all possible alternatives.

Hirji’s second complaint is that two of the three forms of PR on the ballot (dual member proportional and rural-urban proportional) are not in use anywhere in the world. To adopt either of these two system would be to treat the province “as a guinea pig in electoral experimentation”.

In fact, only DMP is truly untested; RUP would use two well-tried electoral systems in different regions of the province. Moreover, the phrase “guinea pig” is obviously loaded—compare Hirji’s description of the province as a lab animal with Justice Louis Brandeis’s memorable description of the members of a federation as “laboratories of democracy”. Laboratories undertake important experimental work. But notoriously, guinea pigs often suffer and die in laboratory experiments. For Brandeis, citizens are like the scientists and technicians who make use of the laboratory to make important discoveries. For Hirji, citizens are like lab animals who are used in the laboratory. Hirji’s analogy is meant to lead readers to feel that if we try out a new electoral system, we too may suffer and die like the guinea pig. However, Hirji declines to elaborate on his anticipated failure modes of proportional representation in BC. How on earth would any of these systems lead to consequences sufficiently dire as to justify the lab animal analogy? Without elaboration, this part of Hirji’s argument is nothing more than a crude appeal to readers’ fear and can thus be dismissed altogether.

Finally, Hirji complains that voters in the referendum will be forced to choose between electoral systems without knowing what the boundaries of the new electoral districts would be. Hirji compares this to choosing between vehicles without knowing any of their specifications. Again, this analogy is highly misleading.

District boundaries change all the time. They will continue to change regardless of whether voters decide that our electoral system should respect the principle of proportionality. Anyone who votes to retain first past the post is doing so without knowing what the district boundaries will look like several years down the road. By the logic of Hirji’s vehicle analogy, this means that these voters will be making a completely blind choice; not knowing the exact boundaries that single member districts will have in the future means not knowing anything about how FPTP will work in practice. That’s utterly absurd. So something must be wrong with Hirji’s analogy.

One problem is that district boundaries are not comparable to the full set of specifications for a vehicle. In fact boundaries are impossible to evaluate without also knowing the district magnitude (i.e. the number of offices to be filled) and the electoral formula. Hirji’s analogy makes more sense if district boundaries are treated as comparable to a subset of a vehicle’s specifications—they’re important, but they’re not the whole story. And if district boundaries aren’t the whole story, then voters can draw some reasonable conclusions about what would happen in the future under FPTP (e.g. majority governments formed by single parties that lack majority support) and under various kinds of PR (e.g. majority governments formed by compromise between parties that collectively won a majority of votes). Contrary to Hirji’s claims, BC voters do not have to make a blind choice.

A further problem with Hirji’s argument is that in BC, like the rest of Canada, district boundaries are drawn up by an independent, nonpartisan boundaries commission. Commissioners are mandated to set boundaries that allow for effective representation given a particular electoral formula, striving to achieve roughly equal voting power for all citizens while also taking into account the various natural communities of interest that comprise the province. The integrity of this process is unquestionable. And BC will continue to use this process whether or not we decide to change our electoral system. Even though we can’t say in advance exactly what the district boundaries will be if we stick with FPTP, we can confidently say that the boundaries will allow for effective representation. And if this is true under FPTP, it’s also true under PR. So to return to the vehicle analogy, we may not know some of the relevant specs in advance, but we do know that they’re guaranteed to meet our standards whatever we happen to choose.

Hazan and Rahat’s proposal, from Democracy Within Parties: Candidate Selection Methods and their Political Consequences, page 174

In previous posts, I argued that parties ought to internally democratic, but grassroots members’ perennial demand for direct, authoritative control over party policy cannot be met. Direct election of party leaders has also been a failed experiment in democratization; in practice, direct elections seem to have concentrated power in the hands of an increasingly remote central leadership. So what can parties feasibly do to improve their internal democracy?

A more promising avenue for reform has been suggested by political scientists Reuven Hazan and Gideon Rahat in their book Democracy Within Parties: Candidate Selection Methods and their Political Consequences. Selection of candidates for public office is the essential function of political parties—the thing that really sets them apart from interest groups, think tanks, industrial associations, service clubs, and other organizations of civil society. Democratizing the process of candidate selection, then, seems like the natural place to start when thinking about democratizing political parties.

But what would democratizing candidate selection actually involve? The simplest approach would just involve open nominations, local control of nomination contests, and inclusive membership rules. However, Hazan and Rahat argue that this maximally inclusive approach can have a perverse effect on intraparty democracy.

Parties are largely volunteer-run organizations, and so need to provide non-monetary incentives to attract and retain committed long-term activists. Traditionally these incentives have included opportunities for positions of power and responsibility within the party, up to and including the prospect of being selected for candidacy. A maximally inclusive candidate selection process precludes such incentives, threatening to reduce parties to a small professional core of officeholders and staff surrounded by a largely ephemeral membership that tends to dissipate almost completely between elections.

Without some degree of exclusivity, then, parties cannot possibly be loci of sustained, high quality democratic participation, independent of both the state and the broader society in which they are embedded. Parties need to strike a balance between inclusivity and exclusivity. Hazan and Rahat argue that in the context of candidate selection, this balance would best be achieved by adapting the classical concept of mixed government (i.e. a regime that incorporates features of monarchy, aristocracy and democracy, so that the particular virtues of each basic regime type compensates for or corrects the defects of the others) to the nomination process.

