Andrew Petter ..................................

Federalism and the Myth of the Federal Spending Power, by
Andrew Petter

MP Intro. Introduction. 1. 2, 3, 4, 5, Conclusion, Postscript, End

II. Constitutional Doctrine

Administrative federalism has been nurtured and sustained by the federal spending power, the power asserted by Ottawa to spend funds on matters falling outside its legislative jurisdiction. Yet this power is mentioned nowhere in sections 91 or 92 of the Constitution Act, 1867. What then is its basis in constitutional doctrine? Legal scholars have proposed a number of theories. Some have suggested that the power flows from the royal prerogative,26 others from federal authority with respect to appropriation,27 and still others from Parliament's power to legislate in relation to the public debt and property.28

None of these theories is particularly convincing. The suggestion that the spending power flows from the prerogative ignores two important facts. The first is that all funds spent by the executive must be expended in accordance with the terms of appropriating legislation, such legislation being subject to the division of powers in sections 91 and 92. The second is that, even if one could view the act of spending independently from its authorizing legislation, it is a well established principle of Canadian constitutional law that the grant of prerogative powers follows that of legislative powers.29

Sections 102 and 106 of the Constitution Act, 1867, establish a Consolidated Revenue Fund of Canada "to be appropriated for the Public Service of Canada". Some have sought to read into these provisions a conferral of power upon Parliament to enact appropriating legislation for any purpose including one which would otherwise fall within provincial jurisdiction. Even if it were accepted that the limiting words "Public Service" in section 106 and "Public Service of Canada" in section 102 do not, of themselves, restrict spending to purposes falling within the federal legislative sphere, it is improbable that the framers of these sections intended them to have so broad an effect. The primary purpose of these clauses is to establish the Consolidated Revenue Fund and to provide the means for its appropriation. To read them as somehow qualifying the division of powers in sections 91 and 92 is to give them a significance that their wording and placement in the Constitution simply do not justify. Had so fundamental a qualification been intended, one would expect to see it spelled out in unequivocal language.30 In the absence of such language, it seems more credible to interpret these appropriation provisions as not disturbing the general pattern of legislative powers provided elsewhere in the Constitution.

Arguments supporting the federal spending power on the basis of Parliament's authority to legislate with respect to the public debt and property are also problematic. Such arguments are founded upon the assertion that, for the purposes of constitutional characterization, a law which provides for an expenditure of funds should be viewed differently from a law which establishes a scheme of compulsory regulation. Proponents of the spending power argue that, while a law regulating a matter ought to be characterized as a law in relation to that matter, a law authorizing funds to be spent on that matter should be characterized as a law in relation to spending (that is, the disposition of public property), a subject that falls within section 91(1A).

An initial problem with this spending/regulating dichotomy is that it finds little support in the words of the Constitution. The opening words of section 92 of the Constitution Act, 1867 bestow upon provincial legislatures exclusive power to make laws "in relation to matters", not to make laws regulating those matters. There is no basis in language or in logic for suggesting that when Parliament authorizes expenditures of funds with respect to some matter it acts any less "in relation" to that matter than when it regulates with respect to the same matter. Put another way, there is no greater justification for characterizing a law that provides for expenditures on hospitals as a law in relation to spending than there is for characterizing a law that regulates hospitals as a law in relation to regulation.

The argument in favour of viewing "spending" as a separate matter embraced within the above classes of subjects is analogous to arguments made in the past by federal and provincial governments with respect to the implementation of treaties and Crown immunity. In the Labour Conventions case,31 the federal government maintained that labour legislation enacted by Parliament in fulfillment of Canada's treaty obligations should be upheld as a law "in relation" to treaties, and therefore outside the scope of powers assigned to the provinces under section 92 of the Constitution Act, 1867. The Privy Council rejected this argument on the ground that it was inappropriate to characterize legislation in such a way as to avoid the division of powers in sections 91 and 92. For the purpose of the federal distribution of legislative powers, said Lord Atkin, "there is no such thing as treaty legislation as such. The distribution is based on classes of subjects; and as a treaty deals with a particular class of subjects so will the legislative power of performing it be ascertained".32 What Lord Atkin said about treaty implementation can be said with equal force with respect to spending. Just as treaties are not implemented for their own sake, Parliament does not appropriate funds for the purpose of giving away property. In each case, there is some underlying purpose which identifies the "matter" of the legislation.

