Canada is the only Westminster jurisdiction to have enacted a near-absolute immunity for Cabinet confidences. Parliament entrenched executive supremacy over the disclosure of Cabinet confidences at the behest of Prime Minister Pierre Elliott Trudeau’s Liberal government, which did not trust the courts to properly protect its political secrets. To ensure the highest level of protection to Cabinet confidences, Parliament adopted a broad and robust statutory scheme, the scope of which went beyond the protection afforded to this type of information under constitutional conventions and the common law. In doing so, Parliament has effectively removed from the courts the power to inspect and order the production of Cabinet confidences, making it extremely difficult to challenge Cabinet immunity claims. Indeed, litigants and judges do not have access to the information required to determine whether such claims are made reasonably and in good faith by the executive branch.
The question at the heart of this article is whether this special statutory regime is constitutional. Under the common law, which applies at the provincial level in Canada, the courts’ power to inspect and order the production of Cabinet confidences is now considered a constitutional imperative. In Carey v. Ontario, the Supreme Court of Canada (SCC) concluded that it would be “contrary to the constitutional relationship that ought to prevail between the executive and the courts in this country” to deprive the judiciary of this power. Despite this statement, courts have held that Parliament could, pursuant to the doctrine of parliamentary sovereignty, provide a near-absolute statutory immunity to the government over Cabinet confidences. Consequently, courts have ruled that this kind of immunity does not violate the Constitution. The courts’ position on this issue appears conceptually inconsistent: it is either constitutional to deprive the judiciary of the power to inspect and order the production of Cabinet confidences, or it is not, as the common law and statute law must both comply with the same constitutional rules.
In its seminal 2002 decision on Cabinet immunity under statute law, Babcock v. Canada (A.G.), the SCC did not address this inconsistency. In nine paragraphs, at the very end of its reasons, the SCC held that executive supremacy over Cabinet immunity did not violate the unwritten principles of the rule of law, the separation of powers, and judicial independence. Based on historical considerations, and the fact that a weak form of judicial review is possible, the SCC further stated that the statutory regime did not fundamentally alter the relationship between the executive and the judicial branches of the state. It thus confirmed that the near-absolute nature of Cabinet immunity at the federal level is not only legal, but also legitimate. The SCC reached its conclusion by embracing a very thin conception of the rule of law and a very narrow interpretation of the separation of powers. In doing so, the SCC disregarded the wisdom of Carey and the common law legacy of Cabinet immunity.
The objective of this article is to show that the near-absolute statutory immunity granted to Cabinet confidences violates the rule of law and the provisions of the Constitution. I will focus on section 39 of the Canada Evidence Act (CEA), which deprives judges of the power to inspect and order the production of Cabinet confidences in litigation. The article is divided into two parts. In Part 1, I will demonstrate that the SCC’s very thin conception of the rule of law is of limited use as a normative framework to assess the legality of section 39. To that end, I will turn to the thicker theory of law as justification. In Part 2, I will demonstrate that section 39 breaches the requirements of procedural fairness and intrudes upon the core, or inherent, jurisdiction and powers of provincial superior courts. I will argue that to comply with the rule of law and the provisions of the Constitution, Cabinet immunity claims should be decided by independent and impartial judges, following a fair decision-making process, in the course of which judges can inspect the documents and assess the public interest.
To determine whether section 39 of the CEA is consistent with the rule of law, it is first necessary to define what is meant by “rule of law”. While the rule of law is an important political ideal, with deep historical roots going back to the Greeks and the Romans, there is no broad consensus on its precise meaning. All conceptions of the rule of law seek, to various degrees, to limit the powers of the state, to replace the often-arbitrary rule of men by a more rational rule of law. Beyond this basic idea, legal theorists tend to distinguish thin from thick conceptions of the rule of law. Thin conceptions are associated with positivist approaches to the rule of law while thick conceptions are associated with natural law approaches. In an influential article on the rule of law, Paul Craig describes the main differences between thin (or formal) and thick (or substantive) conceptions:
Formal conceptions of the rule of law address the manner in which the law was promulgated (was it by a properly authorised person, in a properly authorised manner, etc.); the clarity of the ensuing norm (was it sufficiently clear to guide an individual’s conduct so as to enable a person to plan his or her life, etc.); and the temporal dimension of the enacted norm, (was it prospective or retrospective, etc.). Formal conceptions of the rule of law do not however seek to pass judgment upon the actual content of the law itself. They are not concerned with whether the law was in that sense a good law or a bad law, provided that the formal precepts of the rule of law were themselves met. Those who espouse substantive conceptions of the rule of law seek to go beyond this. They accept that the rule of law has the formal attributes mentioned above, but they wish to take the doctrine further. Certain substantive rights are said to be based on, or derived from, the rule of law. The concept is used as the foundation for these rights, which are then used to distinguish between “good” laws, which comply with such rights, and “bad” laws which do not.
This part of the article will define the scope of the rule of law conception that will then be used in Part 2 to determine whether Parliament can confer on the government a near-absolute immunity over the production of Cabinet confidences through section 39 of the CEA. Part 1 is divided into two subparts. In the first subpart, I will argue that the SCC has thus far adopted a very thin conception of the rule of law as an unwritten constitutional principle, one which is of limited use to assess the legality of legislation. In the second subpart, I will argue that the thicker theory of law as a culture of justification, which is implicit in the Canadian legal order, provides a better normative framework to this end because it imposes meaningful constraints on the state which, in turn, illuminate the flaws afflicting section 39.
What conception of the rule of law has the SCC adopted to date? In the Canadian legal order, the rule of law is an “unwritten constitutional principle”, the existence of which stems implicitly from the preamble to the Constitution Act, 1867 and explicitly from the preamble to the Constitution Act, 1982. The SCC has characterized the rule of law as a “fundamental postulate of our constitutional structure” lying “at the root of our system of government.” This depiction calls for two comments. First, despite its importance, the rule of law is only one of several unwritten constitutional principles, distinct from the separation of powers, judicial independence, federalism, democracy, and respect for minorities. The fact that the SCC has distinguished the rule of law from these principles suggests that it has a narrow conception of the rule of law. Second, Canada has a written Constitution, setting out its constitutional structure and outlining fundamental rights and freedoms. For reasons of certainty, predictability, and legitimacy, the SCC considers that judicial review of legislative action should be grounded in the text, not the unwritten principles, of the Constitution. As such, it gives more importance to the Big-C Constitution over the small-c constitution.
The SCC’s conception of the rule of law contains at least three elements. The first requires “the creation and maintenance of an actual order of positive laws which preserves and embodies the more general principle of normative order.” The second recognizes that “the law is supreme over officials of the government as well as private individuals, and thereby preclusive of the influence of arbitrary power.” The third demands that “the relationship between the state and the individual must be regulated by law.” In short, the rule of law presupposes: the existence of a legal order; within that legal order, legal rules must apply equally to the state and legal subjects; and the actions of the state must be authorized by legal rules. The SCC has recognized that there may be other components to the rule of law, but has not yet ruled on what else may be involved. While the three elements identified thus far are prerequisites to the existence of the rule of law, they do not impose any meaningful constraint on the state. They incorporate the idea of legal limits on state power, without attempting to define the boundaries imposed by these limits. These elements thus set out a “skeletal version of the rule of law,” which corresponds to a thin positivist conception of the rule of law, known as “rule by law”. Under this conception, a legal rule is valid simply because it has been enacted by the proper authority under the proper procedure.
The SCC’s conception of the rule of law, as an unwritten constitutional principle, does not yet incorporate the principles of formal legality. Like “rule by law”, “formal legality” is a thin conception of the rule of law. However, it goes beyond “rule by law” by imposing certain constraints on the state. With the aim of ensuring that the law can guide the conduct of legal subjects efficiently, formal legality requires that legal rules be general, clear, public, prospective, coherent, possible to comply with, and relatively stable over time. In addition, it requires that state action be congruent with legal rules. While these principles are widely accepted among legal theorists, the SCC has so far refused to include them in its conception of the rule of law. It will only recognize and enforce the principles of formal legality where they are explicitly entrenched in the text of the Constitution. Hence, in British Columbia v. Imperial Tobacco Canada Ltd., the SCC upheld the validity of provincial legislation that targeted tobacco companies and held them retroactively liable for the healthcare expenses triggered by tobacco consumption, although the legislation was neither general nor prospective. According to the SCC, nothing in the text of the Constitution explicitly prohibited such measures in civil matters.
To what extent can the unwritten rule of law principle, however understood, be used, as a normative framework, to assess the legality of legislation? The answer is threefold. First, the SCC has so far suggested that the rule of law cannot be used, on its own, to invalidate legislation. This is in line with the position defended by respected constitutional scholars. Peter Hogg and Warren Newman argue that the rule of law can be used to support the interpretation of constitutional provisions, but it should not be relied upon as the sole basis to strike down legislation. The rule of law must be reconciled with the principles of constitutionalism and democracy. If a legal rule is consistent with the text of the Constitution (constitutionalism), and has been adopted under the proper procedure by the legislative branch (democracy), judges should enforce it, whether or not they agree with the underlying policy. This is coherent with the very thin conception of the rule of law adopted by the SCC. Under this view, the remedy against legislation that is perceived as unjust or unfair does not lie in the unwritten principles, but in the text of the Constitution and the ballot box.
This approach has prevailed in several important cases, including cases dealing with the constitutionality of Cabinet immunity. In Singh v. Canada (A.G.), the Federal Court of Appeal dismissed the claim that section 39 of the CEA violates the rule of law, finding it to be consistent with the SCC’s three elements: it establishes a legal order for the protection of Cabinet confidences; it applies equally to the government and legal subjects; and it allows the government to protect Cabinet confidences in litigation. Whether section 39 is good or bad policy is irrelevant, the Federal Court of Appeal said, for “the rule of law does not preclude a special law with a special result dealing with a special class of documents which, for long standing reasons based on constitutional principles such as responsible government, have been treated differently.” The SCC later quoted this passage with approval in Babcock. While section 39 may be a “draconian” rule, the SCC added, it is one that Parliament has the power to adopt. The SCC has consistently declined to invalidate legislation on the basis that it violates the unwritten rule of law principle, thus suggesting that the rule of law does not have normative force as an independent base to challenge the validity of legislation.
Second, while the rule of law has not been used, as a standalone principle, to invalidate legislation, it has been relied upon to bolster the interpretation of provisions of the Constitution which, in turn, have been used to invalidate legislation or justify a specific remedy. This form of argument has been considered more legitimate because the declaration of invalidity is based on the text of the Constitution. For example, in the Manitoba Language Rights Reference, the rule of law was used in two ways: in conjunction with section 23 of the Manitoba Act, 1870, to sustain the decision that the unilingual statutes enacted by the legislature since 1890 were invalid, as they had not been enacted in French; and in conjunction with section 52 of the Constitution Act, 1982, to suspend the effects of the decision while the statutes were translated and re-enacted, in order to maintain a positive legal order. In R. v. Nova Scotia Pharmaceutical Society, the SCC held, based on the rule of law and section 7 of the Charter, that vague and unintelligible legal rules are invalid. In the Reference Re Secession of Quebec, the rule of law was used to support the conclusion that Quebec could not unilaterally secede from Canada, even after a clear majority vote on a clear question, as this would require an amendment to the Constitution which could not lawfully take place without negotiation with the federal government and the other provinces under Part V of the Constitution Act, 1982. In Trial Lawyers Association of British Columbia v. British Columbia (A.G.), the SCC invalidated regulations imposing court hearing fees on the basis that, without proper exemptions, these fees could impede access to justice and violate the core, or inherent, jurisdiction of superior courts under section 96 of the Constitution Act, 1867. In Trial Lawyers Association, the rule of law served to bolster the nexus between the core jurisdiction of superior courts and access to justice. In light of the foregoing, the rule of law may thus have some normative force in connection with the text of the Constitution.
Third, it may be open to the courts to conclude that a legal rule is, at the same time, inconsistent with the unwritten rule of law principle and constitutionally valid. This type of situation could occur where a legal rule violates a requirement of the rule of law without violating the text of the Constitution. The obvious example is Imperial Tobacco. There, had the SCC incorporated the principles of formal legality as part of the unwritten rule of law principle in its reasons, it would have concluded that the impugned legislation was inconsistent with the rule of law because the measures were not general and prospective. Yet, as the written Constitution does not require that rules be general and prospective in civil matters, the legislation would still have been constitutionally valid. In adopting this midway position, the SCC would have upheld its duty to protect the rule of law in a way that is respectful of its role in the Canadian legal order. This would have signaled to Parliament and the government that there was something wrong with the legislation and opened the way for corrective measures through an institutional dialogue. The government would still have been able to apply the legislation, but would have suffered political costs for doing so. Perhaps, in the future, the SCC may be persuaded to use the rule of law in this manner.
To sum up, the SCC has so far adopted a very thin, not to say weak, conception of the unwritten rule of law principle, which embraces three elements: there must be a legal order; within that legal order, legal rules must apply equally to the state and legal subjects; and state action must be authorized by legal rules. The SCC has suggested the rule of law could not be used, as a standalone principle, to challenge the validity of democratically-enacted legal rules. To be legitimate, judicial review of legislation should be grounded on the text of the Constitution—the big-C Constitution. Thus, to challenge the validity of legislation, the rule of law should be used in conjunction with specific provisions of the Constitution. This approach is consistent with the SCC’s positivist conception of the rule of law and its neglect of small-c constitutionalism. The problem with this approach is the following: because the SCC’s current conception of the rule of law is very thin, its normative force is limited. At present, the rule of law cannot breathe much life into the text of the Constitution. While it does have some normative potential, that potential is locked. Hence, we are in a cul-de-sac: the rule of law can be used to bolster the interpretation of the text of the Constitution, but the SCC’s conception of the rule of law is too thin to play that role in a meaningful manner. To unlock the full normative potential of the rule of law principle, I will turn to an alternative conception of the rule of law, the theory of law as justification, which is implicit is the Canadian legal order.
The theory of law as justification offers a compelling alternative conception of the rule of law, in contrast to the “rule by law” conception, as it fosters the establishment of a “culture of justification” in legal orders. The first to coin this term was South African scholar Etienne Mureinik. In 1994, in an influential article, he presented the “culture of justification” as an ideal for the new South African Constitution, as opposed to the “culture of authority” that prevailed during Apartheid. For Mureinik, a culture of justification is “a culture in which every exercise of power is expected to be justified; in which the leadership given by government rests on the cogency of the case offered in defence of its decisions, not the fear inspired by the force at its command.” This simple idea, that is, the idea that state actors must provide substantive justification for all their actions, was subsequently developed into an elaborate rule of law theory by David Dyzenhaus, as part of what he calls the “rule-of-law project.” The aim of this project is to ensure that state power is exercised within the limits of the rule of law by requiring state actors to embrace a culture of justification. The presence of a culture of justification in a legal order requires three key elements: the recognition of fundamental legal principles; the judicial review of state action; and the imposition of an onus of justification on the state.
a. Fundamental Legal Principles
The first element of a culture of justification is that judges must take the view that the rule of law has content, that is, it is made of fundamental legal principles. These principles can be enshrined in a written or unwritten constitution. Law as justification promotes a thicker conception of the rule of law, as compared to rule by law. The eight principles of formal legality, set out by Lon Fuller, are at the centre of the rule-of-law project. The first seven principles are directed primarily to the legislative branch because they pertain to the inner qualities that legal rules must possess to count as law. Legal rules must be general, public, prospective, clear, coherent, possible to comply with, and relatively stable over time, so that legal subjects can be guided by them. A rule that fails to meet one of these principles entirely, or several of them substantially, cannot claim legal legitimacy. The eighth principle, congruence, is directed primarily to the executive branch: it requires state actors to act in a manner consistent with legal rules. Compliance with these principles infuses the law with an “inner morality”. Law as justification focuses on “a kind of justice located within the law” in the design and administration of the law.
