Parliament cannot delegate its legislative power: A British constitutional reality or myth? Dennis Morris1
With a view to answering the question posed in its title, this article discusses whether or not there is any British constitutional principle that would prevent the British Parliament from enacting the following hypothetical section:
Regulations implementing the United Nations’ Multinational Corporations (Controls) Convention
(1) The Secretary of State may make regulations providing for such matters as are necessary to enable the performance of the obligations of the United Kingdom, or to obtain for the United Kingdom any advantage or benefit, under the Convention.
(2) In particular (but without limiting subsection (1)), the regulations may deal with the following matters:
(a) the procedure for dealing with applications for international registration of multinational corporations that are to be filed with the International Bureau through the intermediary of the Multinational Corporations Registration Office;
(b) the procedure for dealing with requests to extend to the United Kingdom the protection resulting from international registration of multinational corporations;
(c) the protection given to protect multinational corporations in the United Kingdom;
(d) the circumstances in which such protection ceases and the procedures to be followed in cases of cessation;
(e) the cancellation of an international registration at the United Kingdom’s request, as contemplated by Article 6 of the Convention;
(f) the effect of cancelling an international registration.
(3) Regulations made for the purposes of this section—
(a) may be inconsistent with this Act; and
(b) prevail over this Act (including any other regulations or other instruments made under this Act), to the extent of the inconsistency.
This hypothetical section closely follows section 189A of the Australian Trade Marks Act 1995, which was inserted by the Trade Marks Amendment (Madrid Protocol) Act 2000 (Cwlth). Only the words in Italics differ from those of section 189A. However, this article does not deal with issues arising under Australian law but rather is concerned with whether the British Parliament would be free to enact a provision in an analogous form should it so wish. For the sake of argument, I have assumed that if a Bill containing a similar provision were to be introduced into the British Parliament, its provisions would—
(a) neither include a provision making the imaginary convention referred to in the hypothetical section part of the UK’s domestic law,
(b) nor include that convention’s text, either in a schedule or elsewhere.
Comparison with Henry VIII clauses
In a paper read to the Third Commonwealth Conference on Delegated Legislation, the late Lord Rippon of Hexham, QC, said:
In recent years some of us in the UK have been expressing concern over the ever increasing quantity and decreasing quality2of our legislation. .... The fault lies with Government and Parliament. Although our Bills get longer and longer, increasingly they are merely skeletal, leaving vital details to be settled by others and regulations.
Lord Rippon went on to deal in particular with Henry VIII clauses and the UK constitutional problem their ever-increasing use had by then created. The use of this drafting device has been criticised with increasing ineffectiveness ever since the Donoughmore Report of 1926. Indeed, whereas not a single Henry VIII clause was enacted by the British Parliament during World War 23, this now frequently employed legislative device hardly raises an eyebrow at Westminster. The use of such clauses is probably part of the “spin off” resulting from the shift of the United Kingdom’s most significant political forum from Parliament to the television studio.
A Henry VIII clause has been described as “a provision in a Bill that enables primary legislation to be amended or repealed by subordinate legislation with or without further parliamentary scrutiny”.4 In contrast, section 189A of the Trade Marks Act 1995 (Cwlth)5seems to be even more far reaching in that, as far as regulations made under the section are concerned, it dispenses with the need to slot the regulations into that Act’s scheme by way of amendment. Thus, under the section, the Governor-General of Australia in Council is free to make whatever regulations are thought to be appropriate (irrespective of whether they are consistent with the enabling Act) so long as they are necessary to enable the relevant treaty obligations to be performed. Having regard to subsection (2) of the hypothetical section, those regulations could, at the very least, have very serious financial consequences for a multinational corporation that was affected by the legislation.
