Stephen Harper's conservative coalition conforms with the game-theoretic ideal of a minimum connected winning coalition and, as such, should be internally stable and difficult for opponents to break up.
The 2011 federal election saw the emergence of a majority Conservative electoral coalition. Built gradually over the last four elections, as shown in table 1, it may dominate Canadian politics for years to come.
When he reentered electoral politics in 2002, Stephen Harper wanted to reconstitute Brian Mulroney’s coalition of western populists, traditional Tories and francophone nationalists; but when the francophone pillar of the coalition proved unstable, he was able to replace francophones with sizable elements of Canada’s ethnic communities. The resulting coalition conforms with the game-theoretic ideal of a minimum connected winning coalition and, as such, should be internally stable and difficult for opponents to break up. After examining the process by which the new coalition was formed, I will do a simple game-theoretic analysis to highlight its desirable properties.
In May 1996, David Frum and Ezra Levant organized the Winds of Change conference in Calgary to discuss a possible unification of the Reform and Progressive Conservative (PC) parties. The most important thing that came out of the meeting was a statement by Stephen Harper of how a conservative party could regain power in Canada. Harper’s speech turned into a road map that he followed faithfully once he became leader of the Canadian Alliance; and, with one important modification, it led to the Conservative parliamentary majority elected on May 2, 2011.
The basic idea that Harper laid out at the Winds of Change conference was to reconstitute Brian Mulroney’s electoral coalition, which Harper analyzed in tripartite terms: populists in western Canada and rural Ontario (who then supported the Reform Party); traditional Tories in Ontario and Atlantic Canada (who were still voting PC); and francophone nationalists in Quebec (who were then voting for the Bloc Québécois [BQ]). Harper argued compellingly that all previous Conservative majorities in the 20th century, whether led by Mulroney, Diefenbaker, Bennett or Borden, had been constituted in this way. By implication, this tripartite electoral coalition was the only way to form a Conservative majority at the federal level in Canadian politics.
At the same time, Harper was very aware that previous victorious Conservative coalitions had not lasted very long, because they had an ends-against-the-middle quality. The francophone nationalists decamped after a few years, when they found their demands could not be met by the rest of the country; think of the way that Diefenbaker’s huge 1958 majority disintegrated in 1962, and Mulroney’s equally huge 1984 majority disappeared in 1993. Harper hoped that a future reincarnation of the Conservative winning formula could avoid a Mulroney-style overreach; that is, it could recruit only francophone nationalists who were not separatists, and who shared at least some elements of economic and social conservatism with western populists and traditional Tories.
As soon as he was chosen leader of the Canadian Alliance in March 2002, Harper sought a meeting with Progressive Conservative Leader Joe Clark to discuss a possible unification of the two parties. In Harper’s mind, that would have been the essential first step in rebuilding a winning coalition. Clark wasn’t interested, but Harper didn’t give up. As soon as Peter MacKay was elected PC leader in May 2003, Harper proposed that they start discussing cooperation of the two parties; and he pursued that project relentlessly until a merger was achieved late in 2003. Harper insisted on only one condition — that the merged party be named the Conservative Party, not the Progressive Conservative Party — while yielding on all other issues. He knew that if he could win the leadership of the merged party, he would have the freedom to shape it as he wished.
Bringing the western populists and traditional Tories back together had the immediate effect of making the new Conservative Party competitive with the Liberals. With very little time after the leadership race to prepare for the June 2004 election, the Conservatives were still able to bring Paul Martin’s Liberals down to a minority government — a foretaste of things to come. But the winning formula was still not complete, for the Conservatives won no seats at all in Quebec and got less than 10 percent of the popular vote in that province. The merged party had inherited odds and ends of PC and Canadian Alliance support, while the BQ had a lock on the francophone vote and the Liberals were still the dominant force, and default federalist choice, among anglophones and allophones.