Law as justification goes beyond formal legality as it is “both liberal and democratic in inspiration.” Thus, it seeks to reconcile the need to place the development of legal rules in the hands of the elected representatives of the people—the value of participation—with the need to ensure that rules do not unjustifiably violate the rights of individuals—the value of accountability. Dyzenhaus accepts that legal legitimacy is incomplete without democratic legitimacy. Furthermore, he views legal subjects as “bearer[s] of human rights”, who must be treated with equality and dignity, and whose liberty must be respected. Beyond these fundamental attributes and the requirements of procedural fairness, Dyzenhaus has resisted the urge to define more precisely the substantive content of the rule of law. This may be explained by the fact that the focus of law as justification is on the connection between process and substance; the intuition is that a good decision-making process should lead to a substantively good decision. That said, a culture of justification is more likely to prosper in a legal order that values democracy and human rights.
b. Judicial Review
The second element of a culture of justification is that judges must be empowered to review legislative and executive actions to ensure that they comply with the aforementioned fundamental legal principles. This implies the existence of a separation of powers between the legislative, judicial, and executive branches, under which independent and impartial judges have the ultimate authority to interpret and apply the law in cases of controversy. Although the rule-of-law project requires the cooperation of the three branches of the state in the maintenance of the rule of law, judges have a special role in this regard. Because of their obligation of fidelity to the law, judges must hold the legislative and executive branches to account when they fail to uphold the rule-of-law project. There are many ways in which this may happen. Lawmakers may enact legal rules limiting access to the courts, depriving litigants of procedural fairness, or give state actors a broad and unreviewable discretion on a given matter, which could then be abused. Privative clauses are the archetype of such legal rules. The danger with privative clauses is that they create “legal black holes”, that is, zones which are not controlled by the rule of law. These legal black holes are inconsistent with a culture of justification because they insulate state action from judicial review.
Judicial review is also essential to the fulfilment of Fuller’s congruence principle. The essential purpose of judicial review is to ensure that state actors act in a manner that is consistent with legal rules. There could be no rule of law, even on a bare rule by law understanding of that concept, if the executive branch was free to ignore the legal rules enacted by the legislative branch. While the first seven principles of legality enable the conversion of public policy objectives into legal rules, the last one enables the conversion of legal rules into what Fuller describes as “claims of right or accusations of fault.” It is the function of independent and impartial judges to review executive action for congruence, in line with the rules of procedural fairness. In public law, “adjudication” is one of the main devices under which legal subjects can hold the state accountable. It is a “form of social ordering,” which enables legal subjects to participate in decisions that affect them, through the presentation of “proofs and reasoned arguments” that are “supported by a principle.” When lawmakers shield executive action from judicial review, that is, when they create a legal black hole, they prevent the application of the congruence principle. This is inconsistent with a culture of justification as it ensures “that there is no law with which official action has to be congruent.” In doing so, lawmakers fail to address legal subjects with the dignity they deserve as bearers of human rights.
c. Onus of Justification
The third element of a culture of justification is that the legislative and executive branches must bear the onus of justifying their actions by reference to the fundamental legal principles. The onus of justification is triggered when state action affects individual rights, privileges, or interests. In such cases, legal subjects will inquire: “But how can that be law for me?” In answering the question, the state ought to satisfy what Bernard Williams describes as the “Basic Legitimation Demand”; to do so, the state must “offer a justification of its power to each subject.” In other words, the state must provide reasons for its action. State actors must be transparent and identify the source of the legal power they are asserting, as well as the way that power has been interpreted and applied, in view of the legal subject’s specific circumstances. Only in light of the reasons provided by state actors to justify their actions can legal subjects, and judges, determine whether these actions are consistent with fundamental legal principles. By discharging their “duty to give reasons”, state actors enable legal subjects to know that their dignity and equal status under the law have been respected, and that the actions were taken in good faith, free of bias, based on the appropriate legal considerations.
Yet, the duty to give reasons makes sense only if the reasons provided do in fact justify the action. Hence, it is not enough that reasons be provided; they must also be reviewed by a judge to assess whether they show, or are capable of showing, that the action is justifiable in view of the fundamental legal principles. “Justifiable” must not be confused with “justified.” When a judge inquires whether a decision is “justified”, he or she is asking whether he or she would have made the same decision, which constitutes the correctness standard in administrative law. When a judge inquires whether a decision is “justifiable”, he or she is asking whether the decision is defensible in light of the appropriate legal considerations, which constitutes the reasonableness standard in administrative law. The reasonableness standard requires that the decision be rational and proportional. By demanding that state action be “justifiable”, as opposed to “justified”, law as justification recognizes that state actors have a legitimate role in the interpretation of the law. This implies that judges should not quash an executive decision just because they would have reached a different conclusion, provided that the executive decision falls within a range of defensible outcomes. Dyzenhaus uses the concept of “deference as respect” to communicate the idea that judges should defer to executive interpretation of the law, within expertise, if the reasons provided to support that interpretation are “good enough”.
Now that the three elements of the theory of law as justification have been outlined, I will demonstrate that the theory is implicit in the Canadian legal order, even though the courts have not yet recognized it. I will then argue that the theory of law as justification also underpins the approach taken by the courts under the common law to determine the validity of public interest immunity (PII) claims, also known as claims of Crown privilege, at the provincial level in Canada. Finally, I will identify the normative conditions that a statutory regime of PII, such as section 39 of the CEA, must fulfil to comply with the theory of law as justification.
a. Law as Justification and the Canadian Legal Order
The three elements of the theory of law as justification are embedded in the Canadian legal order. First, the Canadian legal order contains written and unwritten fundamental legal principles. The written Constitution establishes democratic governance in the form of an elected House of Commons and protects a broad range of human rights including equality, dignity, and liberty. Moreover, the Constitution contains unwritten principles, such as the rule of law, democracy, and respect for minorities, which are broad expressions of the nature of the Canadian legal order and can be used to fill in the gaps in the written Constitution. Similarly, in administrative law, the SCC has declared that executive discretion should be exercised “in accordance with the boundaries imposed in the statute, the principles of the rule of law, the principles of administrative law, the fundamental values of Canadian society, and the principles of the Charter.” Democracy and human rights are therefore important features of both law as justification and the Canadian legal order. Lastly, important principles of formal legality, such as the principles of generality, clarity, publicity, prospectivity, and congruence have a clear textual basis in the Constitution and ordinary statutes.
Second, the Canadian legal order establishes a separation of powers which affirms the power of the courts to review the validity of legislative and executive actions and strike down invalid actions. While there may be no genuine separation of powers between the legislative and executive branches, given that ministers are members of both branches, the judicial branch is clearly segregated from the other two. The institutional independence of the courts stems from the unwritten principle of judicial independence as well as paragraph 11(d) of the Charter and section 96 of the Constitution Act, 1867. Section 96, especially, constitutionalizes the core, or inherent, jurisdiction and powers of provincial superior courts, and protects their authority to review executive action for jurisdictional errors. As for the courts’ power to invalidate legislation, this stems from section 52 of the Constitution Act, 1982, as well as the principle of constitutionalism. Rules of procedural fairness under the common law, the Bill of Rights, and the Charter typically impose on public decision-makers a duty to act fairly in coming to decisions that affect individual interests. The Canadian legal order is thus designed to ensure that fundamental legal principles are recognized and enforced in a manner consistent with the congruence principle and a culture of justification.
Third, the Canadian legal order imposes on the legislative and executive branches the onus of justifying their actions when they affect legally protected individual interests. As a matter of constitutional law, the government bears the onus of justifying any legal rule that limits a fundamental right or freedom under section 1 of the Charter. Section 1 involves an inquiry into the proportionality of a law. The government also bears the onus of justifying, pursuant to the principles of fundamental justice, any legal rule that violates the right to life, liberty and security of the person protected under section 7 of the Charter. A similar onus of justification exists as a matter of administrative law. The rules of procedural fairness impose on the government a legal duty to provide reasons, especially when it makes a decision which has “important significance” for an individual. As shown by Roncarelli v. Duplessis and Baker v. Canada (Minister of Citizenship and Immigration), awareness of the reasons supporting executive actions is key to a successful judicial review application. Courts examine reasons to determine whether or not the executive decision is reasonable, a standard of review which focuses on the “existence of justification, transparency and intelligibility” in the decision-making process. The onus of justification, and the concept of “deference as respect”, are accordingly part of the Canadian legal order.
b. Law as Justification and PII under the Common Law
The requirements of “judicial review” and “onus of justification” are essential to assessing whether the various common law approaches to PII in Westminster jurisdictions comply with the theory of law as justification. For almost three decades, as a result of the House of Lords’ 1942 decision in Duncan v. Cammell, Laird and Co. Ltd., the common law deprived judges of the power to assess the substantive validity of PII claims made in the proper form by the responsible minister. This situation led to instances of abuse of power, as ministers claimed PII for tactical reasons in cases where the government was a party to litigation, in breach of the fundamental rules of procedural fairness. Duncan was inconsistent with the congruence principle and the second element of law as justification as it prevented meaningful judicial review of state action. This flaw was corrected in Conway v. Rimmer, when the House of Lords overruled Duncan and affirmed the power of the courts to: inspect the documents subject to a PII claim; weigh and balance the competing aspects of the public interest; and order their production when appropriate. Since Conway, the common law has enabled judges to assess the substantive validity of PII claims, in accordance with the second element of law as justification—judicial review.
This development did not, however, have the effect of making the common law of PII consistent with the third element of law as justification. Courts in the United Kingdom and Australia have taken the position that judges should not inspect Cabinet documents that fall within the discovery standard unless the litigant persuades them that an inspection is necessary. However, it is almost impossible for litigants to meet this burden as they are not privy to the documents’ contents. Cabinet immunity claims therefore remain, to some extent, unreviewable in these jurisdictions. This is problematic from a rule of law perspective, especially when it is alleged that the government has acted illegally. The “public interest … in the maintenance of legality,” cannot be vindicated if judges refuse to inspect relevant documents subject to a Cabinet immunity claim. In this context, judges cannot determine whether executive action is truly congruent with legal rules. For this reason, courts in Canada, at the provincial level, inspired by the position taken in New Zealand, have reversed the onus: the default position is that judges should inspect Cabinet documents that fall within the discovery standard unless the government persuades them that it is not necessary. Thus, Canada’s and New Zealand’s approach under the common law also complies with the third element of law as justification—the onus of justification.
c. Law as Justification and PII under Statute Law
In this context, we can identify two normative conditions that a statutory PII regime, such as the one set out in section 39 of the CEA, must follow to comply with the theory of law as justification. These conditions are triggered each time an executive decision to exclude information has the effect of depriving litigants of evidence relevant to the fair disposition of their case. The first condition combines the fundamental legal principles with the onus of justification, while the second deals with the requirement of judicial review.
- The statutory regime should establish a fair decision-making process. This implies that the ultimate decision to exclude relevant Cabinet confidences in litigation should be made by an independent and impartial decision-maker. In addition, it implies that state actors should have the obligation to properly justify Cabinet immunity claims.
- Executive decisions to exclude relevant Cabinet confidences in litigation should be subject to meaningful review by the courts to ensure that executive action is congruent with legal rules and state actors do not exceed the limits of their legal powers. This implies that the courts should have the power to inspect Cabinet confidences and to overrule immunity claims where they fail to meet the standard of reasonableness.
To sum up, the theory of law as justification provides a compelling alternative conception of the rule of law as compared to the SCC’s rule by law conception. Because it is already embedded in the Canadian legal order, law as justification can legitimately be relied upon by the courts to guide their interpretation of the Constitution. What does it imply for section 39 of the CEA? In a culture of justification, if the congruence principle and the integrity of the adjudicative process are to be maintained, the ultimate decision to deprive litigants of relevant evidence in litigation must be made pursuant to a fair process, by unbiased judges, with the power to review both the justification and the information. Yet, at the same time, under a culture of justification, judges must defer to an executive decision to shield Cabinet confidences in litigation if it is reasonable, in accordance with the idea of “deference as respect”.
Is the regime of Cabinet secrecy set forth in section 39 of the CEA consistent with the theory of law as justification and the provisions of the Constitution? Part 2 will answer this question. It is divided into two subparts. In the first subpart, I will inquire whether the statutory decision-making process is procedurally fair. I will focus on two requirements of procedural fairness, namely, the litigant’s right to: have his or her case determined by an independent and impartial decision-maker; and have access to the reasons underpinning the decision. These requirements, essential to law as justification, stem from the common law, paragraph 2(e) of the Bill of Rights, as well as section 7 and paragraph 11(d) of the Charter. In the second subpart, I will inquire whether section 39 unduly limits the authority of provincial superior courts to: control the admissibility of evidence in litigation; and review the legality of executive action. Judicial review, a key element of law as justification, is protected under section 96 of the Constitution Act, 1867. I will demonstrate that section 39 is procedurally unfair and encroaches upon the core, or inherent, jurisdiction and powers of superior courts. It is thus inconsistent with the rule of law as a culture of justification and, because it violates provisions of the Constitution, it is also unconstitutional.
In civil cases against the federal government, litigants have challenged—unsuccessfully—the validity of section 39 of the CEA, and the provisions that preceded it, on the basis that it violates the “right to a fair hearing” guaranteed under paragraph 2(e) of the Bill of Rights. Paragraph 2(e) provides that no law of Canada must be construed or applied so as to “deprive a person of the right to a fair hearing in accordance with the principles of fundamental justice for the determination of his rights and obligations.” The SCC has held that paragraph 2(e) requires that decision-makers who determine individual rights and obligations “act fairly, in good faith, without bias and in a judicial temper, and must give to [the litigant] the opportunity adequately to state his case.” In their fight against section 39, litigants have focused on the last element of that statement: the right to adequately state one’s case, which is part of the natural justice principle audi alteram partem—“hear the other side”, also known as the “right to be heard”. The gist of their claims has been that section 39 deprived them of access to evidence that was prima facie relevant to the fair disposition of their cases, thus preventing them from adequately stating their cases. Courts have rejected these claims on the basis that the exclusion of relevant evidence because of a privilege or immunity does not violate the “right to be heard”. The rationale underlying the courts’ position was summarized as follows by Justice Iacobucci, then of the Federal Court of Appeal, in Central Cartage:
Many questions of privilege such as solicitor-client, priest-penitent, or rules on hearsay evidence can operate to cut down on the ability to state one’s case by denying admissibility into evidence even though relevance may be established. The issue of Crown privilege attaching to Cabinet confidences is firmly established as one of these exceptions and I believe it has not been ousted by the wording of paragraph 2(e) of the Canadian Bill of Rights.
While the litigants were on to something, the thrust of their argument was misplaced. What makes the decision-making process set out under section 39 procedurally unfair is not the fact that it may lead to the exclusion of relevant evidence for public policy reasons; rather, it is the fact that the decision to exclude the evidence is made by someone who is seemingly biased, namely, “a minister of the Crown or the Clerk of the Privy Council.” Subsection 39(1) gives members of the executive branch a very broad discretion to decide whether relevant evidence should be excluded in proceedings where the government is a party, in breach of the natural justice principle nemo judex in sua causa—“no one may be judge in his own cause”, also known as the “rule against bias”. This attribute differentiates section 39 from the other existing privileges and immunities. The minister or the Clerk is not just “objecting” to the production of information, he or she is finally and conclusively “deciding” the matter. No other privilege or immunity enables a litigating party to decide what evidence should be excluded from the proceedings. This is normally a matter for the judge to decide. Hence, the problem is not so much that section 39 prevents a party from adequately stating his or her case, the problem is that the individual who has the power to exclude the evidence is not “without bias”. I will flesh out this argument by answering three questions: What are the requirements of the rule against bias? Why does section 39 violate that rule? What are the consequences of the violation?
a. Requirements of the Rule Against Bias
The rule against bias provides that decision-makers and decision-making processes must not grant undue preferential treatment to one party over another or be driven by preconceived notions. It seeks to maintain the fairness of the process for litigants and the confidence of the public in the administration of justice. Thus, a mere perception of bias, if that perception is reasonable, may nullify a decision: “[It is] of fundamental importance that justice should not only be done, but should manifestly and undoubtedly be seen to be done.” In civil cases, the rule against bias applies to any decision-maker who must decide matters affecting individual rights. In such cases, a litigant is legally entitled to an independent and impartial decision-maker. The required degree of independence and impartiality varies depending on whether the decision-maker performs judicial or quasi-judicial functions, which will require a higher degree, or administrative functions, which will require a lower degree.
The notion of “impartiality” refers to the decision-maker’s state of mind in relation to the issues and the parties in a case. An impartial decision-maker is one that can decide with an open mind and who does not have an interest with, or connection to, the matter in dispute. A decision-maker will be disqualified if there is actual evidence of bias or if the circumstances give rise to a “reasonable apprehension of bias”. In the latter case, the question to ask is:
[W]hat would an informed person, viewing the matter realistically and practically—and having thought the matter through—conclude. Would he think that it is more likely than not that [the decision-maker], whether consciously or unconsciously, would not decide fairly.
A reasonable apprehension of bias can be individual or institutional. An individual bias arises where the decision-maker has: a pecuniary interest in the outcome of the case; a relationship with one of the parties; a prior involvement in the case; or an attitudinal predisposition toward a specific outcome. An institutional bias arises where the structure or internal practices of a decision-making body give rise to a “reasonable apprehension of bias in the mind of a fully informed person in a substantial number of cases.”
