Do parties dump their own PM less often in Westminster systems?

Building on a discussion in one of the Australia post-election threads, I checked the data for Presidents, Parties, and Prime Ministers to see if PMs in “Westminster” systems were any less frequently dismissed due to “intra-party” politics than PMs in parliamentary systems overall. To make a long story short, they are (except, apparently, in Australia).

The question came up because Julia Gillard, the current PM of Australia, replaced a co-partisan PM between elections, and Alan noted that her predecessor was the third out of the last ten so replaced. At the time I pointed out that 3 out of 10 was precisely the rate at which PMs in parliamentary systems overall are dismissed by their own parties. More precisely, we found that 30.2% of 354 PMs surveyed left office on account of internal party politics. (The remainder was about evenly split between election defeats and coalition breakdown.) So, Australia seemed “normal,” not a case where PM termination between elections was overly frequent by the comparative yardstick.

But what about compared to similar parliamentary democracies, where majority government is the norm (or at least used to be!)? That is, let’s leave out parliamentary systems where coalitions are the norm, and where perhaps some cases of intra-party conflict are really generated by the tensions of governing in coalition, and not by purely intra-party matters.

My first look at the data suggested that the rate was not much different in this subset than overall (I found 22 of 78). However, quite a few of these were coded in the detailed data as something like “Intraparty — left office voluntarily.” Now, that is clearly an intra-party replacement, so I am not second-guessing our own data coding! However, there is quite a lot of difference between a party undertaking an inter-electoral leadership change because the former PM decided on his or her own accord to leave, and a challenge to a sitting leader who goes involuntarily.

So let’s look at things again, with cases of “voluntary” departure relegated to a residual category.

Now things look rather different. There are only nine of 78 PMs in Westminster systems who leave for reasons that might be termed intra-party conflict. That’s 11.5%, which I have to agree is a good deal less than 3 of 10!

What about in the non-Westminster cases? We should also remove the “voluntary” departures from this subset. When we do, we are left with about 24% leaving due to intra-party conflict in the entire parliamentary data. In the non-Westminster subset, it’s 28%.

So there you have it. Roughly one eighth of Westminster PMs leave due to intra-party conflict, compared to well over a quarter of non-Westminster PMs. That seems pretty significant. And I thank Alan for prompting me to look deeper at the data!


The countries taken as “Westminster” for the purposes of this exercise, and their own rate of dismissals for intra-party reasons, are:

Australia (2/9)*
Bangladesh (0/3)
Canada (1/12)
India (1/14)**
Jamaica (0/6)
New Zealand (1/12)***
Sri Lanka (0/10 during its parliamentary regime)
U.K. (4/12)

The real question, of course, would be to break down majority vs. coalition (and minority) governments, not Westminster vs. non-Westminster. But the two categories are quite closely aligned.

Obviously, another big difference across parliamentary subtypes is in the percentage of PMs who leave office due to electoral defeat. Elections account for about a third of all parliamentary PM terminations, but 48.1% (38 of 79) in these Westminster systems. Electoral defeat is what we might expect to be the most common mechanism of PM termination in a Westminster system, of course. Just as obviously, inter-party conflict is almost never a cause of PM termination in Westminster systems: just five PMs in India and one in Bangladesh left for such reasons. (Perhaps we could add Whitlam, in Australia, who is coded as an “other” due to his dismissal by the Governor General, but the underlying reason for conflict was his lack of support in the second chamber.)


* This was before Rudd’s forced resignation.

** The Indian case is Desai, who headed the unwieldy (and not very Westminster) anti-Congress coalition that collapsed in 1979.

*** I excluded PMs in the PR era, since 1996. Since then, our dataset has only Shipley, who served out her term. We could add Clark, who also survived to see electoral defeat. The two of them would leave NZ, through 2008, at 1/14.

33 thoughts on “Do parties dump their own PM less often in Westminster systems?

  1. We will not even mention the great state of New South Wales where 2 premiers have been removed by their own party in the current parliament.

  2. I recall that the LDP in Japan had a policy of rotating its Prime Ministers every two years, though some could stay for another two year term. I’m not sure where this shows up in your analysis.

    The famous rotating governments in the Italian First Republic was really something similar to the Japanese system, just done more informally.