In Amax Potash Ltd. v. The Government of Saskatchewan,33 the Saskatchewan government argued that a statute enabling it to retain taxes collected unconstitutionally was legislation in relation to Crown immunity and could be supported on the basis of a number of heads of section 92, including amendment of the constitution of the province (section 92(1)), "property and civil rights in the province" (section 92(13)), "administration of justice in the province" (section 92(14)), and matters of a local nature in the province (section 92(16)). The Supreme Court of Canada rejected Saskatchewan's claim on the basis that Crown immunity was not a valid characterization of the legislation. Rather, the court held that one had to look behind the immunity to ascertain the purpose for which the immunity was being invoked. Again, the same can be said with respect to spending.

Another problem with the alleged dichotomy between spending and regulating is its tenuous and artificial nature. The monies that governments spend are, for the most part, monies that have been generated through compulsory measures such as taxation. Thus it is unrealistic to characterize spending as though it were an independent governmental function. Rather, spending should properly be viewed as one component of a larger reallocative activity. This is a point that has been crossed over by many writers, yet it is key to understanding the significance of the limited case law on the spending power.

There is only one case concerning the federal spending power which has reached the Supreme Court of Canada or the Privy Council.34 The Employment and Social Insurance Act Reference35 concerned the constitutionality of a statutory scheme of unemployment insurance proposed by the depression-era government of R.B. Bennett. Under the scheme, a fund was to be created by means of compulsory contributions from employers and employees. The purpose of the fund was to provide a pool of monies from which insurance benefits could be paid to contributing employees who became unemployed. The federal Government argued before the Supreme Court that the scheme could be sustained on the basis that the compulsory contributions constituted a form of taxation within the meaning of section 91(3), and that the fund comprised public property which Parliament had an unfettered right to dispose of as it wished under section 91 (1A).

The Supreme Court majority rejected this argument, but their reasons for doing so are confusing and contradictory. On the one hand, the majority voiced support in principle for unlimited powers of taxation and spending. Kerwin J. (with the concurrence of Rinfret and Crockett JJ.) said:36

It is quite true that Parliament, by properly framed legislation may raise money by taxation and dispose of its public property in any manner that it sees fit. As to the latter point, it is evident that the Dominion may grant sums of money to individuals or organizations and that the gift may be accomplished by such restrictions and conditions as Parliament may see fit to enact.

On the other hand, the majority refused to invoke these powers to uphold the legislation, deciding instead that the combination of mandatory collection and spending, made the statute, in pith and substance, a law in relation to employment services and unemployment insurance rather than one in relation to taxation and public property. Rinfret J. (with the concurrence of Kerwin and Crockett JJ.) put it this way:37

The object of the Act, the end sought to be accomplished by it is a scheme for employment service and unemployment insurance; the contributions levied from the employers and employees are only incidents of the proposed scheme, and, in fact, merely means of carrying it into effect. The Act does not possess the character of a taxing statute, but it is 'legislation intending to do precisely what the title says: to establish an employment insurance commission, to provide for a national employment service, for insurance against unemployment, for aid to unemployed persons, or other forms of social insurance and security and for purposes related thereto.

It being well understood and, in fact, conceded that these are subject-matters falling within the legislative authority of the provinces, the Dominion Parliament may not, under pretext of the exercise of the power to deal with its property, or to raise money by taxation, indirectly accomplish the ends sought for in this legislation. If it were otherwise, the Dominion Parliament, under colour of the taxing power, would be permitted to invade almost any of the fields exclusively reserved by the Constitution to the legislatures in each province.