The notion of “independence” is connected to the notion of “impartiality” in the sense that independence strengthens the public’s perception of impartiality. The essence of judicial independence is
the complete liberty of individual judges to hear and decide the cases that come before them: no outsider—be it government, pressure group, individual or even another judge—should interfere in fact, or attempt to interfere, with the way in which a judge conducts his or her case and makes his or her decision.”
Judicial independence therefore connotes a status of relationship to other branches, especially to the executive, which rests on objective conditions or guarantees like security of tenure, financial security, and administrative control. To protect the courts’ independence from improper executive interference, judges can only be removed following a judicial inquiry at which they are given a full opportunity to be heard, their salary and pension are fixed by law, and they have control over how their affairs are managed. The SCC has ruled that these conditions apply, to some extent, to administrative bodies, even if they are part of the executive branch. That position begs the question of how can administrative bodies be, at the same time, a part of, and independent from, the executive branch? What matters here is not so much that these bodies be independent from the executive branch, but rather that they be independent from the parties to the litigation. As such, a reasonable apprehension of bias arises where one of the parties has control over the tenure, remuneration or modus operandi of the decision-maker.
b. Consistency of Section 39 with the Rule Against Bias
Three conditions must be met for the rule against bias to be engaged in the context of section 39 of the CEA. First, the litigation must involve a litigant and the government. The central concern is that one of the parties to the litigation, the government, may improperly exclude relevant information. This concern is less acute in litigation where the government is not a party. Second, the information excluded must be “relevant” to the fair disposition of the case. It is the exclusion of relevant information that would undermine the litigant’s right to state his or her case and lead to a denial of justice. In contrast, the exclusion of irrelevant information would have no impact on the fairness of the proceedings. However, given that no one can inspect the information excluded by the government, it is reasonable to assume that any information that falls within the standard of discovery is prima facie relevant to the fair disposition of the case. Third, the power to exclude relevant information in litigation, for public policy reasons, must be judicial in nature. Because it involves the interpretation and application of legal standards, in a manner that is meant to be final and conclusive, and which will affect the litigant’s rights, the power to exclude relevant documents is judicial in nature whether it is exercised by a judge, a minister, or a senior public servant. Hence, to prevent the existence of a reasonable apprehension of bias, the decision-maker must have a high level of independence and impartiality.
Would the exclusion of relevant Cabinet confidences in litigation by a minister or the Clerk give rise to a reasonable apprehension of bias? Is the decision-maker sufficiently independent and impartial to perform this function in a fair manner?
With respect to the question of independence, neither the minister nor the Clerk are independent from the executive branch as such. But this is not the issue: the issue is whether a party to the litigation has direct or indirect control over their tenure, remuneration, or modus operandi. The answer is “yes”. Ministers and Clerks are appointed under the prime minister’s recommendation and hold their position at pleasure, meaning that they can be removed at any time, for any reason. Let us focus on the Clerk who has the primary responsibility for the application of section 39. The Clerk performs three functions: deputy minister to the prime minister; secretary to the Cabinet; and head of the Public Service of Canada. Given his or her unique position in the public service, the Clerk must serve the prime minister and Cabinet with loyalty. The Clerk’s main role is to ensure the efficiency of the collective decision-making process and to assist his or her political masters to develop and implement their policy goals. At the same time, the Clerk must decide issues related to the application of section 39. When deciding if Cabinet confidences should be disclosed in litigation, the Clerk is acting for the executive branch and is accountable to the prime minister and Cabinet for his or her decisions.
The disclosure of Cabinet confidences is naturally a question of concern for the prime minister and Cabinet, especially if the information pertains to their own ministry, as it could have major political consequences. It could weaken the candour and the efficiency of Cabinet deliberations as well as ministerial solidarity. It could even jeopardize the ministry’s capacity to maintain the confidence of the House of Commons and electorate. These are the kind of negative effects that Parliament tried to prevent when it made Cabinet immunity almost absolute. The disclosure of Cabinet confidences could also have major legal consequences as the information could be used to: attack the legality of government policies; establish governmental liability for civil wrongs; or uncover the mismanagement of public funds. There are strong incentives within the executive branch to keep Cabinet documents secret. It would be unprecedented for anyone to allow the disclosure of an official Cabinet document without the prime minister’s consent. This could explain why Cabinet documents are typically not voluntarily disclosed in litigation before the expiry of the 20-year moratorium. Hence, a reasonable apprehension of bias arises from the prime minister’s power over the decision-maker’s tenure and his or her “political self-interest” in maintaining Cabinet secrecy. An informed person would likely conclude that the minister or the Clerk cannot decide the matter fairly.
With respect to the question of impartiality, the process set out under section 39 of the CEA may give rise to a reasonable apprehension of individual and institutional bias. A perception of individual bias may arise from the minister’s or the Clerk’s prior involvement in the government decision under scrutiny. Ministers and Clerks should not make a final and conclusive decision on the disclosure of Cabinet confidences in litigation if they have been involved in making the impugned decision.
Assume that the Minister of Environment and the Clerk have been involved in the collective decision-making process leading to the decision to approve the construction of a pipeline, despite significant adverse environmental effects. Assume that the legality of the decision is now being challenged by First Nations bands on the basis that their rights were not properly considered. Would an informed person likely conclude that the minister or the Clerk has an interest in excluding Cabinet confidences that may be used to nullify the decision? In these circumstances, a reasonable apprehension of bias would arise from the direct connection between the decision-maker and the impugned decision.
The same reasoning would apply in cases where allegations of breach of contract or fiduciary duties, or unconscionable conduct, are made against the government. If the minister or the Clerk has been involved in making the impugned decision, he or she cannot finally and conclusively decide whether relevant evidence should be excluded in related proceedings, without giving rise to a reasonable apprehension of bias.
On a final note, a perception of institutional bias may arise from the role played by Counsel to the Clerk in the decision-making process under section 39 of the CEA. In relation to any major litigation against the government, Counsel is called to perform two seemingly incompatible functions: on the one hand, provide legal advice to, and seek instructions from, the Clerk, the prime minister, and the Cabinet, on which position should be taken in the litigation and, on the other hand, provide legal advice to the Clerk on whether evidence that is relevant to the fair disposition of the case should be excluded. These functions are incompatible because the government’s ability to successfully defend against a lawsuit may be undermined by the disclosure of Cabinet confidences. The problem is that one person is involved in both developing the government’s litigation strategy and deciding whether evidence that could weaken that strategy should be excluded. Lawyers involved in the litigation process in a specific case should not participate in the adjudicative process, as it would give rise to a “reasonable apprehension of bias [in the mind of an informed person] in a substantial number of cases.” When Counsel advises the Clerk to issue a certificate on the basis of Cabinet immunity, he or she is in effect participating in the adjudicative process because the certificate will deprive the litigant and the judge of evidence that may have been relevant to the fair disposition of the case. In doing so, Counsel is effectively acting both as judge and party in the same cause.
c. Consequences of the Violation of the Rule Against Bias
The decision-making process established by Parliament under section 39 of the CEA is thus inconsistent with the rule against bias, which is part of both the common law and the theory of law as justification. The wording of section 39 suggests that Parliament wanted to overrule the common law by enabling the minister or the Clerk to exclude relevant Cabinet confidences despite any reasonable apprehension of bias. Absent constitutional constraints, the courts would have to enforce Parliament’s intention. Yet, at the federal level, the rule against bias has been given a quasi-constitutional status pursuant to paragraph 2(e) of the Bill of Rights. As such, the rule against bias should not, in the absence of an express declaration, be displaced by an ordinary statute like the CEA. In the end, the fact that the prime minister controls the appointment of ministers and Clerks, and could exert an influence over the application of Cabinet immunity, leads to the conclusion that section 39 does not meet the requirement of independence. As previously discussed, there are also concerns with respect to individual and institutional bias in the decision-making process. Consequently, section 39 should be declared “inoperative” insofar as it violates paragraph 2(e) of the Bill of Rights.
Until now, this article has focused on the question of independence and impartiality in the context of civil cases, but the same reasoning would apply in criminal cases. Where an individual faces true penal consequences, the analysis would focus on section 7 and paragraph 11(d) of the Charter. In R. v. Stinchcombe, the SCC held that an accused has a constitutional right to make full answer and defence as per section 7. To that end, the principles of fundamental justice, which incorporate the duty of procedural fairness, require that all relevant evidence be provided to the accused. In addition, where a dispute arises as to the admissibility of relevant evidence, the accused is constitutionally entitled to have an “independent and impartial tribunal” decide the matter as per paragraph 11(d). Thus, it is doubtful that the issuance of a final and conclusive certificate under section 39 of the CEA in a criminal case would survive Charter scrutiny. This could explain why the government has so far refrained from relying upon section 39 in this context. In cases in which Cabinet confidences were relevant to the defence of former ministers, the information was voluntarily disclosed to the accused. Considering the importance of the interest at stake in criminal cases, a choice must be made between prosecuting the accused and protecting Cabinet confidences. Section 39 should not be used to deprive the accused of relevant evidence given the risk of wrongful conviction.
2. The Decision-Maker Is Not Required to Give Reasons
There is another reason for concern, from a procedural fairness perspective, with the way section 39 of the CEA has been interpreted and applied. This concern goes to the heart of the theory of law as justification, as it relates to the reasons given by the decision-maker for excluding relevant information in litigation. The duty to give reasons is now firmly established in the Canadian positive law as an element of the natural justice principle audi alteram partem. I will argue that, in their current form, the certificates issued by the government to exclude Cabinet confidences are structurally deficient because they fail to address a fundamental issue: Why does the public interest require that the information be excluded in the circumstances of the case? The minister or the Clerk is bound to answer that question before issuing a section 39 certificate, but his or her reasoning is not communicated to the litigant and the judge, as it would be for any other PII claim. To support this argument, I will address the following three questions: What are the requirements of the duty to give reasons? Why are the reasons provided in current section 39 certificates insufficient? What are the consequences of the failure to provide sufficient reasons?
a. Requirements of the Duty to Give Reasons
As a rule, under the common law, procedural fairness did not historically require that reasons be provided for administrative decisions. The reluctance to impose a general duty to give reasons stemmed from numerous concerns. Courts considered that it would place an undue burden on administrative decision-makers, thus triggering additional cost and delay, and possibly even inducing a lack of candour from decision-makers. Nevertheless, these concerns have been gradually overshadowed by the benefits of recognizing a general duty to give reasons. The provision of reasons fosters transparency and accountability, and bolsters public confidence in the integrity of the decision-making process. It also promotes better decisions as it forces decision-makers to address the critical issues in dispute and to carefully articulate their reasoning. As a result, the provision of reasons enables affected individuals to assess: whether their submissions and all the appropriate factual and legal considerations have been taken into account; and whether the decision should be challenged by way of statutory appeal or judicial review. If and when the decision is challenged, it enables the appeal or reviewing body to assess whether the decision should be sustained or not. All in all, the provision of reasons ensures that the affected individual is treated with fairness, dignity, and respect. This explains why the SCC has recognized in Baker a duty to give reasons “where the decision has important significance for the individual, when there is a statutory right of appeal, or in other circumstances [such as to facilitate judicial review].”
To address the traditional concerns regarding the duty to give reasons, and maintain the efficiency of administrative decision-making processes, the SCC has shown flexibility as to how this duty can be performed. When the duty to give reasons applies, procedural fairness requires that the decision-maker provide some form of reasons, but these reasons do not necessarily need to be as detailed as those drafted by judges. At minimum, to meet the Basic Legitimation Demand, the reasons must answer the question “Why?” The litigant is entitled to know why the decision-maker has reached a specific conclusion. To answer the Demand, the decision-maker must do more than simply recite the parties’ submissions, summarize the evidence and state a conclusion. The decision-maker must address the critical issues in dispute and explain why he or she was persuaded by some arguments and not by others. In doing so, the decision-maker must identify his or her key findings of fact and the evidence upon which they are based. In addition, if his or her conclusion is based on the interpretation and application of legal provisions or precedents, the decision-maker is expected to explain his or her legal reasoning. In short, he or she must show that there is a “logical connection” between the conclusion and the basis for the conclusion. Only in light of the decision-maker’s factual and legal justification will the litigant and the appellate or reviewing body be able to assess whether all the relevant considerations have properly been taken into account. As the SCC held in Dunsmuir, the validity of a decision depends in part on the “justification, transparency and intelligibility” of the reasons articulated to support it.
b. Consistency of Section 39 Certificates with the Duty to Give Reasons
Would a government’s decision to protect Cabinet confidences under section 39 of the CEA trigger the duty to give reasons? The answer is “yes”. A decision of this nature is bound to have “important significance” for the litigant if it deprives him or her of evidence that is relevant to the fair disposition of the case. That is why the litigant should be entitled to understand the rationale behind the decision beyond the obvious fact that information falls within the statutory class of “Cabinet confidences”. Plus, while a section 39 certificate is meant to be final and conclusive, and thus is not subject to a “statutory right of appeal”, it is not immune from judicial review on jurisdictional grounds. However, for such challenge to be meaningful, the litigant and the judge need to have access to the justification supporting the claim. The fact that the plain wording of subsection 39(1) imposes on the government the duty to “[certify] in writing that the information constitutes a confidence of the Queen’s Privy Council for Canada” suggests that Parliament itself envisioned that some form of reasons be provided to justify the use of Cabinet immunity. This is consistent with the courts’ longstanding position under the common law. Therefore, the issue is not so much whether the government has a duty to give reasons when claiming Cabinet immunity; rather, it is the scope of that duty.
In Babcock, the SCC stated that the government—that is, the minister or the Clerk—must answer two questions before issuing a certificate: (1) do the documents contain “Cabinet confidences” within the meaning of subsection 39(2) of the CEA; and (2) should the documents be protected in the circumstances of the case considering the competing aspects of the public interest? One would expect the certificate to address both questions. Unfortunately, it does not. The SCC has imposed on the government the duty to give reasons regarding the first but not the second question. To show that it has duly addressed the first question, the government must provide a sufficient description of the documents in the certificate. To that end, it must: identify each document’s date, title, author and recipient; and indicate under which paragraph of subsection 39(2) each document falls—memoranda to Cabinet, discussion papers, agenda, minutes, decisions, briefing notes, letters or draft legislation. If the descriptions provided clearly establish that the documents contain Cabinet confidences, the duty to give reasons will be satisfied regarding the first question. Turning to the second question, the SCC stated that the public interest assessment is a “discretionary element [that] may be taken as satisfied by the act of certification.” This implies that the government is not required to justify why the documents must be protected in the public interest. In fact, certificates issued in the aftermath of Babcock do not address this issue. This is an important flaw in the manner in which section 39 was interpreted by the SCC: without justification, there is simply no way to know whether the government has properly addressed the second question.
The SCC’s interpretation does not flow from the plain wording of the statutory provision: Parliament has not expressly superseded the relevant common law principles in enacting section 39 of the CEA. At most, one could say that the statutory provision is silent on whether reasons must be provided with respect to the public interest assessment. It is thus open to the courts to refer to the common law to fill in the gap in the statutory provision. What does the common law say about the issue? Under the common law, to make a successful claim, the government must not only describe the documents in the certificate, but explain why the public interest require that they be excluded. It must provide an assessment of the documents’ degree of relevance and degree of injury. The public interest assessment is the critical part of certificates. Under the common law, the debate is seldom about whether the documents contain Cabinet confidences; it is about whether the interest of justice outweighs the interest of good government. In assessing the public interest, the government must take into consideration factors such as the documents’ probative value and materiality, on the one hand, and the sensitivity of their contents and the timing of their production, on the other hand. Failure to provide a rational justification would defeat the claim. As the government is already under a duty to perform a public interest assessment under section 39, why should it be exempted from the duty to share its justification with the litigant and the judge?