  3. The LDP, historically more a collection of factions, is sort of Exhibit A in our analysis, as a parliamentary party that regularly changed PMs between elections. And the new ruling party, DPJ, seems to be doing its best to catch up with its predecessor.

  4. Regarding Westminster PMs, the question might arise as to why they are so much more likely to leave “voluntarily.” And I suppose the answer is simple: if they are less likely to be dumped by their own parties and highly unlikely to lose office due to conflict with other parties, then those that survive a while without losing elections wind up leaving because they are old or in ill health.

  5. Although it does not affect the main argument above, the case of Paul Martin could be re-coded. We have him as “lost election,” but the 2006 election was called only because Martin’s minority government lost a no-confidence vote in the Canadian House of Commons in December, 2005. So, it was indeed an election that caused him not to continue as PM, but it was “inter-party conflict” that forced the election. Martin’s was the first Canadian government to fall on a no-confidence motion.

    While on the topic, as might be expected, the only Indian cases of PM termination due to inter-party conflict were minority administrations after the demise of normal Congress majorities but before the emergence of the two competing pre-electoral blocs (Singh 1990, Shekhar 1991, Vaypayee 1996, Gujral 1998).

    The other “Westminster” PM to have left office due to inter-party politics was Khaleda Zia in Bangladesh in 1996, following boycotted elections and subsequent turmoil.

    So, indeed PMs losing power due to inter-party politics in Westminster systems can be considered extraordinary. Even more so than losing due to intra-party politics, which remains sort of “normal” parliamentary party politics, even if it accounts for less than 12% of Westminster PM terminations.

  6. > “… wind up leaving because they are old or in ill health”

    Until about five years ago, it was very rare for Australian chief ministers to retire voluntarily – Bob Menzies and Bob Askin (NSW 1965-75) are among the few who come to mind. Then, in quick succession, in the past half-decade, Bob Carr, Peter Beattie and Steve Bracks retired as Premiers of NSW, Qld and Victoria (not sure in that order but googling would be cheating). All three were relatively young (50s) and in good health. They’d just had 2-3 terms in the job and “been there, done that”.

    Geoff Gallup also resigned as WA Premier because he had difficulty with depression (as had the Liberal leaders in Vic and NSW, although they didn’t come out about it until after they were deposed).

    Jim Bacon “retired” as Tasmanian Premier shortly before he passed away from cancer.

  7. … Further to this: Bracks and Beattie managed a smooth transition to a successor (John Brumby – ironically, an electoral failure as Bracks’ predecessor in the Victorian Labor leader’s position, but a success as his successor – an Anna Bligh respectively) who was widely-known in advance – sort of like Hawke/Keating or Blair/Brown but without the bloodied knives.

    In both cases, presumably as a result of this, B2 and B2 have managed (… so far) to continue B1’s and B1’s electoral success.

    Bob Carr, on the other hand, resigned very suddenly, and his replacement Morris Iemma was little-known.Post (and presumably ergo propter) hoc, NSW Labor is in trouble at the polls… putting it mildly. (Someone noted in The Australian this morning that there are now 14 ex-Ministers from the current NSW parliamentary term who are either in jail or else out of office, as opposed to only 13 still in the Cabinet. It’s like Illinois with a harbo[u]r bridge).

  8. I don’t have time to research the Indian states today. But there is something odd about a list that includes U.K. (62 million), Canada (33), Australia (21), Sri Lanka (21), New Zealand (4), and Jamaica (3), while excluding Uttar Pradesh (197 million), Maharashtra (112), Bihar (100), Bangla (90), Andhra Pradesh (84), Madhya Pradesh (71), Rajasthan (68), Tamil Nadu (66), Karnataka (59), Gujarat (59), Orissa (41), Kerala (34), Jharkhand (31), Assam (30), Punjab (28), Haryana (25), Chhattisgarh (23), Delhi (19, chief minister Sheila Dikshit), and Jammu and Kashmir (12), never mind the 16 minor states.

  9. The biggest reason for excluding the Indian states from this sort of analysis (even if I had the data) would be that the parties at that jurisdiction level do not fully control the selection of their own candidates for the position of Chief Minister. Those decisions (at least for Congress) are taken at the center level.

    That is only one example of the difficulty of comparing jurisdictions at different levels. That is not to say that it can’t be, or should not be, done. But the theoretical reason for why national population would “contaminate” the relationships being studied here is far less clear to me than is the case federal politics contaminating the politics of the sub-units (in all federations to some degree, but especially given party organization in India).