The earliest statutory instruments6seem to date from the 1820s-1830s and for almost a century those responsible for preparing subordinate legislation, other than “machinery legislation” such as commencement orders or regulations fixing fees, charges and so on, ensured that legislative power was delegated but not abrogated. Until recently at least, the extent of the relevant power granted to make subordinate legislation was regarded as being one of the crucial considerations in judicially determining the validity of such legislation. This flowed from the general acceptance, as being axiomatic, that under the British constitution, Parliament’s ability to make laws could not be delegated to anyone else. This principle may have been first established at the beginning of the 1880s when Gladstone’s first Irish Home Rule Bill was introduced. The Bill generated much discussion (sometimes heated) as to whether or not the British Parliament could transfer sovereignty and it seems likely that it was then that the question of whether Parliament could delegate its legislative function first arose.7 But irrespective of whether this was in fact the case, the principle appears to have been settled before the end of the nineteenth century. So if power to make subordinate legislation were to be included in proposed legislation, it was (because of Parliament’s then accepted inability to delegate its legislative function) thought crucial to devise a constitutional safeguard, namely the ultra vires principle setting parameters within which the delegated power must be exercised. For once an enactment circumscribed the grantee of the power—and thus exposed the exercise of the power to legal scrutiny—Parliament must be regarded as having fallen short of delegating (or as “semi-abrogating”)8its sovereign power to legislate on any matter to which subordinate legislation made under the enactment could relate.
The relevant case law
As far as the case law is concerned, the two earliest cases in which the inability of Parliament to delegate its ability to make laws was an issue were both heard by the High Court of Australia. In Roche v. Kronhiemer,9 it was submitted (inter alia) that there is no Commonwealth power to hand over to the Commonwealth Executive the whole of the power of that Parliament to legislate as regards a particular matter. But the argument that the Commonwealth Parliament was so circumscribed seems to have been based on the assumption that the Australian Constitution10separates Commonwealth executive and legislative functions in the sense expounded by Montesquieu. Unfortunately, neither of the judgments in Kronhiemer expressly deals with the submission referred to. In the second Australian case, The Victorian Stevedoring and General Contracting Company Proprietary Ltd v. Dignan and another,11questions broadly similar to those in Kronhiemer were in issue and the Court delivered long and detailed judgments.12 In that case, the Montesquieu doctrine of the separation of powers was again discussed. Although holding that the Commonwealth Parliament could not delegate as described, the Court said it was not because that doctrine applied in Australia but rather because, as in the United Kingdom (but not in the United States), the Executive is accountable to Parliament. However, none of the dicta in Dignan is as forthright as that of Hanna J in Pigs Marketing Board v. Donnelly (Dublin) Ltd., who said (as regards Irish constitutional law):
It is axiomatic that powers conferred upon the Legislature to make laws cannot be delegated to any other body or authority. .... But the Legislature may, it has always been conceded, delegate to subordinate bodies or departments not only the making of administrative rules and regulations, but the power to exercise, within the principles laid down by the Legislature, the powers so delegated and the manner in which the statutory provisions shall be carried out.13 [Emphasis added]
This passage was later approved of by Hedderman J in The State (Gilliland) v. Governor of Mountjoy Prison (Irish Supreme Court).14 Although neither Hanna J nor Hedderman J expressly said so, I think it can be inferred, as regards Irish legislation at least, that where powers are given to make subordinate legislation, parameters must be included in the enabling Act by reference to which it would be possible to decide whether or not, in a particular case, this or that might be included in the subordinate legislation. So in reposing authority in the Executive to legislate by means of subordinate legislation, even if he believed the constitutional principle mentioned in this article’s title is now academic, I think a prudent draftsman should, in the absence of clear judicial authority, nonetheless “weave” the principle of ultra vires into the relevant Bill, even if that Bill provided that the validity of an instrument made under the Bill were not to be open to question in legal proceedings. Take the following case. Assume the British Parliament had enacted legislation giving a Secretary of State a power to make regulations. Assume also that, irrespective of the desirability of having the ultra vires principle apply to the regulation-making power, the relevant section were couched in terms that in fact disregard it. Then if, in legal proceedings, the validity of regulations made under the section were questioned solely on the ground that it was unconstitutional for Parliament to ignore, or alternatively to override, the ultra vires principle when enacting the section, should the court find in favour of the applicant, it would surely do so by declaring the purported grant of the regulation-making power to be constitutionally invalid rather than by impugning the regulations purporting to have been made in exercise of the power?