The common law doctrine of procedural fairness also supports the recognition of a broader duty to give reasons under section 39. Statutory provisions should not lightly be interpreted to displace the common law in this regard. Access to the reasons would assure the litigant that the government has considered all the appropriate factors, and adequately weighed and balanced the competing aspects of the public interest. This is imperative not only to ensure that the litigant is treated with respect, but also as a means of bolstering the public confidence in the integrity of the decision-making process. Access to the reasons may bring to light bona fide or mala fide errors that may have been made during the public interest assessment. Knowledge of these errors would enable the litigant to exercise his or her constitutional right to judicially challenge the decision. Without this knowledge, the litigant would be unable to do so in a meaningful way, for he or she would not know whether the decision is reasonable and within jurisdiction. His or her ability to participate in the review process by making submissions to the court would be impeded. Moreover, without this knowledge, the reviewing court would be unable to assess whether the government’s decision to exclude relevant Cabinet confidences can be justified. Courts should thus rely on the common law to broaden the duty to give reasons as part of section 39 to include the public interest assessment. In Babcock, the SCC took a first step by “reading in” the public interest assessment in section 39, although it was not explicitly required by the plain wording of the provision. Courts should now compel the government to disclose its public interest assessment in certificates, as a condition of validity of Cabinet immunity claims, considering that this information can be provided without revealing any legitimate Cabinet confidence. This approach can be implemented by way of statutory interpretation without any legislative modification.
c. Consequences of the Violation of the Duty to Give Reasons
What are the consequences of the government’s failure to communicate its public interest assessment to the litigant and the judge? Beyond violating a key requirement of the theory of law as justification, the government’s failure would invalidate the certificate, and the Cabinet immunity claim, whether the failure is seen as a procedural or substantive defect. The procedural question—subject to the correctness standard—is whether the government has given reasons to sustain the decision. In contrast, the substantive question—subject to the reasonableness standard—is whether the government’s reasons are adequate to sustain the decision. The government’s failure to communicate its public interest assessment would likely be viewed as a substantive rather than a procedural defect. If the court considers that the issuance of a certificate fulfils the procedural requirement to give reasons, it would turn its mind to the issue of whether the reasons provided in the certificate are adequate. However, as the certificate would not contain the government’s public interest assessment, the court would be unable to assess the reasonableness of the decision to exclude relevant Cabinet confidences. This deficiency would stem from the government’s decision not to provide the information. In this context, the court would be powerless to palliate the lack of reasons as judges sometimes do. The court would not be able to “supplement” the reasons provided in the certificate as it could not inspect documents subject to section 39 and conduct its own public interest assessment. As such, the court would have no other option than to declare the certificate invalid and remit the matter to the government, so that a minister or the Clerk could issue a new certificate, containing his or her public interest assessment, or abandon the claim.
To sum up, in this section, I have developed arguments supporting the position that the way section 39 has been designed, interpreted, and applied is procedurally unfair and inconsistent with the theory of law as justification. First, I have argued that the decision-making process gives rise to a reasonable apprehension of bias, as the minister and Clerk do not have the independence and impartiality needed to fairly adjudicate Cabinet immunity claims in cases involving the government. Second, I have argued that the courts are not able to assess the reasonableness of certificates because the minister and Clerk do not provide reasons explaining why the public interest requires that relevant Cabinet confidences be excluded in the circumstances of the case. These rule of law problems pertain to the identity of the decision-maker and his or her relationship with the prime minister, and the way the decision-maker exercises his or her statutory discretion. I will now turn to another fundamental flaw of section 39 relating to the separation of powers between the executive and the judicial branches of the state.
It is often said that there “is no general ‘separation of powers’ in the Constitution Act, 1867,” in the sense that it “does not separate the legislative, executive and judicial functions and insist that each branch of government exercise only ‘its own’ function.” This assertion is based on two considerations. First, in a system of responsible government, the executive branch is responsible to the legislative branch, and such a system is often described as the “antithesis of separation of powers.” Second, the legislative branch may confer non-judicial functions to the courts or judicial functions to bodies which are not courts. While the Canadian Constitution does not insist on a rigid separation of powers, there is nonetheless a “functional” separation of powers between the three branches of the state under which the legislature is primarily responsible to make the law, the judiciary to interpret and apply the law, and the executive to administer and implement the law. The aspect of the separation of powers which concerns us is the separation between the judiciary, on the one hand, and the legislature and the executive, on the other. The rule of law is meaningless without the existence of independent and impartial courts with jurisdiction to review the legality of legislative and executive action. This aspect of the separation of powers is entrenched in section 96 of the Constitution Act, 1867, which protects the core, or inherent, jurisdiction and powers of provincial superior courts from legislative encroachments.
There are two ways in which legislatures may encroach upon the core jurisdiction and powers of superior courts. First, they may grant to an inferior court or tribunal a jurisdiction or power that has traditionally belonged to superior courts. Second, they may remove from superior courts a jurisdiction or power that has traditionally been within their purview. Courts have devised different legal tests whether there is a “grant” of authority to an inferior court or tribunal, or a “removal” of authority from superior courts. The first test prevents provincial legislatures from creating bodies competing with superior courts while the second test prevents provincial legislatures as well as Parliament from restricting the jurisdiction and powers of superior courts. When examining the validity of section 39, the question is not whether Parliament can enable public officials to claim Cabinet immunity on behalf of the government. The common law has long recognized that public officials can, subject to judicial supervision, object to the production of sensitive information on public policy grounds. The question is whether Parliament can make Cabinet immunity claims final and conclusive; the effect of which is to prevent superior courts from inspecting and ordering the production of Cabinet confidences in litigation. The problem is thus the removal of the superior courts’ jurisdiction and powers over Cabinet immunity claims.
The rationale supporting section 96 of the Constitution Act, 1867 is the “maintenance of the rule of law through the protection of the judicial role.” Governance by the rule of law requires the existence of a judicial system that can ensure that legal rules and processes are upheld. In Canada, superior courts are the “foundation of the rule of law.” Removing part of their core jurisdiction or powers would emasculate them: it would make them something less than a superior court. In MacMillan Bloedel, Chief Justice Lamer stated that “[t]he core jurisdiction of the provincial superior courts comprises those powers which are essential to the administration of justice and the maintenance of the rule of law.” The concepts of “jurisdiction” and “powers” should not be conflated. The core “jurisdiction” of superior courts is their non-statutory authority to hear and determine disputes. Superior courts, as the only courts of general jurisdiction, have inherited their core jurisdiction from the English superior courts. In addition, “a court which is endowed with a particular jurisdiction has powers which are necessary to enable it to act effectively within such jurisdiction.” These core “powers” are procedural in nature and dependent on the exercise of jurisdiction; they enable superior courts to control their procedure, maintain the fairness of their processes, and award remedies. The core power which concerns us is the power to control the admissibility of evidence in litigation; the core jurisdiction which concerns us is the jurisdiction to review the legality of executive action.
These aspects of superior courts’ core jurisdiction and powers are linked to another legal doctrine that the SCC has merged into section 96 of the Constitution Act, 1867. In Trial Lawyers Association, the SCC invalidated regulations imposing “court hearing fees” because these fees could, in some cases, impede access to justice. In doing so, the SCC stressed that the “power to impose hearing fees must be consistent with s. 96 of the Constitution Act, 1867 and the requirements that flow by necessary implication from s. 96.” One such requirement is the need to maintain access to justice. It may be argued that section 39 of the CEA impedes access to justice by depriving litigants of evidence that would enable them to enforce their rights and contest the legality of executive action. As noted in Trial Lawyers Association, “[i]f people cannot challenge government actions in court, individuals cannot hold the state to account—the government will be, or be seen to be, above the law.” Discussions about the validity of section 39 often include an analysis of Commission des droits de la personne v. Canada (A.G.). In this case, the SCC held that Parliament had jurisdiction over federal PII and, pursuant to the doctrine of parliamentary sovereignty, could make the immunity absolute. However, this position is flawed: Parliament’s jurisdiction is, and has always been, subject to section 96. Commission des droits de la personne stands for the position that an absolute PII is consistent with the division of powers between the two levels of government, federal and provincial, but it does not stand for the position that an absolute PII is consistent with the separation of powers between the three branches of the state. That issue was not tackled in 1982 and has since only been addressed in a cursory manner by the SCC.
The question of whether section 39 of the CEA violates section 96 of the Constitution Act, 1867 was summarily addressed by the SCC in Babcock. There, Department of Justice lawyers working in Vancouver sued the federal government for breach of contract and breach of fiduciary duty as they were not paid as much as their Toronto counterparts. The rates of pay had been fixed by the Treasury Board, which is a committee of the Privy Council. The lawsuit was filed in the Supreme Court of British Columbia, which is a superior court. On discovery, the Clerk relied on section 39 to exclude 51 documents setting out the Treasury Board decision’s rationale. The lawyers challenged the constitutionality of section 39 of the CEA on the basis that it violated section 96 of the Constitution Act, 1867. They submitted that the power to regulate the admissibility of evidence in litigation is part of the “core jurisdiction”, or, more specifically, the “core powers”, of superior courts. The government replied that superior courts did not have the power to require the production of Cabinet confidences in 1867 and, as such, that power was not protected by section 96. The Supreme Court of British Columbia and the SCC both came to this conclusion. The SCC insisted that:
[T]here is a long common law tradition of protecting Cabinet confidences. In Canada, superior courts operated since pre-Confederation without the power to compel Cabinet confidences. Indeed, at the time of Confederation, no court had any jurisdiction regarding actions against the Sovereign.
I will challenge these conclusions on three grounds: the Constitution should be interpreted progressively; superior courts did have the power to overrule PII claims in 1867; and the power to overrule PII claims is constitutional in nature.
a. Progressive Interpretation of the Constitution
In Babcock, the SCC stated that there was “no clear test” to determine the core jurisdiction and powers protected under section 96 of the Constitution Act, 1867. However, as a starting point, when inquiring whether a grant or removal of authority is lawful, the SCC has established that the inquiry shall be into whether superior courts had that authority in 1867. While the SCC has not explained why it is necessary to travel back in time to delineate the core jurisdiction and powers of superior courts, its position is likely based on the desire to respect the framers’ intent. The problem with this “originalist” approach is that there is no evidence that the framers intended the concepts of “core jurisdiction” and “core powers” to stay frozen in time. Moreover, whether that was the framers’ intent or not, originalism is inconsistent with the prevailing approach to constitutional interpretation: the “living tree” doctrine. It is well established that the Constitution must be interpreted in a progressive manner to adapt to changing times. The SCC has often rejected the claim that the Constitution should be interpreted statically:
The “frozen concepts” reasoning runs contrary to one of the most fundamental principles of Canadian constitutional interpretation: that our Constitution is a living tree which, by way of progressive interpretation, accommodates and addresses the realities of modern life.
With the expansion of state activities and the rise of human rights instruments such as the Charter, the role of superior courts has significantly evolved since 1867. Why should their core jurisdiction and powers remain the same? Kent Roach rightly points out that the SCC’s interpretative approach to section 96 has the effect of transforming important separation of powers issues into narrow historical questions: “Section 96 emerges as a weak device to protect the separation of powers.” What would it entail to interpret section 96 progressively? It would entail asking one key question: Is the jurisdiction or power “essential to the administration of justice and the maintenance of the rule of law?” I will show that, whether we ask the question in 1867 or today, the power of superior courts to control the admissibility of evidence—and overrule PII claims—is, and has always been, essential.
b. Longstanding Judicial Power to Overrule PII Claims
Working with the hypothesis that the core jurisdiction and powers of superior courts are frozen in time, we must inquire whether they had the power to overrule PII claims in 1867. In Babcock, the SCC held that they did not, but its reasoning contained two flaws: it mischaracterized the issue; and it conflated Crown immunity and Crown privilege. First, the SCC asked whether superior courts had “the power to compel Cabinet confidences” in 1867. While the characterization of a purported section 96 power should be narrow, it should not be so narrow as to make the inquiry pointless. Given that the organized system of Cabinet records that is currently in place did not exist in 1867, and the issue of their production did not arise, it is pointless to ask whether superior courts could compel “Cabinet confidences” at that time. Rather, we must “search for analogous, not precisely the same, jurisdiction” and focus on the “type of dispute” at issue. The dispute in Babcock pertained to the production of “government documents” in litigation: the specific class of documents sought was immaterial. Second, the SCC stated that “no court had any jurisdiction regarding actions against the Sovereign” in 1867. This is true, but irrelevant. The Crown immunity and Crown privilege doctrines are not one and the same. Although no one could sue the Crown without its consent, as per the Crown immunity doctrine, the courts could still compel the production of government documents in proceedings over which they had jurisdiction, whether or not the Crown was a party to the proceedings. Litigants could subpoena public officials at trial to compel them to testify and produce documents. In such cases, the Crown’s right to object to the disclosure of sensitive information on public policy grounds was not a matter of Crown immunity, it was a matter of Crown privilege or PII.
The relevant authorities do not support the position taken by the SCC in Babcock. Rather, they support the position that, as a matter of law, the courts have historically had the power to inspect and order the production of government documents; however, as a matter of practice, they have traditionally been reluctant to exercise it. The most relevant English case pre-1867 is Beatson v. Skene. In Beatson, the Court of Exchequer ruled on the validity of a PII claim over the production of documents in a defamation suit. While sustaining the claim, Chief Baron Pollock, for the majority, agreed that “cases might arise where the matter would be so clear that the Judge might well ask for [a document] in spite of some official scruples as to producing it,” but added that judges should only do this in “extreme cases”. Baron Martin, in comparison, took a more liberal approach: “[W]henever the Judge is satisfied that the document may be made public without prejudice to the public service, the Judge ought to compel its production, notwithstanding the reluctance of [public officials] to produce it.” In the end, the difference between them was only a matter of degree, as they both agreed that judges could, in some cases, overrule PII claims. Beatson was applied by the Court of Queen’s Bench—in what was then Canada East—in Gugy v. Maguire, the sole relevant Canadian case pre-1867. There, the majority upheld a PII claim in a defamation suit. In doing so, it did not explicitly deny the possibility that a PII claim could be overruled in extreme cases, but did not consider that Gugy was such a case. Justice Mondelet, in dissent, would have overruled the claim. Given that a copy of the document had been made public, he could not see how its production would injure the public interest. To him, Gugy was exactly the type of case warranting judicial intervention.
Before the House of Lords’ decision in Duncan, the English authorities supported the position that the courts had a reserve power to overrule PII claims. In 1888, the High Court asserted the power to privately inspect documents subject to a PII claim to verify that the public interest was the real reason for the claim, but did not exercise it. In 1916, the English Court of Appeal upheld the decision of an application judge who exercised the power to inspect the documents sought before upholding the claim. In 1931, the Judicial Committee of the Privy Council confirmed that the Supreme Court of South Australia, a superior court, had a “reserve” power to inspect and order the production of government documents despite the government’s objection. This decision opened the way to the production of thousands of documents to the plaintiffs. In 1933, the High Court overruled a PII claim after inspecting the documents and concluding that their production would not injure the public interest. The state of the law in 1941 was summed up by Lord Justice MacKinnon as follows: “[T]heoretically the Court has a right to look at the documents, notwithstanding the claim made by the minister, in order to form its own view of the validity of the claim, but, by the practice of the Court, the power will only rarely be exercised.” This historical survey shows that before Duncan, “the judiciary [was] the ultimate arbiter of the public interest.” In 1942, the House of Lords parted ways with this line of authorities by deciding that the courts were bound to accept PII claims made in the proper form. Yet, it mitigated its position by stating that PII claims were only final and conclusive in civil, not criminal, cases. Furthermore, it insisted “it is the judge who is in control of the trial, not the executive” and, as such, “the decision ruling out such documents is the decision of the judge.” Hence, even then, the House of Lords did not entirely deny the courts’ power to overrule PII claims.
Duncan was rejected by the highest courts in Australia, New Zealand, and Canada. The SCC has established that PII claims can be overruled in criminal and civil cases. In a case arising out of Scotland, Lord Radcliffe said that it would be a “very great pity” if the courts “abdicated … a right of control which their predecessors in earlier centuries have been insistent to assert.” Duncan was disapproved of by the English Court of Appeal in 1964, before being overturned by the House of Lords in Conway v. Rimmer. In Conway, the Law Lords did not purport to create a new judicial power to overrule PII claims; they rather re-established an enduring power that had been improperly constrained. Speaking of the PII doctrine, Lord Upjohn said that it was time for the judiciary to “regain its control over the whole of this field of the law.” D.H. Clark noted that Conway “restore[d] to the judiciary … its inherent residual power … to overrule a formally unimpeachable objection made on behalf of the Crown, in the name of the public interest, to the disclosure of documentary or oral evidence in legal proceedings.” When the issue regarding the production of Cabinet documents was first adjudicated in the wake of Conway, the courts held that they had the power to inspect them and order their production. Thus, the position that Canadian courts did not have the power to compel government documents, whatever their nature, before or after 1867, is not supported by the authorities. The fact that the courts have traditionally been reluctant to exercise the power to compel government documents and overrule PII claims, as a matter of practice, did not divest them of that power, as a matter of law.
c. Constitutional Nature of the Judicial Power to Overrule PII Claims
The key issue is whether the judicial power to control the admissibility of evidence, including the power to overrule PII claims, is “essential to the administration of justice and the maintenance of the rule of law.” The case law on this issue is inconsistent. As previously noted, in Carey, the SCC held that “it would be contrary to the constitutional relationship that ought to prevail between the executive and the courts in this country” to deprive the judiciary of the power to inspect and order the production of Cabinet confidences. Yet, in Babcock, the SCC stated that section 39 of the CEA, which deprives the judiciary of the power to inspect and order the production of Cabinet confidences, “does not fundamentally alter or interfere with the relationship between the courts and the other branches of government.” While Carey was decided under the common law and Babcock was decided under statute law, these positions cannot both be accurate. Depriving judges of the power to inspect and order the production of Cabinet confidences either interferes with the proper constitutional relationship between the judiciary and the executive or it does not. Common law and statute law rules must both comply with section 96 of the Constitution Act, 1867. The SCC’s position in Babcock suggests an increased judicial deference to statute law, which is unwarranted when dealing with the core powers of superior courts.
