Separation of Powers in Canada
Separationism and Parliamentarism
Canada's system of government is based on a parliamentary model quite distinct from the presidential system operating in the United States. One of our leading constitutional writers said that Canada’s retention of the British system of responsible government is “utterly inconsistent with any separation of the executive and legislative functions;” (Hogg, Constitutional Law of Canada, 1999 student ed., p. 321). While this is one important view, it has never been approved by the Supreme Court of Canada. Indeed, the Supreme Court of Canada has made passing reference to the doctrine of the separation of powers in several cases, including Fraser v. P.S.S.R.B.,  2 S.C.R. 455, 479 and Provincial Judges Reference,  3 S.C.R. 3 at para. 108. On occasion, the Court has used muscular language, as in Operation Dismantle v. The Queen,  1 S.C.R. 441, 491 when the Court referred to the doctrine as one of the “essential features of our constitution”. In R. v. Power, below, the separation of powers was actually harnessed by the court for use as an operative doctrine to reinforce the independence of Crown Attorney decisions as against judicial interference with prosecutorial decisions. The “rule of law” is a highly textured expression [...] conveying, for example, a sense of orderliness, of subjection to known legal rules and of executive accountability to legal authority.
Because Canadian parliamentary democracy increasingly trends towards power concentration in the executive branch – a tendency that has disturbed many observers – it may be time to reconsider the corrective role that could be played by the separation of powers theory in Canadian constitutional doctrine.
Constitutional Convention, Branch Fusion and the Democratic Deficit
First, we need to understand how it is that parliamentary government fuses the legislative and the executive branches. In a parliamentary system the executive springs from the legislature, is part of it and is responsible to it as a confidence chamber.
The Lieutenant‑Governor is part and parcel of the Legislature (B.N.A. Act, s. 71; the Legislature Act, R.S.Q. 1977, c. L‑1, s. 1). He appoints members of the Executive Council and Ministers (B.N.A. Act, s. 63; Executive Power Act, R.S.Q. 1977, c. E‑18, ss. 3 to 5) and these, according to constitutional principles of a customary nature referred to in the preamble of the B.N.A. Act as well as in some statutory provisions (Executive Power Act, R.S.Q. 1977, c. E‑18, ss. 3 to 5, 7 and 11(1); Legislature Act, R.S.Q. 1977, c. L‑1, s. 56(1)), must be or become members of the Legislature and are expected, individually and collectively, to enjoy the confidence of its elected branch. There is thus a considerable degree of integration between the Legislature and the Government; (Blaikie v. A.G. Quebec (No. 2) (1981), 123 D.L.R. (3d) 15 at 122 (S.C.C.)).
Although Blaikie dealt specifically with the provincial executive power, the Court’s description applies equally to the federal executive. The Court's observations in Blaikie are interesting because the Court focuses on the institutions of parliamentary government established by constitutional convention, particularly the institutions of responsible government. It is at the conventional level that integration between the executive and legislative branches occurs.
Constitutional convention enhances integration between the legislature and executive in two respects. First, the formal executive, the Governor General, is controlled by responsible ministers of the Crown, creatures unknown to the formal constitution. Second, the legislature's powers and priorities are in practice controlled by other executive instrumentalities unknown to the formal constitution – the PMO (office of the Prime Minister), PCO (Privy Council Office) and Cabinet. These institutions, particularly PMO and PCO, act as a clutch that meshes the gears of formal constitutional institutions into the full force of operating political power. Donald Savoie, Governing From the Centre: The Concentration of Power in Canadian Politics (1999) describes the real situation.
Central agencies stand at the apex of the machinery of government.... they have a licence to roam wherever they wish and to raise whatever issue they may choose; (p. 5) ... The prime minister alone thus has access to virtually every lever of power in the federal government, and when he put his mind to it he can get his way on almost any issue; (p. 87).
In other words, the central agencies, particularly PMO, PCO and, to a lesser extent, Cabinet, are the conventional executive. It is the conventional executive which in practice controls the legislature, and which allows the writers to speak about the integration between the executive and legislature.
The Queen of Canada is our head of state, and under our Constitution she is represented in most capacities within the federal sphere by the Governor General. The Governor General's executive powers are of course exercised in accordance with constitutional conventions. For example, after an election he asks the appropriate party leader to form a government. Once a government is in place, democratic principles dictate that the bulk of the Governor General's powers be exercised in accordance with the wishes of the leadership of that government, namely the Cabinet. So the true executive power lies in the Cabinet. And since the Cabinet controls the government, there is in practice a degree of overlap among the terms “government”, “Cabinet” and “executive”. In these reasons, I have used all of these terms, as one or another may be more appropriate in a given context. The government has the power to introduce legislation in Parliament.
In practice, the bulk of the new legislation is initiated by the government. By virtue of s. 54 of the Constitution Act, 1867, a money bill, including an amendment to a money bill, can only be introduced by means of the initiative of the government; (Reference Re Canada Assistance Plan,  2 S.C.R. 525, per Sopinka, J.).