It is difficult to know how to reconcile these two positions. What the majority seems to have failed to grasp is that the only reason that Parliament need resort to spending powers is to achieve indirectly ends that fall within provincial legislative competence. Nor does it help to emphasize, as the majority did elsewhere,38 the fact that the conditions imposed by the Act (that is, the contributions) were made compulsory. Taxation is by its nature compulsory.

When the case reached the Privy Council, it fell to Lord Atkin to try to resolve the confusion created by the conflicting views contained within the majority opinions of the Supreme Court. He did so in the following terms:39

That the Dominion may impose taxation for the purpose of creating a fund for special purposes, and may apply that fund for making contributions in the public interest to individuals, corporations or public authorities could not as a general proposition be denied. Whether in such an Act as the present compulsion applied to an employed person to make a contribution to an insurance fund out of which he will receive benefit for a period proportionate to the number of his contributions is in fact taxation it is not necessary finally to decide. . . But assuming that the Dominion has collected by means of taxation a fund, it by no means follows that any legislation which disposes of it is necessarily within Dominion competence.

It may still be legislation affecting the classes of subjects enumerated in s. 92, and, if so, would be ultra vires. In other words, Dominion legislation, even though it deals with Dominion property, may yet be so framed as to invade civil rights within the Province, or to encroach upon the classes of subjects which are reserved to Provincial competence. It is not necessary that it should be a colourable device, or a pretence. If on the true view of the legislation it is found that in reality in pith and substance the legislation invades civil rights within the Province, or in respect of other classes of subjects otherwise encroaches upon the provincial field, the legislation will be invalid. To hold otherwise would afford the Dominion an easy passage into the Provincial domain.

Some writers have sought to minimize the impact of this passage. La Forest, for example, describes Lord Atkin's words as having, "a some- what delphic character" and argues that they must be read in light of the Supreme Court's judgment.40 Yet the basic thrust of the passage seems straight forward enough. Unlike the Supreme Court, the Privy Council rejected the notion that money raised by means of taxation could be disposed of by Parliament "in any manner that it sees fit" - While allowing that the contributions required under the legislation might be regarded as taxation, Lord Atkin went on to state that legislation disposing of these contributions would nevertheless be unconstitutional if its pith and substance fell within provincial heads of power. The clear implication is that laws governing spending must be characterized according to the underlying purpose that the spending seeks to achieve. On this basis, Lord Atkin was able to agree with the Supreme Court majority that "in pith and substance this Act is an insurance Act affecting the civil rights of employers and employed in each Province, and as such is invalid".41

The federal government appears to have interpreted the Privy Council decision as being confined to situations in which federal spending is directly related to a specific revenue raising measure. Smiley has noted that:42

. . . the effect of the Privy Council judgment . . . has been to inhibit the federal government from supporting an activity within provincial jurisdiction partly or wholly from a federal levy made for that purpose. Thus, when the government decided to finance its program of old age pensions to persons 70 and over partly through a special income tax levy, an amendment to the B.N.A. Act was secured with unanimous provincial consent to give Parliament concurrent jurisdiction in that field. If the policy had been to make these payments from the Consolidated Revenue Fund, as in the case of family allowances, no such amendment would have been necessary according to the understanding of the federal spending power which has prevailed in Ottawa in recent years.

What this federal position amounts to is an assertion that legislation appropriating monies from the Consolidated Revenue should be characterized differently for constitutional purposes from legislation appropriating monies from other funds. The federal Government seems to believe that if the relationship between the spending measure and the taxing measure is not made explicit, the constitutional problem will simply evaporate. Not only does this interpretation disregard Lord Atkins' emphasis upon the disposition of tax revenues, but it treats as mindlessly formalistic the objections expressed in the Supreme Court and the Privy Council that Parliament not be able to use taxing and spending as means of circumventing the division of legislative responsibilities set out in sections 91 and 92. Indeed, if the federal position were correct, it would mean that precisely the same scheme of unemployment insurance that was struck down by the Supreme Court and the Privy Council would have been upheld had the contributions been collected under separate legislation and channelled through the Consolidated Revenue Fund before being spent in the form of insurance benefits.