In a seminal article, I.H. Jacob stated that the “essential character of a superior court of law necessarily involves that it should be invested with a power to maintain its authority and to prevent its process being obstructed and abused.” Such powers are the “life-blood” of a superior court as they enable it to “fulfil itself as a court of law.” What is the function of the courts if not to resolve disputes by the interpretation and application of the law? As the law is not applied in a vacuum, to perform this function the courts first need to determine the relevant facts. In our adversarial system, it is the parties’ role to present the relevant oral or documentary evidence to the court. In the proper administration of justice, all relevant evidence should be available to the parties within the framework of the rules of evidence. By exception, the parties can claim some privileges for confidential information or object to the admissibility of unreliable or unduly prejudicial evidence. Yet, to maintain the integrity of the adjudicative process, the final decision on the production or admissibility of evidence is made by the judge. If this kind of decision was left to the parties, it would be in their self-interest to suppress unfavourable evidence. As a consequence, judges would no longer be in control of the adjudicative process and would be unable to prevent abuses. This would, in turn, undermine the public confidence in the administration of justice.
John Wigmore described the courts’ power to assess the admissibility of evidence as an “indestructible judicial function”. Rules preventing courts from fulfilling this function trespass “upon their exclusive province.” The highest courts in common law jurisdictions support Wigmore’s position. In 1954, Justice Rand held that it would be inconsistent with “basic conceptions of our polity” to forbid judges from determining PII claims, as this power is necessary to prevent “executive encroachments upon the administration of justice.” In 1964, Lord Denning stated that PII was a constitutional law matter as it deals with the proper relationship between the respective powers of the executive and the courts. As “guardians of justice”, Lord Denning wrote, judges must have the final word on the production of evidence in court. This point was later confirmed by the House of Lords. In 1974, Chief Justice Burger held that the doctrine of executive privilege should be considered in light of the United States’ “historic commitment to the rule of law” and rejected President Nixon’s claim of absolute executive privilege for the reason that it would “gravely impair the basic function of the courts” under the United States Constitution. Decades earlier, Chief Justice Vinson had said that “[j]udicial control over the evidence in a case cannot be abdicated to the caprice of executive officers” as it would lead to “intolerable abuses”. Associate Chief Justice Gibbs and Justices Stephen and Mason of the High Court of Australia also held that judges should have the exclusive authority to determine which evidence shall be produced in court. In 1982, Justice Wilson of the SCC denied that the executive could be the “arbiter of [its] own immunity,” as this power “is for the courts.” The authorities support the position taken in Carey. Judges, who have the final responsibility for maintaining the rule of law, should also have final responsibility for deciding what evidence should be available to enable them to do justice. In Allan Manson’s words:
[A]n absolute privilege from disclosure and production will always impede the functioning of the courts. If one accepts that the separation of powers thesis is premised on the need to ensure the independent and unimpaired functioning of different branches of government, it is logically inconsistent to argue that the Constitution empowers Parliament to legislate in a way that can, in serious cases, emasculate the judiciary.
In Babcock, the SCC overlooked the wisdom of Carey and the common law when dismissing the claim that section 39 of the CEA violates the core powers of superior courts. While the SCC is free to change its mind, in a culture of justification, it should at least explain why it did so. It should explain why the authorities cited above are mistaken in holding that the power to control the admissibility of evidence and overrule PII claims is constitutional in nature. The explanation should go beyond the issue of whether superior courts had that power in 1867; it must address the substantive question, that is, whether the superior courts’ power to control the admissibility of evidence is essential to the administration of justice and the maintenance of the rule of law. Babcock has given a negative answer to that question. This answer, if taken to its logical extreme, could have major consequences: Parliament could adopt a near-absolute immunity not just over Cabinet confidences, but over all government information, and abolish the right to discovery against the state. If the government could control access to evidence in this way through statute law, citizens’ capacity to hold it to account, and by extension the rule of law, would be undermined. It is troubling that the SCC’s position, could ultimately lead to such an absurd outcome.
2. Superior Courts Cannot Meaningfully Review the Legality of Executive Action
Parliament and the provincial legislatures cannot totally deprive superior courts of their jurisdiction to review the legality of executive action. While they can limit the scope of judicial review through the use of privative clauses, they cannot prevent superior courts from reviewing executive action for jurisdictional errors. In Crevier, the SCC held that judicial review for jurisdictional errors was the “hallmark of a superior court” and was constitutionally protected under section 96 of the Constitution Act, 1867. The rule of law is maintained by ensuring that superior courts have the last word on jurisdiction and that individuals have meaningful remedies to protect themselves from unlawful executive action. If section 39 of the CEA was interpreted to totally exclude judicial review, it would be unlawful. In Babcock, the SCC held that this was not the effect of section 39. While section 39 can be viewed as a “draconian” privative clause, it does not, in the SCC’s view, prevent judicial review for jurisdictional errors. Judges can review Cabinet immunity claims to determine: whether the certified documents fall within the scope of section 39; and whether the certificate was issued in bad faith. As judges cannot inspect the documents, any challenge must be made based on the information available on the face of the certificate or external evidence. Even though “these limitations may have the practical effect of making it difficult to set aside a section 39 certification,” section 39 has not, according to the SCC, “substantially altered the role of the judiciary from their function under the common law regime.”
The issue is whether section 39 of the CEA, as interpreted by the SCC in Babcock, really enables superior courts to review Cabinet immunity claims on jurisdictional grounds. In other words, does it enable meaningful judicial review of executive action, as required by the congruence principle, the theory of law as justification, and the Canadian Constitution? I will show that, contrary to the SCC’s position, it does not. At the outset, it is important to stress that the certification of documents under section 39 is not just a question of fact, as stated by the Federal Court of Appeal in Singh. The certification involves mixed questions of fact and law. The decision-maker must first inspect the documents to assess whether they fall within the legal meaning of the term “Cabinet confidences”, which is partly a matter of statutory interpretation. He or she must then weigh and balance the competing aspects of the public interest to assess whether the documents should be excluded. This requires an analysis of the documents’ degree of relevance and their degree of injury. I will argue that superior courts cannot review whether Cabinet immunity has been claimed mistakenly or abusively without access to the documents and the justification for their protection.
a. Judicial Review of Executive Mistakes
There are three types of bona fide mistakes that the administrative decision-maker could make when issuing a certificate under section 39 of the CEA.
First, the minister or the Clerk could claim the immunity for documents that do not fall within the statutory definition of “Cabinet confidences” under subsection 39(2) because: (1) the documents are not sufficiently connected to the collective decision-making process; or (2) the information they contain has been made public. This first problem arises because subsection 39(2) is open-ended: it does not provide an exhaustive list of documents which contain Cabinet confidences. Moreover, the list of documents provided in subsection 39(2) is very broad: it is not limited to official Cabinet documents; it includes departmental documents that have a tenuous connection to the collective decision-making process. With respect to departmental documents, the descriptions provided in section 39 certificates—namely, the date, title, author, and recipient—may not be sufficient to clearly establish, on the face of the certificate, that the documents contain Cabinet confidences. This is a serious problem given that an increasing number of documents excluded by the government based on Cabinet immunity are departmental documents such as emails, briefing notes, and PowerPoint slides. The second problem relates to the confidentiality of the information. If the information has been made public, it can no longer be protected under section 39. Yet, the government sometimes claims section 39 for documents the contents of which have been made public. Judges cannot know if the information found in certified documents is truly confidential without inspecting them.
Second, the minister or the Clerk could claim the immunity although the public interest requires that the documents be produced in the specific circumstances of the case. The public interest weighing and balancing process is a fundamental part of the analysis. Without access to the “public interest” justification underpinning Cabinet immunity claims, judges cannot know if the decision-maker has reasonably weighed and balanced the competing aspects of the public interest. The decision-maker could underestimate the documents’ degree of relevance and overestimate the degree of injury that would result from production.
Third, the minister or the Clerk could misapply the “discussion paper exception”. Because of modifications to the Cabinet paper system, no discussion paper was disclosed from 1984 to 2003. This exception was revived following the intervention of the Information Commissioner and a Federal Court of Appeal judgment. However, in 2012, new modifications to the Cabinet paper system had the effect of significantly narrowing the scope of this exception. In practice, the application of this exception depends on a careful review of the documents to determine whether they contain: “a corpus of words the purpose of which is to present background explanations, analyses of problems or policy options to [Cabinet] for consideration … in making decisions.” Any part of a document which falls within this definition must be disclosed if the Cabinet decision has been made public or, if not, four years have passed since it was made. But no one can determine if this exception applies based on the description of the documents without reading them. Without inspection powers, judges cannot know if the documents listed in section 39 certificates are subject to the discussion paper exception. This is troubling given that the government has often both circumvented and misapplied this exception in the past.
b. Judicial Review of Executive Abuses of Power
In addition, the minister or the Clerk could claim Cabinet immunity abusively, that is, to thwart a public inquiry or to gain a tactical advantage in litigation. History contains many examples in which PII was used for such improper purposes. Bad faith is rarely apparent on the face of a certificate as the decision-maker can misrepresent the true nature of the documents and the outcome of the public interest assessment. Bad faith can be uncovered through: external evidence; or a judicial inspection of the documents. The first method is problematic as external evidence is very difficult to obtain. It requires one of three scenarios to play out. First, the decision-maker could publicly reveal the true motives underpinning his or her decision without realizing their improper nature. Second, a whistleblower could leak the true motives underpinning the decision. Third, an external body with subpoena power could investigate and find that the government has used PII improperly. These scenarios all involve an element of luck. Technically, the litigant cannot cross-examine the decision-maker to probe his or her motives because a certificate is not an affidavit. The only effective way to ensure that Cabinet immunity is consistently claimed in good faith, in line with the rule of law as justification, is the second: to enable judges to inspect documents to confirm that they contain Cabinet confidences which should be excluded in the public interest.
c. Cabinet Immunity as a Legal Black Hole
As the interpretation and application of section 39 of the CEA is almost exclusively in the government’s hands, there is a risk that Cabinet immunity could be overclaimed. The dangers of executive auto-interpretation are well-documented. Decision-makers are often subject to subtle pressure from their political masters, who wield power over them. The legal analysis behind their decisions may be self-serving or mistaken. The decision-maker could end up exceeding his or her jurisdiction and, without meaningful review, this unlawful conduct would remain undetected. Scrutiny of executive auto-interpretation demands a level of transparency, justification, and intelligibility that section 39 forbids. Usually, when judges review the legality of administrative decisions, they have access to the record of tribunal and the reasons for the decision, but this is not the case under section 39. Contrary to the SCC’s position in Commission des droits de la personne, the problem stems from the design of the legislation, not only the way in which it is applied. Section 39 conceals possible unlawful executive action by blinding judges. By precluding judges from inspecting documents, section 39 prevents the application of the congruence principle. Left in the dark, deprived of the light that would enable them to detect unlawful conduct, judges are ill-equipped to uphold the rule of law. A judge cannot confirm whether a PII claim is valid or not without inspecting the documents at issue; a judge cannot make this confirmation any more than a house inspector can confirm that a house is free of defect without setting foot inside it. Reading a certificate is like looking at a house’s façade: it is a very poor way of assessing the quality of the thing under scrutiny.
In enacting section 39 of the CEA, Parliament has created a legal black hole, that is, a zone which is not controlled by law. To put it bluntly, it has taken “the risk of an abuse of power that lies beyond judicial review.” The SCC should have ruled that Parliament has no constitutional authority to create legal black holes. Under section 96 of the Constitution Act, 1867, it cannot remove from superior courts the core jurisdiction to review executive action for jurisdictional errors. Unfortunately, rather than taking this course of action, the SCC turned a blind eye to section 39. In Babcock, it sanctioned the myth that judges can meaningfully review Cabinet immunity claims without access to the documents and the justification for the claim. If that were true, there would be no difference, from a rule of law perspective, between a relative and a near-absolute immunity. The SCC has denied that section 39 produces a legal black hole. In doing so, it has created something more dangerous than a legal black hole, namely, a legal grey hole. Dyzenhaus explains that a legal grey hole is a zone which appears to be controlled by law, and thus garners the legitimacy of the rule of law, but in fact is not. Section 39 is a legal grey hole in the sense that the government can overclaim Cabinet immunity, without being detected, and still pretend that it is acting in accordance with the rule of law. This is nothing more than a smokescreen given that no one outside the government can confirm whether Cabinet immunity is properly applied.
To sum up, in this section, I have challenged the SCC’s position in Babcock that section 39 of the CEA does not interfere with the proper relationship between the executive and the judicial branches. First, I have argued that the power of superior courts to control the admissibility of evidence in litigation is essential to the administration of justice and the maintenance of the rule of law. By severely curtailing that power, Parliament has undermined the ability of judges to ensure the fairness of the proceedings and to remedy abuses of process. Second, I have argued that section 39 overly restricts the jurisdiction of superior courts to review executive action for jurisdictional errors by preventing them from inspecting certified documents. Without inspection powers, judges cannot assess whether a claim has been made reasonably and in good faith. By creating a zone in which the executive branch is free from judicial scrutiny, section 39 weakens the separation of powers. A judge cannot be a judge, and fulfil his or her constitutional responsibility to maintain the rule of law, without the tools necessary to hold the executive branch accountable.
The objective of this article was to show that section 39 of the CEA violates the rule of law and the provisions of the Constitution. To this end, I have adopted, as a normative framework, the theory of law as justification, which is implicit in the Canadian legal order. The theory of law as justification imposes meaningful constraints on the state and, in contrast to the SCC’s very thin conception of the rule of law, it illuminates the flaws afflicting section 39. I have relied on the theory of law as justification to give substance to the unwritten rule of law principle and guide the interpretation of the relevant provisions of the Constitution. I have built the arguments around two principles: procedural fairness; and the separation of powers. As such, I have examined whether the statutory regime established by Parliament to regulate Cabinet immunity claims violates the duty of procedural fairness and the core, or inherent, jurisdiction and powers of superior courts. I have concluded that it does. Consequently, the statutory regime is not only incompatible with the theory of law as justification, but also unconstitutional.
Procedural fairness is a fundamental legal principle both under the theory of law as justification, the common law, and the Constitution. A government decision to deprive a litigant of relevant evidence in litigation triggers two aspects of the duty of fairness: the litigant’s right to have the matter decided by an independent and impartial decision-maker; and, the litigant’s right to be informed of the reasons for the decision. First, it is trite law that “no one may be judge in his own cause.” Hence, a member of the executive branch should not be allowed to make a final and conclusive decision to exclude relevant evidence in cases where the government is a party. Indeed, a party to the litigation should not have control over the tenure of the decision-maker. In addition, the decision-maker should not decide questions of disclosure if he or she has been involved in the development of the impugned government policy. These situations raise a reasonable apprehension of bias. Second, claims of Cabinet secrecy are not made in a manner consistent with an important aspect of law as justification: the onus of justification. The decision-maker is not currently required to justify why the public interest demands that Cabinet confidences be excluded. Without this information, the litigant and the judge cannot determine whether the decision-maker has properly weighed and balanced the interests of justice and good government. They cannot assess whether the claim is rational, proportional, and reasonable. This lack of transparency prevents meaningful judicial review.
Judicial review is fundamental to the maintenance of a culture of justification and the rule of law in Canada. Section 39 of the CEA impedes judicial review, and the function of superior courts under the separation of powers, as it unduly limits: their core power to control the admissibility of evidence; and their core jurisdiction to review the legality of executive action. First, the authorities support the position that the power to control the admissibility of evidence is essential to the administration of justice and the maintenance of the rule of law. If the parties could decide what evidence is admissible or not, it would lead to grave abuses. Self-interested parties would suppress any evidence that is unfavourable to their case. Judges would no longer be in full control of the judicial process and would be unable to remedy abuses of process. Their capacity to search for the truth would be undermined along with the public confidence in the administration of justice. Second, by depriving superior courts of the power to inspect documents, section 39 undermines their capacity to review executive action for jurisdictional errors. Under normal circumstances, judges cannot detect whether a claim has been made mistakenly or abusively without inspection powers. As such, judges cannot assess if the executive action is, in Fuller’s words, “congruent” to the legal rules. To the extent that it insulates executive action from rule of law constraints, section 39 is a legal black hole. Unfortunately, the SCC has legitimized that hole by denying its existence.
