There are three main ways to deal with this dividend. One is to increase federal expenditures, a second is to cut federal taxes, and the third is to pay down the federal debt. Doing nothing—making no policy changes—would, of course, divert all surpluses to debt retirement.

At the outset, there are complex questions of definition here. For example, should a new tax credit directed towards child care be deemed an expenditure, or a tax cut? It seems to me that any federal transfer such as a tax credit should be deemed as a tax cut only if it is totally unconditional (and they almost never are), and otherwise as an expenditure. Why so? Because the imposition of conditions imposes federal government priorities on the recipient, who is not otherwise free to use the money.

This distinction is extremely important, as the use of tax credits is large and growing. The morality of this, in an accounting sense, is a good subject for discussion as well. Tax credits tend to hide costs to the taxpayers, costs that would otherwise be highlighted by a straight payment or subsidy. The issue is similar to that of tariff protection versus overt subsidies to producers.

In the tripartite allocation debate, this article will focus on the issue of tax cuts versus increased expenditures. Prior to that, some portion of any surplus should undoubtedly be dedicated to reducing the federal debt, and indeed, it should be a first and fixed charge on the fiscal dividend, as our debt is dangerously high by international and prudential standards. What follows refers to the division of what remains after that first charge.

My position is that essentially all of that residual fiscal dividend should go to tax reduction. That position flows directly from the interests of Canadian federalism.

The objectives of new expenditure that the federal government has been contemplating are very worthy—health, education, and social policy. However all of these areas are squarely in the provincial jurisdiction, with minor exceptions. The federal government has the legal right to attempt to influence provincial activities in these areas by using conditional, cost-shared programs, as it has in the past. The problem is, this practice invariably distorts provincial priorities and undermines democratic accountability. Indeed, we should be backing out of inter-governmental transfers, not adding to them.

There are three main reasons for this, turning on efficiency, accountability, and unity.

In the first place, with regards to efficiency, the circumstances of every province are different. Each has different demographics as to age and educational achievement, different industrial and resource structures, and so on. All of these differences will influence the appropriate policy mix in health, education, and social policy, which is of course why the provinces, rather than the federal government, were given responsibility for these areas under our constitution.

Locally developed solutions, informed by "best practice" experience from around the world, are likely to be best suited to local needs. In addition, a part of the theoretical justification for federal systems is the "laboratory" effect of experimenting with different approaches to problem solving. For example, medicare has become a Canadian icon, but it was developed in the 1960s by Saskatchewan in a way that would not have been possible had there then been a Canada Health Act and cost-sharing financial enforcement to lock in the conventional wisdom of that time.

As to accountability, governments closer to home have more to lose from faulty policy solutions than does the senior government, and also find bond dealers less tolerant of unsound financial management. They therefore get stronger feedback signals than would a more distant government. For example, if provincial governments over the years had had responsibility for aboriginal affairs, unemployment insurance, and fisheries, which are all federal jurisdictions, it is inconceivable that local electorates would have tolerated the immense human and financial costs of these sadly failed programs.

The issue of unity is central to Canada today. A strong concern not just of Quebec but of all the four largest provinces that contain over 85 percent of our population, is federal intrusion into local priority-setting by way of tantalizing, cost-sharing programs financed by the federal government. To the extent that the foreshadowed new federal activity in health, education, and social services is structured in this way, we will have a further strain on unity.

In fact, the federal government should be moving in the opposite direction, to phase out the existing $12.5 billion CHST (Canada Health and Social Transfer) payment to the provinces. The CHST is unfairly allocated (i.e., not on a per capita basis) which is a major irritant, and in addition serves as the "enforcement" mechanism for the Canada Health Act, over the objections of provinces such as Alberta.

Any phase-out of the CHST would, of course, require a symmetrical lowering of federal taxes, in order that the provinces could recoup the lost revenue through their own increased taxation should they choose to do so. That, again, is consistent with the concept of accountability, that is, governments raising what they spend.

If the fiscal dividend was used to cut federal taxes, then put towards implementing a phased withdrawal from the CHST, the net result would be to confer a large amount of "tax room" to the provinces, to be used as local voters saw fit. Thus, we would have 10 provincial laboratories developing the best way to use the famous fiscal dividend, rather than central desk planning.

Does all of this make the federal government irrelevant in social policy? Not at all, but the federal government's proper relationship with its citizens is direct, rather than through the provinces. This, as well as needed adjustments in the large equalization program will be the subject of a future article.