Fiscal Rules for the Control of Government (BP358e)

Fiscal restraint has become the order of the day for governments around the world. Whatever their political stripe, they are finding it virtually impossible to support ever- expanding government through the use of deficit financing. This new mood has become so entrenched that we have heard talk of fiscal restraint in times of poor economic growth and even recession, something that would have been unthinkable as recently as a decade ago.

The fact that government rhetoric now promotes restraint does not mean that widespread success has been achieved in this area. Indeed, many governments today face a fiscal situation substantially worse than in the 1981-82 recession. The difference today is that the public seems to have largely accepted that continued increases in debt are undesirable and counterproductive, and is demanding measures to curtail growing public sector debt. One way of doing this is by the use of fiscal rules, defined here as any legislative or constitutional constraints on governments with respect to deficits/debt, aggregate spending or revenues, tax rates, etc. Such rules remove legislators' discretion in making spending and taxation decisions and often include provisions for monitoring compliance and a series of penalties for any breaches.

Governments have reacted in different ways to the demands for such fiscal rules, the most famous response being Proposition 13, which in 1978 placed limits on the extent to which California municipalities could levy property taxes. There have also been repeated calls from "conservative" groups in the United States for a constitutional amendment requiring balanced federal budgets. Of the required 34 states, 32 have passed resolutions calling for a constitutional convention to force the federal government to achieve a balanced budget.(1) Just recently, the House of Representatives failed by only nine votes to achieve the required two-thirds majority to pass such a resolution.(2)

The American government has also passed several bills attempting to set a schedule for balancing the budget. The most famous of these was the Balanced Budget and Emergency Deficit Control Act of 1985, better known as Gramm-Rudman-Hollings (GRH), which was succeeded in 1990 by the Budget Enforcement Act.

Today, virtually every American state has either a constitutional prohibition against deficits or a legislative control over deficit spending. The Maastricht Treaty in Europe has established a set of fiscal rules to govern member nations as those countries move to greater economic and political union. In Canada, the federal government has established legislation for the control of government program spending and two provinces have passed legislation mandating balanced budgets.

Much in the news today is the "Social Contract" imposed by the government of Ontario on the public sector in that province. This does not constitute a set of fiscal rules, however. It does not attempt to impose long-range fiscal constraints to generate some predetermined fiscal result, and it does not remove discretionary powers from legislators. Rather, it is a short-term and temporary response to fiscal problems.

The use of true fiscal rules to curtail the activities of government represents an attempt to tie the hands of legislators in the belief that they have proven their incapacity to curtail their appetite for spending. Some think that this has come about because their constituents do not reward such fiscally prudent behaviour; legislators have merely responded to pressures to provide ever more programs. Others see the need for such controls as arising from legislators' failure to deliver the prudent fiscal policy really desired by constituents.

The issue now is whether such fiscal rules can curtail "excessive" spending and deficits or whether they merely enable politicians to appear serious and responsible, while allowing them to continue their profligate ways. Professor Aaron Wildavsky notes that whereas such strategies have always been part of the political process, the move to budget stringency in the United States has led to a substantial increase in their use.(3)

This paper will examine the conceptual rationale for the use of fiscal rules, describe certain existing and proposed rules and draw conclusions that might be instructive for Canadian fiscal policy.

When the Spending Control Act was being examined by the House of Commons Standing Committee on Finance, two contrasting views were expressed by its supporters. One view was that we have high levels of spending and large deficits because this is what a very wide range of Canadians want; undesirable as this situation might be, Parliament is responding to the wishes of its constituents and giving them what they demand. The opposing view was that Canadians do not collectively want such a degree of spending and such deficits, but that these are the inevitable result of modern democratic governments, which feel they are compelled to respond to the demands of various special interest groups. Supporters of both views thought that legislated controls could be of value in generating sound fiscal policy that would benefit all Canadians in the long run.

