At the time this article was written John Wilson was Professor of Political Science at the University of Waterloo. On September 25, 1995 he appeared as a witness before the Special Senate Committee on the Pearson Airport Agreements. This article is a revised and expanded version of his testimony to the Committee.

The autumn issue of the Review included an article by Professor Andrew Heard on the general nature of constitutional conventions and discussed the kinds of constraints that face a government in its last days. On September 25, Professor Heard appeared before the Special Senate Committee on the Pearson Airport Agreements. At that time the Committee also heard from Professor John Wilson and Professor James Mallory. In this article Professor Wilson provides a different perspective on the nature of constitutional conventions during election periods.

Many different issues are involved in the agreements between the Government of Canada and the Pearson Development Corporation which were authorized by the Prime Minister on October 7, 1993 and signed on the same day by representatives of the Government and of the Corporation. The Special Senate Committee has examined these in some detail but it has given hardly any consideration at all to what many observers regard as the most important element of the whole question – the status of the constitutional conventions surrounding the event and the importance which should be accorded them.¹

Indeed, this may be the only issue which has any significance in the debate which has been taking place since the new Liberal Government moved to cancel the agreements. Nearly every other aspect covers an area where there is legitimate room for different opinions about what is acceptable and what is not – in short, about different views based on different value systems – if only because such differences always exist in our society and in politics nobody is necessarily right. We are all, in our way, partisan.

But when it comes to questions of constitutional propriety we go beyond the particular merits or otherwise of the agreements and how a decision was reached. We are dealing instead with the practice of government itself – the rules and forms for decision-making in our society – and here there ought to be no room for argument. If we know what our constitutional practice is and should be then we should always insist on its observance.

In what follows I therefore want to address the character of the constitutional conventions which appear to me to apply in this instance, that is to say, what is regarded as appropriate behaviour by a Canadian government in the period following the dissolution of parliament and leading up to the conclusion of the ensuing general election.

Convention and the Constitution

It is important to start by reminding ourselves of the distinction between questions of law and questions of convention, custom, and our usual practice. In discussing the decision-making process with respect to the Pearson Airport Agreements I am concerned only with the latter.

I am therefore not going to refer to anything which is written in our various constitutional documents but only to certain unwritten customs and conventions which we have observed over the years and which come to us as a consequence of having inherited a parliamentary system of government from Great Britain. The distinction is of enormous importance, if only because so much of what is fundamentally significant to the successful operation of a parliamentary system of government is based on custom and convention.

If one were to believe what is written in the Constitution Act one would have to conclude that the Governor General has absolute power. There is no mention of the Prime Minister (except in the most minor way since the 1982 amendments) nor of the Cabinet. Perhaps the most fundamental principle of all – that a government which has lost the confidence of the House of Commons must resign or seek a dissolution – is not written down anywhere. The rules, if they may be given that title, which are involved in a decision by the Governor General to grant dissolution (or not to grant it) remain unclear to this day, despite the conviction of some scholars that they were settled in 1926. Many of the most important rules of parliamentary procedure are also in place by custom only. The list is endless, and yet all of these matters are essential elements of a parliamentary system of government. The usage in each case is only customary and conventional, but it is part of our constitutional practice.

We cannot make sense of how we are governed without understanding these unwritten rules.

But debates about these kinds of questions cannot be settled by reference to the law. Indeed, settling them at all involves an understanding of practice over time. I suppose that the principal distinction between law and convention is that there is an agency which determines what the law is – the courts – but there is no similar agency to deal with custom and convention. Sir Ivor Jennings has set out more clearly than most commentators what is involved in these matters.