Section 39 of the CEA is the antithesis of the rule of law and the principle of access to justice in Canada. While it seeks to protect a principle that is important to the functioning of the system of responsible government, that is, Cabinet secrecy, it does so in a manner that is inconsistent with the provisions of the Constitution, interpreted in light of the theory of law as justification. Even if the courts concluded, based on positive law considerations, that section 39 does not breach any specific provision of the Big-C Constitution, the controversy would remain. Indeed, under the theory of law as justification, legislation may be considered in breach of the rule of law although it does not clearly breach the Big-C Constitution. In such cases, while the courts cannot strike down the legislation, they should openly acknowledge the rule of law breach in order to deprive the offending legislation of legal legitimacy. This would give rise to an institutional dialogue between the three branches of the state as it would warn Parliament and the government that the legislation should be corrected to comply with the rule of law.
Cabinet secrecy can be protected at the federal level in Canada as well as it is at the provincial level and elsewhere, without depriving litigants of basic procedural fairness and superior courts of their core jurisdiction and powers. Any legal system that incorporates the principles of procedural fairness and the separation of powers in its conception of the rule of law would likewise find the idea of a near-absolute PII immunity objectionable. From a theoretical perspective, to comply with the rule of law, Cabinet immunity claims should be decided by independent and impartial judges who have unfettered access to the justification and the information. This is not to say that judges should never defer to executive decisions to claim Cabinet immunity, but deference should be triggered by the quality of the reasons supporting the decision, not blind submission to the assumed wisdom of public officials. The rule of law rejects absolute rules of secrecy or disclosure in favour of a more contextual assessment in which all the aspects of the public interest are duly considered.
Appendix: Canada Evidence Act, RSC 1985, c C-5
Confidences of the Queen’s Privy Council for Canada
Objection relating to a confidence of the Queen’s Privy Council
39 (1) Where a minister of the Crown or the Clerk of the Privy Council objects to the disclosure of information before a court, person or body with jurisdiction to compel the production of information by certifying in writing that the information constitutes a confidence of the Queen’s Privy Council for Canada, disclosure of the information shall be refused without examination or hearing of the information by the court, person or body.
(2) For the purpose of subsection (1), a confidence of the Queen’s Privy Council for Canada includes, without restricting the generality thereof, information contained in
(a) a memorandum the purpose of which is to present proposals or recommendations to Council;
(b) a discussion paper the purpose of which is to present background explanations, analyses of problems or policy options to Council for consideration by Council in making decisions;
(c) an agendum of Council or a record recording deliberations or decisions of Council;
(d) a record used for or reflecting communications or discussions between ministers of the Crown on matters relating to the making of government decisions or the formulation of government policy;
(e) a record the purpose of which is to brief Ministers of the Crown in relation to matters that are brought before, or are proposed to be brought before, Council or that are the subject of communications or discussions referred to in paragraph (d); and
(f) draft legislation.
Definition of Council
(3) For the purposes of subsection (2), Council means the Queen’s Privy Council for Canada, committees of the Queen’s Privy Council for Canada, Cabinet and committees of Cabinet.
(4) Subsection (1) does not apply in respect of
(a) a confidence of the Queen’s Privy Council for Canada that has been in existence for more than twenty years; or
(b) a discussion paper described in paragraph (2)(b)
(i) if the decisions to which the discussion paper relates have been made public, or
(ii) where the decisions have not been made public, if four years have passed since the decisions were made.
 For an overview of Cabinet immunity under statute law, see Yan Campagnolo, “The History, Law and Practice of Cabinet Immunity in Canada” (2017) 47:2 RGD 239 [Campagnolo, “History of Cabinet Immunity”].
  2 SCR 637 at 654, 35 DLR (4th) 161 [Carey].
 See especially Commission des droits de la personne v Canada (AG),  1 SCR 215 at 228, 134 DLR (3d) 17 [Commission des droits de la personne SCC]; Canada (AG) v Central Cartage Co,  2 FC 641 at 652, 664–65, 71 DLR (4th) 253 (FCA), leave to appeal to SCC refused,  1 SCR vii [Central Cartage]; Singh v Canada (AG),  3 FC 185 at paras 29, 36, 42, 20 Admin LR (3d) 168 (FCA) [Singh]; Babcock v Canada (AG), 2002 SCC 57 at paras 23, 61,  3 SCR 3 [Babcock SCC].
 Babcock SCC, supra note 3 at paras 53–61.
 RSC 1985, c C-5, s 39 [CEA 1985] (reproduced in the Appendix to this article).
 See Canadian Bill of Rights, SC 1960, c 44, s 2(e) [Bill of Rights]. See also Canadian Charter of Rights and Freedoms, Part 1 of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (UK), 1982, c 11, ss 7, 11(d) [Charter].
 See Constitution Act, 1867 (UK), 30 & 31 Vict, c 3, s 96, reprinted in RSC 1985, Appendix II, No 5 [Constitution Act, 1867].
 Paul Craig, “Formal and Substantive Conceptions of the Rule of Law: An Analytical Framework”  Public L 467 at 467. On the distinction between thin and thick conceptions of the rule of law, see also Allan C Hutchinson & Patrick Monahan, “Democracy and the Rule of Law” in Allan C Hutchinson & Patrick Monahan, eds, The Rule of Law: Ideal or Ideology (Toronto: Carswell, 1987) 97 at 100–02; Brian Z Tamanaha, On the Rule of Law: History, Politics, Theory (Cambridge: Cambridge University Press, 2004) at 91–126 [Tamanaha, Rule of Law].
 Supra note 7. The preamble establishes that Canada is to have “a Constitution similar in Principle to that of the United Kingdom.”
 Being Schedule B to the Canada Act 1982 (UK), 1982, c 11 [Constitution Act, 1982]. The preamble establishes that “Canada is founded upon principles that recognize the supremacy of God and the rule of law.”
 Roncarelli v Duplessis,  SCR 121 at 142, 16 DLR (2d) 689, Rand J [Roncarelli].
 Reference Re Secession of Quebec,  2 SCR 217 at para 70, 161 DLR (4th) 385 [Secession Reference].
 Ibid at para 32 (for the principles of federalism, democracy, the rule of law, and respect for minorities); Reference Re Remuneration of Judges of the Provincial Court of Prince Edward Island,  3 SCR 3 at paras 106–07, 163, 150 DLR (4th) 577 [Provincial Judges Reference] (for the judicial independence principle). For a thorough analysis of unwritten constitutional principles, see Robin Elliot, “References, Structural Argumentation and the Organizing Principles of Canada’s Constitution” (2001) 80:1&2 Can Bar Rev 67.
 See Secession Reference, supra note 12 at para 53; Provincial Judges Reference, supra note 13 at paras 93, 314–16.
 See David Mullan, “Not in the Public Interest: Crown Privilege Defined” (1971) 19:9 Chitty’s LJ 289 at 291.
 Re Manitoba Language Rights,  1 SCR 721 at 749, 19 DLR (4th) 1 [Manitoba Language Rights Reference]. The SCC relied on the thought of Joseph Raz, a positivist legal philosopher, to articulate its own conception of the rule of law. See Joseph Raz, The Authority of Law: Essays on Law and Morality (Oxford: Clarendon Press, 1979) at 212–14.
 Manitoba Language Rights Reference, supra note 16 at 748.
 Secession Reference, supra note 12 at para 71. See also Provincial Judges Reference, supra note 13 at para 10.
 See British Columbia (AG) v Christie, 2007 SCC 21 at paras 20–21,  1 SCR 873.
 Mark Carter, “The Rule of Law, Legal Rights in the Charter, and the Supreme Court’s New Positivism” (2008) 33:2 Queen’s LJ 453 at 464.
 In other words, a legal rule is valid if it satisfies the “rule of recognition”. See HLA Hart, The Concept of Law (Oxford: Clarendon Press, 1961) at 92.
 See Lon L Fuller, The Morality of Law, revised ed (New Haven: Yale University Press, 1969) at 46–81 [Fuller, Morality of Law].
 Ibid at 81–91.
 See Tamanaha, Rule of Law, supra note 8 at 119. See also Brian Z Tamanaha, “The History and Elements of the Rule of Law”  2 Sing JLS 232; Jutta Brunnée & Stephen J Toope, Legitimacy and Legality in International Law: An Interactional Account (Cambridge: Cambridge University Press, 2010) at 6–7.
 2005 SCC 49 at paras 57–77,  2 SCR 473 [Imperial Tobacco]. For a forceful criticism of this case, see FC DeCoste, “Smoked: Tradition and the Rule of Law in British Columbia v Imperial Tobacco Ltd” (2006) 24:2 Windsor YB Access Just 327.
 In contrast, in criminal matters, this type of measure would have clearly violated ss 7 and paragraph 11(g) of the Charter, supra note 6.
 See Imperial Tobacco, supra note 25 (“it is difficult to conceive of how the rule of law could be used as a basis for invalidating legislation such as the Act based on its content” at para 59). To support this statement, Justice Major cited Elliot, supra note 13 at 114–115, who argued that none of the three rule of law elements identified by the SCC has the normative potential to limit legislative, as opposed to executive, action.
 Peter W Hogg & Cara F Zwibel, “The Rule of Law in the Supreme Court of Canada” (2005) 55:3 UTLJ 715 at 727; Warren J Newman, “The Principles of the Rule of Law and Parliamentary Sovereignty in Constitutional Theory and Litigation” (2005) 16:2 NJCL 175 at 187.
 See Imperial Tobacco, supra note 25 at para 66.
 Singh, supra note 3 at para 36.
 Supra note 3 at para 56.
 Ibid at para 57.
 See especially Imperial Tobacco, supra note 25 at paras 59–60.
 See Elliot, supra note 13 at 115, 141–42; Hogg & Zwibel, supra note 28 at 723, 727; Newman, supra note 28 at 289–90; Carter, supra note 20 at 457, 485.
 See Manitoba Language Rights Reference, supra note 16 at 754–58.
  2 SCR 606 at 626–27, 643, 93 DLR (4th) 36 [Nova Scotia Pharmaceutical Society].
 Supra note 12 at paras 76, 84–91. See also Constitution Act, 1982, supra note 10, Part V.
 2014 SCC 59 at paras 24–43,  3 SCR 31 [Trial Lawyers Association].
 For a discussion of constitutional constraints of this kind, see David Dyzenhaus, The Constitution of Law: Legality in a Time of Emergency (Cambridge: Cambridge University Press, 2006) at 6, 201 [Dyzenhaus, Constitution of Law].
 Concerning institutional dialogue between the three branches of the state, see generally Kent Roach, The Supreme Court on Trial: Judicial Activism or Democratic Dialogue, revised ed (Toronto: Irwin Law, 2016).
 Etienne Mureinik, “A Bridge to Where? Introducing the Interim Bill of Rights” (1994) 10:1 SAJHR 31 at 32. See also David Dyzenhaus, “Law as Justification: Etienne Mureinik’s Conception of Legal Culture” (1998) 14:1 SAJHR 11 [Dyzenhaus, “Law as Justification”].
 Dyzenhaus, Constitution of Law, supra note 39 at 3.
 Ibid at 139.
 Fuller, Morality of Law, supra note 22 at 46–81.
 See David Dyzenhaus, “Process and Substance as Aspects of the Public Law Form” (2015) 74:2 Cambridge LJ 284 at 294, 297 [Dyzenhaus, “Process and Substance”].
 See Fuller, Morality of Law, supra note 22 at 81–91.
 See Dyzenhaus, “Process and Substance”, supra note 45 at 294.
 Dyzenhaus, Constitution of Law, supra note 39 at 12.
 David Dyzenhaus, “Deference, Security and Human Rights” in Benjamin J Goold & Liora Lazarus, eds, Security and Human Rights (Oxford: Hart, 2007) 125 at 138 [Dyzenhaus, “Deference, Security and Human Rights”].
 Dyzenhaus, “Process and Substance”, supra note 45 at 297.
 Dyzenhaus, Constitution of Law, supra note 39 at 13.
 David Dyzenhaus, “The Politics of Deference: Judicial Review and Democracy” in Michael Taggart, ed, The Province of Administrative Law (Oxford: Hart, 1997) 279 at 307 [Dyzenhaus, “Politics of Deference”].
 See generally David Dyzenhaus, “Dignity in Administrative Law: Judicial Deference in a Culture of Justification” (2012) 17:1 Rev Const Stud 87.
 See generally David Dyzenhaus, “Preventive Justice and the Rule-of-Law Project” in Andrew Ashworth, Lucia Zedner & Patrick Tomlin, eds, Prevention and the Limits of the Criminal Law (Oxford: Oxford University Press, 2013) 91 [Dyzenhaus, “Rule-of-Law Project”].
 Dyzenhaus gives the following account of the rule of law requirements: “[L]egislation must be capable of being interpreted in such a way that it can be enforced in accordance with the requirements of due process: the officials who implement it can comply with a duty to act fairly, reasonably and in a fashion that respects the equality of all those who are subject to the law.” See Dyzenhaus, Constitution of Law, supra note 39 at 12–13.
 See generally Dyzenhaus, “Process and Substance”, supra note 45.
 See Dyzenhaus, Constitution of Law, supra note 39 at 10–11 (noting that “judicial reasoning remains the main site for articulating the principles of the rule-of-law project” at 11). Dyzenhaus uses the metaphor of judges as “weathermen” to describe this role, that is, the “role of alerting the Commonwealth to storm clouds on the horizon when the rule of law which secures the fabric of civil society is put under strain” (ibid at 201).
 Ibid at 55. See also Lon L Fuller, “Positivism and Fidelity to Law—A Reply to Professor Hart” (1958) 71:4 Harv L Rev 630.
 “Privative clauses” are statutory provisions seeking to limit or preclude judicial review of executive action.
 The term “legal black hole” was first used in R (Abbasi) v Secretary of State for Foreign Affairs,  EWCA Civ 1598, and then by Johan Steyn in an article entitled “Guantanamo Bay: The Legal Black Hole” (2004) 53:1 ICLQ 1.
 See Dyzenhaus, Constitution of Law, supra note 39 at 3, 42, 50.
 See Fuller, Morality of Law, supra note 22 at 81–83.
 Lon L Fuller, “The Forms and Limits of Adjudication” (1978) 92:2 Harv L Rev 353 at 369.
 Ibid at 357, 369. See also Dyzenhaus, “Rule-of-Law Project”, supra note 54 at 96.
 Dyzenhaus, “Process and Substance”, supra note 45 at 303.
 See David Dyzenhaus, “Proportionality and Deference in a Culture of Justification” in Grant Huscroft, Bradley W Miller & Grégoire Webber, eds, Proportionality and the Rule of Law: Rights, Justification, Reasoning (New York: Cambridge University Press, 2014) 234 at 242, 254 [Dyzenhaus, “Proportionality and Deference”].
 Dyzenhaus, “Process and Substance”, supra note 45 at 304.
 Bernard Williams, In the Beginning was the Deed: Realism and Moralism in Political Argument, ed by Geoffrey Hawthorn (Princeton: Princeton University Press, 2005) at 4 [emphasis in original]. See also Dyzenhaus, “Process and Substance”, supra note 45 at 304–06.
 Dyzenhaus, Constitution of Law, supra note 39 at 139–40.
 Dyzenhaus, “Law as Justification”, supra note 41 at 27–28.
 See Dunsmuir v New Brunswick, 2008 SCC 9 at paras 46–47,  1 SCR 190 [Dunsmuir]; Doré v Barreau du Québec, 2012 SCC 12 at paras 55–58,  1 SCR 395; Loyola High School v Quebec (AG), 2015 SCC 12 at paras 32, 35–42,  1 SCR 613.
 See Dyzenhaus, “Proportionality and Deference”, supra note 66 at 255.
 Dyzenhaus, “Deference, Security and Human Rights”, supra note 49 at 131. “Deference as respect” must be distinguished from “deference as submission”, that is, the blind acceptance of executive interpretation of the law by judges based on legislative intention. See Dyzenhaus, “Politics of Deference,” supra note 52 at 303–04.
 The PII doctrine empowers the government to object to the disclosure of relevant yet sensitive information in the context of litigation on the basis that such disclosure would be injurious to the public interest.
 See Constitution Act, 1867, supra note 7, s 37; Charter, supra note 6, s 3.
 See Bill of Rights, supra note 6, ss 1–2; Charter, supra note 6, ss 2, 6–12, 15.