After the 2004 election, Harper made it his top priority to find electoral support among francophones in Quebec. He hired more French staff and visited Quebec frequently to make contacts and recruit candidates. He dictated that the party’s March 2005 policy conference would be held in Montreal, and that the policy book should contain certain items necessary for success in Quebec (such as bilingualism, supply management and fiscal imbalance). Those of us charged with preparing the next campaign reviewed what we had done in 2004 and realized that we could not win in Quebec simply by constructing a campaign narrative in English and translating it into French. We would have to run a made-in-Quebec campaign, finding Quebec communications and advertising people to build a narrative in which the BQ was the major opponent, not the Liberals.
All this work seemed to produce few results at first, but Harper persisted, even though there was considerable pressure within the party to de-emphasize Quebec and put more stress on Ontario. The reward came in the election campaign of 2005-06. Harper’s December 19, 2005, speech in Quebec City outlining the Conservatives’ Quebec platform was extraordinarily well received, and his polling numbers shot up thereafter. In the end, the Conservatives won 10 seats and 25 percent of the vote in Quebec, which made a big contribution to beating the Liberals and establishing a minority Conservative government.
As soon as Peter MacKay was elected PC leader in May 2003, Harper proposed that they start discussing cooperation of the two parties; and he pursued that project relentlessly until a merger was achieved late in 2003. Harper insisted on only one condition — that the merged party be named the Conservative Party, not the Progressive Conservative Party — while yielding on all other issues.
The 2006 result seemed to validate Harper’s theory of how to reconstitute a Conservative coalition of majority dimensions. With a beachhead of 10 seats in Quebec, Harper thought he was only one election away from harvesting a much larger number of seats in the province. After fulfilling campaign promises to Quebec — higher transfer payments to correct the fiscal imbalance, representation at UNESCO — Harper called another election in September 2008, hoping this time to win a majority with much greater representation from Quebec.
Things seemed to be on track initially. After the first two weeks of the campaign, the Conservatives were running neck-and-neck with the BQ in Quebec, each party commanding about 30 percent of the vote. That looked like 25 or 30 seats for the Conservatives, until the BQ counterattacked effectively over the so-called “culture cuts—” small budgetary reductions to certain cultural programs financed by Ottawa. When the Bloc portrayed these as an insult to Quebec and a threat to the French language, Conservative support went back down to the 20 percent range, and at the end of the campaign the Conservatives were lucky to escape with 10 Quebec seats, the same number they had won in 2006. They ended up with 143 seats nationally, enough for an enhanced minority government, but not the majority they had hoped for.
This reversal in the 2008 campaign proved to be a turning point in Conservative electoral strategy. Although the party did not abandon hope of making gains, Quebec ceased to be pivotal in Conservative electoral calculations.
Increasingly, the role of third and essential pillar to complement the western populists and traditional Tories would be played by ethnic voters, new Canadians, mostly in Ontario rather than Quebec. This would involve a direct assault on the Liberals’ ethnic fortress of Toronto rather than on the BQ’s hegemonic hold over francophones in Quebec.
This strategic shift was not a totally new departure but rather an increased emphasis on what had been taking place since early 2005. At that time, when same-sex marriage was under heated debate in Parliament, Harper decided to use it as a wedge issue to approach ethnic voters. He ordered the party to spend about $300,000 on print advertisements in Canadian ethnic newspapers, running in early 2005, to point out that the Conservatives were the only party opposed to same-sex marriage. From that point on, the party put more and more effort into courting ethnic voters.
Many new Canadians are socially conservative, believing in stable traditional families rather than the lifestyle obsessions of Liberal elites. Most are religious; a surprising number, especially among Chinese, Vietnamese, Koreans and Filipinos, are Christian. Many are economically conservative and entrepreneurial, running small businesses and concerned about the tax burden. In other words, many immigrants look like Conservative core voters, except that they may have a different skin colour and mother tongue.