The fact that there is no authoritative tribunal for the determination of conventions does, however, create difficulties. They grow out of practice and their existence is determined by precedents. Such precedents are not authoritative, like the precedents of a court of law. There are precedents which have created no conventions, and there are conventions based on precedents which have fallen into desuetude ....Every act is a precedent, but not every precedent creates a rule .... Precedents create a rule because they have been recognized as creating a rule. It is sometimes enough to show that a rule has received general acceptance. Persons of authority for nearly a century have asserted the right of the Prime Minister to choose his colleagues .... Persons of authority have never, so far as is known, asserted the duty of the monarch to grant a dissolution on request.2

It is important for what follows to recognize Jennings’ assertion that it is "sometimes enough to show that a rule has received general acceptance." In other words, conventions are not founded only on a series of prior events which may be regarded as precedents; they may also stem from the absence of particular forms of behaviour. To make the point in a different way it may be said that the fact that a request for dissolution has not been refused by a Governor General since 1926 should simply be taken as evidence of the understanding of successive Canadian prime ministers that inappropriate requests ought not to be made, rather than seen as proof that the Governor General no longer has the power to refuse.3

Simply because nothing has happened – or because nobody has said or done anything – there is no reason to assume that there are no rules in place about what is acceptable. The rule in question may say, in effect, that nothing should happen unless something ultimately to be regarded as unacceptable takes place.

I am aware, of course, that this interpretation of Jennings’ meaning is generally not accepted by most constitutional theorists. Instead, they insist that there must be at least one precedent to establish the existence of a convention,4 and they quote with approval another, more succinct passage from the master. Recognizing the existence of a convention, wrote Jennings, depends on three very simple criteria.

We have to ask ourselves three questions: first, what are the precedents, secondly, did the actors in the precedents believe that they were bound by a rule; and thirdly, is there a reason for the rule?5

Clearly, therefore, there must have been prior events to establish the beginning of the alleged convention, and without them nothing which we can reasonably call a constitutional convention can exist. As Eugene Forsey wrote:

A constitutional convention without a singe precedent to support it is a house without any foundation. . . . At least one precedent is essential. If there is no precedent there is no convention.6

But the fact is that these statements are not only not consistent with other observations by Jennings, they make very little rational sense.

If we are able to conceive of what I will call "negative" practices or customs – that is to say, the continued absence of specific kinds of behaviour – then there is every reason to believe that in certain very special circumstances there might well be constitutional conventions founded on exactly the same history. After all, a history of nothing having happened when one might have expected something to happen is as much a precedent as the more positive occurrences favoured by the authorities. In such cases denying convention status to such rules seems rather perverse. Our primary concern should be to discover whether there are any rules at all which are accepted as applying to the events with which we are concerned rather than spending our time quibbling about what they can be called. Peter Hogg’s view of the matter seems to strike the most appropriate balance:

Very little turns on the question whether a practice is a usage or a convention, because a convention is as unenforceable as a usage. The most that can be said is that there is a stronger moral obligation to follow a convention than a usage, and that departure from convention may be criticized more severely than departure from usage.7

But whatever label we may use it seems very clear that it is probably quite irrelevant that there appear to be no examples in practice to illustrate the play of the conventions governing the period following a dissolution of parliament. If the principal convention exists in the form in which I will state it we ought not to expect to see any examples at all.

The Caretaker Convention

What is known in some quarters as the "caretaker convention" is easily described although it is nowhere written down. It is a well-established principle of parliamentary government that once parliament has been dissolved and an election campaign is under way the government’s freedom of decision-making is firmly restricted and should be confined to dealing with only routine matters of administration – apart, of course, from any emergency situation which may arise.

More specifically, it is said that three areas of decision-making in particular should be avoided in this period – matters involving considerable controversy, matters which are not urgent (that is to say, matters which can wait for a later decision without causing irreparable damage), and matters where a decision would unreasonably bind the freedom of decision-making of a future government. It is occasionally said as well that matters involving the expenditure of very large sums of public money should also be avoided.

There are two reasons for this view, both of which would appear to be obvious. The first is that if parliament is not in session (and may not be for a somewhat extended period of time) the ordinary mechanisms for scrutinizing the government’s behaviour – question time, a debate on the adjournment, motions for supply, and debate generally – are not available.