 See Provincial Judges Reference, supra note 13 at para 104; Secession Reference, supra note 12 at para 53. For an application of the unwritten principle of democracy to fill in the gap regarding access to government information, see Vincent Kazmierski, “Something to Talk About: Is There a Charter Right to Access Government Information?” (2008) 31:2 Dal LJ 351 at 372–92.
 Baker v Canada (Minister of Citizenship and Immigration),  2 SCR 817 at para 56, 174 DLR (4th) 193 [Baker].
 See Charter, supra note 6, s 15.
 Ibid, s 7. See also Nova Scotia Pharmaceutical Society, supra note 36; Ontario v Canadian Pacific Ltd,  2 SCR 1031, 24 OR (3d) 454.
 See Publication of Statutes Act, RSC 1985, c S-21.
 See Charter, supra note 6, ss 11(g), 11(i). See also R v KRJ, 2016 SCC 31,  1 SCR 906 at paras 20–27.
 See Constitution Act, 1867, supra note 7, s 96. See also Roncarelli, supra note 11 at 142; David Dyzenhaus, “The Deep Structure of Roncarelli v Duplessis” (2004) 53 UNBLJ 111.
 See Fraser v Canada (Public Service Staff Relations Board),  2 SCR 455 at 469–70, 23 DLR (4th) 122 [Fraser]; Dunsmuir, supra note 71 (discussing the power of courts to strike down unlawful administrative action at paras 27–33); Constitution Act, 1982, supra note 10, s 52.
 See Provincial Judges Reference, supra note 13 at paras 82–109.
 See Crevier v Quebec (AG),  2 SCR 220 at 237–38, 127 DLR (3d) 1 [Crevier].
 See Reference re Secession of Quebec, supra note 12 at para 72.
 See e.g. Nicholson v Haldimand-Norfolk Regional Police Commissioners,  1 SCR 311 at 324, 88 DLR (3d) 671; Cardinal v Director of Kent Institution,  2 SCR 643 at 653, 24 DLR (4th) 44; Knight v Indian Head School Division No 19,  1 SCR 653 at 677, 69 DLR (4th) 489; Baker, supra note 78 at para 28; Dunsmuir, supra note 71 at paras 79, 90.
 Supra note 6, ss 1(a), 2(e), 2(f).
 Supra note 6, ss 7, 11(d).
 See R v Oakes  1 SCR 103 at 139, 26 DLR (4th) 200. Proportionality is also paramount in a culture of justification. See Moshe Cohen-Eliya & Iddo Porat, “Proportionality and the Culture of Justification” (2011) 59:2 Am J Comp L 463 (“[a]t its core, a culture of justification requires that governments should provide substantive justification for all their actions … justification in terms of the rationality and reasonableness of every action and the trade-offs that every action necessarily involves, i.e., in terms of proportionality” at 466–67).
 See Reference Re BC Motor Vehicle Act,  2 SCR 486 at 515, 24 DLR (4th) 536.
 See Baker, supra note 78 at para 43.
 Dunsmuir, supra note 71 at para 47.
 Baker, supra note 78 at para 65 (concerning “deference as respect” in particular).
 For an analysis of Cabinet immunity under the common law, see Yan Campagnolo, “A Rational Approach to Cabinet Immunity under the Common law” (2017) 55:1 Alta L Rev 43 [Campagnolo, “Common Law”].
  AC 624 at 641–43,  1 All ER 587 (HL (Eng)) [Duncan UKHL].
 See Ellis v Home Office,  2 QB 135,  2 All ER 149, (CA) [Ellis]. See also Campagnolo, “Common Law”, supra note 96 at 49–51.
  AC 910 at 951–53, 958,  1 All ER 874 (HL (Eng)) [Conway].
 See Burmah Oil Co Ltd v Bank of England (1979),  AC 1090 at 1117,  3 All ER 700 (HL (Eng)) [Burmah Oil]; Air Canada v Secretary of State for Trade,  2 AC 394 at 434–35, 439,  1 All ER 910 (HL (Eng)) [Air Canada]; Commonwealth v Northern Land Council (1993), 176 CLR 604 at 619–20 (HCA) [Northern Land Council].
 TRS Allan, “Abuse of Power and Public Interest Immunity: Justice, Rights and Truth” (1985) 101:2 Law Q Rev 200 at 206. See also TRS Allan, “Discovery of Cabinet Documents: the Northern Land Council Case” (1992) 14:2 Sydney L Rev 230 at 236–39.
 See Smallwood v Sparling,  2 SCR 686 at 703, 141 DLR (3d) 395 [Smallwood]; Carey, supra note 2 at 678, 681–83; Fletcher Timber Ltd v Attorney-General,  1 NZLR 290 at 295, 301, 305, 308 (NZ CA) [Fletcher Timber].
 See Federal Court Act, RSC 1970, c 10 (2nd Supp), s 41(2); Canada Evidence Act, RSC 1970, c E-10, s 36.3 [CEA 1970], as amended by An Act to enact the Access to Information Act and the Privacy Act, to amend the Federal Court Act and the Canada Evidence Act, and to amend certain other Acts in consequence thereof, SC 1980–81–82–83, c 111, s 4.
 See Commission des droits de la personne c Canada (PG),  RJQ 67 at 74, 93 DLR (3d) 562 (QCCA) [Commission des droits de la personne QCCA]; Ouvrage de raffinage de métaux Dominion Ltée c Énergie Atomique du Canada Ltée,  RJQ 2232 at 2238, JQ no 2680 (QL) (QC Sup Ct); Central Cartage, supra note 3 at 661–66; Wedge v Canada (AG) (1995), 102 FTR 311 para 13, 34 Admin LR (2d) 237 (FC).
 Duke v R,  SCR 917 at 923, 28 DLR (3d) 129.
 Supra note 3 at 664–65 [footnotes omitted].
 See 2747-3174 Québec Inc v Quebec (Régie des permis d’alcool),  3 SCR 919 at para 45, 140 DLR (4th) 577 [2747-3174 Québec Inc].
 R v Sussex Justices (1923),  1 KB 256 at 259,  All ER Rep 233, Lord Hewart CJ.
 See Valente v R,  2 SCR 673 at 685, 52 OR (2d) 779 [Valente]; Wewaykum Indian Band v Canada, 2003 SCC 45 at paras 57–58,  2 SCR 259; Yukon Francophone School Board, Education Area #23 v Yukon (AG), 2015 SCC 25 at paras 20–23,  2 SCR 282.
 Committee for Justice and Liberty v National Energy Board,  1 SCR 369 at 394, 68 DLR (3d) 716.
 R v Lippé,  2 SCR 114 at 144, 61 CCC (3d) 127 [Lippé, emphasis in original]. The SCC confirmed the applicability of this test in the administrative context in Canadian Pacific Ltd v Matsqui Indian Band,  1 SCR 3 at para 67, 122 DLR (4th) 129 [Matsqui Indian Band].
 See Lippé, supra note 111 at 139.
 R v Beauregard,  2 SCR 56 at 69, 30 DLR (4th) 481.
 See Valente, supra note 109 at 685.
 Ibid at 697–98. The government cannot dismiss a judge because it does not like his or her decisions. See Provincial Judges Reference, supra note 13 at para 115; Re Therrien, 2001 SCC 35 at para 66,  2 SCR 3.
 See Valente, supra note 109 at 704. The government cannot alter judges’ salaries for arbitrary reasons in a manner that could affect judicial independence, such as discontent with decisions rendered. In addition, judges’ salaries must be sufficient to keep them from seeking alternative sources of income. See Provincial Judges Reference, supra note 13 at para 222.
 See Valente, supra note 109 at 708–12. The government cannot decide which judge should hear a particular case (administrative independence) or how a case should be decided (adjudicative independence).
 See Matsqui Indian Band, supra note 111 at paras 80, 83–84.
 See Ocean Port Hotel Ltd v British Columbia (General Manager, Liquor Control and Licensing Branch), 2001 SCC 52 at paras 24, 32,  2 SCR 781 [Ocean Port Hotel] (discussing how administrative tribunals span the “constitutional divide” between executive and judicial branches).
 See Ann Chaplin, “Travelling in Constitutional Circles: The Paradox of Tribunal Independence” (2016) 36:1 NJCL 73 at 109–10.
 Ibid at 107, 110. See also Matsqui Indian Band, supra note 111 at paras 93–99.
 That is, whether the documents contain Cabinet confidences within the meaning of subsection 39(2) of the CEA 1985, supra note 5, and whether the public interest requires that they be protected.
 Under s 39 of the CEA 1985, supra note 5, the minister or the Clerk is not merely objecting to the production of Cabinet confidences and letting the judge decide the matter. Rather, he or she is deciding the matter in a final and conclusive manner (and there is no appeal procedure built into s 39). The power to exclude evidence in litigation is judicial in nature.
 See Chaplin, supra note 120 at 107.
 See Canada, Commission of Inquiry into the Sponsorship Program & Advertising Activities, Restoring Accountability: Recommendations (Ottawa: Public Works and Government Services Canada, 2006), ch 8; ADP Heaney, “Mackenzie King and the Cabinet Secretariat” (1967) 10:3 Can Public Administration 366 (describing the Clerk as “the servant of the Prime Minister, as the Prime Minister is the ‘master’ … of Cabinet business” at 373).
 For the public interest rationales underpinning Cabinet secrecy, see Yan Campagnolo, “The Political Legitimacy of Cabinet Secrecy” (2017) 51:1 RJTUM 51 at 66–72 [Campagnolo, “Legitimacy of Cabinet Secrecy”].
 See e.g. Air Canada, supra note 100 at 430; Environmental Defence Society Inc v South Pacific Aluminium Ltd (No 2),  1 NZLR 153 at 156–57 (NZ CA).
 See e.g. Burmah Oil, supra note 100 at 1108; Fletcher Timber, supra note 102 at 291; Carey, supra note 2; Northern Land Council, supra note 100 at 619–620; Babcock SCC, supra note 3.
 See e.g. Canada, Commission of Inquiry into the Sponsorship Program and Advertising Activities, Who is Responsible? Summary (Ottawa: Public Works and Government Services Canada, 2005) at 2–3.
 See Christopher Berzins, “Crown Privilege: A Troubled Exclusionary Rule of Evidence” (1984) 10:1 Queen’s LJ 135 at 161, 163.
 2747-3174 Québec Inc, supra note 107 at para 54.
 See Ocean Port Hotel, supra note 119 at para 24.
 See Bell Canada v Canadian Telephone Employees Association, 2003 SCC 36 at para 28,  1 SCR 884.
 See 2747-3174 Québec Inc, supra note 107 at paras 67–68. The SCC’s decision suggests that, to meet the quasi-constitutional requirement of independence, the decision-maker must at least have been appointed for a fixed term and must only be removable for cause. He or she must thus have some security of tenure.
 See MacBain v Lederman,  1 FC 856 at 880–81, 22 DLR (4th) 119 (FCA). See also Hassouna v Canada (Minister of Citizenship and Immigration), 2017 FC 473 at para 195, 20 Admin LR (6th) 207.
 S 7 of the Charter, supra note 6, would also apply in any proceeding, criminal or not, where the “right to life, liberty and security of the person” is engaged, such as extradition proceedings. See United States v Burns, 2001 SCC 7 at para 59,  1 SCR 283; Kindler v Canada (Minister of Justice),  2 SCR 779 at 831, 84 DLR (4th) 438.
  3 SCR 326 at 336, 83 Alta LR (2d) 193 (in criminal cases). See also Charkaoui v Canada (Citizenship and Immigration), 2007 SCC 9 at para 61,  1 SCR 350 (in cases where a person’s life, liberty and security is at stake); Hamish Stewart, Fundamental Justice: Section 7 of the Canadian Charter of Rights and Freedoms (Toronto: Irwin Law, 2012) at 249–59.
 See Suresh v Canada (Minister of Citizenship and Immigration), 2002 SCC 1 at para 127,  1 SCR 3.
 See Canadian Assn of Regulated Importers v Canada (AG),  2 FC 130 at 140–41, 87 DLR (4th) 730 (FCA).
 For example, the disclosure of Cabinet confidences was authorized in the course of the criminal prosecutions against former Progressive Conservative minister André Bissonnette (see Privy Council, Order in Council PC 1987-2284 (6 November 1987)) and former Liberal minister John Munro (see Privy Council, Orders in Council PC 1990-2228–30 (11 October 1990)). Based on these precedents, the government should authorize the disclosure of the Cabinet confidences that are relevant to the case of Vice-Admiral Mark Norman, who is currently being prosecuted for breach of trust pursuant to s 122 of the Criminal Code, RSC 1985, c C-46.
 See David J Mullan, Administrative Law (Toronto: Irwin Law, 2001) at 306–18 [Mullan, Administrative Law].
 See e.g. Public Service Board of New South Wales v Osmond (1986), 159 CLR 656 at 668 (HCA).
 See Northwestern Utilities Ltd v Edmonton (City of),  1 SCR 684 at 706, 89 DLR (3d) 161.
 See Baker, supra note 78 at para 39.
 See National Corn Growers Assn v Canada (Import Tribunal),  2 SCR 1324 at 1383, 75 DLR (4th) 449, Gonthier J. See generally Baker, supra note 78.
 Supra note 78 at para 43.
 Ibid at paras 40, 44.
 See Wall v Independent Police Review Director, 2013 ONSC 3312 at para 46, 362 DLR (4th) 687 (Ont Div Ct), aff’d 2014 ONCA 884, 123 OR (3d) 574. See also Williams, supra note 68 and accompanying text.
 Gray v Ontario (Disability Support Program, Director) (2002), 59 OR (3d) 364 at para 22, 212 DLR (4th) 353 (CA) citing VIA Rail Canada Inc v National Transportation Agency,  2 FC 25 at paras 16–22, 193 DLR (4th) 357 [Via Rail].
 See Via Rail, supra note 150 at para 22.
 See Mullan, Administrative Law, supra note 141 at 314–15.
 See R v REM, 2008 SCC 51 at paras 17, 35,  3 SCR 3.
 See Dunsmuir, supra note 71 at para 47.
 See generally Société des services Ozanam inc v Québec (Commission municipale),  RJQ 364, 1994 CanLII 6507 (Sup Ct); Future Inns Canada Inc v Nova Scotia (Labour Relations Board) (1997), 160 NSR (2d) 241, 4 Admin LR (3d) 248 (CA); Canadian Union of Public Employees, Local 301 v Montreal (City),  1 SCR 793 at paras 72–87, 8 Admin LR (3d) 89.
 See e.g. Robinson v South Australia (State of) (No 2),  AC 704 at 721–22,  All ER Rep 333 (PC) [Robinson].
 Supra note 3 at para 22.
 Ibid at para 28.
 See e.g. Pelletier v Canada (AG), 2005 FCA 118 at para 12,  3 FCR 317 [Pelletier].
 See Cooper v Wandsworth (Board of Works) (1863), 143 ER 414 at 420, 14 CBNS 180 (CP).
 See Fletcher Timber, supra note 102 at 295; Campagnolo, “Common Law”, supra note 96 at 73–74.
 See generally Carey, supra note 2.
 For a similar argument, see Kent Roach, “Constitutional and Common Law Dialogues between the Supreme Court and Canadian Legislatures” (2001) 80:1&2 Can Bar Rev 481 at 523.
 Litigants and judges need to know whether the certified documents are relevant (considering their probative value and their materiality to the issues in dispute) and sensitive (considering the nature of the information they reveal—that is, core or noncore secrets—and the timing of their production before or after a decision has been made public). This information can be provided without breaching any legitimate Cabinet confidence. By analogy, the government is already bound to provide such information when claiming PII under ss 37 and 38 of the CEA 1985, supra note 5.
 See Newfoundland and Labrador Nurses’ Union v Newfoundland and Labrador (Treasury Board), 2011 SCC 62,  3 SCR 708, Abella J (“[i]t strikes me as an unhelpful elaboration on Baker to suggest that alleged deficiencies or flaws in the reasons fall under the category of a breach of the duty of procedural fairness and that they are subject to a correctness review” at para 21).
 Ibid (“the breach of a duty of procedural fairness is an error in law. Where there are no reasons in circumstances where they are required, there is nothing to review. But where, as here, there are reasons, there is no such breach. Any challenge to the reasoning/result of the decision should therefore be made within the reasonableness analysis” at para 22 [emphasis in original]).
 See e.g. Alberta (Information and Privacy Commissioner) v Alberta Teachers’ Association, 2011 SCC 61 at paras 53–54,  3 SCR 654.
 It would be inappropriate for the court to order the production of the documents without first giving the government the opportunity to justify its decision by providing its public interest assessment. Ibid at para 55.
 Peter W Hogg, Constitutional Law of Canada, student ed (Toronto: Thomson Reuters, 2017) at 7.3(a) [Hogg, Constitutional Law].
