(a) Distinguishing delegated legislation from primary legislation

Delegated legislation can look very much like primary legislation. Indeed, it can shape the scope of power of executive authorities, prescribe legal obligations, create offences, and stipulate sentences for offences. The question of how primary legislation and delegated legislation are distinct took on a particular salience in the UK context – whether Acts of the Scottish Parliament ought to be characterised as primary or delegated legislation had important implications for the scope of judicial review available against them.Footnote ³² While this issue has since been settled by the UK Supreme Court in AXA General Insurance Ltd v Lord Advocate,Footnote ³³ the literature that developed in response to this issue is nevertheless quite useful in understanding how one ought to characterise the difference between primary and delegated legislation.

Aileen McHarg's account of the distinction between primary and delegated legislation is particularly instructive.Footnote ³⁴ She first identified a common method of characterising this distinction: primary legislation is the ultimate source of authority as expressions of the sovereign Parliament, while delegated legislation owes its existence and authority to primary legislation.Footnote ³⁵ This might be thought to be an uncontroversial proposition, especially within the context of jurisdictions such as the UK which adhere to the doctrine of parliamentary sovereignty.

However, calling this method a ‘source-based approach’, McHarg argued that there were several issues with taking such an approach to distinguish the two types of legislation. For one, a source-based approach would encounter difficulties when applied to jurisdictions possessing supreme written constitutions, since it would suggest that in such jurisdictions where law-making power is delegated by the constitution, there can be no primary legislation at all.Footnote ³⁶ As McHarg pointed out, even Acts of Parliament would be characterised as delegated legislation on such an approach in these jurisdictions.

McHarg argued that the better view was that delegated legislation was distinct from primary legislation on two grounds. First, delegated legislation, being authored by executive authorities, does not carry the same weight of democratic legitimacy as primary legislation.Footnote ³⁷ Secondly, and more crucially, delegated legislation is susceptible to the constraint of fulfilling the specific purposes for which power was conferred upon the executive authorities – indeed, even broad powers to legislate for ‘peace, order and good government’ are not entirely unfettered and have to be exercised for such purposes, as Laws LJ held in R (on the application of Bancoult) v Secretary of State for the Foreign and Commonwealth Office.Footnote ³⁸ This constraint is not applicable to primary legislation. On this basis, McHarg concluded that Acts of the Scottish Parliament ought to be characterised as primary legislation.Footnote ³⁹

It is clear that McHarg's approach relies primarily on the constraints upon delegated power to specify the difference between primary and delegated legislation – delegated powers must be used to further the purposes of the delegator. Building upon this idea, Ahmed's recently-proposed delegation theory of judicial review is particularly useful in substantiating what the precise constraints upon delegated powers are. Ahmed offered an interpretive theory of judicial review – in her view, the grounds of judicial review in administrative law can be explained as a means of holding administrators to their moral duties as delegates of Parliament or the Crown.Footnote ⁴⁰ She argued that there are three types of delegation relationships, which each impose different moral duties upon delegates: (1) mandate delegation, which requires the specific instructions of delegators to be fulfilled with little discretion accorded to delegates; (2) effectuate delegation, which calls upon delegates to give concrete form to the general or abstract purposes of the delegator; and (3) expressive delegation, which requires delegates to express the character, beliefs, or attitudes of the delegator.Footnote ⁴¹

Ahmed proposed that the various grounds of judicial review can be understood as directed at holding administrators to the moral duties expressed in each of these types of delegation relationships, and argued that this theory of judicial review coheres well with existing judicial review doctrine and has predictive and normative value, inter alia.Footnote ⁴² To be clear, Ahmed's argument was not directed specifically at the context of delegated legislation. Nevertheless, should one accept that the key distinction between delegated legislation and primary legislation is that delegated legislation must further the specific purposes of the delegator, as McHarg argued, then it becomes readily apparent that Ahmed's delegation theory of judicial review is particularly apt for translation to this context.

Accordingly, we have discerned that the principal distinction between primary and delegated legislation is that delegated legislation must further the specific purposes of the delegator. These purposes can constrain the exercise of delegated power by imposing moral duties on delegates. These constraints do not apply to primary legislation. This is helpful in framing how delegated legislation ought to be approached for the purposes of judicial review. Indeed, as will be observed later, this account of the distinction between primary and delegated legislation justifies the present judicial practice of reviewing delegated legislation on a similar basis as other types of executive action.