Everyone knows that in a parliamentary system the power of the executive is potentially enormous. It is the existence of these mechanisms of "constructive obstruction"8 that takes the edge off that power and assures the people that the government is being kept responsible. But if they cannot be used there must be a compensating reduction in the usual extent of the power of the government. The caretaker convention addresses that issue.

The fundamental significance of this observation - when it comes to the importance of maintaining responsible government in our system - is hardly removed by the distinction which is made between a government which has lost the confidence of the House of Commons and one which has merely dissolved parliament in the ordinary course of approaching the legal end of its term in office.9 What on earth can be the relevance – constitutional or otherwise – of the fact that the government retains the support of a substantial majority of the members of the House of Commons if that House is never going to be called on to pass judgment?10

The second reason is, of course, that an election campaign always entails the possibility of the government’s defeat, and therefore the possibility that its leaders will not be able to take responsibility for the consequences of their decisions. It follows that only the most routine administrative decisions ought to be made in this period - decisions, in short, which any government might make - again excepting an emergency.

Some people object to this argument on the ground that during an election campaign a government is surely very much in the public view, and hardly able to do as it pleases. Professor Heard, for example, puts it this way.

The dissolution of the legislature is conducted to hold an election. The government is then being directly held accountable by the electorate. There is no greater accountability than what happens on election day.11

One may wonder, however, whether the daily media conference - or the mindless scrum which occurs at each bus stop on the leader’s tour - is any substitute for the cut and thrust of the House of Commons, led by knowledgeable and skilled practitioners of the art of opposition, as a means of reigning in the excesses of the government.12 The prospect seems every bit as absurd as the suggestion which surfaces from time to time that the press is just as able as the real opposition to maintain a proper watch on the government.

There really is nothing to match the effectiveness of the House of Commons itself as an agency for monitoring and limiting the power of the executive. The purpose, after all, of "constructive obstruction" is to persuade the government to back down, or at least to take a second look at its proposals, and neither media scrums nor the electorate nor leading editorials have ever had much success with that.

The British Practice

Although these observations seem intuitively to make some kind of sense, on the face of things they are difficult to substantiate even in a general way when we examine the practice of the leading parliamentary democracies. In the United Kingdom, for example, there does not appear to be any evidence of the caretaker convention, at least in the sense that the customs associated with it are not part of the ordinary language of British politics.13

Jennings makes it quite clear that it is not common practice to appoint a formal caretaker government with very restricted powers during the period of an election, although he cites the exception made by Churchill in 1945, who chose this way of breaking up the wartime coalition – and getting rid of the Liberal and Labour ministers in the process – in preparation for the party fight which was bound to occur in the forthcoming general election.14

But he does not explicitly address the more general question of the behaviour of governments during the caretaker period, saying only that even a government which has been defeated in the House of Commons will stay in office until a new government materializes, so that what he calls the "King’s service" may be carried on. I think, however, that a moment’s reflection on the course of British politics since, let us say, the end of the First World War will show that there has never been an occasion when a British government made anything other than routine decisions in the period following a dissolution of parliament.

Of course, the matter may well be addressed in Great Britain by the principle of parliamentary sovereignty – the rule that no parliament may bind a future parliament. It is easy to see how something resembling the caretaker convention could be derived from that more general understanding. If parliament cannot be permanently bound by the actions of a prior parliament then surely it cannot be bound by the executive acting without parliamentary consent.

The Australian Practice

In sharp contrast to the United Kingdom the Australian practice has for some years been to limit the power of the government in the caretaker period – but not by anything more than a set of customs to which all parties have willingly subscribed. Their existence is justified in almost exactly the same terms as set out above – the absence of a popular chamber to which the government can be held responsible and the possibility of a change of government as a result of the election. The Australian conventions are described in the following way:

The basic caretaker conventions require a government to avoid implementing major policy initiatives, making appointments of significance or entering major contracts or undertakings during the caretaker period and to avoid involving departmental officers in election activities.