Still others see high levels of spending and deficit financing as a reflection of poor management by government. Thus, fiscal rules would be merely administrative tools with which politicians could better enact fiscally sound policies. Fiscal rules would perform an administrative function similar to that of comprehensive government bookkeeping.

Establishing such rules would be one of several management devices for the control of government spending, such as sunset clauses, zero base or performance budgeting, indexation of the income tax system, and the requirement for super majorities in legislatures for the passage of tax-raising legislation. Others, however, see a more serious problem.

The strongest proponents of constitutional limits on the size of government and its use of deficit financing, are adherents of the "Public Choice" subdiscipline of economics. This approach assigns utility-maximizing behaviour to the activities of individuals in public institutions, including their decision-making on public issues. Rather than assuming that governments and legislatures are impersonal and benign institutions, Public Choice theory assumes that the individuals working in these institutions are generally no different from those working in private institutions or those who engage in a wide range of market activities.

In a democracy, politicians must be elected. They purchase public support from voters through their promises and actions, and voters sell that support in exchange for policies they believe will increase their own utility. This is as much an economic transaction as the buying and selling of an automobile, though it takes place in a different environment with different rules. In this view, collective actions of government are nothing more than the attempt by one group of people to sell something to another group of people, just as is done in other markets.

Government is involved with collective action. In such a market voters tend to spend little time and effort to inform themselves about the issues at hand, especially as such information is costly to acquire. Individuals have little impact on voting outcomes, and each government policy tends to have only a minor impact on individual voters. Where their interests are concentrated, however, voters have a tendency to be more informed and to make their wishes better known. The same is true of politicians. They are relatively ill informed over a wide range of issues but are better informed over issues important for their re-election, which are the issues about which voters feel strongly.

These features of the political marketplace make it particularly prone to capture by special interest groups. Professor Aaron Wildavsky offers the view that modern technological changes have greatly lowered the cost of communication, making lobbying less expensive and making the role of interest groups more prominent in political life. The situation has completely changed from the time when the Canadian and American constitutions were written.(4) These special interest groups have been able to organize and communicate their wishes to politicians, who see little to be gained in spurning these, especially when they are widely publicized by the media. And as long as special interest groups are not competitive in nature, politicians can gain votes (avoid losing votes) by yielding to their wishes.(5) Those voters who are not members of a favoured group lose little by this governmental support; should they think such support is unfair, their reaction might well be to undertake lobbying for their own benefit. Thus the political process can be seen as a kind of logrolling exercise on both the demand and supply side. A legislator may attempt to garner support for a group he or she endorses by voting to support another group, which is important to a different legislator. Similarly, interest groups could form "umbrella group" coalitions of their own.

In a more general view, Public Choice economists argue that the post-war history of persistent deficits could not have happened without the complete reversal of citizens' and policy makers' attitudes to public debt that resulted from the widespread acceptance of Keynesian economics. Once the old notions against deficit financing had been removed, the democratic political process inevitably led to government growth fed by debt accumulation, as politicians, bureaucrats, interest groups and ordinary voters engaged in their politico-economic transactions. Prior to the Keynesian revolution, public debt had been viewed as undesirable because it reduced the capital stock of society, and ultimately made citizens poorer. Many now believe, however, that such a debt can actually produce prosperity. Since this approach to public policy allows increased spending without corresponding tax increases, it removes a natural constraint on government growth.

The political process, and the cost structure of information gathering, generates incentives on both sides to promote special interests at the expense of the public good. If politicians, bureaucrats and voters are maximizers of individual utility rather than selfless individuals, politico-economic transactions can lead, and some would argue that they inevitably do lead, to economic policies with negative consequences for society, even though the participants in transactions see them as mutually beneficial. This is the ultimate in negative externalities. Hence, Public Choice economists see the need to curtail these transactions and suggest fiscal rules as one way of doing so.