So can that constitutional assumption still be made as regards the United Kingdom? Before grappling with this difficult question, it is appropriate to consider two questions that the hypothetical section raises. One relates to the ultra vires principle and the other to the relationship between the common law and international law.
The ultra vires principle
Regarding the first question, in Dignan, Evatt J in referring to executive powers to make subordinate legislation said:
It is true that the extent of the power granted will often be a very material circumstance in the examination of the validity of the legislation conferring that power.
And earlier Dixon J, as that particularly distinguished future Australian Chief Justice then was, said, having conceded a Legislature’s ability to confer such executive powers:
This does not mean that a law confiding authority to the Executive will be valid, however extensive or vague the subject matter may be... There may be such a width or such an uncertainty of the subject matter to be handed over that the enactment attempting it is not a law with respect to any particular head or heads of legislative power. Nor does it mean that the distribution of powers can supply no considerations of weight affecting the validity of an Act creating a legislative authority.15
So can the parameters set by the hypothetical section be faulted on a ground either Dixon or Evatt JJ mentioned? I rather think not. For not only are they precise but, in addition, they clearly state the purpose for which the power is being granted. Also, as will be noticed, the section has been cast in a form that would not require Parliament to debate the imaginary UN convention. However, I do not think this matters. Indeed, there is an excellent precedent. In 1972, when the European Communities Bill was being debated in the House of Commons, strenuous and furious efforts by the Opposition to have the EEC Treaties themselves discussed (discussions that the Government ardently wished to avoid) were thwarted. The Government was well-served by parliamentary counsel for, as those counsel must have foreseen, the Chairman of Ways and Means ruled the discussion out of order because of the Bill’s non-inclusion of a provision “enacting” those treaties.
The relationship between the common law and international law
As to the second of the two questions referred to, assume that the object of the hypothetical section is to enable regulations to be made whose purpose would be to ensure that the United Kingdom’s treaty obligations as regards imposing agreed treaty restrictions or controls regarding specified activities of multinational corporations would be discharged. Also assume that, but for the regulations, these activities could be freely pursued by anyone anywhere in the kingdom. And thirdly, bearing in mind the principle that in making domestic law, the Legislature cannot be fettered by any inconsistent provisions of international law.16 So while the Sovereign has prerogative power to accept treaty obligations these are neither binding on, nor (where appropriate) can they be relied on, by any person in the realm in the absence of legislation to that effect. Nowadays, the most common method employed in the United Kingdom to extend treaty obligations so that they affect persons other than the State is to enact legislation incorporating a (usually scheduled) treaty or other international agreement into the domestic law. But is formal legislative incorporation of such an agreement always necessary? Having regard to the European Communities Act 1972, it seems not. For under that Act a new cause of action, an “enforceable Community right” was created. And as Sir Geoffrey Howe, who had steered the Bill for that Act through the House of Commons, said in 1973, the EEC Treaties had not been “incorporated into or made identical with our domestic law. Our courts are simply required to give direct effect to Community law according to its own nature.”17 But as the early EEC Treaties are the basic law or grundnorm from which those rights and European Union law, according to which they are to be enforced in the UK, both spring, the relevant provisions of those instruments must, I suggest, impliedly enjoy a recognition of some sort within the UK’s corpus juris? So if this be correct, merely referring to the (supposed) UN Convention and its implementation to set parameters as regards the scope of any regulations to be made under the hypothetical section would, I believe, suffice.18
The question of delegation
As regards the question of delegation, by enacting the hypothetical section, the British Parliament would delegate the whole of its power to legislate as regards implementation of the imaginary UN convention. For the Australian enactment on which the hypothetical section is modelled, section 189A of the Trade Marks Act 1995 (Cwlth), rather suggests that those responsible for it foresaw possible conflicts between that Act and what would be necessary to include in regulations implementing the Madrid Protocol in Australia. A subsection in identical terms to those of subsection (3) above was therefore included. This has the appearance of going much further than a “classical” Henry VIII clause.19 But is there really any difference in principle? The Henry VIII formula enables a statute to be amended or repealed, its usual purpose being to avoid potential conflict, whereas the hypothetical section itself operates to remove the conflict. Indeed, one could constitutionally knit pick by arguing that under it no power to amend or repeal legislation is given and so it is constitutionally less sullied! And anyway subsection (3) above is no more radical than, say, section 108(9) of the Children Act 1989 (UK), which enables the Lord Chancellor to amend or repeal any (other) enactment in a manner that he considers necessary or expedient in consequence of any provision of that Act. Both powers are circumscribed: one is limited to doing what is necessary to implement a convention and the other to avoiding conflict between provisions of the 1989 Act and other enactments. Moreover, subsection (3) above is “small constitutional beer” as compared with, say, section 17(4) of the 1989 Act, which enables the Secretary of State to add by order any further power or duty to those mentioned for the time being in Part I of Schedule 2 to that Act. So were a provision similar to the hypothetical section’s subsection (3) to appear in a British Bill, it seems unlikely that any Member of Parliament would turn a hair.
