I first uncovered the correspondence in which Governor General Lord Aberdeen dismissed Prime Minister Sir Charles Tupper from office in July 1896 all the way back in 2012 while conducting research on something else. The documents have proven more fruitful and given me more ideas than I could ever have imagined at the time. But the material does not end with the correspondence between Aberdeen and Tupper. The Debates in the House of Commons between Tupper, then in opposition, and the new Prime Minister Sir Wilfrid Laurier, provided another fascinating take on these questions.

• Sessional Papers (1896), Correspondence Between Lord Aberdeen and PM Tupper
• Tupper’s Apology in the 8th Parliament

This latest article draws together all my analyses on this crucial precedent from 1896 with respect to the Caretaker Convention, when Governors General can dismiss Prime Ministers, the manner in which Governors General appoint Prime Ministers, whether Prime Ministers should resign so that the Governor General can appoint the new Prime Minister before a new parliament meets or not, and, more broadly, the nature of constitutional conventions themselves.

Bowden, J.W.J. “The Origins of the Caretaker Convention: When Governor General Lord Aberdeen Dismissed Prime Minister Tupper in 1896.” Journal of Parliamentary and Political Law 16, no. 2 (2022): 391-444.

First, this precedent shows without any doubt that Prime Ministers must resign when the Governor General rejects their constitutional advice and now forms a chain of other similar precedents in Canada stretching from George Brown in 1858 to Christie Clark in 2017. This is a practical matter. Prime Ministers and ministries as a whole must not merely command the confidence of a majority in the House of Commons but also that of the Governor General, who commissions the Prime Ministers to form a ministry in the name of the Queen of Canada. The Governor General represents Her Majesty the Queen and thus a higher authority than a mere Prime Minister. When I first re-discovered this old idea that prevailed from the 1840s to the 1970s but which recent political scientists have obscured around 2012, my presentations were met with skepticism at best and scorn at worst. What business did I, then a vicenarian with only an undergraduate degree, have in contradicting the luminaries of my field? Even now, I only have an MA and not PhD, and I’m not a professor, so how dare I continue to display such audacity? In reality, facts remain facts irrespective of who highlights them. An argument should stand or fall on its own merits instead of propping itself up against post-nominals and social prestige.  Furthermore, these luminaries of the 2010s gave me a niche to exploit through their own refusal to defer to their predecessors like R. Macgregor Dawson and Eugene Forsey and John George Bourinot and Alpheus Todd. (The fallacy of Appeal to Authority cuts both ways, you see.) In any case, after Lieutenant Governor Guichon dismissed Christie Clark as Premier of British Columbia in 2017 by rejecting her constitutional advice to dissolve the legislature, these wayward scholars came back into the fold and re-affirmed what all Canadian scholars had known from the 1840s to 1970s. The raw truth of what happened in June 2017 left them no other choice. As of August 2021, even Peter Russell now agrees with me and all our predecessors from before the mid-20^th century that Governors General force Prime Ministers to resign by rejecting their constitutional advice.

Second, Aberdeen articulated a strong rationale for what we now call the Caretaker Convention all the way back in 1896 when he refused to sign off on summoning senators and making other patronage appointments in the civil service. Tupper presented Aberdeen with these prospective appointments about two weeks after Laurier led the Liberals to their first parliamentary majority since 1874. Aberdeen believed that outgoing ministries should limit themselves to necessary public business and that they should avoid embarrassing the incoming ministry with partisan appointments and cited some precedents in his favour. In other words, Aberdeen articulated an argument for what we now call the Caretaker Convention. Tupper replied back with some other precedents which did, in fact, support his case – but they were all at least 20 years old. Already by 1896, trends had changed from the 1860s and 1870s, and Aberdeen decided to enforce the new emerging understanding instead of enabling classic mid-19^th century spoils. Tupper understood that Aberdeen would force him to resign by rejecting these appointments because their written correspondence took place in between a handful of meetings in person. Tupper therefore also argued that he did not need to resign before the new parliament met at all but could instead remain in office test the confidence of the new House of Commons and resign only after losing the Address-in-Reply to the Speech from the Throne, which would also put some distance between the partisan appointments and resigning from office. On this question, Aberdeen and Tupper also both cited legitimate precedents which broke on both sides. In other words, precedents supports both of their arguments, so Aberdeen simply had to decide which precedents proved more persuasive and whether to throw his support behind the new convention emerging out of the United Kingdom and the Canadian precedents from 1873-1874 and 1878, or whether to defer to the older British precedents that Tupper cited. In the end, Aberdeen held firm, rejected Tupper’s advice, returned those minutes of council unsigned, and Tupper thereupon resigned. Tupper says unequivocally that he resigned because Aberdeen rejected his advice.

