Of course, that we are on the whole doing much better than has been usual does not mean that we can relax. The price of fiscal propriety is constant vigilance. In particular, you will have to be vigilant against those who will appear before you asserting that now is the time to return to more program spending—that such spending would be in the public interest. In fact, the evidence is quite to the contrary.

Before I give the evidence, I will make one comment about recent developments in the United States which this committee should consider. As you may know, the recent budget package agreed to in the United States involved the establishment of a top capital gains tax rate of 20 percent. This compares to a rate in British Columbia of just over 40 percent. While the federal government cannot be held responsible for these provincial tax extraction practises of which a Byzantine emperor would have been embarrassed, it is nevertheless the case that the Canadian capital gains tax rate varies from 34.6 to more than 40 percent. It is very necessary to address this large difference between the Canadian and US treatment of capital gains. If a left-of-centre government of the most powerful economy and most effective job-producing machine on the face of the earth has found it advisable to cut its capital gains tax rate, it surely behooves us to follow suit as we attempt to reduce the four percentage point gap between their unemployment rate and our own.

The main issue that I would like to raise with the committee is the targeting of federal fiscal policy and the principles which should guide you in your deliberations. Yogi Bera, most famous of contemporary philosophers, reminded us that, "if you don't know where you are going you will end up some place else," or else, when you come to the fork in the road you might well take it. It is particularly important to know where we are trying to go at this time, because unlike the recent past we now have a range of alternatives, and the government will be pressured by special interest groups to convey benefits on them at the expense of the general taxpayers and the long-term interests of the population—particularly those who are unemployed.

Size of government

The crucial question which we in Canada must now address is how big should the government sector be? Debt repayments aside, this question is equivalent to asking, what is the total tax rate the average Canadian should be forced to bear? This is a question which, if not addressed directly, will be determined as a by-product of other discussions. It is not a question which should be determined by default. It should be asked and answered directly. At the moment, owing to the fact that a substantial number of low-income Canadians bear no significant tax burden, the average family pays nearly half its income to government. While, of course, the federal government collects only 48 percent of this, as the senior level of government, the federal government ought to take the lead in addressing the issue.

What should be the tax rate faced by the average Canadian family? Evidently, it should be the ideal rate. That is, the rate of tax that is in some way better than any other. What is the ideal rate of tax? Presumably it is the rate of tax that produces just the right level of government expenditure. If the tax rate were lower than this ideal rate, there would be too little government spending; if it were greater, there would be too much. How can we determine what this ideal rate of government spending is? This question is the subject of an extensive Fraser Institute paper which will be published later this year. However, the flavour of the findings can be summarized as follows.

There are essentially three empirical approaches to the question, only two of which I will address. The first was conducted by Ludger Schuknecht and Vito Tanzi at the International Monetary Fund. In this approach, the authors analyzed 17 countries during the period 1870 to 1990. They treated government like a factor of societal production and asked the question, at what point does the addition of further government to the society-wide production process lead to diminishing returns? Or, put another way, at what point does the addition of further government spending cease to produce any further improvement in the social and economic objectives which are presumably the intent of government spending to influence? I can do no better than to quote these two eminent government economists directly. They conclude, first, "the expansion of public expenditure and the welfare state during the past three decades has yielded limited gains in terms of social objectives" (Tanzi and Schuknecht, p. 25). Then they go on to note that, "we have argued that most of the important social and economic gains can be achieved with a drastically lower level of public spending than what prevails today. Perhaps the level of public spending does not need to be much higher than, say, 30 percent of GDP to achieve most of the important social and economic objectives that justify government interventions (Tanzi and Schuknecht, p. 34).

A second approach to the question of the optimal size of government arises currently from the work of Gerald Scully (Scully, 1991) but actually has a root that goes back to the conjecture of an Australian economist, Colin Clark. As early as the 1950s, Clark said that the maximum size of government shouldn`t exceed 25 percent of the GDP or else there would be a high price to pay in terms of foregone economic growth. The work of Scully for various countries suggests that the total tax rate ought not to exceed about 23 percent. Scully arrives at this result by asking the question, what tax rate will maximize the rate of economic growth? He approaches this question in three different ways and the answer he provides is the average of the three. Two Fraser Institute economists, Joel Emes and Dexter Samida, have provided the calculation for Canada. Their unsurprising conclusion is that the growth-maximizing size of government in Canada is 29.8 percent (see figure 1).