It is at the conventional level, not the formal level or the text of the constitution, that the operation of Canada’s constitution exhibits a high degree of integration between the executive and legislative branches of government. At the conventional level, where the constitution actually functions, it is accurate to say that Canadian government is characterized by a high degree of control by the executive over the legislative branch, particularly as contrasted with presidential systems. It is perhaps this situation that was in the mind of the Supreme Court of Canada when it commented that “the Canadian Constitution does not insist on a strict separation of powers;” (Reference Re Secession of Quebec, S.C.C. Aug. 20, 1998, para 15).
It is also at the level of the operating conventional constitution that the writers observe worrisome signals of a deterioration in accountability and transparency of governmental processes – what is commonly referred to as the democratic deficit. The democratic deficit is enhanced by the extensive use of executive federalism to coordinate the actions of the thirteen governments in the federation. Executive federalism, the negotiation of issues of the day between senior officials and ministers, takes place behind closed doors; it is not visible. For this reason Canadian governance tends to be unduly secretive, and lacking in reliable structures of accountability.
Given the simultaneous growth of Canada’s democratic deficit and the concentration of political power in somewhat mysterious central agencies, it is not surprising that there should be a renewed interest in the separation of powers. It is at root a concept designed to guard against tyrannical concentrations of power and to protect political liberty.
The Formal Constitution and Separationism
The conventional machinery that integrates the executive and legislative branches in Canada obscures the very real structural separation of powers that the text of the constitution ordains. The Constitution Act, 1867 sets out separate and divided powers that, at least textually and formally, has close parallels to presidentialism. Blaikie drew attention to this. After describing the conventional machinery which integrates the legislative and executive branches, the Supreme Court went on to observe:
The Government of the province is not a body of the Legislature's own creation. It has a constitutional status and is not subordinate to the Legislature in the same sense as other provincial legislative agencies established by the Legislature (Blaikie v. A.G. Quebec (No. 2) (1981), 123 D.L.R. (3d) 15 at 122 (S.C.C.)).
It is useful to elaborate further on the Court’s observations about formal separation and conventional integration. The Constitution Act, 1867 establishes executive power by ss. 9‑16. These provisions vest the executive power in the Queen, and call for its exercise by the Governor General and Privy Council. The Constitution Act, 1867 establishes significant power in the executive branch, including, by s. 15, the command of the armed forces. The Constitution Act, 1867 identifies and organizes separate constitutional status as well for the legislature (sections 17‑52) and judiciary (sections 96‑101) and specifies their respective powers and limits.
This is why it is accurate to say that, at least textually and formally, the Constitution Act, 1867 has close parallels to presidentialism. Although the realities of conventional integration have made Canada’s formal separation of powers little noticed, it is worth remembering that within the text of the Constitution Act, 1867, powers are formally and structurally separated, as we find in presidential systems. This provides a textual basis for any court that in future decides to improvise a separation of powers doctrine specific to Canada’s parliamentary system.
It is also worth remembering that within the text of the Constitution Act, 1867 the three branches of government are connected functionally “as to give to each a constitutional control over the others.” Parliament is invested with constitutional power to enact all federal laws and to establish federal courts. Parliament is checked by the power of the executive to call the House of Commons into session (s. 38) and by the power of the judiciary to declare laws enacted unconstitutional. Parliament is also checked by power in the executive to reserve Bills passed by the Houses of Parliament and to disallow laws enacted (secs. 55-7). These veto-like powers, designed for British control of Canadian law-making, have long since fallen into disuse, but they still exist in the text and structure of the Constitution. The Judicial branch has constitutional power to try all cases, to interpret the laws in those cases and to declare any law or executive act unconstitutional. The judiciary is checked by power in the executive to appoint its members; by power in the legislature to enact amendments that overturn judicial decisions, including many constitutional decisions (Charter of Rights, s. 33); and also by the combined power of the executive and legislative branches to remove judges.
The Constitution of Ceylon is drawn from the same British colonial sources as the Constitution of Canada. It is interesting to observe in that Constitution’s structure a closely similar formal constitutional separation of powers. It was this structural separation, which, setting out executive, legislative and judicial powers in separate chapters that motivated the Privy Council to find “an intention to secure in the judiciary a freedom from political, legislative and executive control.” Because of that intention their Lordships overturned special legislation that would have intruded the legislative power too deeply into the judicial sphere ( Liyanage v. The Queen, [1967 1 A.C. 259). Perhaps this ruling is further evidence that the separation of powers doctrine is capable of more operational development in parliamentary systems, including Canada. Indeed, there is some dicta in this case which shapes Canada’s important doctrine of judicial independence out of separationist language.