(b) Distinguishing delegated legislation from executive action

Yet, as an account of the proper normative orientation of judicial review of delegated legislation, the account offered in the previous section remains incomplete. Indeed, this account of what distinguishes delegated legislation from primary legislation does not distinguish delegated legislation from other types of executive action. What then makes delegated legislation distinct from other types of executive action, for the purposes of judicial review?

The answer is relatively straightforward. Delegated legislation is distinct from other types of executive action principally on the basis that it is are similar in substance to primary legislation – indeed, it comprises published general rules which are capable of shaping the scope of legal powers and specifying legal rights and duties, extending even to the provision of criminal offences and punishments.Footnote ⁴³

In view of the fact that delegated legislation is distinct from other types of executive action on this basis, two important implications follow. First, the capacity of delegated legislation to serve as general rules of conduct means that it attracts the concern of democratic legitimacy in a unique manner as compared to other types of executive action. Indeed, Lockean political theory suggests that the legislative function, which involves the creation of general rules binding on all citizens on pain of sanctions, represents a particular kind of intrusion on individual liberty which would have to be legitimised by being promulgated through elected decision-making bodies.Footnote ⁴⁴ This would ensure that the decision-making bodies responsible for such general rules are accountable to the people for their decisions.Footnote ⁴⁵ Indeed, as Charleton J held in the Irish Supreme Court decision of Náisiúnta Leictreach Contraitheoir Eireann Cuideachta Faoi Theorainn Ráthaíochta v The Labour Court, The Minister For Business Enterprise & Innovation, Ireland and the Attorney General (NECI), ‘[e]lected representatives owe a responsibility to the Irish people to legislate for them and this cannot be abdicated by any indolent or ill-defined delegation to a subordinate authority’.Footnote ⁴⁶ Delegated legislation, as a product of executive authorities, has a more distant link with the sources of democratic legitimacy as compared to primary legislation, thereby raising unique concerns as to its democratic credentials.

Second, the distinctive nature of delegated legislation also implicates rule of law concerns in a manner unique from other types of executive action. As Eoin Carolan has pointed out, the rule of law requires that the creation of general rules of conduct must comply with certain principles in order to safeguard against arbitrary government action.Footnote ⁴⁷ These principles are neatly captured by Lon Fuller's well-known account of the eight requirements of the internal morality of law, which include requirements that laws have to be made at the appropriate level of generality, be publicly accessible, and cannot have retroactive effect, inter alia.Footnote ⁴⁸ Being general rules of conduct, delegated legislation also implicates the rule of law concern of certainty more strongly than other types of executive action, so much so that the courts have exercised their discretion not to quash delegated legislation for procedural defects on the ground that doing so may lead to uncertainty in administration.Footnote ⁴⁹ Accordingly, in addition to democratic legitimacy, compliance with the rule of law serves as an additional and related source of legitimacy for such general rules.Footnote ⁵⁰

To be clear, this is not to say that ordinary executive action is incapable of raising concerns of democratic legitimacy and the rule of law. Regular executive action can certainly raise serious concerns in this regard – for instance, unchecked executive discretion over a wide range of conduct is severely corrosive to the rule of law. Rather, the point sought to be made here is simply that delegated legislation, as a unique species of executive action, raises democratic legitimacy and rule of law concerns in a manner distinct from other types of executive action, given its nature as published general rules of conduct capable of shaping legal powers and specifying legal rights and obligations.

(a) The major questions doctrine?

One candidate that arises for attention is the major questions doctrine – a doctrine that has gained traction in US Supreme Court jurisprudence. The major questions doctrine was recently discussed and affirmed by the US Supreme Court in West Virginia v Environmental Protection Agency Footnote ⁶⁵ (West Virginia). In this case, the key issue was whether the US Environmental Protection Agency (EPA) had the power, under its discretion to determine the ‘best system of emission reduction’ in the Clean Air Act, to promulgate rules which would amount to a fundamental restructuring of the overall mix of electricity generation in the US to reduce reliance on coal.