Prior to the enactment of Gramm-Rudman-Hollings in the United States, some economists and politicians had been calling for constitutional limits to federal spending and taxation. They wanted a rule to constrain the ability of legislators to spend without taxing and to prevent them from spending too much. The broad outlines of such a proposal were presented by James Buchanan, who subsequently received the Nobel Prize in economics, and Richard Wagner.(6)

In their view, such a rule should be transparent, so that it could be understood by both politicians and the public and so that voters could recognize when it had been breached. It should apply to the net deficit position of the budget, not to the level or composition of spending, decisions that would be left to the political process. Nevertheless, since the use of debt financing is believed to have led to high growth in government spending, such a rule would by its very nature curb government's appetite for spending by removing the ability to defer the tax consequences of doing so.

This proposal required both the President and the Congress to present and to approve budgets that projected zero deficits. Should the projections prove to be in error, automatic spending cuts would return the budget to balance within three months. Any surplus would have to be used to repurchase outstanding debt. The proposal offered a five-year transition period to budget balance and allowed for a one-year override if supported by two-thirds of Congress and the President.

This version seems straightforward. But so do others. The important feature is that it provided for after-the-fact reconciliation and automatic across-the-board, spending cuts. Sequestration [cuts] need not be initiated by some legislator, and thus legislative collusion could not be used to override the spirit of the law. The authors also saw an advantage in that the rule would not merely be another piece of legislation, but a constitutional amendment.

Other forms of automatic penalties could also be used. Non-compliance could result in automatic tax increases, reductions in legislators' salaries or an election call, with currently sitting legislators, or those voting for a non-complying budget, prohibited from seeking re-election.

The following comprehensive proposal has recently been suggested for the federal government in Canada:(7)

This represents a set of rules that would govern a stationary state, once budget balance had been achieved. It requires, in addition, a set of transition rules leading to budget balance as well as a determination of the appropriate level of federal spending. Because of the important role played by federal fiscal transfers to the provinces, any determination of appropriate federal spending should be undertaken with some degree of provincial consultation.

In 1985 the American federal government passed into law the Balanced Budget and Deficit Reduction Act, better known as the Gramm-Rudman-Hollings bill (GRH). That law presented a five-year schedule for reducing the deficit, with a balanced budget set for the sixth year. It was hoped this would be accomplished through a series of spending cuts. If the President of the United States and the Congress could not agree on the cuts needed, a series of mandatory cuts ("sequesters") would automatically take place. Thus in the absence of Congressional and Executive agreement, an automatic mechanism was to ensure that the schedule was in fact met.

In fiscal year 1986, the federal deficit was $221,000 million U.S.,(8) while the GRH target was set at $172,000 million. For 1991, the original GRH target, labelled GRH1 in Table 1, was zero, whereas the deficit came in at $270,000 million. But in 1987 the targets were revised upwards when it became evident that they could not easily be met. Comparing the actual budget deficit with this second round target of $64,000 million, labelled GRH2, indicates a miss of only $206,000 million.

While deficits actually grew during the years of GRH, some commentators have argued that they could have been even higher in its absence. Nevertheless, the American budgetary results failed to meet their targets and did not come close to balancing the budget. On balance, it appears that GRH was a failure.

There are several reasons for this failure. In the first place, the targets applied only to the deficits projected in the budget – they did not apply to the actual deficits. By understating expenditures, by overstating revenues and by forecasting excessively optimistic economic performance, budgetary figures could easily meet targets yet without any real hope of actually fulfilling these expectations at the end of the fiscal year. Moreover, legislators became adept at budgetary "slights of hand," re-arranging fiscal expenditure periods, moving items off budget, etc. In 1986, for example, Congress advanced a payment of $680 million so that it would take place prior to the coming into effect of GRH. This advance payment resulted in extra borrowing costs, prompting one Representative to admit: "Isn't it ironic that we are wasting half a million dollars trying to fool the taxpayer into believing that we are getting the deficit down, when in fact we are actually increasing it?"(9)