A Bill implementing a convention or other international agreement would of course probably contain many provisions of an administrative or machinery nature. In passing, perhaps I should mention that I doubt whether regulations under the hypothetical section could create offences. But as regards the hypothetical treaty obligations imposed on the United Kingdom, certain steps, and all and only those steps, would need to be taken in exercise of the power to make regulations were the United Kingdom to be enabled to perform those obligations. In other words, a failure to exercise all of the powers that section gives would leave the Government unable to fully discharge those obligations. Accordingly, in creating executive power to do everything necessary to implement the Convention, the whole of Parliament’s legislative power as regards that implementation is thus delegated.20 So could the British Parliament so legislate?
Could the British Parliament so legislate?
Before 1973 at least, by adapting a dictum of Evatt J in Dignan, it could have been argued that the constitutional impediment to delegating the legislative function arose—
(a) not because Parliament is bound to exercise or perform any of its legislative powers or functions,21
(b) nor because the Montesquieu doctrine operates and so prevents its granting authority to anyone else to make instruments having legislative effect,22
but because every one of the Acts enacted by Parliament must, to some extent, establish principles as regards, or otherwise regulate, the particular subject matter or branch of the law to which the Act relates. So a law by which, as regards a particular matter, Parliament gave all its law-making authority to a Secretary of State or other person would be bad merely because it failed to pass the test mentioned. In other words, it would have failed to meet the requirements that the connotation of the word “law” itself envisages. For example, an Act that merely provided that contracts of employment are to be fair and then went on to give a Secretary of State power to make such regulations as are necessary to give effect to the Act would, in my view, fail such test. Therefore, before 1973 at least, the law would be of no effect. However, the European Communities Act 1972 seems to have destroyed this argument. This is because Parliament, through that Act, as well as delegating to what is today the European Commission a power to make laws for the United Kingdom, also delegated a power to make laws relating to a range of unspecified matters. Indeed, if the Commission’s powers continue to expand as they have done during the past 5 years, then by 2010 the British Parliament may have less control over how revenue is to be raised for the Exchequer than that currently enjoyed by a local authority in relation to funding its services.
While the 1972 Act does not accord with the constitutional principle23that this article is concerned with, a court might hold that the constitutional circumstances in which that Act was enacted and the consequences of its enactment were exceptional in that the United Kingdom was thereby able to enter a union of states founded on a separate and distinct body of law. Accordingly, a court might hold that the principle of non-delegation must give way in those circumstances. But would such a distinction have any real constitutional significance? I doubt it for two reasons. First, even though membership of the European Union has engendered the application of the primacy of European Union law principle, British Judges still appear to be reluctant to question the validity on other grounds of any enactment. Indeed, as recently as 1995 Lord Mustell could say:
Parliament has a legally unchallengeable right to make whatever laws it likes.24
So were a court pressed to strike down the hypothetical section on the basis of its being repugnant to the constitutional principle referred to in the title to this article, I feel certain it would refuse, if only because the courts will go to great lengths to avoid creating a “political crisis”. Indeed, it might, by invoking my second reason, apply the de minimus principle! And that reason is the ease with which a draftsman can “outflank” the application of the non-delegation principle. For example, if the hypothetical section were to provide that the imagined UN convention was to have the force of law and then to empower the making of such regulations as would be necessary to give effect to the convention, the constitutional principle could not then be raised.