Drawing from precedents such as Food and Drug Administration v Brown & Williamson Tobacco Corp,Footnote ⁶⁶ Chief Justice Roberts affirmed the existence of a major questions doctrine. This doctrine was based on a presumption that Congress would intend to make major policy decisions for itself; a presumption in turn grounded both in the separation of powers and ‘a practical understanding of legislative intent’.Footnote ⁶⁷ The doctrine required that when faced with a claim of a highly consequential scope of delegated power, the relevant agency ‘must point to “clear congressional authorization” for the power it claims’.Footnote ⁶⁸

On the facts, Roberts CJ found that the major questions doctrine was indeed implicated. The power claimed by the EPA amounted to a transformative expansion of its authority, such that the EPA could even force coal power plants to stop generating power altogether. Indeed, Roberts CJ thought that it was not plausible that Congress would leave to agency discretion the decision of how much coal-based generation there should be over the coming decades, being a decision of great economic and political significance.Footnote ⁶⁹ Accordingly, any claims to possess powers with such economic and political significance would have to be buttressed by clear congressional authorisation.Footnote ⁷⁰ Unfortunately for the EPA, such congressional authorisation could not be found.Footnote ⁷¹

Described as such, the major questions doctrine is an interpretive principle. It is also a prima facie plausible solution to the problem identified in this paper. It would directly further the normative concern of democratic accountability. Indeed, by limiting the permissible scope of delegated law-making power, the existence of such a doctrine would help ensure that decisions bearing major political or economic consequences would be taken by the legislature. It would also help to safeguard the constitutional separation of powers by policing a clear demarcation between legislative and executive powers.

However, what is a ‘major’ question that would trigger the application of the doctrine? A determination of which questions are major and which are not ultimately rests in an assessment of which functions are properly reserved to the legislative branch and which are the domain of the executive branch. The major questions doctrine is premised then on the possibility of drawing a clear demarcation between legislative and executive power. Yet, the articulation of such a distinction may be less straightforward than one might think.

A pair of decisions from common law jurisdictions will help to illustrate this point. The High Court of Australia in The Commonwealth v Grunseit Footnote ⁷² (Grunseit) was confronted with a ministerial direction under the National Security Act requiring male refugee aliens to perform service which the Minister for the State of the Interior was of the opinion that they would be capable of performing. A key issue the court had to decide was whether this direction had legislative character and therefore had to comply with procedural requirements on presentation before Parliament. In deciding this question, the court held that ‘the general distinction between legislation and the execution of legislation is that legislation determines the content of a law as a rule of conduct or a declaration as to power, right or duty, whereas executive authority applies the law in particular cases’. Applying this test, the court held that the ministerial direction at hand amounted merely to an application of a general rule to particular circumstances and was therefore not legislative in nature – it did not change the content of the law by extending or limiting it and did not create a new rule of conduct.Footnote ⁷³

The Singapore High Court in Cheong Seok Leng v Public Prosecutor Footnote ⁷⁴ (Cheong Seok Leng) accepted this test but reached a different conclusion on the executive action at issue. In this case, the Minister for Home Affairs had assigned civil defence duties to the Vigilante Corps. This assignment was subsequently challenged on the ground, inter alia, that it had legislative effect and therefore had to comply with the procedural requirements applicable to delegated legislation. Applying the Grunseit test, the court held that this assignment had legislative effect – ‘it was intended to charge the Vigilante Corps with additional functions and duties which, but for the assignment, it would not have had’, and amounted to a determination of ‘the content of the Vigilante Corps Act as a rule of conduct or a declaration of the duties of the Vigilante Corps’. This meant, therefore, that the assignment was invalid, since the procedural requirements which it would have to comply with were not fulfilled on the facts.Footnote ⁷⁵

It should be apparent from these decisions that it can be challenging to articulate precisely just where the boundary between making law and executing the law lies. Indeed, as Richard White has argued, it is difficult to draw clear distinctions between legislative and executive power – it is difficult to ‘execute the law without in some way also legislating and adjudicating’.Footnote ⁷⁶ Along similar lines, Ivor Jennings noted that ‘administrative action so often partakes of both legislative and executive characteristics’, leading to his conclusion that there is no real distinction between these powers.Footnote ⁷⁷ The challenges of distinguishing between these two functions, exacerbated by the complexity of the modern administrative state, means that it will be difficult to specify judicially manageable standards for a doctrine framed in such a manner, as Carolan has argued.Footnote ⁷⁸