Since the GRH bill contained no provision for after-the-fact reconciliation of deficits with the GRH targets, it was the forecasts that were important in determining a possible breach of the law, not the results. Indeed, once the GRH was put in place, budgetary forecasts became extremely poor predictors of actual deficits, something that had not been true prior to the law's introduction.(10)

In addition, the sequestration process was extremely limited, with almost one-half of all expenditures exempt from automatic cuts and another 24% of total spending subject to only limited automatic cuts. In 1986, 48% of spending was exempt, of which almost one-half comprised social security spending.(11) The next large chunk of exempt spending, at 15% of the total, was interest on the public debt. Consequently, all the automatic cuts had to come from items comprising only 27% of the budget, and almost two-thirds of this was from defence spending; this led to intense political pressures not to make extensive use of the sequestration process.

According to Professor Aaron Wildavsky, the primary reason for the failure of GRH was the fact that it actually provided legislators with a strong disincentive to cut program spending when they knew that a sequestration process would ignore the very programs that were the sources of the deficit.(12)

GRH was replaced in 1990 by the Budget Enforcement Act (BEA). This new law shifted the focus of fiscal rules away from the deficit and towards expenditure control. While deficit reduction was expected to be a result of the BEA process, and maximum deficit levels were set in the law (See Table 1), it was no longer the primary goal.

The BEA established caps on discretionary spending, which for the first three years was divided into three categories: defence, international spending and domestic spending. After the third year, the program places limits on total discretionary spending. While the caps are meant to be permanent, certain circumstances, such as revised inflation forecasts, exceptionally slow growth (defined as growth for two successive quarters of less than 1% on an annual basis) or some kind of natural emergency, could lead to their alteration. Table 1 shows the dramatic increase in the targets for 1994 and 1995 resulting from these technical adjustments (labelled BEA2). They are twice as high as the original GRH limits for 1986, the year that was to have seen the deficit peak. Moreover the sequester provision with respect to the deficit does not apply for the first three years and then applies to what has effectively become a moving target.

There are several important features of the caps on appropriations.(13) Spending cannot exceed the cap, even if the additional expenditures are self financing or if taxes are increased. During the first three years, when three categories exist, spending cannot be transferred from one basket to another.

The caps described above apply to discretionary federal spending; they do not touch entitlement programs. The BEA legislation also establishes rules for revenues and mandatory spending programs, known as the pay-as-you-go rules. At the time of the BEA's passage, the baseline projections for revenues and mandatory spending were enacted in law. Any subsequent legislative effort to lower taxes or increase mandatory spending must now be deficit-neutral. This has enshrined the current entitlement shortfalls in the structure of the BEA.

The BEA, like the GRH, contains sequester provisions which take effect if the pay-as-you-go provisions are breached. But there are as many exemptions as in the GRH, so that cuts can be made in only a small subset of areas.

The BEA, like the GRH and any other piece of legislation, can be changed via the legislative process, though so far there appears to have been no attempt to do this. If the law remains in place, the legislative caps will shortly become binding, as outlays are now close to their appropriations limits. To date, however, the BEA has not been much more effective than the GRH in controlling the deficit, nor has it been particularly effective at curtailing growth in government spending.

The American budgetary process also contains a procedural rule that performs many of the functions of a fiscal rule, although it is not automatic. Both Houses of Congress divide spending appropriations among a vast number of Committees and Sub-Committees. In the House of Representatives, a Sub-Committee may exceed its total appropriations as long as its senior Committee does not; that is, aggregate limits apply to spending appropriations for the various Committees. In the Senate, on the other hand, a member may raise a point of order whenever the appropriations for a particular sub-Committee exceed its limits; this point of order can be overturned only by a three-fifths majority.