The basic conventions are directed to the taking of decisions, and not to their announcement. Accordingly, the conventions are not infringed where decisions taken before the caretaker period are announced during the caretaker period. However, it is desirable, if the decisions concern significant initiatives, that they be announced in advance of the caretaker period in order to avoid controversy.15

These customary and conventional practices at the federal level in Australia go back to the time of Sir Robert Menzies in the 1950s – who was responsible for formally and explicitly initiating them – but in fact there is some evidence that they may have been accepted as existing informally for some time before that.16

They are, moreover, to be distinguished from the more restrictive rules which generally apply to explicit cases of caretaker governments such as that of Malcolm Fraser following the Governor General’s dismissal of Gough Whitlam in 1975. Such governments are ordinarily expected to do little more than attend to one or two specific tasks before seeking a dissolution and a general election.

The general Australian practice goes some considerable distance beyond these kinds of cases by laying down an appropriate pattern of behaviour for any government which has dissolved parliament and entered the election period.

The Canadian Practice

The Canadian practice would appear to fall somewhere in between what happens in the United Kingdom and what happens in Australia. As in Great Britain there is no formal reference to anything which looks like the caretaker convention in the literature. But in the summer of 1953 there was an exchange of letters in The Ottawa Journal between Arthur Beauchesne, former Clerk of the House of Commons, and Professor J.R. Mallory, where the two competing views were clearly set out. Although it is rather overstated Beauchesne outlined the position which I would regard as broadly correct in these terms.

It has always been the practice in British democracies that the cabinet, in the period between dissolution and general elections, only acts in matters that are absolutely necessary for the ordinary conduct of affairs.

Our ministers are not now members of parliament for the good reason that there is no parliament. They are private citizens ...In the United Kingdom the administration which happens to be in office during that period is called the "caretaker government" and its actions are limited to mere departmental routine....

The doctrine is well-known and has been respected by John A. Macdonald, Laurier, Borden, Mr. Meighen and Mackenzie King.17

Professor Mallory’s response dismissed this view more or less out of hand, but in a way which seems to the contemporary reader to have missed Beauchesne’s point entirely. He wrote:

I am unaware of any example of this practice which, in any event, could only be destructive to effective constitutional government. It appears to be Dr. Beauchesne’s view that ministers of the Crown are, after the dissolution of parliament, only private citizens and that they thus have no right to carry out their responsibilities of office, save for "mere departmental routine."

But ministers are not just private citizens. They are the Crown’s ministers, responsible to the Crown for the conduct of government. For them to lay down their responsibilities as the Crown’s confidential advisors for the period of two months which must elapse between the dissolution of Parliament and a general election, would be for them flagrantly to dishonor their oaths as privy councillors and ministers of the Crown.

It was never contemplated that we should have in effect no government at all for a period of over two months just because of the calling of a general election.18

This is, to say the least, a peculiar argument in the circumstances. The issue is not whether the government continues to govern during the caretaker period – that is to say, continues to make decisions – but whether those decision are confined to the merely routine.

In other words, nothing extraordinary is supposed to happen. Given these expectations following from the practice Beauchesne described one may wonder what kind of event Professor Mallory might have expected to see as a demonstration of the validity of Beauchesne’s argument. Indeed, the very fact that he cannot point to anything at all by way of example suggests that Beauchesne was right. For the fact of the matter is that a quick examination of the behaviour of Canadian governments in the caretaker period since, again, the end of the First World War shows no case at all of decision-making out of the ordinary. Surely it is worth asking whether this is entirely accidental?

Professor Mallory has expressed his current view with much greater precision, although by implication traces of the earlier view seem to persist.

When a government has been defeated at the polls or in the House of Commons, it becomes an obligation of all party leaders to assist in the formation of a new government. Until a new government can be formed, it is the duty of the old one to remain in office. While in office it still has the duty and the authority to govern, though a government which has lost the confidence of the people or of the House of Commons can only make routine decisions until a government which has the support of the House can be formed.19

This is, of course, what I would call the "conventional wisdom" on the question. After the election is completed the course is clear on the basis of the result – either the existing government has won and therefore carries on or it behaves with very great circumspection until the new government can take office. Alternatively, the existing government is only limited when it has lost the confidence of the House of Commons. But we live in a democratic age, and the point made earlier is worth repeating. It is surely a very short road between a government which has lost its authority to govern because it has lost an election or has been defeated in the House of Commons and a government which, in the nature of the case, cannot answer to parliament because parliament does not exist.