If the United Kingdom had a wholly written constitution and if that constitution had vested what might be called the state’s supreme legislative function in the Queen in Parliament, then the dictum of Hanna J quoted earlier in this article25would26be as true of the British constitution as it is of that of Ireland.27 However, in view of the foregoing, it seems reasonable to regard Parliament’s inability to delegate its legislative power as being truly mythical. But that is not to say that references to it will fail to earn brownie points in parliamentary debate. For, like “the necessity” of including laying provisions as regards statutory instruments, obeisance will be made to it in the name of the sovereignty of Parliament, even though those so acting will be aware that in reality it is nothing but an empty formula. Thus has a stage in British constitutional development now been reached when it is open to coin a new British constitutional maxim: Vox sublegis vox Dei28?
1 Legislative consultant, formerly of the Irish Parliamentary Draftsman’s Office (now the Office of Parliamentary Counsel to the Government) and the Law Drafting Division of the Attorney General’s Chambers, Hong Kong.
2 I question this.
. The paper is published in  10 Stat LR 205.
3 However, one should bear in mind the very extensive powers conferred on the Executive by the Defence of the Realm Act 1914 (UK).
4 Review (October 1994) of the first year’s work of the House of Lords’ Select Committee on the Scrutiny of Delegated Powers, para. 16.
5 I.e. the enactment that the hypothetical section all but replicates.
6 Sometimes statutory instruments are called statutory rules or legislative instruments, but whatever they are called they are a form of subordinate legislation. (Sometimes the latter is called subsidiary or delegated legislation.)
7 It is possible that the well-known constitutional lawyer A.V. Dicey also raised this issue at that time.
8 Because it had no power so to delegate, Parliament could always legislate as regards matters covered by a previously granted power to make subsidiary legislation. But perhaps this may now no longer be the case.
9  29 CLR 329.
10 Then wholly contained in the Commonwealth of Australia Constitution Act (1900).
17 Sir Geoffrey Howe: The European Communities Act 1972 (1973) International Affairs, 1.
18 It seems unlikely that Parliament would forego an opportunity to discuss and either approve of or reject the international agreement in question were it to consider proposed legislation analogous to the hypothetical section.
19 Which is limited to amendments or repeals.
20 The fact that the delegated power is not exclusive, i.e. Parliament could itself implement the Convention, would, in the circumstances being considered, be irrelevant.
, Op.cit. notes 7, 121.In that case, Evatt J had this to say at p. 123:
On final analysis therefore, the Parliament of the Commonwealth is not competent to 'abdicate' its powers of legislation. This is not because Parliament is bound to perform any or all of its legislative powers or functions, for it may elect not to do so; and not because the doctrine of separation of powers prevents Parliament from granting authority to other bodies to make laws or by-laws and thereby exercise legislative power, for it does so in almost every statute; but because each and every one of the laws passed by Parliament must answer the description of a law upon one or more of the subject matters stated in the Constitution. A law by which Parliament gave all its law-making authority to another body would be bad merely because it would fail to pass the last test mentioned.
21 For it may elect not to do so, but clearly this would not happen as regards supply.
22 For this is now done as a matter of course.
23 If indeed there is such a principle.
, A matter about which neither the Treaty of Rome, nor any other EEC/EU treaty, has anything to say.
24 R. v. Secretary of State for the Home Department, ex parte Fire Brigades Union  2 All ER 244, 267 j.
25 See p. 52.
26 Leaving aside the constitutional implications of British and Irish membership of the European Union.
27 See Article 15.2 of the Irish Constitution.
28In other words “Whatever says the subordinate legislation is divinely revealed truth.”