Such offset provisions provide a strong notion of opportunity cost at the Senate level. Whenever a politician or lobbyist wishes to enhance a program or create a new one, he or she must find offsetting spending reductions elsewhere. And because this provision can apply at the Sub-Committee level, it can force an offset, even though overall spending caps might not be binding.(14)

Fiscal Year	GRH1	GRH2	BEA1	BEA2	Actual
1986	172				221
1987	144				150
1988	108	144			155
1989	72	136			153
1990	36	100			220
1991	0	64			270
1992		28	360		290
1993		0	292		327
1994			162	345
1995			122	354

Source: S.E. Collender, The Guide to the Federal Budget, Fiscal 1993, The Urban Institute Press, Washington, 1992;

Economic Report of the President, Transmitted to the Congress January 1993, Washington, 1993; and

Congressional Budget Office, Sequestration Preview Report For Fiscal Year 1994, Washington, April 2, 1993.

Virtually every American state has a provision requiring a balanced budget, although the specifics are so complicated and varied that there appears to be some disagreement as to the extent of these controls. In some cases the governor must present a balanced budget, in others the legislature must pass a balanced budget, while in still others the budget must actually balance at the end of the year.

According to the United States General Accounting Office (GAO), Vermont and Wyoming are the only two American states without a legislative or a constitutional requirement for the state government to balance its budget, although in practice there is such an expectation for the government of Wyoming. Of the states with rules, 13 rely primarily on legislative controls while 35 use constitutional restrictions. On the other hand, the U.S. Advisory Commission on Intergovernmental Relations (ACIR), a body created by the United States Congress, reports that the state of Wyoming is bound by its constitution to balance its budget while the state of Ohio does not require the governor to submit a balanced budget nor does it require the legislature to pass a balanced budget. Instead it requires that "sufficient funds be made available for expenditures," an expression that would appear to permit borrowing as a means of providing sufficient funds.

According to the GAO, ten states do not require that their budgets balance at the end of the year, but, of the 40 that do, 11 allow for a carryover of a deficit if necessary, usually through the issuance of short-term debt instruments. According to the ACIR, only nine states may carry over a deficit, but most of these nine are not the states cited by the GAO. Thus it appears that balanced budget requirements may not always be what they seem and there is a difference of opinion as to the rules in place at the state level.

In most states, the "budget" to be balanced is not the consolidated budget of the government; it is not, therefore, as comprehensive as the budget of the United States federal government. Unlike the federal government, the states generally separate their accounts into capital and current (operating) budgets. They generally finance large-scale capital spending, accounting for about 9% of total expenditures, through borrowing. Mandatory spending may also run a deficit in some states. As a rule, the general budget consists of discretionary spending on one side and general tax revenues on the other and it is only this budget that needs to be balanced. Such "general fund" spending accounts on average for 54% of total spending, ranging from a high of 74% to a low of 21%, according to the GAO. Where other parts of the budget are required to balance, receipts from long-term borrowing are often treated as revenue.

These rules do, nevertheless, figure prominently in the fiscal actions taken by state governments to balance their budgets. State governments have, in response to unforseen budgetary difficulties, taken actions to curb potential deficits to the tune of about $40 billion in the past couple of years. And figures from the United States Department of Commerce indicate that the state government sector is far healthier than the provincial government sector in Canada. Nevertheless, states with balanced budget requirements still run deficits. The GAO study indicates that in 1992 Illinois and Maryland ran deficits, while in the past three years another six states had to carry over a deficit. New York, a state with a long history of circumventing its balanced budget rules, ran total deficits of $2,387 million in the past three years. Connecticut issued short-term bonds of $965 million to cover two years of deficits and Massachusetts borrowed $1,400 million to cover a deficit. All of these states had balanced budget provisions.

Moreover, our own analysis of U.S. Department of Commerce data indicates that 24 states had general expenditure levels in excess of general revenue levels for fiscal year 1991,(16) where this is calculated by excluding insurance and trust fund expenditures and revenues from the totals. This does not mean that these same governments had a deficit under their particular accounting and legislative regimes; however, it does indicate that they suffered a revenue shortfall. The year 1991 was, however, an unusual one for state governments; data for earlier years indicate that revenue shortfalls were very much the exception rather than the rule.