In fact, however, there does now seem to be some indication that the caretaker convention is at work in Canada. The evidence given to the Special Senate Committee on the Pearson Airport Agreements by the present Clerk of the Privy Council, Madam Jocelyne Bourgon, was abundantly clear on the question in several different places. She had this to say, for example, about the process of decision-making in the case of the Pearson Airport Agreements, from her perspective on the situation as Deputy Minister of Transport at the time, as the end of the negotiating period was approaching:

But there was a need in my judgment at that point in time to satisfy ourselves that it was indeed the wish of the government to proceed. And that is not unusual by the way. I don’t want to leave the impression there is anything unusual about that. There is a general rule of conduct to act with caution as soon as Parliament is dissolved. The purpose of seeking guidance is to make sure that those who have the power of making these decisions are the ones making these decisions as opposed to those who do not have the authority. So there was a need to ascertain that it was the wish of the minister to proceed, and that was clarified, and his will was very clear. And later on the same thing was sought from the Prime Minister.20

And again, describing the different phases of the decision-making process as the summer of 1993 wound down to the fall:

From the end of August till dissolution of Parliament, September 8, we were still at the stage of converting this general agreement into all the component pieces and agreements required to give effect.

After Parliament was dissolved, what happens in terms of conduct for officials is that there is this general rule. It’s not a law. There is a general rule that from that point on, you must act with caution. So the question comes, who is going to make the judgment as to whether or not you’re cautious. Well, that’s not a judgment to be made for officials. You go to your minister or the first minister, the Prime Minister, depending on the circumstances.21

In these observations Madam Bourgon seems only to be describing the appropriate caution which should be exercised by public servants in a period when partisan feeling is likely to be more in the air generally in Ottawa. But when pressed she came closer to the concerns I am addressing. She described "two events" which caused her to pause.

There was a statement by the Leader of the Opposition requesting publicly the Prime Minister to put everything - I think he used the expression - in the freezer. The day after, ...I believe there was also a statement by the Leader of the Opposition to the effect that he would wish, should he form the government, to review the approach.

These two events raised in my mind the need to receive guidance on the appropriateness of proceeding further, which is closure on the 7th, but this time from the Prime Minister. Because the Prime Minister is responsible for the behaviour of government during a period of election. And the call having been made at the level of the Leader of the Opposition, in my mind, it was not sufficient to simply ask guidance from the minister at that point in time....

Now, it’s not for the Deputy Minister of Transport to get on the phone and call the Prime Minister and say, "I would wish to get guidance." You refer the matter to the clerk, whose job it is to make sure that we respect tradition and values and due process and so on. And when I raised my view with the clerk, the clerk was also of the view that it was appropriate to seek guidance from the Prime Minister. He did and gave me my instructions.22

It was the sudden evidence that the issue was not a matter of routine decision-making – brought on by the Leader of the Opposition’s forceful intervention – that caused Madam Bourgon to pause. But there is more to it than that, as the following elaboration in response to yet another question shows.

I think controversy is not the only factor, ... I think the general rule of conduct to act with caution during an election means that you would consider factors such as: Is it a transcription that is going to bind future governments? What is the - are there alternatives? Are there urgencies in the matter? Is there an obligation to act? Is there controversy? Controversy is certainly a factor, but I would simply want to say that it is not the only one you would consider, in which cases you would seek guidance. There would be more than one factor which would be considered.23

With that observation, it seems to me, Madam Bourgon effectively stated the existence of the caretaker convention in Canada, or at the very least the existence of a very firm practice amongst senior public servants in Ottawa. I have deliberately provided extended quotations from her testimony because she has given us – for the first time – a fascinating and important insight into the real character of decision-making in these special circumstances. What she has said will surely become an important part of the literature for students of our system, even if there is some difficulty in trying to find examples in practice of the operation of the caretaker rules.