Further, the major questions doctrine fundamentally rests on a problematic picture of administrative governance – that is, that executive authorities serve as a mere ‘transmission belt’ for the legislature, executing and carrying out policy decisions taken by the legislature.Footnote ⁷⁹ However, as Justice Stevens of the US Supreme Court pointed out in Whitman v American Trucking Associations, Inc, it is clear that administrative agencies play a much bigger role in modern administration than this model suggests – they do make rules and engage in policy-making.Footnote ⁸⁰ As Jeff King has argued, ‘“everyday policy-making” is plainly expected under delegated powers’, with the consequence that it is quite widely-accepted in practice for delegated legislation to also play a policy-making function.Footnote ⁸¹ Modern administration is now so complex that the zone of ‘technical detail’ traditionally entrusted to administrative authorities has expanded significantly, such that the exercise of what one may have thought to be a mere ‘filling up of detail’ now in truth involves ‘the same kinds of policy choices, interest balancing, and power struggles that attend real lawmaking by real legislatures’.Footnote ⁸²

The question that arises then is whether it is possible for legal doctrine to further the important interests served by the major questions doctrine, yet avoid the fundamental difficulty of having to draw distinctions between the functions properly exercised by the legislative and executive branch.

It is proposed that a possible way of doing this is to avoid a focus on the nature of the functions which can properly be delegated, but rather to direct one's attention towards ensuring that the conferral and exercise of delegated law-making power complies with rule of law standards. Indeed, if important law and policy-making functions will inevitably be exercised by executive authorities, then a more fruitful emphasis for judicial review would be to ensure that the conferral and exercise of these functions is performed in accordance with the rule of law. This can be done by way of two separate doctrines: a doctrine directed at ensuring that delegations of law-making power comply with the rule of law, and another doctrine directed at ensuring that exercises of delegated law-making power comply with the rule of law.

(b) Conferring delegated law-making power in accordance with the rule of law

Doctrines directed at ensuring that delegations of law-making power must comply with rule of law standards can be found in various common law jurisdictions. These doctrines have been usually termed non-delegation doctrines, capable of rendering impermissible certain delegations of law-making power.Footnote ⁸³ While the ‘non-delegation’ descriptor may prima facie suggest some overlap with the major questions doctrine, the analytical focus of such doctrines is quite different. Non-delegation doctrines are focused not so much on the nature of the function sought to be exercised by the executive authority – as the major questions doctrine is – but rather on the manner by which law-making power is delegated. No matter the subject matter of the delegated power, non-delegation doctrines hold that delegations of law-making power cannot be unconstrained and vague, and that they must specify sufficient guidance for executive authorities in their exercise of delegated law-making powers.

There are a variety of ways by which a doctrine of non-delegation can be articulated. One notable example can be found in Irish Supreme Court jurisprudence. Indeed, the Irish courts have held that the exclusive vesting of legislative power in the Irish legislature found in Article 15 of the Irish Constitution mandates the acceptance of a non-delegation doctrine.Footnote ⁸⁴ The prevailing requirements of this doctrine were first articulated in the seminal case of Cityview Press v an Chomhairle Oilina (Cityview Press).Footnote ⁸⁵ In Cityview Press, one of the arguments made before the Irish Supreme Court was that the Oireachtas had impermissibly delegated its law-making powers to the Industrial Training Authority by empowering it to impose a levy upon employers in certain designated industries to fund industrial and commercial training. In addressing this argument, the Supreme Court formulated the ‘principles and policies’ approach: ‘the test is whether that which is challenged as an unauthorised delegation of parliamentary power is more than a mere giving effect to principles and policies which are contained in the statute itself’.Footnote ⁸⁶ If so, then any purported exercise of such delegated power would be unconstitutional. Applying this test, the court in Cityview Press found that there had been no unconstitutional delegation of law-making authority, since the primary legislation in question had contained ‘clear declarations of policies and aims’, and the levy power granted to the Industrial Training Authority amounted to no more than ‘adding the final detail which brings into operation the general law which is laid down by the section’.Footnote ⁸⁷