    Comparing the Indian states to one another, on the other hand, would be very interesting. If only the data were readily available. (Some day it may be, given the increased interest over the years in subnational politics in general, and India in particular.)

  10. @Bancki

    As far as I know no premier has become prime minister of Australia since shortly after Federation when almost all MHRs were former state MPs. There were 2 attempts to promote premiers of Queensland as federal labor leaders in the 1930s but neither succeeded. The experience in other Westminster federations may be different.

    I have rather a soft spot for Ben Chifley, prime minister 1945-9 who used to drive himself about 250 kilometres from Canberra to Bathurst to chair the shire council once a month.

  11. Good question, Bancki. In the research for the book, David and I found that it was much less common in parliamentary than in presidential systems for the head of government at the national level to have been previously head of a state/provincial government.

    The reason is that “outsiders” are relatively more attractive to parties in presidentialism, where they have to win popular votes as individuals. Governors can run on their record as executives at state/provincial level. Governors are relative outsiders to the national party.

    By contrast, parliamentary parties prefer relative “insiders” who have built their career in the national party. So PMs in parliamentary systems are much more likely to have been MPs, cabinet ministers, and to have had various party leadership positions. So is it somewhat rare for subnational executives to become PM–not only in Australia.

    (See Chapter 3 of the book.)

  12. Sorry for my bad english (promoVAte)
    Such state-federal-swaps do occur in my own Belgium (but that’s a special case of federalism) and, I think, in Germany.

  13. Do provincial/federal swaps happen in Canada? I’m wondering because they tend to elect party leaders by convention rather than legislative caucus. In Australia the Greens are the only parliamentary party where the leader is not elected and removed by the caucus.

  14. [What Alan said, +] The former (?) Aust Democrats had a system where their leader was elected directly by the membership, but had to be an MP (if the Dems held any seats). Moreover, a new leadership election (does anyone outside Oz call those a “spill”?) could be triggered either by a majority of the party’s MPs, or by some percentage (10%? 15%?) of the membership.

  15. The Democrats destroyed themselves, largely be setting their recall trigger too low, it was 100 signatures to recall the federal leader. No vote just a petition.

  16. In Québec, it seems far more common for a federal MP to switch to provincial politics! The current premier, Jean Charest, was the leader of the federal Progressive Conservatives until he jumped to the provincial Liberals (which is not affiliated with the federal Liberal party). He even sat in provincial opposition for awhile, which would certainly seem like a demotion. The previous elected premier was Lucien Bouchard, and he was first a cabinet minister in the federal Provincial Conservatives, then founder and leader of the federal Bloc Québecois, and finally leader of the provincial Parti Québecois. Lesage was briefly an MP as well.

    Outside Québec, Tommy Douglas went from Saskatchewan premier to leader of the federal NDP. Bob Rae was NDP premier of Ontario, then an MP and candidate for the federal Liberal leadership.

    Since my ability to recall lists of premiers is limited, Wikipedia is actually a reasonably good data source for this, if imperfect. Just use the intersection tool, eg: Prime ministers who were also premiers: Tupper & Thompson. Or other PMs who were once provincial legislators: Mackenzie, Bennet & Campbell.

    You can also ask for federal party leaders who were at some point provincial MPPs or MNAs:
    Conservative, Liberal, NDP/CCF, Bloc Québécois, Social Credit. Some names that stand out: Bob Stanfield, Stockwell Day, Alexa McDonough & Michel Gauthier.

    Finally, provincial premiers who were also federal MPs, including Barrett, Dosanjh, Tobin, Mercier & Binns.

  17. Bob Rae started as a federal MP (NDP), switched to provincial politics (NDP), then wound back in federal politics (Liberal).

    Until recently, the Ontario legislature used the same districts as the federal legislature, so it wasn’t that difficult to make the switch, and I wonder why it hasn’t happened more.

    In the US, the state legislatures and other local offices are viewed as a sort of a minor leagues, which will produce candidates for the federal legislature. I wonder what this says about the real nature of federalism in the U.S. I’ve noticed this is even the case in New York, even though the local legislatures are fairly powerless (the NYC Council), or the individual legislators are fairly powerless (state legislators), and neither gets much media coverage.