The process Hazan and Rahat envision consists of three stages, proceeding from the most exclusive to the most inclusive. At the first stage, a screening committee would create a longlist of prospective candidates. This committee would be composed of senior party figures including retired politicians and veteran activists who do not have a direct stake in nomination contest, randomly selected rank and file members, and perhaps representatives of organized internal party factions, affiliates, and stakeholder groups. The exclusivity of the screening committee together with its balanced composition improves the conditions for serious deliberation; however, Hazan and Rahat stress that the function of the screening committee is only to rule out those candidacies that would harm the party, not to fix the outcome of the race.

The longlist would then be submitted to a nominating agency composed of party delegates elected for the purpose of amending the screening committee’s recommendations. By giving a special opportunity for influence to active party members, this second stage preserves incentives for members to commit to activism on the party’s behalf, yet it is also reasonably inclusive given that delegates responsible for the shortlist must be selected by the general membership.

At the third, maximally inclusive stage, the nominating agency’s shortlist is submitted to the whole membership in the nominating district, who have the final say on who should receive the party’s nomination. This third stage preserves incentives for party membership, creates an opportunity for member recruitment to activist roles, and maintains the benefits of mass participation (such as democratic socialization, opportunities for non-citizens to exercise political influence, and providing a counterweight to more ideologically motivated activists).

Above all, Hazan and Rahat stress that the final stage of the nomination process should be the most inclusive. In Canadian political parties, however, the final stage is the most exclusive—the party leader, a selectorate of one, has an absolute veto over the local party’s choice. In a sense, then, Canadian parties already use a kind of mixed regime, albeit one in which the mix seems to have been arranged so as to emphasize the defects and suppress the virtues of each basic regime type. But the principle is familiar enough. We just need to apply it properly.

Here’s an argument you might have seen before:

1. According to campaigners for proportional representation, my vote doesn’t count if I cast a vote for a losing candidate.
2. But in the last election, my vote for a losing candidate was counted.
3. So campaigners for proportional representation are wrong; a vote for a losing candidate still counts.

And here’s another one:

1. “No taxation without representation” is widely accepted as a fundamental democratic principle.
2. My MLA does not represent my views.
3. So the status quo violates a widely accepted fundamental democratic principle.

Both of these arguments—one which you might have heard from supporters of first past the post, the other from supporters of proportional representation—are bad. They’re bad because they both commit the fallacy of equivocation.

The fallacy of equivocation is committed when the definition of a key term shifts over the course of the argument. In the first example, this is done with the term “count”. When “count” appears in the first premise, it’s used to mean something like “matter”. In the second premise, it just means that one’s vote was included in the numerical totals of votes cast. So the first argument can be restated as follows:

1. According to campaigners for proportional representation, my vote doesn’t matter if I cast a vote for a losing candidate.
2. But in the last election, my vote for a losing candidate was included in the numerical totals of votes cast.
3. So campaigners for proportional representation are wrong; a vote for a losing candidate still matters.

And this argument is obviously invalid; the second premise is a complete non sequitur and does not support the conclusion. The pro-PR campaigners cited in the first premise might be wrong that votes cast for losing candidates do not matter; the claim would have to be argued in any case. However, this argument does not provide good reason to believe that the pro-PR campaigners are wrong.

The second argument also commits the fallacy of equivocation, although perhaps in a less obvious way. The error is less obvious because political representation is a complex concept. On Hanna Pitkin’s highly influential account, for example, there are four basic views of political representation: formalistic, symbolic, descriptive and substantive. Descriptive representation requires resemblance between the representative and the represented (e.g. in terms of their political views), and this appears to the sense invoked in the second premise. However, it’s not clear that descriptive representation is the kind of representation demanded by the slogan “no taxation without representation”. In fact this slogan is often interpreted as a demand for formalistic or substantive rather than descriptive representation. So the argument can be restated as follows:

1. “No taxation without [formalistic or substantive] representation” is widely accepted as a fundamental democratic principle.
2. I do not have descriptive representation.
3. So the status quo violates a widely accepted fundamental democratic principle.

Again, once the ambiguity of a key term is resolved, the conclusion no longer follows from the premises. Of course, one might argue that “No taxation without [descriptive] representation” is a fundamental democratic principle. If one were prepared to do so, this argument could be made against the status quo:

1. No taxation without descriptive representation is a fundamental democratic principle.
2. I do not have descriptive representation.
3. So the status quo violates a fundamental democratic principle.

And this argument is logically valid. But the appeal of the original (fallacious) version of the argument is that it invokes a principle that practically everyone already endorses. The revised (valid) version lacks this virtue.

Fallacious arguments might convince someone, but they ought to convince no one. This is reason enough to avoid making them. But it’s not the only reason.

In this referendum, only one side can come out the winner. But ideally, both sides will come out of it feeling that the debate leading up to the vote itself involved a meaningful exploration of the political values at stake, thus conferring legitimacy on the result whatever it happens to be. Both of the examples of fallacious arguments in this post represent missed opportunities for this kind of exploration—in the first case, a missed opportunity for supporters of the status quo to explain what they believe makes votes matter, and in the second, for supporters of proportional representation to explain why it’s important to have representatives who reflect their views. It’s not every day that basic principles of our democracy are brought under this level of scrutiny. Let’s make the most of it.