The focus of the ‘principles and policies’ test on ensuring that delegations of law-making power are not unconstrained became clearer in the wake of the recent Irish Supreme Court decision of NECI.Footnote ⁸⁸ Faced with a challenge to Chapter 3 of the Industrial Relations (Amendment) Act 2015 alleging that it was an unconstitutional delegation of law-making authority, MacMenamin J affirmed the applicability of the ‘principles and policies’ test, explaining that it is ‘best understood as a means whereby, recognising the present day realities of the administrative state, the Oireachtas, when it delegates power by law, does so only in a manner that the delegate or subordinate body does not, itself, become a legislator’.Footnote ⁸⁹ MacMenamin J suggested that the proper inquiry that the test involves would benefit from a ‘negative recasting’ – specifically, the ‘principles and policies’ test should require a court to ask ‘whether the absence of such principles and policies actually trespasses on the power of the Oireachtas’.Footnote ⁹⁰ In answering this question, judges ought to adopt a holistic consideration of the entire Act, and ultimately direct their minds to the issue of ‘whether there are sufficient limiting principles and policies so as to confine the area of choice’.Footnote ⁹¹ In a similar vein, Charleton J in a concurring judgment held that the ‘principles and policies test’ essentially requires ‘[l]imited delegation for an express purpose and with sufficient guidance to an assured outcome, or an outcome within a defined and limited range, or discretion within a limited assessment’.Footnote ⁹² Applying this approach, both judges agreed that the challenged Act did not fall foul of Article 15, on the basis that the power vested in the Labour Court by the Act was constrained within a prescribed legislative framework and subject to safeguards and guidelines – all of which served to narrow the choices available to the decision-maker.Footnote ⁹³

Doctrines similar to the Irish ‘principles and policies’ test can be found elsewhere in the common law world – for instance, the ‘excessive delegation’ doctrine formulated by the Supreme Court of India is substantially similar to the ‘principles and policies’ test.Footnote ⁹⁴ Further, the Irish ‘principles and policies’ test was formulated by reference to the ‘intelligible principle’ test of the US Supreme Court.Footnote ⁹⁵ The ‘intelligible principle’ test holds that a delegation of law-making power would be constitutional as long as Congress has laid down an intelligible principle to which the delegate is required to conform.Footnote ⁹⁶ This test, however, has not been applied with great vigour in the US. As the majority opinion of the US Supreme Court in Gundy v United States noted, the standard required by the doctrine was ‘not demanding’, with the effect that the court had ‘over and over upheld even very broad delegations’.Footnote ⁹⁷ Scholars have remarked that the non-delegation doctrine in the US is essentially moribund – while the doctrine remains on the books, it is hardly enforced by the courts.Footnote ⁹⁸

Suggestions as to how a non-delegation doctrine ought to be framed have also been made in academic literature. For instance, Appleby has argued within an Australian context that the courts can perform a meaningful role in checking the usage of delegated legislation by placing substantive and procedural limits upon the delegation of law-making power.Footnote ⁹⁹ She proposed, as a substantive limit, that delegations of law-making power in primary legislation have to be sufficiently preciseFootnote ¹⁰⁰ – a proposal which clearly bears some resemblance to the ‘principles and policies’ and ‘intelligible principle’ tests. The procedural limit she proposed was for the court to instate a requirement for parliamentary approval before delegated legislation comes into force.Footnote ¹⁰¹

A non-delegation doctrine, no matter what specific form it may take, helps to further democratic accountability, though admittedly in a less direct fashion than the major questions doctrine – indeed, to the extent that a non-delegation doctrine is grounded upon safeguarding the separation of powers, such a doctrine would help ensure that the primary responsibility for law-making rests in the legislature. But the link between the rule of law and the non-delegation doctrine would be more direct. The doctrine ensures that delegated law-making powers are meaningfully constrained by law, furthering the fundamental rule of law principle that legal powers must have legal limits. Such a doctrine also furthers the rule of law interest of clarity by requiring that delegations of law-making power cannot be unacceptably vague.