Constitutional Dialogue and the Separation of Powers
Still, beyond protecting the independence of the judiciary, it is unusual to conceive of the separation of powers doctrine as an operative doctrine that controls any undue mixing of the three branches in parliamentary systems of government. However, it has become common in constitutional doctrine to conceive that the institutions of government have proper roles to play in Canadian democracy. In carrying out their functions, each branch should have proper regard and “mutual respect” for the role of the other branches.
respect by the courts for the legislature and executive role is as important as ensuring that the other branches respect each others' role and the role of the courts; (Vriend v. Alberta,  1 S.C.R. 493, para. 136).
In modern constitutional doctrine, the rise of this mutual respect “gives rise to a more dynamic interaction among the branches of government,” what is also called “a dialogue” between the institutions of government.
In reviewing legislative enactments and executive decisions to ensure constitutional validity, the courts speak to the legislative and executive branches. ... most of the legislation held not to pass constitutional muster has been followed by new legislation designed to accomplish similar objectives ... By doing this, the legislature responds to the courts; hence the dialogue among the branches; (para. 138).
An important value of this mutual respect, dynamic interaction and dialogue is that “each of the branches is made somewhat accountable to the other.”
This dialogue between and accountability of each of the branches have the effect of enhancing the democratic process; (para. 139).
In Mills, this perspective of dynamic interaction was applied to alter the normal suspicious posture reviewing courts sometimes adopt with respect to legislative acts alleged to be unconstitutional.
Courts do not hold a monopoly on the protection and promotion of rights and freedoms; Parliament also plays a role in this regard and is often able to act as a significant ally for vulnerable groups.... If constitutional democracy is meant to ensure that due regard is given to the voices of those vulnerable to being overlooked by the majority, then this court has an obligation to consider respectfully Parliament’s attempt to respond to such voices; (R. v. Mills,  3 S.C.R. 668).
Dialogue and dynamic interaction may be newly minted constitutional doctrine to explain to the citizenry why courts are sometimes obliged to overturn policies enacted into law by the representative branch. This is legitimacy theory, consciously meant to blunt attack on the constitutional review function repeatedly heard from the right and left. It is interesting to note that this theory is wrapped in separationism concepts. Is this really part of the sculpting of a separation of powers theory appropriate to parliamentary systems? To ask the question in other terms, is there anything in this separation of powers talk from the courts that can respond to the real problem that Canadian governance confronts today – democratic deficits being rung up by the fusing of political power in executive agencies?
In the Provincial Judges Reference,  3 S.C.R. 3 at para 139 the Supreme Court observed:
These different components of the institutional financial security of the courts inhere, in my view, in a fundamental principle of the Canadian Constitution, the separation of powers. As I discussed above, the institutional independence of the courts is inextricably bound up with the separation of powers, because in order to guarantee that the courts can protect the Constitution, they must be protected by a set of objective guarantees against intrusions by the executive and legislative branches of government.
The separation of powers requires, at the very least, that some functions must be exclusively reserved to particular bodies: see Cooper, supra, at para. 13. However, there is also another aspect of the separation of powers -- the notion that the principle requires that the different branches of government only interact, as much as possible, in particular ways. In other words, the relationships between the different branches of government should have a particular character. For example, there is a hierarchical relationship between the executive and the legislature, whereby the executive must execute and implement the policies which have been enacted by the legislature in statutory form: see Cooper, supra, at paras. 23 and 24. In a system of responsible government, once legislatures have made political decisions and embodied those decisions in law, it is the constitutional duty of the executive to implement those choices.
What is at issue here is the character of the relationships between the legislature and the executive on the one hand, and the judiciary on the other. These relationships should be depoliticized. When I say that those relationships are depoliticized, I do not mean to deny that they are political in the sense that court decisions (both constitutional and non-constitutional) often have political implications, and that the statutes which courts adjudicate upon emerge from the political process. What I mean instead is the legislature and executive cannot, and cannot appear to, exert political pressure on the judiciary, and conversely, that members of the judiciary should exercise reserve in speaking out publicly on issues of general public policy that are or have the potential to come before the courts, that are the subject of political debate, and which do not relate to the proper administration of justice.
To be sure, the depoliticization of the relationships between the legislature and the executive on the one hand, and the judiciary on the other, is largely governed by convention. And as I said in Cooper, supra, at para. 22, the conventions of the British Constitution do not have the force of law in Canada: Reference re Resolution to Amend the Constitution, supra. However, to my mind, the depoliticization of these relationships is so fundamental to the separation of powers, and hence to the Canadian Constitution, that the provisions of the Constitution, such as s. 11(d) of the Charter, must be interpreted in such a manner as to protect this principle.
This is a more developed idea of the role of the courts as resolver of disputes, interpreter of the law and defender of the Constitution that the Supreme Court explained earlier. In R. v. Beauregard,  2 S.C.R. 56, 73 the Court required that as a result of these functions, the courts be completely separate in authority and function from all other participants in the justice system.