Thinking presidentially, Canadian debate edition

Canada’s party leaders–or rather four of them–are holding their only pre-election debate in English this evening at 7:00 p.m., Eastern Time. (A debate in French is tomorrow, the date having been changed for the thoroughly Canadian reason of conflict with hockey.)

In Monday’s Globe and Mail, John Ibbitson says, in explaining why leaders, and by extension debates, matter:

while they [“political analysts”] know that Canada is on paper a parliamentary system, in reality voters think presidentially. They don’t vote for the candidate they want to represent them in the House of Commons; they vote for the leader and party they want to form the government

There are actually two statements in there, one fairly accurate, and the other way off. So let’s unpack them.

In virtually all parliamentary democracies, voters vote mainly for the “leader and party” they want to form the government–or gain a share of it, where coalitions are expected–and not for the candidate they want to be their local or regional representative. That’s parliamentary politics 101.

However, the first part of the quoted passage makes a much bolder claim, that voters vote as if it were a presidential election. This is a very different claim from simply asserting, as the second sentence implies, that the leader might be a key part of the reason for a voter’s party choice. It actually suggests that voters go a step farther, and vote for a given party if they want the leader to head the government–full stop. That has to be what “thinking presidentially” means. For what makes presidentialism distinct from parliamentarism in the voting process is precisely that voters can vote for a presidential candidate without also voting for that candidate’s party in the legislative election.

There is a literature in political science on “presidentialization” of prime ministers. It may be clear from these remarks what I think of it. If it is not, please consider reading (better yet, buying and reading!) Presidents, Parties, and Prime Ministers, because I could quite literally write a whole book on the topic (with lots of help; thanks, David!).

The short(-ish) version is that personalization is not the same as presidentialization. The personal characteristics of the executive candidate (or the legislative candidate, for that matter) might be a major factor in vote choice, perhaps more important than the party program or simple party loyalty. But unless you can show me that the candidate was selected to appeal to voters who would not otherwise vote for the party, and with the party having accepted that said candidate may not actually share the collective preferences of the party, then please do not tell me it is presidentialization. Because, one can show that parties in presidential systems make these sorts of tradeoffs all the time, and often to the detriment of the party as a whole. But one can also show that it simply is not the case very often in parliamentary systems. Yes, it happens. Blair and Koizumi come to mind. But I do not believe any of the about-to-debate Canadian leaders represent parties that have made such presidentializing tradeoffs, and certainly not Stephen Harper, who is running his fourth consecutive campaign and whose party has won a plurality, but not a majority, in the last two.

12 thoughts on “Thinking presidentially, Canadian debate edition

  1. My guess is that (in Australia at least) voters give somewhere between 60% and 90% weighting to whom they want to be prime minister, and the rest to how well they like the local candidate. Anecdotally, I’ve come across a lot of people who’ll say “I don’t really like X’s party leader but X is a hard-working local member who knows the area”. If X is an independent, “party leader” ceases to be a factor, of course.

    Many commentators have concluded that John Howard hung on as PM in 1998 (2 years 7 months after he first became PM) mainly because a lot of the new Liberal MHRs who got in at the 1996 landslide managed to hold their seats on a personal vote. Howard actually lost the two-party preferred. (In Australia, the Senate vote – which is almost solely party-line – allows a useful comparison against the party’s local candidate’s personal vote. Unfortunately, unlike MMP, Senate and House votes are on separate pieces of paper!)

  2. I can see that because of Canada’s proximity to the United States, and because of Canadian voters’ exposure to American television (and how much of the Canadian elite attend US universities?), many Canadian voters fall into the habit of thinking about their system in American terms.

    This would explain the wide swings in the votes of the different parties that sometimes occur in Canadian elections, among other things. Canadian parties also use an American-style convention system to select their leaders, its perfectly possible that the leader of one of the two big parties be someone without much ministerial experience, or even much experience in federal politics, or in extreme cases someone who has not even lived long in the country.

    The Liberals are apt to combat the NDP with appeals to low information potential NDP voters to not split the anti-Conservative vote, which has apparently resulted in the past in NDP losses in NDP-Conservative marginal ridings where the Liberal vote had been minimal. Voters may be thinking they are voting for the party leader, not a local MP, in these cases.