In fact, that difficulty may not be as great as has been suggested. One of the areas in which there has been said to be a need for caution is the making of appointments after parliament has been dissolved. Here the record over the years since the end of the First World War is instructive. Mackenzie King was a major offender is this respect - in 1945 he recommended a total of 18 senatorial appointments during the caretaker period (he had recommended 14 in 1940) – but after that the practice effectively died out, and no such recommendation has been made at all since 1962.24 Although there is not the same complete absence of activity during the caretaker period in recent years with lesser appointments – to diplomatic posts, the senior public service, the bench, and various agencies boards and commissions – the record suggests that what was done was for the most part routine.

With contracts, on the other hand, the record shows nothing except minor and routine confirmations of obligations already accepted by the government, again going back to the end of the First World War. Certainly there is nothing even remotely resembling the magnitude of the Pearson Airport Agreements.25 It is difficult to avoid the conclusion that this is due to a general reluctance on the part of successive Canadian governments to consider such decisions during the caretaker period. Some confirmation of this possibility may be found in Joe Clark’s refusal in 1979 to go ahead with a decision to purchase over $2 billion worth of new fighter aircraft. "It is my judgment," he said on the day after his government was defeated, "that a government that has lost the confidence of Parliament does not have the authority to make that decision."26

These limited observations, together with Madam Bourgon’s testimony, make it clear that there must be an understood system of rules at least within the Privy Council Office which address the need for caution during the caretaker period. But this implies the existence of mechanisms devoted to that purpose. Madam Bourgon has said that it is the job of the Clerk "to make sure that we respect tradition and values and due process." Others have described the Clerk as the custodian of parliamentary conventions, customs, and practices. Not only is this obviously the case, but it is clear as well that the Machinery of Government Group within the Privy Council Office is specifically charged with ongoing responsibility for monitoring the conduct of government from this perspective.²⁷ In fact, it is not too much to say that the Machinery of Government Group, under the direction of the Clerk, has a responsibility to, in effect, teach new prime ministers and new cabinet ministers what the rules of proper government are. Where, one may ask, are these people going to get such lessons if not from those whose job it is to know the answers?²⁸

But in the end it is always up to the Prime Minister to decide what will be done. Conventions, customs, and practices, simply because they are not in any way linked to legality, are always political in character. Their application depends upon judgments made in particular circumstances, and even the most thoroughgoing convention - such as, let us say, the federal representative character of the Canadian cabinet – can be ignored on occasion if that becomes absolutely necessary.29 The best that may be hoped for at such times is that the advice of a Clerk with many years of experience will be enough to tilt the balance towards appropriate and responsible behaviour. But however much the Clerk of the Privy Council may be the custodian of all proper governmental practice he or she cannot compel the Prime Minister to walk away from a decision the Prime Minister has decided to make.

Beyond Convention - A Different Political Dimension

That reflection naturally leads back to the final decision to go ahead with the Pearson Airport Agreements. We have no way of knowing what advice the then Clerk of the Privy Council, Glen Shortliffe, gave to Prime Minister Campbell on that occasion. It is clear, however, that while he is willing to recognize a general case for restricted decision-making in the caretaker period Mr. Shortliffe regarded the Pearson Airport Agreements as a "done deal" prior to the calling of the election, and therefore the final authorization given on October 7th was merely pro forma and effectively of no consequence.30 That is to say, the issue of appropriate decision-making simply did not arise.