American state governments have much less access to tax revenue than the federal government and their bond ratings are consequently very sensitive to the risk perceptions of bond markets. It is this sensitivity that, on the one hand, controls deficit financing but that, on the other, could lead to the use of fiscal rules to restrain legislators. A recent study found that state governments with more prudent fiscal policies (defined as low debt and low deficit trends combined with strong fiscal rules) do pay lower borrowing costs than states with less prudent fiscal policies, other things being equal.(17)

Evidence suggests that state government spending is partly driven by forecasts of revenues, which tend to be underestimated. Such practice has moderated the effects of unexpectedly low revenues at the end of a fiscal year and has lowered the probability that balanced budget rules need to be invoked.(18) Similar forecasting requirements have been put in place in the Alberta Deficit Elimination Act.

In the past, some state governments tried to circumvent their own rules by redefining some current expenditures as capital spending and financing them via deficits. Financial markets quickly saw through such measures leading to fiscal crises.(19) State government bond ratings are very sensitive to their fiscal positions, and their potential for change appears to provide strong motivation to balance budgets.(20)

In the American federation, the federal government plays a stronger economic role than its counterpart in Canada. In the United States the federal government controls about two-thirds of all public spending, whereas in this country the federal and provincial governments are about equal. Comparisons between provinces here and states in the U.S. should bear these jurisdictional differences in mind.

During the 1980s, when President Reagan occupied the White House and the Democrats controlled the Congress, conservative Republicans called repeatedly for a "line item veto" power as a way of curtailing excessive expenditures and the deficit. At present, though the President can veto a bill, he cannot veto parts of a bill and it is common for a bill to have all sorts of unrelated spending measures attached.

At the state level, governors tend to have a great deal of discretion as to how they deal with a bill. In 40 states, the governor may veto or alter a line item. In 32 states he may veto funding for a program or agency. In other states he may veto the language of a bill, which governs the way funds are to be spent.(21) These gubernatorial powers, rather than formula-based cuts, are generally how balanced budget rules are enforced.(22)

This particular feature, while important in a Congressional form of government, has little relevance to the parliamentary form of government in place in Canada.

In addition to balanced budget rules, several states also face legislative or constitutional limits on the growth of expenditures and /or tax revenues. Since 1976, such limits have come into force in a large number of states,(23) with four states adding or amending such limits in 1991. As with the budget balance rules, these limits can have either legislative or constitutional authority, and they can be written in a wide variety of ways. The most common way is to limit expenditure growth so as to keep it from growing in real per capita terms, or to keep it constant in terms of state personal income. In some cases, there is a numerical limit that can become outdated with changing circumstances. Rhode Island, for example, restricts expenditure growth to 6% per annum and Oklahoma restricts it to 12%, adjusted for inflation, with the added proviso that appropriations be no more than 95% of certified revenue. In today's low inflation climate, the Rhode Island restriction is not very limiting and the Oklahoma restriction can provide no limit at all if revenues grow quickly enough.

Also, like the balanced budget rules, these TELs apply to only a subset of revenues and expenditures, usually those described above as general revenues and expenditures. While these count for 60% of totals on average, they range from a low of 25% to a high of 86% for the states covered by TELs. Moreover, as with all sorts of rules, legislators can be creative in circumventing them through creating "special funds," "tax relief funding" or other such means. Thus one might suggest that the limits have really not been effective.(24)

There are several possible reasons for this. These limits have generally been crafted by the same legislators that the voters had already deemed to be fiscally irresponsible. As well, some limits include relatively painless escape clauses and thus have been rather easy to change. The evidence suggests that revenue and expenditure growth in states with TELs have grown at about the same rate as in those without. This first-glance assessment may be wrong, however, since it is also true that the real rate of growth of state governments has declined substantially since TELs were introduced. The existing TELs, and the current demand in some states for even tougher variants, could in fact be effectively prompting all state governments to curtail their growth.