A few policy innovations in the 2005-06 platform — lowering the immigration landing fee, an apology for the Chinese head tax, a judicial inquiry into the Air India bombing — signalled to multicultural communities that the Conservatives cared about their issues. Beyond that it was mainly the patient effort of establishing contact — visits by Harper and other leading Conservatives, notably Jason Kenney, to ethnic events; recruiting multicultural candidates and political organizers; printing political materials in languages other than English and French. The steps are easy to enumerate, but they took years of effort to carry out.
The underlying assumption of the Conservative outreach was that many ethnic voters “ought” to be voting Conservative rather than Liberal. Many new Canadians are socially conservative, believing in stable traditional families rather than the lifestyle obsessions of Liberal elites. Most are religious; a surprising number, especially among Chinese, Vietnamese, Koreans and Filipinos, are Christian. Many are economically conservative and entrepreneurial, running small businesses and concerned about the tax burden. In other words, many immigrants look like Conservative core voters, except that they may have a different skin colour and mother tongue. The task, as Conservative strategists saw it, was not to win over these voters by promising them a potpourri of new benefits; it was to help them realize that their convictions and interests would be better represented by the Conservatives than by any other party.
Following Harper’s decision to make a serious play for ethnic voters, there was a steady rise in the Conservative share of the multicultural vote through the elections of 2006, 2008 and 2011, to the point where the Liberal dominance of cultural communities has been definitively broken. According to the 2011 Ipsos Reid exit poll (n > 36,000), Conservatives received the support of
- 43 percent of immigrants who have been in Canada longer than 10 years,
- 42 percent of Canadian voters born outside Canada,
- 31 percent of visible minority voters,
- 52 percent of Jewish voters, but
- only 12 percent of Muslim voters.
Although much more analysis will be required, it is clear that the Conservatives, once a largely “white bread” party, are now dominant or competitive among most categories of immigrant, ethnic and minority voters. The one obvious exception is Canadian Muslims, among whom foreign policy issues raise barriers.
This increase in ethnic support released a treasure trove of seats for the Conservatives. In the Greater Toronto Area, once the Liberal equivalent of the Tory Fortress Alberta, the Conservatives won 30 of 45 seats in the recent election, including many in areas such as Brampton that are heavily ethnic. It was this batch of new seats in Ontario, mainly in the GTA, that gave the Tories their majority in 2011, for they actually had a net loss of seats outside Ontario. Harper’s road map finally took him to the goal of a majority government, but only after francophone nationalists were replaced by Ontario ethnics as the third pillar of the coalition. Francophones provided an essential boost in 2006, but their support did not prove large and durable enough to constitute an essential pillar of the victorious coalition. The Conservatives would still have a majority in 2011 even if they had won no seats at all in Quebec, where the rising tide of support for Jack Layton peaked on election day at 43 percent and took out five Conservatives who looked set to survive had the NDP risen no higher than the low 30s.
Game theory predicts that, in inperson voting games, rational actors will seek to assemble a minimum winning coalition (MWC), that is, a coalition barely large enough to win. The theorem is counterintuitive, for politicians normally speak as if they would like to have everyone’s support. However, if the purpose of a coalition is to deliver benefits to the included at the expense of the excluded, it follows that the winning coalition should be as small as possible if it is to maximize benefits to the participants per capita.
In Canadian federal politics, the MWC for the House of Commons is 155 (50 percent + 1 of 308 seats). In the real world, of course, you want to have more than a bare majority of 155 to guard against the possibility of resignations, deaths or defections of caucus members to other parties. The MWC for the popular vote cannot be defined so precisely but seems to be slightly less than 40 percent, given the current state of Canada’s multiparty system. By both standards, the current Conservative coalition (166 seats, 39.6 percent popular vote) is ideal.