The key advantage of a non-delegation doctrine is that it is capable of achieving these objectives in a manner that does not require the court to enter into the difficult territory of distinguishing between functions which are properly legislative or executive in nature. By shifting the focus away from complex questions of institutional balance to the concept of legal constraint, a non-delegation doctrine falls more properly within the expertise of judges, who would already be accustomed to thinking in such terms in the context of judicial review.

One might argue, however, that one difficulty with a non-delegation doctrine is that it is principally targeted at identifying impermissible delegations. In other words, it renders impermissible certain delegations of law-making power in primary legislation, which then naturally bears consequences for delegated legislation purportedly made under such power. One might, therefore, argue that a non-delegation doctrine will have less salience in a jurisdiction adhering to the doctrine of parliamentary supremacy such as the UK.Footnote ¹⁰² Indeed, it is telling that the jurisdictions examined earlier which saw the articulation of a doctrine of non-delegation – Ireland and the US – both adhere to the doctrine of constitutional supremacy.

This difficulty is not an insuperable one. Indeed, a doctrine of non-delegation can be adapted for use within the UK by effectuating it as an interpretive principle akin to the principle of legality – for instance, the courts could presume that Parliament will not delegate unfettered legislative powers unless it makes express its intent to do so. The courts in the UK have already adopted similar interpretive principles to address problems raised by the usage of delegated legislation. For example, the courts in the UK have scrutinised Henry VIII clauses strictly and accorded them narrow constructions, resolving any ambiguities against the executive.Footnote ¹⁰³ Accordingly, such a doctrine can be a valuable device in ensuring that delegations of law-making power comply with rule of law standards across a variety of constitutional contexts.

(c) Exercising delegated law-making power in accordance with the rule of law

In addition to ensuring that the conferral of delegated law-making power is performed in accordance with the rule of law, it is worth considering whether the law of judicial review can also contribute to ensuring that the exercise of such law-making powers coheres with the rule of law. The focus in this regard would shift from the permissibility of delegations of law-making power at the level of primary legislation, and towards the lawfulness of specific exercises of law-making power by executive authorities. The classic grounds of judicial review of executive action, discussed earlier, are obviously also applicable in this space – the question here is whether more can be done to better account for the unique concerns raised by delegated legislation.

One possibility would be the acceptance of what has been called the ‘new non-delegation doctrine’.Footnote ¹⁰⁴ This doctrine, originating in the US, would require delegated law-making powers to be exercised ‘in a manner that promotes the rule of law, accountability, public responsiveness, and individual liberty’, as Lisa Schultz Bressman described.Footnote ¹⁰⁵ Accordingly, as long as there are clear and specific standards guiding and limiting the exercise of delegated law-making powers, then even if these standards are self-imposed, the authority's exercise of law-making powers ought not to be impugned.Footnote ¹⁰⁶

Before exploring the specifics of how such a doctrine may operate, it is important to first address the criticisms which have been directed against it – after all, should the criticisms prove fatal, the doctrine would be a non-starter to begin with. Most notably, this ‘new non-delegation doctrine’ was rejected by the US Supreme Court in Whitman v American Trucking Associations, Inc, on the ground that such a doctrine was constitutionally unsound: ‘[t]he idea that an agency can cure an unconstitutionally standardless delegation of power by declining to exercise some of that power seems to us internally contradictory’.Footnote ¹⁰⁷ The response to this criticism is that this doctrine, in the form that is proposed here, would do no such curing of ‘unconstitutionally standardless’ delegations. Indeed, the doctrine proposed here would apply in tandem with the non-delegation doctrine discussed earlier. The non-delegation doctrine would render ‘standardless’ delegations of law-making power unlawful, while the doctrine discussed here would be directed at lawful delegations of law-making power which were nevertheless exercised in a manner contrary to the rule of law.

