The consequences of electoral reform

Canadians will see higher spending if system changes, writes Barry Cooper.

A minor plank in the Liberals’ election platform aimed to ensure “that 2015 will be the last federal election” conducted with the “first-pastthe-post” electoral system. It was enshrined in the throne speech. The objective is “to make sure that every vote counts.”

The ensuing media and blogosphere discussion has recapitulated a lot of postSecond World War analysis by political scientists on the effects of electoral systems on the new governments of Europe and of Europe’s former colonies. The high point came in 1951 with the publication of Kenneth Arrow’s Social Choice and Individual Values, for which, in part, he received the 1972 Nobel Memorial Prize in economics.

Arrow’s argument, summarized in his impossibility theorem, reveals that, in democratic societies where individuals hold distinct preferences, if there are more than two choices, it is impossible to order those preferences consistently. The relevant conclusion to his complex argument is this: There is no best electoral system.

So why are Canadians repeating a 70-year-old debate now? And why have the Liberals decided against a referendum on the question?

To answer these questions, first jettison the malarkey about every vote counting. Every vote counts now and has done so since the Constitution Act of 1791. Second, even if no electoral system can produce a rational outcome, different systems contain different incentives.

Here’s why: Parties, whatever the electoral system, are coalitions that want to rule. The present electoral system incentivizes parties to keep the coalition within the party and win a parliamentary majority. In contrast, all other electoral systems provide incentives for singleissue parties (or in Canada, regional parties) to run on their own and form a governing coalition in Parliament, not within a big-tent party.

This is not news. But additional implications are often overlooked. The first is that, when several parties form a governing coalition in Parliament, the largest has to buy off smaller single-issue parties. From around the world, the evidence is overwhelming that this practice entails higher government spending, increased deficits, lowered ability to deal with financial crises and greater bureaucratic control, all of which leads to increased government instability.

The second is that, while electoral changes always have consequences for specific interests, and notwithstanding the truth of Arrow’s theorem, centrist parties typically think they will be the shortterm beneficiaries by being the second choice of left- and right-wing voters.

The first implication explains why Elizabeth May and the Greens favour change; the second explains why the Liberals do.

But now things get weird. Those who advocate changing the electoral system, who mouth the bogus claim about wanting “every vote to count,” do not want to apply that principle to a yes-no vote on electoral change, even though it would produce a genuine majority vote. And in fact, when Canadian electorates in British Columbia, Ontario and Prince Edward Island were asked to approve electoral changes, they refused.

Even if we were all so stupid as to trust any government, the Liberals may have no choice about a referendum. The legal logic is identical to that which required Pierre Trudeau to consult the provinces in 1981. As in 1981, precedent matters. After 225 years, the existing electoral system amounts to a constitutional convention.

In short, the Liberals have neither political mandate nor constitutional right to change the electoral system without a referendum that they likely would lose.