Larger-than-necessary coalitions tend to be unstable because of the difficulty of satisfying too many participants. In international relations, think of the breakup of the victorious alliances after the Second World War, the First World War and the Napoleonic Wars. The paradigmatic cases in Canadian Conservative history are the collapse of the Diefenbaker and Mulroney supermajorities earned in 1958 and 1984. The 2011 Conservative majority will be much less unwieldy and should not be subject to the stresses arising from unnecessary size.
An even better concept than MWC for analyzing political coalitions is MCWC — the minimum connected winning coalition. In the real world, coalitions are costly to form and maintain, and it is easier to bring together participants who are closer together on relevant dimensions of difference than participants who are farther away from each other. Thus, in the aftermath of the most recent British election, the governing coalitions that received serious consideration were Conservative-Liberal Democrat and Labour-Liberal Democrat, but not Conservative-Labour (the latter would also have been much larger than necessary). A Conservative-Labour coalition would have had to jump over the centrist position of the Liberal Democrats.
In the Canadian Conservative electoral coalition, the new multicultural pillar seems well connected to the older western populist and traditional Tory pillars. The multicultural voters who have now been attracted to the Conservative Party (not all, to be sure) seem to be demographically and psychographically similar to other Conservative voters — middle-aged or older, married with children, imbued with family values, respectful of religion, distressed about the impact of crime, oriented toward the private sector and concerned about taxes and the general business climate. They make no demands on government other than those that Conservatives generally make. They may have some racial or linguistic differences, but their location in policy space is very close to other Conservative voters. That makes the coalition “connected.”
In contrast, francophone nationalists always present a problem, even when they can be brought to offer support to the Conservatives. They tend to have an instrumental orientation toward the federal government, seeing it primarily as a source of benefits for Quebec. This raises resistance among other Conservatives, who fear they will have to pay for these benefits to Quebec. Also, in the unique political culture of Quebec, the position of the median voter is substantially to the left of the median Canadian voter on many issues. Francophone nationalists are, therefore, less likely to share the economic and social world view of Conservatives in the rest of Canada. Also, francophone nationalists, even if they are not separatists, usually see the separation of Quebec as a legitimate option, whereas anglophones tend to see it as disloyal. For all these reasons francophone nationalists in Quebec seem less connected to the rest of the Conservative coalition.
In practice, French support in Quebec for the Conservatives was always fragile. Think how easily the issue of “culture cuts” dislodged Conservative supporters in Quebec in 2008, even after the Conservatives had built a beachhead into the province in 2006. By comparison, the support of ethnic voters in Toronto and other metropolitan areas seems more likely to be stable, precisely because the Conservatives have attracted those ethnic voters who were already most like themselves in terms of demographics and politics.
It has not always been a straight path for Harper and the Conservatives since the party was refounded in 2003. Harper’s road map to power took him most of the way to majority government, but not quite all the way. It did not get him there until he made a crucial substitution of multicultural for francophone voters. But now that it is constructed, the Conservative electoral coalition of western populists, traditional Tories, and ethnic voters who share Conservative economic and social values should be difficult to break up. It fulfills all the requirements of a minimum connected winning coalition. This does not mean the Conservatives will win every election, but it does mean they should be able to hold their core voters together in bad times as well as good, adding and subtracting other supporters as conditions vary. Unlike the Diefenbaker and Mulroney coalitions, the Harper coalition will not be constantly threatened with breakup from internal tensions. It has the potential to dominate Canadian federal politics for a long time.
The main threat to the dominance of the Conservative coalition will be the difficulty of governing Quebec. With only five Conservative members from that province, Quebec will not be very well represented in caucus and cabinet. And Quebec issues may move front and centre for the federal government if, as expected, the Parti Québécois wins the next provincial election and once again starts talking openly about sovereignty. Hence the Conservatives cannot afford to write off Quebec, for fear of another crisis of national unity. But in their attempts to rebuild support in Quebec, it will be reassuring to the Conservatives to know that support in that province is not pivotal to their majority coalition.