Carolan also expressed some concerns about the ‘new non-delegation doctrine’. In his view, the trouble with this doctrine is that it focuses exclusively on the rule of law concern raised by delegated legislation, while neglecting ‘the issues of democratic accountability and institutional balance’.Footnote ¹⁰⁸ The response to this concern is – again – that on the approach suggested by this paper, this doctrine would not operate on its own. It would operate in conjunction with a non-delegation doctrine which would help to further the normative concern of democratic accountability. Insofar as considerations of institutional balance are relatively de-emphasised by both of the doctrines proposed here, it is suggested that this is a necessary compromise – indeed, any adventure into complex questions of institutional balance would run squarely into the problems outlined earlier in articulating a clear distinction between legislative and executive functions. The proposals in this paper in fact align well with Carolan's conclusion on the proper way forward for judicial review of delegated legislation: he ultimately argued that looking forward, ‘the courts could concentrate on the more familiar task of enforcing the substantive values which have been partially attributed to the non-delegation doctrine: the rule of law; principles of publicity, certainty and generality in legislation; fairness in the exercise of discretionary powers; and so on’.Footnote ¹⁰⁹

Since the ‘new non-delegation doctrine’ is not rendered a non-starter by these criticisms, then given that it is based on ensuring that exercises of law-making power cohere with the rule of law, the next issue that arises for consideration is what standards of the rule of law ought to be adopted. It is suggested that one can have useful recourse to Fuller's well-known eight requirements of the internal morality of law as a starting point. Fuller argued that law possesses an internal morality distinguishing the rule of law from arbitrary power, and that this internal morality can be specified in eight moral requirements: laws must have generality of application; laws must be promulgated and accessible; laws should not have retroactive effect; laws must be clear; there should be no contradictions between laws; laws cannot require the impossible; laws must be certain and constant through time; and there must be congruence between official action and the declared rule.Footnote ¹¹⁰ These moral requirements are inherent in the nature of law – indeed, if there were a total failure of any of these requirements, Fuller suggested that the resultant legal system could not properly be characterised as a legal system at all.Footnote ¹¹¹

These requirements are clearly applicable to delegated legislation. For instance, the concern of constancy through time is captured by the courts’ willingness, mentioned earlier, to find that delegated legislation is capable of giving rise to legitimate expectations. While all the requirements are relevant in principle, three of them can be focused on for a rule of law-based ground of review directed against delegated legislation: the requirements of clarity, accessibility, and non-retroactivity. These three requirements address the rule of law problems posed specifically by the usage of delegated legislation, and can provide the doctrinal substantiation of such a ground of review – that is, whether the usage of delegated legislation is unlawful will depend upon whether any of these requirements have been contravened.

It is useful at this juncture to sketch out briefly how such a ground of review can be substantiated. The doctrinal content of such a ground of review can draw upon existing jurisprudence in common law administrative law. Commencing with the requirement of clarity, the courts in the UK have already accepted that if delegated legislation is so vague that its meaning cannot be ascertained with reasonable certainty, it will be invalidated.Footnote ¹¹² Indeed, the High Court in Kay Tabernacle v Secretary of State for Defence recognised that the principle of vagueness ‘is established in both Strasbourg and domestic jurisprudence’, with the effect that delegated legislation that is ‘so vague or imprecise as to have unforeseeable application’ would be unlawful.Footnote ¹¹³ The proposed ground of review would supplement this already-accepted legal principle with a refreshed theoretical grounding.

As for accessibility, the importance that the common law has traditionally accorded to the accessibility of law was recognised by the Singapore High Court in Cheong Seok Leng in the context of delegated legislation.Footnote ¹¹⁴ Holding that the assignment of civil defence duties to the Vigilante Corps indeed had legislative effect, the High Court emphasised the importance of ensuring that delegated legislation had to be published – the public has ‘a right to know or be informed of subsidiary legislation affecting them and this right may be said to be inherent in the legal system’ since a person's liberty should not be ‘indirectly curtailed by laws and regulations unknown or inaccessible to him’.Footnote ¹¹⁵ Accordingly, the High Court held that the requirement of publication of delegated legislation specified in section 23(1) of the Interpretation Act was a mandatory one, such that non-compliance with the requirement would render the legislation invalid.Footnote ¹¹⁶

The court's reasoning here suggests that where there is a statute specifying a requirement of publication of delegated legislation, such a requirement ought to be interpreted as a mandatory requirement instead of a directory one. Going further, however, even in the absence of a statutory requirement of publication, it is suggested that the rule of law concern of accessibility justifies the imposition of a requirement of publication even at common law, with the effect that delegated legislation must be published and made reasonably accessible unless there are strong reasons justifying non-publication. The requirement of accessibility can be seen as related to the preceding one of clarity. Indeed, even published delegated legislation would not be accessible in a meaningful sense if its contents are excessively vague. It is also conceptually aligned with Appleby's proposed procedural limit upon the usage of subsidiary legislation – that the courts should impose a requirement of parliamentary approval before delegated legislation comes into force – insofar as it is also concerned with ensuring that delegated legislation receives a reasonable degree of publicity.