But there is an entirely different kind of consideration that could be brought to bear on the question. The decision to go ahead was clearly made at a time when the Prime Minister and those around her must have known that her government was likely to be defeated. The Gallup Poll published on September 22nd showed the Liberals at 37 percent and the Conservatives at 30 percent (down from 36 percent in August). A month later the Gallup Poll had the Conservatives at 16 percent - a massive decline. One does not need to see internal party polls to know that on October 7th, roughly midway between these two public polls, the government must already have been close to only 20 percent. That is to say, a major defeat was effectively unavoidable.31

Whatever was being said for public consumption at the time it is impossible to believe that the Prime Minister was not aware of this catastrophic political situation. The hard facts of the case must therefore be that she chose to authorize the signing of the Pearson Airport Agreements at a time when she knew that she would be unable to take responsibility for that decision. This looks very close to the work of a government which has already lost the moral authority to govern.

How, then, are we to assess the dimensions of the decision that was made? Clearly the issue had by October 7th met all the criteria for caution described above. It was clearly a matter of considerable controversy - the then Leader of the Opposition had vowed to cancel the agreement if his party won the election - which ought to have been enough on the basis of the examples I have described to stop the process. An enormous amount of public money was involved, there was no demonstrated urgency to settle the question, and the arrangement locked the government of Canada into a 57-year leasing agreement with no cancellation clause, all of which should equally have made it an inappropriate candidate for decision-making in the caretaker period. No doubt the Pearson Airport Agreements can be overturned (and, indeed, are in process of being overturned) but that reversal comes, apparently, at a very high price.

A responsible Prime Minister ought to have refused to have anything to do with such a proposal even if she thought she was going to be re-elected. Knowing that she was not should have been enough to put the lid on it. To say that her decision was a constitutionally inappropriate exercise of power is, in my view, to put it mildly. We must hope that the wise men and women who judge these things will say in future that the Conservative defeat in the 1993 federal election can be taken as a precedent demonstrating that our people will not accept such a gross violation of the rules for appropriate government behaviour - and that, indeed, the caretaker convention is alive and well in Canada.

Notes

1. Aspects of this question were touched on in passing during Professor Andrew Heard’s testimony regarding Bill C-22 before the Standing Senate Committee on Legal and Constitutional Affairs at the end of 1994, but not in sufficient detail to explore the whole range of what may be at stake. See Proceedings of the Standing Senate Committee on Legal and Constitutional Affairs, no. 18, December 8, 1994.

2. Cabinet Government, 3rd edition (Cambridge, 1961), 5-7.

3. See my "A Quite Constitutional Prayer: Reflections on the Character of the Royal Power of Dissolution in Canada," paper presented to a Department of Political Science colloquium, University of Waterloo, February, 1983.

4. See, for example, Andrew Heard, Canadian Constitutional Conventions: The Marriage of Law and Politics, (Toronto, 1991), 13.

5. The Law and the Constitution, 5th edition (London, 1959), 81.

6. "The Courts and the Conventions of the Constitution," UNB Law Journal, vol. 33 (1984), 34. These passages from Forsey and Jennings are the ones usually employed to demonstrate the absolute need for a precedent to establish the existence of a convention.

7. Constitutional Law of Canada, 2nd edition (Toronto, 1985), 16.

8. I owe this wonderful phrase to the late Eugene Forsey.

9. Virtually all of the authorities make this distinction. See, for example, Andrew Heard, "Constitutional Conventions and Election Campaigns," Canadian Parliamentary Review, vol. 18, no. 3 (Autumn, 1995), 10; and J. R. Mallory, The Structure of Canadian Government (Toronto, 1971), 75-76.

10. No doubt constitutional conventions cannot be held to depend on such changeable things as public opinion polls, but in this context one would hope that the likelihood of the government being returned to office might have a bearing on our judgment.

11. Proceedings of the Standing Senate Committee on Legal and Constitutional Affairs, no. 18, December 8, 1994, 16.

12. For an elaboration of these points see my "On the Virtue of Being a Nuisance," Past and Present (April, 1985), 2-7 (a publication of the Faculty of Arts of the University of Waterloo).