This brings us back to a point raised earlier. Underlying the federal position, and the position of many constitutional writers, is the belief that spending is somehow different from, and less in need of constitutional containment than, other forms of governmental activity. Scott43 and Driedger44 equate spending with gifting, and express shock at the suggestion that private persons could make gifts but that governments could not. Hogg states:45

There is a distinction, in my view, between compulsory regulation, which can obviously be accomplished only by legislation enacted within the limits of legislative power, and spending or lending or contracting, which either imposes no obligations on the recipient . . .or obligations which are voluntarily assumed by the recipient. . . There is no compelling reason to confine spending or lending or contracting within the limits of legislative power, because in those functions the government is not purporting to exercise any peculiarly governmental authority over its subjects.

With respect, the views of these scholars have a sense of unreality about them. What they seem to forget is that governmental spending is not an isolated activity. When a government spends, it must derive the revenue from somewhere. The way that Government usually does this is through the imposition of taxation, something which is, without doubt, the exercise of a "peculiarly governmental authority". The importance of the Privy Council judgment in the Employment and Social Insurance Act Reference is that it recognized this relationship and rejected arguments that it should characterize the spending function as though it were unrelated to the taxing function. Rather, it correctly saw them as part and parcel of a single redistributive activity.

The corollary to the suggestion that Parliament has unlimited spending power with respect to tax revenues is that Parliament has the power to levy taxes for any purposes. Yet this is a proposition that flies directly in the face of the well-established limitation against Parliament levying direct taxation for provincial purposes.46 It also flies in the face of sections 54 and 90 of the Constitution Act, 1867 which, as Laskin pointed out,47 clearly contemplate that Governmental responsibility for the raising of a revenue for provincial purposes by any means of taxation rests exclusively with the provincial authorities.48

It might be asserted that, since non-tax revenues are derived through activities which are not "peculiarly governmental", the federal government ought to have unlimited authority to spend these revenues.49 Put another way, it could be argued that legislation disposing of non-tax revenues should be characterized as a purely allocative activity falling within federal jurisdiction over public debt and property. This, however, is a much narrower claim and one that has little relevance to conditional grants. An examination of the growth of federal spending following the Second World War establishes that the rise in conditional grant programs has been directly related to the rise in tax revenues. Indeed, the federal government has done nothing to disguise this relationship, as evidenced by the fact that, in 1977, when it changed the formula with respect to contributions for post secondary education, hospital insurance and medicare (the "established programs"), it shifted from cash payments to a combination of cash payments and tax points. Perhaps the most obvious indication that conditional grants are largely attributable to tax revenues, however, is the fact that the amount of such grants far exceeds the amount of federal non-tax revenues. In 1984, for example, conditional transfers to the provinces alone exceeded all federal non-tax revenues by more than eleven billion dollars.50

There is a related point that needs to be addressed. It concerns the difficulty of distinguishing between tax and non-tax revenues flowing through the Consolidated Revenue Fund. Trudeau argued in 1957 that this difficulty "makes it impossible, in practice, to know what money may be used without limitations for grants, and what money may not".51 This conclusion may have made sense at a time when conditional grants for provincial programs constituted a small percentage of total federal expenditures and could be attributed solely to non-tax revenues.52 But where, as is the case today, conditional grants comprise a substantial portion of federal expenditures, one that far exceeds non-tax revenues, there can be no justification for shielding such grants from constitutional scrutiny simply because they have been "laundered" through the Consolidated Revenue Fund. On the contrary, there, is ample reason to proceed on the assumption that such grants derive from tax revenues.




See Scott, loc. cit., footnote 22; Hanssen, loc. cit., footnote 25, Driedger, loc. cit., footnote 24.


See Driedger, loc. cit., footnote 24; La Forest, op. cit., footnote 23; Hogg, op. cit., footnote 6.