  18. I really can’t figure out the tally for Canada for intra-party dismissals of a sitting PM (1/12). Only short-lived PM Bowell in 1896 (6th PM of 22) was forced to resign by his Cabinet, but he is not among the last 12 PMs. So it’s either 0/12 or 1/22.

    One could quibble about Chretien (#20), but Chretien could and did choose his own retirement date – in fact, intra-party opposition likely meant he postponed his retirement, out of spite.

    It is practically very difficult for a party to remove a leader or PM in Canada since leaders are selected by convention and removed by losing a leadership review at a convention. [Aside: IMO, the highest-impact yet easy improvement to Canadian federal politics would be to allow leaders to be dismissed by caucus, even if still selected by convention]

    “…the case of Paul Martin could be re-coded. We have him as “lost election,” but the 2006 election was called only because Martin’s minority government lost a no-confidence vote in the Canadian House of Commons in December, 2005. So, it was indeed an election that caused him not to continue as PM, but it was “inter-party conflict” that forced the election. Martin’s was the first Canadian government to fall on a no-confidence motion.”

    In minority-led Canadian parliaments the coding distinction is meaningless. Both non-confidence motions and elections are inter-party conflicts, the former a means of opportunity for the latter. Lost elections tend to trigger leadership reviews, although the criteria varies by party.

    To boot, Martin went into the election of 2005/6 triggered by the non-confidence motion ahead in the polls.

    “Point of order”
    Martin was hardly the first to lose his government to a motion of non-confidence, of which 5 have been lost, 4 leading to loss of government, 2 to immediate resignation as party leader.

    Meighen (Conservative, 1926)
    – King-Byng affair. Meighen loses a non-confidence motion, loses the election, resigns party leadership soon thereafter

    Diefenbaker (Progressive Conservative, 1963)
    – heavy interference by Kennedy Admininistration to force nuclear proliferation into Canada triggers non-confidence motion in House, Diefenbaker loses motion, loses election, bitterly loses leadership challenge in 1967, retires at death in 1979.

    Trudeau (Liberal, 1974)
    – engineers his own minority gov’ts defeat with non-confidence motion, wins election with majority, retains government

    Clark (Progressive Conservative, 1979)
    – mathematically challenged, loses budget vote, loses subsequent motion of confidence, loses election, later calls leadership review voluntarily, wins, voluntarily calls leadership convention anyway, loses on 4th ballot

    Martin (Liberal, 2005/6)
    – behind in polls, loses confidence of the house in May 2005, doesn’t call election, gives a ministry to an opposition MP 9 days later, carries on governing, is ahead in polls in November, loses motion of non-confidence, calls January election, loses, resigns leadership immediately

    So Martin was the first Liberal PM to lose government by non-confidence, which for the first time required non-confidence of parties to the left and right.

  19. Ross, the case coded as intra-party is indeed Chretien. Your point about his staying on longer than welcome is well taken. Something similar could be said about Blair, by the way.

    Obviously I do not agree that the inter-party/intra-party distinction is meaningless in non-majority situations, or a chapter of the book would not have been written around this distinction as a main theme!

    The recounting of the history of Canadian governments to fall is very useful. Thanks. It is worth noting that many of these are cases of losing confidence motions, not no-confidence motions. It was said by CBC at the time that Martin’s loss on a “straight no confidence motion” (that was the expression they used) was unprecedented. However, that is evidently open to interpretation. As you note, he lost owing to left and right combining against him, which was indeed distinctive.

  20. While demystifying Westminster-style parliamentarism, John McGarry and Brendan O’Leary claim :

    “The only effective checks and oppositions to the prime minister’s power during inter-election periods come from within the prime minister’s own faction in the governing party (Brown versus Blair; Martin versus Chrétien; Keating versus Hawke).”

    [R. Taylor (ed.), Consociational Theory – McGarry and O’Leary and the Nortern Ireland conflict, Routledge, 2009, p. 358]

  21. Possibly true for the UK and Canada, but in Australia the Senate (admittedly our senate is much more Washington than Westminster) is a significant check on the prime minister except in those rare cases when the government has a majority in the Senate.

  22. Federal/ State swaps seem rarer in Australia, but do occur – at most, though, from one Cabinet-level position to another (ie, a position other than premier minister at each level.*)


    The most prominent I can think of was Carmen Lawrence, WA premier who became a federal Cabinet Minister.