    I’ll note that even in the US, even after 2000, there is a tendency to treat Presidential elections as being decided by nationwide popular vote when of course the actual system is significantly different.

    • I would not push this “Americanization” theme too far, Ed. It stands up to scrutiny no better than does “presidentialization.” Consider that every Canadian PM has served in the national legislature prior to becoming PM–at least since at least King (1935-1948; our data start with whoever was in power in any democracy as of 1945 or the first year of democracy). Most of them have served several years in the legislature before becoming PM (Harper 8, Martin 6, only Mulroney and Campbell less than four). Almost all of them, although not Harper, have served in the federal cabinet for at least a few years prior to becoming PM. Only Campbell ever served in a provincial legislature and none has been a provincial premier.

      Obviously these patterns are quite a contrast to south of the border, and the most likely explanation to account for these differences is parliamentary vs. presidential systems.

      None of this speaks to party leaders other than the ones who became PM. I agree that Ignatieff is an odd case, for all the time spent out of the country. I am only able to provide data from our project on PMs (and presidents).

  3. In parliamentary systems all over the world, voters can normally expect the party caucus to restrain, control, and even dismiss a leader who goes wrong or becomes unpopular between elections.

    In Canada, however, (almost uniquely in the parliamentary world) leaders are not chosen by the MPs. MPs are accountable to leaders, not vice versa, and leaders have almost unfettered authority between elections; so voting for a Canadian party candidate does feel like making a “presidential” choice, not a party vote.

    • The specifics of the candidate-nomination process, for both MPs and potential PMs, would be a worthy addition to this analysis of the accountability dynamics. Thanks for bringing that up, Christopher.

      The critical consideration for whether the nomination process breaks or even reverses the standard principal-agent relationship in parliamentarized parties (where, in stylized form, it runs MPs–>leader/PM) is the extent to which the selectorates have opposing rather than aligned interests. Parties in presidential systems vary on this dimension, but the presence of separate electorates–presidential candidates usually having to appeal more broadly than even a cohesive party does, and shaping their own campaign strategies–as well as fixed terms of executives undermine this alignment.

      In parliamentary systems, parties have various mechanisms to reduce their danger of adverse selection (ending up with a leader who does not share the collective goals of the party) and moral hazard (being unable to discipline a leader once selected). Among these are “vetting” leadership candidates through requiring them to serve in the legislature and/or cabinet before becoming a PM candidate, and giving them “pensions” in the form of continuing party service after a stint as executive (or leader), which presumably they get to enjoy only if they were good servants while in the leadership.

      In my earlier comment in this thread, I addressed the first of these: a tendency for Canadian PMs to have spent notably longer time in the national legislature and cabinet prior to be elevated to the top job than do US (or most other countries’) presidents. As for the second mechanism, Canadian PMs also tend to remain in parliament, and often return to cabinet, after their stints as PM ends. Presidents in the US never do this, and those in other presidential systems rarely do.

      It seems to me that the evidence is pretty strong that Canadian parties are highly parliamentarized. Nonetheless, more detailed analysis of candidate-selection rules and their impact on these accountability relationships is clearly in order, because there are no doubt relevant variations within each regime type, as well as across parties within individual countries.

      By the way, are Canadian parties really so unusual in the process of selecting leaders? I don’t think so. It is not necessary for MPs themselves to be the sole selectorate in order for leaders to be bound to the party. Again, what matters is whether the selectorate for the leader/executive has distinct interests from those of the parliamentary caucus, as well as whether the party organization, including MPs, has or lacks means to remove a leader who overstays his/her welcome.

  4. I took the trouble of going through Wikipedia, which is pretty good on presenting the dates various politicians held their offices, to get an idea of the resumes of the past Liberal and Progressive Conservative/ Canadian Alliance/ Conservative Party leaders, starting in 1945.

    There was a big gulf between the Liberals, who as MSS noted generally had substantial federal political experience, and the Conservatives, who were generally primarily provincial pols.

    The Liberals selected eight leaders since the end of World War II. Seven of them served in the federal cabinet prior to becoming leader, and four of them had been federal MPs at least ten years prior to becoming party leader. None of them had any provincial political experience, though several had substantial careers in other fields (mainly law and academics( before entering politics.