 Singh, supra note 3 at para 28.
 See Fraser, supra note 84 at 469–70; Wells v Newfoundland,  3 SCR 199 at paras 51–55, 180 Nfld & PEIR 269.
 S 96 of the Constitution Act, 1867, supra note 7, seeks to promote national unity by establishing courts of general jurisdiction whose members are appointed by the federal government.
 See Re Residential Tenancies Act, 1979,  1 SCR 714 at 734–36, 123 DLR (3d) 555.
 See MacMillan Bloedel Ltd v Simpson,  4 SCR 725 at para 27, 130 DLR (4th) 385 [MacMillan Bloedel].
 See Hogg, Constitutional Law, supra note 170 at 7.3(b).
 See MacMillan Bloedel, supra note 175 at para 15.
 See generally Stephen G Linstead, “The Law of Crown Privilege in Canada and Elsewhere: Part 1” (1968) 3:1 Ottawa L Rev 79 [Linstead, “Crown Privilege Part 1”].
 Provincial Judges Reference, supra note 13 at para 88.
 MacMillan Bloedel, supra note 175 at para 37.
 Ibid at para 38.
 See Rosara Joseph, “Inherent Jurisdiction and Inherent Powers in New Zealand” (2005) 11 Canterbury L Rev 220 at 221. According to Joseph, the core jurisdiction of superior courts includes their jurisdiction over parens patrie, judicial review, and punishment for contempt (ibid at 225–28).
 Connelly v Director of Public Prosecutions,  AC 1254 at 1301,  2 All ER 401 (HL (Eng)).
 See Joseph, supra note 182 at 220–21. According to Joseph, the core powers of the courts include their powers to: regulate their own procedure; ensure the fairness in trial and investigative procedures; and prevent abuse of their processes (ibid at 234–38).
 Supra note 38.
 Ibid at para 24.
 Ibid at paras 38–40.
 Ibid at para 40.
 See e.g. Singh, supra note 3 at para 17; Babcock SCC, supra note 3 at para 55.
 Supra note 3 at 228.
 See David J Mullan, “Developments in Administrative Law: The 1981–1982 Term” (1983) 5 SCLR 1 at 12–14.
 See Babcock SCC, supra note 3 at paras 53–61.
 Babcock v Canada (AG), 176 DLR (4th) 417 at 444, 70 BCLR (3d) 128 (BCSC); Babcock SCC, supra note 3 at para 60. At the British Columbia Court of Appeal, MacKenzie JA suggested that s 39 was contrary to the constitutional relationship that should prevail between the judiciary and the executive, but decided the appeal on another basis. See Babcock v Canada (AG), 2000 BCCA 348 at para 14, 188 DLR (4th) 678.
 Supra note 3 at para 60.
 Ibid at para 59.
 See Sobeys Stores Ltd v Yeomans,  1 SCR 238 at 263, 57 DLR (4th) 1 [Sobeys]. See also Léonid Sirota & Benjamin Oliphant, “Originalist Reasoning in Canadian Constitutional Jurisprudence” (2017) 50:2 UBC L Rev 505 at 531–33.
 In contrast, before being modified by the Imperial Parliament in 1875, s 18 of the Constitution Act, 1867, supra note 7, expressly provided that the privileges of the Canadian House of Commons and Senate could not exceed the privileges that the British House of Commons enjoyed and exercised in 1867.
 The doctrine was relied upon, inter alia, to recognize women’s right to be appointed to the Senate. See Edwards v Canada (AG) (1929),  AC 124 at 136,  1 DLR 98 (PC). It was also relied upon to confirm same-sex couples’ right to have a civil marriage. See Reference Re Same-Sex Marriage, 2004 SCC 79 at paras 22–29,  3 SCR 698.
 Ibid at para 22. The living tree doctrine has also been applied in the context of division of powers disputes between the federal government and the provinces. See Reference Re Employment Insurance Act (Can) ss 22 and 23, 2005 SCC 56 at para 9,  2 SCR 669.
 Kent Roach, “‘Constitutional Chicken’: National Security Confidentiality and Terrorism Prosecutions after R v Ahmad” (2011) 54 SCLR (2d) 357 at 383.
 MacMillan Bloedel, supra note 175 at para 38. A similar “necessity test” is employed to identify the privileges of the House of Commons and the Senate. See Canada (House of Commons) v Vaid, 2005 SCC 30 at paras 41–46,  1 SCR 667.
 Babcock SCC, supra note 3 at para 60. Similarly, the Federal Court of Appeal mischaracterized the issue in Singh, supra note 3 at para 42, when it observed that: “The issuance of [section 39] certificates cannot in my view be characterized as a traditional and necessary function of a superior court of a kind contemplated in 1867.” Obviously, it has never been a function of superior courts to claim Cabinet immunity on behalf of the government. Their function has been to decide whether such claims should be sustained or not, in view of the public interest.
 See Sobeys, supra note 196 at 254.
 The organized system of Cabinet records was established following the creation of the Cabinet secretariat in the United Kingdom (1916) and Canada (1940). See Campagnolo, “Legitimacy of Cabinet Secrecy”, supra note 126 at 72–76.
 Sobeys, supra note 196 at 255.
 Babcock SCC, supra note 3 at para 60.
 Before the adoption of the Crown Proceedings Act, 1947 (UK), 10 & 11 Geo VI, c 44, the Crown could not be sued in court. To initiate proceedings against the Crown, a litigant first had to obtain the Crown’s consent through a Petition of Right. When Crown immunity was abolished in 1947, the Crown’s right to object to the disclosure of government documents on public policy grounds (Crown privilege) was explicitly preserved (ibid, s 28). See TG Cooper, Crown Privilege (Aurora: Canada Law Book, 1990) at 8–16.
 Before the House of Lords’ decision in Duncan UKHL, supra note 97, there was no consistent line of authority preventing judges from overruling PII claims. See Linstead, “Crown Privilege: Part 1”, supra note 178 at 98; Louise McIsaac, Laura Campbell & Paul Lordon, “Crown Information Law” in Paul Lordon, ed, Crown Law (Toronto: Butterworths, 1991) 513 at 516.
 See DH Clark, “Administrative Control of Judicial Action: The Authority of Duncan v. Cammell Laird” (1967) 30:5 Mod L Rev 489 at 500–01 [Clark, “Administrative Control”].
 (1860), 157 ER 1415, 5 H & N 838, [Beatson cited to ER] (Ex Ct). The case law before Beatson did not directly address this issue. See Linstead, “Crown Privilege: Part 1”, supra note 178 at 93.
 Beatson, supra note 210 at 1422.
 (1863), 13 LCR 33 at 53–64, Aylwin J (QB App).
 Ibid at 38–39, 45–48.
 See Hennessy v Wright (1888), 21 QBD 509 at 515.
 See Asiatic Petroleum Co Ltd v Anglo-Persian Oil Co Ltd,  1 KB 822, 60 Solicitors’ J & Weekly Reporter 417 (CA).
 See Robinson, supra note 156 at 716–17, 722–23.
 See Spigelman v Hocken (1933), 150 Law Times Reports 256 at 262,  1 TLR 87.
 Duncan v Cammell Laird & Co Ltd,  1 KB 640 at 644,  1 All ER 437 (CA).
 Clark, “Administrative Control”, supra note 209 at 504.
 Duncan UKHL, supra note 97 at 642.
 See Campagnolo, “Common Law”, supra note 96 at 51.
 See R v Snider,  SCR 479, 4 DLR (1st) 483 [Snider cited to SCR].
 See Gagnon v (Quebec) Commission des valeurs mobilières,  SCR 73, 50 DLR (2d) 329.
 Glasgow (City of) v Central Land Board (1955),  Sess Cas 1 at 19 (HL (Scot)).
 See Re Grosvenor Hotel, London (No 2),  3 All ER 354 at 361–62,  3 WLR 992 (CA) [Grosvenor Hotel].
 Conway, supra note 99 at 994.
 DH Clark, “The Last Word on the Last Word” (1969) 32:2 Mod L Rev 142 at 142.
 See Sankey v Whitlam (1978), 142 CLR 1 at 41–42, 63–64, 96, 21 ALR 505 (HCA) [Whitlam] (Australia); Burmah Oil, supra note 100 at 1113, 1134, 1144 (England); Fletcher Timber, supra note 102 at 296–97, 303, 306–07 (New Zealand); Carey, supra note 2 at 670 (Canada).
 MacMillan Bloedel, supra note 175 at para 38.
 Supra note 2 at 654.
 Supra note 3 at para 57.
 IH Jacob, “The Inherent Jurisdiction of the Court” (1970) 23:1 Current Leg Probs 23 at 27.
 Trial Lawyers Association, supra note 38 at paras 32–33.
 See R v Carosella,  1 SCR 80, 142 DLR (4th) 595 (“[u]nder our system, which is governed by the rule of law, decisions as to which evidence is to be produced or admitted is for the courts” at para 56).
 See John Henry Wigmore, A Treatise on the System of Evidence in Trials at Common Law Including the Statutes and Judicial Decisions of All Jurisdictions of the United States (Boston: Little, Brown, and Company, 1904) vol 2 at § 2376.
 Ibid at § 1353.
 Snider, supra note 223 at 485.
 See Grosvenor Hotel, supra note 226 at 360–62.
 See generally Conway, supra note 99.
 United States v Nixon, 418 US 683 at 708, 712, 94 S Ct 3090 (1974) [Nixon].
 United States v Reynolds, 345 US 1 at 8–10, 73 S Ct 528 (1953).
 See Whitlam, supra note 229 at 38, 58–60, 95–96.
 Smallwood, supra note 102 at 708.
 I am paraphrasing a statement made by Lord Woolf in R v Chief Constable of the West Midlands Police,  3 All ER 420 at 438,  3 WLR 433 (HL (Eng)). While Parliament can regulate Cabinet immunity by way of statute, in doing so, it must preserve the judge’s discretionary power to inspect Cabinet documents and order their production when the interest of justice outweighs the interest of good government.
 Allan Manson, “Questions of Privilege and Openness: Proposed Search and Seizure Reforms” (1984) 29:4 McGill LJ 651 at 697.
 See Stephen G Linstead, “The Law of Crown Privilege in Canada and Elsewhere: Part 2” (1969) 3:2 Ottawa L Rev 449 at 462. See also Canada (Attorney General) v Thouin, 2017 SCC 46,  2 SCR 184.
 See Service Employees’ International Union, Local No 333 v Nipawin District Staff Nurses Association,  1 SCR 382, 41 DLR (3d) 6 [Service Employees’ cited to SCR]. The concept of “jurisdictional error” includes “acting in bad faith, basing the decision on extraneous matters, failing to take relevant factors into account, breaching the provisions of natural justice or misinterpreting [the relevant statutory] provisions” (ibid at 389).
 Crevier, supra note 86 at 237. See also MacMillan Bloedel, supra note 175 at para 35; Cooper v Canada (Human Rights Commission),  3 SCR 854 at para 11, 140 DLR (4th) 193; Dunsmuir, supra note 71 at paras 31, 52; WR Lederman, “The Independence of the Judiciary” (1956) 34:7 Can Bar Rev 1139 at 1174.
 See Dunsmuir, supra note 71 at para 30.
 See Babcock SCC, supra note 3 (“the certification of the Clerk or minister under s. 39(1) may be challenged where the information for which immunity is claimed does not on its face fall within s. 39(1), or where it can be shown that the Clerk or minister has improperly exercised the discretion conferred by s. 39(1)” at para 39).
 Ibid at paras 39, 60.
 Ibid at para 40.
 Ibid at para 60.
 Supra note 3 at paras 29, 40–41, 43.
 The degree of relevance is a function of the probative value and materiality of the information. In contrast, the degree of injury is a function of the sensitivity of the information and the timing of its disclosure.
 For specific instances where Cabinet immunity was overclaimed in the context of s 69 of the Access to Information Act, RSC, 1985, c A-1 [ATIA] (s 69 sets out, for all practical purposes, the same definition of “Cabinet confidences” as s 39 of the CEA 1985, supra note 5), see Ken Rubin, Access to Cabinet Confidences: Some Experiences and Proposals to Restrict Cabinet Confidentiality Claims (Ottawa: Ken Rubin, 1986) at 59–64.
 See Yan Campagnolo, “Cabinet Documents Should be Under the Scope of the ATIA”, The Hill Times (6 June 2016) 15, online: <https://www.hilltimes.com>.
 See Pelletier, supra note 160 at paras 22–26.
 Discussion papers are documents whose purpose is to present background explanations, analyses of problems or policy options to Cabinet for consideration in making decisions. Pursuant to s 39(4)(b) of the CEA 1985, supra note 5, discussion papers are no longer subject to Cabinet immunity if the decisions to which they relate have been made public or, where the decisions have not been made public, if four years have passed since the decisions were made. This exception to Cabinet immunity is known as the “discussion paper exception”. The same exception is found under the access to information legislation. See ATIA, supra note 259, ss 69(1)(b), 69(3)(b); Campagnolo, “History of Cabinet Immunity”, supra note 1 at 263–64, 273–79.
 See Canada (Information Commissioner) v Canada (Minister of the Environment), 2003 FCA 68 at paras 23–26, 224 DLR (4th) 498 [Ethyl].
 See Campagnolo, “History of Cabinet Immunity”, supra note 1 at 278–79; Ken Rubin, “Harper’s Cabinet Need not have any Background Facts, Reinforces Greater Cabinet Secrecy”, The Hill Times (14 April 2014) 15, online: <https://www.hilltimes.com> [Rubin, “Greater Cabinet Secrecy”].
 Ethyl, supra note 263 at para 27 [emphasis in original].
 See CEA 1985, supra note 5, s 39(4)(b).
 See Ethyl, supra note 263 at paras 23–26; Rubin, “Greater Cabinet Secrecy”, supra note 264. See also Campagnolo, “History of Cabinet Immunity,” supra note 1 at 273–79.
 See Babcock SCC, supra note 3 at para 25.
 For example, in Robinson, supra note 156 and Ellis, supra note 98, public officials relied on PII to avoid legal liability in civil proceedings. Similarly, in Nixon, supra note 243 and Whitlam, supra note 229, public officials relied on PII in attempts to thwart criminal proceedings.
 See Marc-André Boucher, “L’évolution de la primauté du droit comme principe constitutionnel et sa relation avec le pouvoir exécutif en matière de renseignements confidentiels” (2002) 32:4 RGD 909 at 961, 975.
 See e.g. Roncarelli, supra note 11 at 133–37, where Premier Duplessis admitted that he had cancelled Mr Roncarelli’s liquor licence as a punishment for helping Jehovah’s witnesses. Similarly, in Conway, supra note 99 at 942–43, the Attorney General conceded that PII was sometimes claimed to shield public officials from legal liability in civil proceedings.
 Note, however, that courts typically condemn the unauthorized disclosure of Cabinet confidences by public officials. See Ontario (AG) v Gowling & Henderson (1984), 47 OR (2d) 449 at 463, 12 DLR (4th) 623 (Ont H Ct J); Bruyere v Canada,  FCJ No 2194 (QL) at para 9 (FC).
 See e.g. Ethyl, supra note 263.
 See Peter W Hogg, Patrick J Monahan & Wade K Wright, Liability of the Crown, 4th ed (Toronto: Carswell, 2011) at 135.
 Likewise, there is evidence that the government tends “to exaggerate claims of national security confidentiality.” See Canada (Citizenship and Immigration) v Harkat, 2014 SCC 37 at para 63,  2 SCR 33.
 See generally Jack Goldsmith, “The Irrelevance of Prerogative Power, and the Evils of Secret Legal Interpretation” in Clement Fatovic & Benjamin A Kleinerman, eds, Extra-Legal Power and Legitimacy: Perspectives on Prerogative (New York: Oxford University Press, 2013) 214.
 Supra note 3 at 228–29. See also Little Sisters Book and Art Emporium v Canada (Minister of Justice), 2000 SCC 69 at paras 203–13,  2 SCR 1120, Iacobucci J, dissenting in part.
 For a discussion of legal black holes, see Dyzenhaus, Constitution of Law, supra note 39 at 3, 42, 50.
 Commission des droits de la personne SCC, supra note 3 at 225–26, citing Commission des droits de la personne QCCA, supra note 104 at 73–74.
 Dyzenhaus, Constitution of Law, supra note 39 at 3, 42, 50, 205.
 See Bill of Rights, supra note 6, s 2(e); Charter, supra note 6, ss 7, 11(d).
 See Constitution Act, 1867, supra note 7, s 96; Crevier, supra note 86 at 234, 237–38; Dunsmuir, supra note 71 at para 30; Trial Lawyers Association, supra note 38 at paras 38–40.