    Also Santo Santoro, Qld MP 1989-2001, and Qld Minister for Training and Industrial Relations 1996-98, lost his seat and was appointed to fill a Senate vacancy, where he served as federal Minister for the Ageing, 2006-07.

    And of course, Robert Menzies (PM 1939-41, 1949-66) had previously been Deputy Premier of Victoria (1932-34). But Pig Iron Bob was a class entirely by himself.

    Top of my head, those are the only three I can think of who served in Cabinets at federal and State levels in recent decades. Some others served as Ministers at one level but backbenchers at another. Eg,


    * Tim Fischer, Deputy PM 1996-2001, had previously been a NSW MP 1971-84.

    * John Olsen was a Senator 1990-92, after losing as State Opposition Leader but before becoming SA premier 1996-2001.


    Bob Katter, Qld MP 1974-92, was a Qld Minister but later a federal backbencher.

    On the other hand, Dean Wells, Qld A-G 1989-95, was a Qld Minister but had previously been a federal backbencher 1983-84.

    So apart from some of the early Founding Fathers (Reid, Barton), we have no examples of the common US practice whereby Governor Roosevelt/ Carter/ Clinton/ Bush becomes President, ie, transfers from head of government at the State level to head of government at the Federal level.

    (Admittedly this is harder to do when one has to win a seat to serve in the executive, but it’s not impossible. Bob Hawke was only an MHR for under 3 years before becoming PM. Maxine McKew and Cameron Dick, and some of Steve Bracks’ Victorian Ministers in 1999, were sworn in as Ministers/ Parliamentary Secretaries immediately after they were first elected. I suspect also that the tendency of deposed or defeated leaders to resign their seat immediately – rather than staying in the House for years after, like Ted Heath or Bill Hayden – also reduces the distance between the Westminster system and the to the US model in practice).


    * my new generic term to cover premiers, prime ministers, chief ministers, first ministers, statesministers, chancellors (in the Austro-German sense), and ministers president.

  23. Surely the methods of electing party leaders(*) play a large role here. If only the national convention can elect a new leader, and thereby deposing the former, it’s impossible for the parliamentary party group to remove the party leader. Different political cultures must also be important. Norwegian party leaders are generally elected unanimously after backroom negotiations and deals, since the parties try to avoid contested elections.

    Secondly, it is probably riskier for a party to dump its prime minister in a coalition government than in a one-party government, since it may risk disrupting the coalition.

    *More accurately prime minister candidates. At least in Norway, the prime minister is not always the leader of his or hers party, though it’s usually the case.

  24. Agreed. Australia is quite reactionary in the way the parliamentary parties, apart from the Greens, limit leadership elections to their caucuses.

    The turnover in prime ministers is bad enough, but the turnover in opposition leaders is even more dramatic. There were 3 in the last parliament. There were 21 opposition leaders between 1967 and now (counting Peacock, Howard and Beasley twice).

  25. I deplore ‘premier minister’ because not all heads of government are premiers. Depending on how MSS and his state vote on Proposition 19, and as this is a Californian blog, I propose ‘high minister’.


    John Brumby represented Victorian voters as a federal Labor MHR from 1983 to 1990, then as a State Labor MLA since 1993. He lost both the State elections he contested as Labor leader – in 1996 as Opposition Leader, and in 2010 as Premier (he succeeded after Steve Bracks resigned).

  27. ADDENDUM #2:

    Rob Hulls was a federal Labor MHR for Qld (for Bob Katter Jr’s seat, ironically, during the one-term interregnum between BK Sr passing away and BK Jr pulling the sword from the stone) then migrated to Victoria and ended up as that State’s attorney general.

  28. I was away on a hiking holiday and have only just come across this post and the thirty comments that followed it. I was surprised to see statements such as “no premier has become prime minister of Australia since shortly after Federation”, because Joseph Lyons was Premier of Tasmania for five years (1923-28) and then, later, Prime Minister of Australia for seven years (1932-39).

  29. Nigel’s quite right. Lyons was pre-Menzies which in US terms is like being pre-FDR – ie, a quite different political game.

    In fact I have even cited Lyons in at least two previous F&V threads – d’oh.

  30. Pingback: Spill time? | Fruits and Votes

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.