The third requirement of non-retroactivity is a widely-accepted principle of the rule of law. The most odious form of retroactive laws, as Fuller recognised, would be laws imposing criminal punishment in a retroactive manner – prohibitions against such laws have been enshrined in written constitutions.Footnote ¹¹⁷ Delegated legislation seeking to impose criminal punishment in a retroactive manner should therefore be rendered invalid under this ground of review. How then should delegated legislation having retroactive effect beyond the imposition of criminal punishment be treated? As Appleby noted, delegated legislation has been used in such a manner in Australia to deny permanent visas retroactively to refugees arriving in Australia by boat, with the intention of deterring such asylum seekers.Footnote ¹¹⁸ Due to the retroactive effect of these regulations – pending applications for protection visas were also captured by these regulations – a large number of refugees was affected. Such usage of delegated legislation, while not going to the extent of imposing criminal punishment retroactively, clearly raises rule of law concerns as well. To take into account such usages of delegated legislation, it is suggested that a requirement of non-retroactivity ought to extend beyond the imposition of retroactive criminal punishment. Nevertheless, in recognition of the fact that not all usages of retroactive legislation are equally odious, this requirement ought to be satisfied if the executive authority in question is able to offer a rational justification for its usage of retroactive legislation.

(d) Overall proposal

Thus, in order for judicial review to more directly address the distinctive concerns of democratic accountability and the rule of law raised by delegated legislation, this paper proposes the development of judicial review doctrine to regulate both the conferral of delegated law-making power and the exercise of such power. The non-delegation doctrine and the novel rule of law-based ground of review discussed earlier are directed respectively at these objectives and ought to be accepted in law.

The acceptance of this proposal brings several advantages. First, as compared to the regular administrative law grounds of review, the proposed doctrines would be able to address more directly the unique democratic accountability and rule of law implications that delegated legislation raises vis-à-vis other types of executive action. As a matter of substance, these doctrines would go further than the usual administrative law grounds of review. The non-delegation doctrine specifies an interpretive principle providing additional texture to an assessment of whether delegated legislation is ultra vires. While one might argue that the ground of irrationality may be broad enough to cover the three rule of law requirements of the novel rule of law-based ground of review, the crucial difference is that irrationality would not specifically target these requirements, making the proposed ground of review valuable as a targeted means of addressing the problems with delegated legislation. More fundamentally, such a ground of review would have the advantage of drawing express attention to the special context of delegated legislation, providing a useful foundation for the further development of specific jurisprudence in this regard.

Further, should one indeed accept the importance of imposing constraints on the usage of delegated legislation by way of judicial review, the proposed doctrines bear several advantages over the major questions doctrine. For one, by focusing on how delegated law-making power is conferred and exercised, the proposed doctrines will not be susceptible to the fundamental difficulty that the major questions doctrine encounters – the drawing of boundaries around the nature of functions which can be permissibly delegated and those which cannot. Further, the proposed doctrines avoid the unrealistic assumption of a ‘transmission belt’ model of the administrative state lying at the heart of the major questions doctrine. Instead, they recognise that executive authorities are independent actors within the administrative state entrusted with making important policy decisions, and they seek to ensure that these authorities receive and exercise delegated law-making power in a manner that coheres with the rule of law.

Finally, the proposed doctrines would cohere with the normative orientation of judicial review more broadly. If the law of judicial review is indeed normatively oriented towards upholding the moral duties of delegates, as Ahmed argued, the proposed doctrines cohere well with this normative foundation. If Parliament, the primary institution entrusted with law-making, can be legitimately presumed to legislate in a manner compliant with the rule of law, then it stands to reason that it intends any law-making power that it delegates to comply with the same rule of law requirements, as an expression of its own attitudes towards law-making.