What is surprising is the extent to which we have endured a loss of output as a result of having operated government above the optimal level since 1965. According to the calculations done by Emes and Samida, the fact that we have had above optimal tax rates since 1965 has cost us $3.7 trillion in output. In more directly understandable terms, if we had had the optimal tax rate over the period, we would today have an average per capita income $11,000 higher than it currently is. From the point of view of the lowest income citizens in our country, this would have made quite a difference. The policies we actually followed produced very little change in the position of the bottom fifth of the population. In 1965, this group received 4.4 percent of the total income before tax, while by 1995 they were receiving only 5.7 percent after tax. This group, as well as everybody else, would have been much better off with a lower rate of total tax, and the higher growth and output it would have produced.

The inevitable criticism of this result is that it ignores the fact that our national system of socialized medicine was introduced in 1970 and this, surely, was worth the additional cost. No other program of public expenditure is more widely supported by Canadians than our health care program. Without the expansion of the government sector, would this program have been possible?

The answer to this question is quite interesting. During 1995, the latest year for which we have complete, reliable figures, health care absorbed about 10.5 percent of the national income. This is 63.9 billion inflation-adjusted dollars. If we had pursued the growth maximizing tax rate over the period, GDP would have been $936 billion. Total current spending on health care of $63.9 billion is 6.8 percent of that higher GDP level. Coincidentally, this is just six-tenths of a percent more than the US spends on its government health care programs, Medicare and Medicaid. It is probably not necessary to note that the US government taxes only 33 percent of its total GDP—very close to the optimal rate for Canada.

Another objection of the result is that 1971 brought the great widening of the parameters of the Unemployment Insurance system. Without an expansion in the size of government, the higher cost associated with this extension of federal program spending would not have been possible. In this case, the response has already been provided by the current government. Program parameters for what is now Employment Insurance have been wound back to their pre-1972 level because of their malevolent effects; in due course the outlays on this program will return to more manageable levels. Indeed, they have already begun to do so while the payroll tax to support them remains at its peak levels.

It appears that we have everything to gain and very little to lose by moving to the optimal tax rate. The focus of the committee in evaluating the proposals of the government and others should be, how does the proposed plan affect the achievement of the optimal tax rate? If a proposal would keep us where we are, or push us further above the optimal rate, what are the material advantages which would accrue to the country as a result? The Committee ought, in its deliberations, to consider how the reduction in the burden of taxation can be reduced gradually to get us back to the sort of growth experience we had in the 1960s. If you reject this as a course of action, you should at least adopt another target and articulate why it would be a superior alternative. If you fail to do so, you will leave yourselves open to the rent seeking of supplicants and the vagaries of Yogi Bera's law.

Of course, there are two additional issues that you will also have to sort out. How much of any reduction in the tax rate ought to come at the federal level, and how quickly should the country aim to get there? The often-repeated response is that there is no useful purpose served by the federal government reducing its spending and taxing if the result is simply going to be an increase in both at the provincial level. This response is anachronistic because the provinces have generally been more fiscally responsible in recent years than the federal government. In fact, something of a tax-and-spending reduction contest is under way between the provinces. (The single exception is here in the trade-union leaders' paradise—but even that is beginning to change as economic conditions deteriorate.)

While there are strong arguments for decentralizing the tax and expenditure process in Canada and hence reducing the fraction of the total government sector which is represented by federal spending, an initial distribution for the purposes of goal setting could be to keep the federal portion of the tax burden the same as it is at the present time.

The final issue that must be resolved is what is the time period over which the government should move to its ideal tax rate of 14.4 percent of the national income from the current 18.8 percent. Given the loss in output which we are experiencing as a result, the first answer must be, as fast as possible. The actual pace of adjustment will have to be influenced by developments elsewhere. Our trading partners are beginning to recognize the need for reducing their tax burdens. To remain competitive in a globalizing world, we will have to take note of their progress. One possible course of action would be to resolve to continue the rate of reduction which has actually been achieved during the past few years.

The average reduction in the size of the total tax burden has been 1.4 percent of GDP per year. We know that this rate of reduction is achievable, and we know that it has produced significant benefits. Of course, some of the recent improvement has been due to spending cuts and some due to the growth in GDP. The composition of the reduction is not all that important. Holding the line on spending and taxation during rapid income growth periods will secure the target just as surely as spending cuts. The important thing, as Mr. Martin has shown with his budget deficit reduction targeting, is to establish clear, unambiguous targets, and then not to exceed them.

With cuts to its participation in the GDP of .67 percent per year, the federal government could achieve the optimal tax rate in seven years.