    The Progressive Conservatives chose eight leaders between 1945 and 1993. The Canadian Alliance selected one leader while they were the official opposition (Preston Manning became Reform leader before they were the official opposition), and so far its successor party has chosen one leader. Of these nine leaders, two served in the federal cabinet before becoming leader. However, five were provincial pols, with two serving as provincial premiers. Four first entered the federal parliament the same year they became leader, though to be fair this includes both provincial premiers. Only one, Diefenbacker, was elected to the federal parliament more then ten years before becoming leader, though Harper barely misses this cutoff, having first entered parliament nine years before becoming leader.

    It should be noted that Canadian federal cabinets are unusually large, so if you sit as a MP for the governing party it is not that difficult to become a cabinet member compared to other parliamentary systems.

    I still think federal Canadian politics has had more “dark horses” and “golden boys” than is normal for a parliamentary system, though Liberal leaders did have substantial political experience during the decades when the Liberals held a near-monopoly of the federal government. There is a difference if you look at just the experience of the PMs, as opposed all the leaders of the two major parties.

    I summarized the careers of the sixteen leaders below dashes (this is pretty tedious, so I made my main points above).


    Louis St. Laurent-entered national politics and the federal parliament in 1941, at the age of 60, after having been a nationally prominent lawyer. Party leader in 1949.

    Lester Pearson-Entered the cabinet and the federal parliament in 1948, after a diplomatic career (apparently a serious candidate for UN Secretary General). Leader in 1958.

    Pierre Trudeau-nationally prominent law processor and intellectual, entered federal parliament in 1965, cabinet minster in 1967, party leader in 1968.

    John Turner-federal parliament in 1962, cabinet minster in 1965, party leader in 1984

    Jean Chretien-federal parliament in 1963, cabinet minster in 1967, party leader in 1990

    Paul Martin-federal parliament in 1988 (political family and corporate career), cabinet member in 1993, party leader in 2002/3

    Stephane Dion -federal parliament and cabinet minster in 1996 (prior career as academic and activist), party leader in 2006

    Michael Ignatieff -federal parliament in 2006, party leader in 2008 (prior academic career)

    Progressive Conservative

    George Drew -Premier of Ontario. First elected to the federal parliament in the year he became leader (1948)

    John Diefenbacker -elected to the House of Commons in 1940. Previous experience was as a crown prosecutor and a (pretty unsucessful) provincial pol. Leader in 1956.

    Robert Stanfield -Premier of Nova Scotia. Entered the federal parliament after becoming leader in 1967.

    Joe Clark -Entered federal parliament in 1972, leader of the Progressive Conservatives in 1976. Obscure career prior to entry in Parliament.

    Brian Mulroney -entered the federal parliament after becoming leader in 1984. Well known lawyer party activist (leadership contestant in 1976), but apparently no elective office prior to becoming party leader.

    Kim Campbell -provincial BC legislature in 1983, federal parliament in 1988, cabinet minster in 1990, party leader in 1993

    Jean Charest -federal parliament in 1984, cabinet minister in 1986, party leader in 1993

    Canadian Alliance etc.

    Stockwell Day -Alberta legislature in 1986, provincial cabinet in 1992, party leader in 2000

    Stephen Harper -federal parliament in 1993, party leader in 2002, national visibility for about a decade before becoming party leader

    • Thanks for that, Ed. Your work here points the way towards further work I’d love to do as an extension of Samuels and Shugart: losing party leaders, including losing presidential candidates, of major parties. In my future career, perhaps…

  5. Matthew, it is rare for a Canadian prime minister (or provincial premier) to remain in the legislature after being the boss (your comment 5 above). John Diefenbaker and Joe Clark are the only ones I can think of in national politics back to before the Second World War. As a rule, Canadian PMs and premiers leave the legislature almost the instant their tenure in office ends.

    Britain, Japan, Australia, New Zealand, Ireland, etc., have many examples of party leaders being removed by their caucuses while prime minister. It hasn’t happened in Canada since 1896 (or 1880, depending on how you count it).