Third, that portion of Aberdeen’s and Tupper’s correspondence points to another fascinating question on whether the incumbent Prime Minister should resign before the new parliament meets when another party has won a majority or not. That question today would strike most of us as obtuse – of course the incumbent should resign under such circumstances. But that convention had only begun to emerge by the late 1870s to 1890s, and some British precedents even from those decades suggested that the incumbent should test the confidence of the new House of Commons and only resign after finding it wanting. Tupper and Laurier themselves in September 1896 then debated this question in the House of Commons. On Tupper’s urging, Laurier had secured Aberdeen’s permission to publish the correspondence between Tupper and Aberdeen, which is why I found it in the Sessional Papers of 1896. Tupper thought that it would vindicate him. Tupper and Laurier debated when the incumbent Prime Minister should resign and whether we could legitimately treat the results of the election themselves as a proxy for how the new House of Commons would vote. Ironically, Tupper’s argument that the incumbent should only resign after the new parliament meets and if he loses a vote of confidence corresponds more closely to the procedures for confirmation voting that most parliamentary democracies in Western Europe – including the devolved assemblies of Scotland and Wales – and even what the legislative assemblies of Northwest Territories and Nunavut use here. In any jurisdiction that uses confirmation voting, the incumbent Prime Minister remains in office up to the moment when the elected assembly votes in favour of appointing a new Prime Minister.

Fourth and finally, all these particular questions converge onto something fundamental and ultimately more interesting: the nature of constitutional conventions themselves. Building upon my thoughts on constitutional conventions that I posted here on Parliamentum in October and November 2021, I argue that we can only understand constitutional conventions as dialectical. In other words, we can only sort out the meaning of conventions through a process of debate, asking questions, examining any relevant precedents, parsing out the underlying reasons why a convention exists, and then deciding which interpretation is better. Even the famous Jennings Test on determining if a constitutional convention exists consists of three questions – which shows that we can only understand and parse out conventions through a dialectic and debate.

Anyone who says that we should not or cannot debate constitutional conventions, or dismisses ongoing scholarly disagreement over the limits of a particular convention, does not understand what they are. For example, historians and political scientists have with good reason continued to examine and present competing interpretations of the King-Byng Affairs almost one century after these events took place. Governors who dismiss Prime Ministers by rejecting their constitutional advice almost always cause great controversy, though Guichon’s dismissal of Clark in 2017 stands as an exception. And the debate between Tupper and Laurier proved so compelling precisely because they both made legitimate and reasonable arguments. On the one hand, Tupper argued that incumbent Prime Minister can remain in office after an election and test the confidence of the new House of Commons because elected MPs, not the electorate, ultimately indicate whom the Governor General should either keep or appoint as Prime Minister. The Governor General should therefore by extension not interference with that process and should allow outgoing Prime Ministers their appointments. On the other hand, Laurier argued for popular sovereignty and kept mentioning “the people” in his rebuttal. Laurier himself also admitted that Tupper brought up some legitimate examples and arguments but concluded that Aberdeen acted correctly and took responsibility for Tupper’s dismissal.

We would do well to understand and appreciate this lesser-known precedent from 1896 instead of focusing entirely on the King-Byng Affair of 1926, and I hope that my article helps achieve that goal.

Similar Posts: 

• Barry K. Wilson’s Biography on Sir Mackenzie Bowell (December 2021)
• Review of Gee & Webber’s “What Is a Political Constitution?” and the Nature of Constitutional Conventions (November 2021)
• Some Thoughts on Jonathan Gould’s “Codifying Constitutional Norms” (October 2021)
• Peter H. Russell Now Agrees with Me: What Happens When a Governor General Rejects a Prime Minister’s Constitutional Advice (August 2021)
• 1896: Tupper & Laurier Debate the Role of the Governor General and Popular vs Parliamentary Sovereignty (April 2020)
• Rejecting Constitutional Advice Equals Dismissal from Office: How Governor General Lord Aberdeen Forced Sir Charles Tupper’s Resignation in 1896 (January 2020)