28. See Hanssen, loc. cit., footnote 25; Magnet, loc. cit., footnote 25; La Forest, op. cit., footnote 23; Hogg, op. cit., footnote 6.
29. See Liquidators of the Maritime Bank of Canada v. Receiver-General of New Brunswick, supra, footnote 3. 
30. It is significant that, despite similarly worded provisions in sections 81 and 83 of the Australian Constitution, the framers saw fit to provide explicitly in section 96 for a parliamentary power to "grant financial assistance to any State on such terms and conditions as the Parliament thinks fit".
31. Attorney-General for Canada v. Attorney-General for Ontario, [1937] A.C. 326, [1937] 1 D.L.R. 673 (P.C.).
32 Ibid., at pp. 351 (AC), 682 (D.L.R.).
33 [1977] 2 S.C.R. 576, (1976) 71 D.L.R. (3rd) 1.
34 There have been a number of cases decided by lower courts, the most recent being Winterhaven Stables Ltd. v. Attorney-General of Canada (1988), 53 D.L.R. (4th) 413 (Alta. C.A.). For a discussion of the lower court decisions, see A Lajoie, The Federal Spending Power and Meech Lake, in K.E. Swinton and C.J. Rogerson (eds.), Competing Constitutional Visions: The Meech Lake Accord (1988), pp. 175-185.
35 Attorney-General for Canada v. Attorney-General for Ontario, [1937] AC 355, [1937] 1 D.L.R. 684 (P.C.), affg., [1936] 3 D.L.R. 644.
36 Ibid., at pp. 457 (S.C.R.), 669 (D.L.R.)
37 Ibid., at p. 454 (S.C.R.), 666-667 (D.L.R.)
38 Ibid., at pp. 455 (S.C.R.), 667 (D.L.R.) (per Rinfret J.).
39 Attorney-General for Canada v. Attorney-General for Ontario, supra, footnote 35, at pp. 366-367 (A.C.), 687 (D.L.R.).
40 La Forest, op. cit., footnote 23, p. 48.
41 Attorney-General for Canada v. Attorney-General for Ontario, supra, footnote 35, at pp. 367 (AC), 687-688 (D.L.R.).
42 Smiley, op. cit., footnote 2, p. 23 (notes in text deleted).
43 Loc. cit., footnote 22, at p. 6.
44 Loc. cit., footnote 24.
45 Op. cit., footnote 6, p. 126.
46 Caron v. The King, [1924] AC 999, [1924] 4 D.L.R. 105 (P.C.).
47 B. Laskin, Provincial Marketing Levies: Indirect Taxation and Federal Power (1959-60), 13 U.TL.J. 1, at pp. 20-21.
48 It is ironic that La Forest, recognizing this, has argued that such limitations on federal taxation should be ignored because "[t]hey were made when the device of transfer payments might still have been considered doubtful...": op. cit., footnote 23, p. 52. This is bizarre reasoning. It relies upon a contentious political practice to cast doubt upon an established constitutional limitation. Neither the Privy Council nor the Supreme Court has passed judgment on the constitutionality of transfer payments whereas both institution have held that Parliament does not possess the authority to levy direct taxes for provincial purposes. If a conflict arises, therefore, surely it is more appropriate to invoke the established limitation to cast doubt on the contentious practice, rather than the other way around.
49 See RE. Trudeau, Federal Grants to Universities, in Federalism and the French Canadians (1968), pp. 79 and 99.
50 Canadian Tax Foundation, op. cit., footnote 19, pp. 3:2 and 16:26, and letter from McLarty, supra, footnote 19.
51 Trudeau, op. cit., footnote 49, p. 98. See also La Forest, op. cit., footnote 23.
52 In 1957, federal expenditures on conditional transfers stood at $145 million, just 2.6 per cent of total budgetary expenditures and less than one third of the $426 million derived from non-tax revenues: Leacy, op. cit., footnote 10, series H16, H34 and H489.