  6. Rumours are going around the commentariat that Kevin Rudd is positioning himself for a possible comeback as PM if Julia Gillard is seen to have failed. Rudd was less unpopular among voters than among his fellow Labor MPs (in public, he comes across like Kevin Spacey playing the Milky Bar Kid but in private he is alleged to be a control freak with a bad temper) and his sudden removal caught most of Australia by surprise.

    Now, with Gillard under fire from her right flank over the newly-resurrected carbon tax plan and from her Left over her opposition to SSM and a speech attacking the Greens, she is looking vulnerable and it is not inconceivable that Rudd could make a comeback as leader (as Menzies, Jeff Kennett and John Howard have pulled off in the past). The interesting question for our purposes is whether Rudd can amass enough goodwill and/or nostalgia at the grassroots to overcome the personal antipathy his colleagues hold towards him, so that they re-install him with their teeth gritted. In a presidential system, of course, he would have a much better chance of a Nixon-style reinvention.

  7. PM, Years MP after (since 1960)
    Diefenbaker, 12
    Person, 0
    Trudeau, 1
    Clark, 17
    Mulroney, 0
    Turner, 6
    Chretien, 0
    Campbell, 0
    Martin, 2

    So, my “tend to remain” was perhaps an overstatement. But in comparison to presidential systems, where this tendency is far more rare (and the USA where it has never happened since 1945), Canada indeed gives us some executives with continuing legislative service.

    All the same, I have to concede that Canada is clearly below the average for parliamentary systems, where we found 82% (!) continue being legislators (compared to 20% of presidents), and that the average PM hangs around for over 7 years (compared to just over a year for the average president). As I mentioned above, this line of research would really benefit from more comparison of countries within regime types. All David and I have done so far is compare aggregate career paths, by regime type. We have over 900 executive bios, so we could do much more! (All the better if Ed wants to be our RA/coauthor on losing party leaders!)

    Also of interest: Only one Canadian PM during this time period has remained in the cabinet–Clark for one year–so this type of “pension” is indeed rare in Canada. For all PMs we find 26% remain in the cabinet after their executive service, compared to no presidents in pure or semi-presidential systems.

  8. Interesting example of this in the recent Irish election. The newly elected Taoiseach, Enda Kenny, had a long-standing credibility problem, which almost cost him his job as party leader. The leader of the Labour Party and newly appoint Tainaiste, Eamonn Gilmore, had very high poll ratings and ran a blatantly “presidential” campaign. Kenny regrouped from almost losing his partys leadership by tacitly admitting that he himself was no great leader, emphasising instead his role as chief of a strong team that could resolve the country’s problems. Gilmore and his advisers pursued the “Gilmore for Taoiseach” slogan up until the dying days of the campaign even as Labours poll ratings slid further-a series of poor performances by key Labour spokespeople making the party seem increasingly like a one-man band, allowing Fine Gael to claim the high ground of “credibility”.

    That isnt the complete explanation for the reversal by Fine Gael of what seemed like a strong challenge from Labour. But it does illustrate the limits of a “presidential” camaign in a parliamentary system-the collective being elected does matter.

  9. A few points:
    – I think Ibbitson (or his editor) is only guilty of using shorthand when he mentions presidentialization. I realize you are trying to teach us something, but it seems ill-placed. Regardless, I enjoyed the thrashing.;)

    -Clark remained in cabinet throughout the two Mulroney terms (9 years). A short-lived PM, he turned out to be an excellent External Affairs minister.

    – The fact that caucus members can’t sack party leaders in any major Canadian parties (since the late ’60s?) IMO is a major drawback. And you’re right to draw the distinction between hiring and firing – MPs could do the firing, while conventions and/or members do the hiring. While this could be problematic in theory, in practice it would make leaders much more accountable to caucus and their constituents. It also hurts that party leaders have a veto on nomination papers, so that leaders completely control the fortunes of MPs. This is why the Canadian Parliament is largely an appendix to Canadian politics.

    Ed: Canadian and American leadership selection are only superficially similar. Canadian leaders are elected increasingly by direct voting by all members, although brokered convention dominated until recently. This is a big contrast to the primary system and convention-as-advertisement in the two modern-day dominant American parties.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.