13. A careful search of the principal texts produces no reference to the practice at all. On the other hand, chapter 15 of Jennings’ Cabinet Government is so laced with a view of the relationship between the Government and Parliament which depends on the need for the Government to respect the House of Commons and, in particular, the Opposition, that it is difficult to believe he does not accept some version of the caretaker convention. Elsewhere he says "Democratic government has its Marquess of Queensberry rules, and public opinion is the referee." Cabinet Government, 3rd edition (Cambridge, 1961), 16. Perhaps it is simply that British election campaigns are considerably shorter than in Canada - the usual length is about 30 days - and ministers are generally not in London but in their home constituencies. In short, there is nothing going on at the centre.

14. Cabinet Government, 3rd edition (Cambridge, 1961), 86 (note) and 531. Churchill simply tendered his resignation as Prime Minister, after only perfunctory consultation with the Liberal and Labour ministers, and was then appointed anew by the King to continue as Prime Minister.

15. "Caretaker Conventions and Other Pre-Election Practices," Annual Report of the Department of the Prime Minister and Cabinet, 1986-1987, 39.

16. Letter to the author from M. S. Keating, Secretary to the Cabinet, Government of Australia, November 21, 1995.

17. The Ottawa Journal, July 1, 1953.

18. The Ottawa Journal, July 7, 1953.

19. The Structure of Canadian Government (Toronto, 1971), 75.

20. Proceedings of the Special Senate Committee on the Pearson Airport Agreements, no. 19, September 14, 1995, 57.

21. Ibid, 59.

22. Ibid., 59-60.

23. Ibid. 100. It is worth noting that the report Robert Nixon prepared for the Prime Minister stated the caretaker convention in almost exactly similar terms. See Pearson Airport Review, November 29, 1993 (Office of the Prime Minister), 8.

24. John Diefenbaker recommended the appointment of two senators (J. Campbell Haig and Harry A. Willis) on June 15, 1962; election day was June 18. It is sometimes wrongly claimed that John Turner violated what now appears to be accepted practice on this score in 1984. In fact, the three individuals in question (Eymard Corbin, Tom Lefebvre, and Charles Turner) - whom the Canadian Parliamentary Guide shows as appointed to the Senate on the day of dissolution - were actually appointed the day before.

25. Because, in the nature of the case, an example which might meet the precedent test would not have seen the light of day, I have asked a number of individuals much closer than I am to the history of decision-making by the federal government since 1945 for their recollections on this point. I have agreed to leave their names off the record but I can report that none of them can think of a single case matching the significance of the Pearson Airport Agreements.

26. "The jet fighters stall near target," Financial Post, December 22, 1979. This is not, of course, a strict application of the caretaker convention as I have described it, since Clark’s government had been defeated in the House of Commons. But it is surely interesting evidence of the awareness of at least one Prime Minister that there are special rules for these occasions. Although it also deals only with the question of appropriate decision-making for a defeated government an article from the same period by Eugene Forsey is instructive. See "Defeated government should exercise restraint in its actions," Ottawa Citizen, January 10, 1980.

27. I have been much helped in understanding this process of decision-making by a number of people formerly associated with the Privy Council Office.

28. This aspect of the work done by the Machinery of Government Group within the Privy Council Office does not seem to have received the recognition it is due. It should perhaps be added to the comprehensive descriptions found in Kenneth Kernaghan and David Siegel, Public Administration in Canada, 3rd edition (Toronto, 1995), 200; and Robert J. Jackson and Doreen Jackson, Politics in Canada: Culture, Institutions, Behaviour and Public Policy, 3rd edition (Scarborough, Ontario, 1994), 305-308.

29. Absolutely fundamental conventions, such as the rule that a government which has lost the confidence of the House of Commons must resign, obviously cannot be treated in this way, and a government which sought to circumvent these kinds of rules could expect to hear very quickly from the Governor General.

30. See his testimony on these points in Proceedings of the Special Senate Committee on the Pearson Airport Agreements, no. 24, September 25, 1995, 60-99. One might ask why, if this was the case, the Prime Minister’s agreement was needed at all, but that question seems to have eluded the Special Senate Committee.

31. No political party has ever won a Canadian election - nor even minority government status - with that level of support.