STATE	LIMIT ON: SPENDING/ REVENUE	COVERAGE: (% of revenue/spending subject to limits)	LIMIT RULE: expressed as growth rate, unless specified otherwise
Alaska	Spending	86%	inflation & population
Arizona	Spending	73%	7% of state personal income
California	Spending	42.5%	inflation & population
Colorado	Spending	44%	was 7%, changed in 1991
Connecticut	Spending, with 4 exceptions	N/A	greater of inflation or personal income
Delaware	Spending	N/A
Hawaii	Spending	57%	personal income
Idaho	Spending	45%	5.33% of state personal income
Louisiana	Spending – Revenue limited until 1991	39.5%	personal income
Mass.	Revenue	N/A	wages & salaries
Michigan	Revenue	70%	personal income
Missouri	Revenue	70%	personal income
Montana	Spending	57.5%	personal income
Nevada	Spending	70%	inflation & population
North Carolina	Spending	N/A	7% of state personal income
Oklahoma	Spending	N/A	12% + inflation, but not more than 95% of revenues
Oregon	Spending	25%	personal income
Rhode Island	Spending	74%	6%
South Carolina	Spending	67.5%	greater of personal income growth or 9.5% of state personal income
Tennessee	Spending	67.5%	personal income
Texas	Spending	57%	personal income
Utah	Spending	75%	85% of increase in personal income
Washington	Revenue	79%	personal income

Source: Table 2 is compiled from information in the following publications: ACIR (1993);

D.G. Bails, "The Effectiveness of Tax-Expenditure Limitations: A Re-Evaluation," American Journal of Economics and Sociology, Vol. 49, No. 2, April 1990, p. 237; and

D.A. Kenyon and K.M. Benker, "Fiscal Discipline: Lessons From the State Experience," National Tax Journal, Vol. XXXVII, No. 3, September 1984, p. 437.

The 1991 budget of the Government of Canada contained a number of items designed to control the extent of government, limit the growth of the debt and better co-ordinate fiscal and monetary policy. The budget also included a set of fiscal rules under the auspices of Bill C-56, An Act Respecting the Control of Government Expenditures, better known as the Spending Control Act, which were designed to provide an upper limit to much, but not all, program spending.

In addition to its deficit projections, the federal government established legislation controlling the amount that it could spend on programs. The intent of the legislation was not to control total spending, although the draft legislation had contained "floating" targets for this. Ultimately, the government aimed to limit "controlled program spending," amounting to on average, 85% of baseline program spending and 62% of projected total spending. (These are shown in Table 3 under the column C-56.) Uncontrolled program spending consists essentially of Unemployment Insurance benefits, while debt service charges, not being program spending, are uncontrolled.

When the draft legislation was put out for comment, the House of Commons Standing Committee on Finance held public hearings on the subject and made recommendations, some of which the government accepted.

During the investigation, the Canadian legislation was frequently compared with the American. The Canadian legislation does not attempt to control the deficit directly; it controls only spending, though it is part of a fiscal package containing five-year deficit projections. It was felt at the time that spending is inherently controllable, whereas the deficit is not. The legislation requires that the limits be met on an ex post as well as on an ex ante basis. Not only must the Minister of Finance place before Parliament budgets that conform to the legislation, he must take corrective action if the targets are not met in any year. Such ex post reconciliation does not take place in the American legislation.

The law differs from its American counterpart in two other respects as well. The American BEA sets spending limits in three broad categories and thus limits flexibility in spending. Moreover, the offset provisions of the US Senate rules, if used to their full extent, make for even less flexibility. This is not the case with the Spending Control Act. Within the category of controlled program spending there is full flexibility to shift resources from one program to another. In addition, the government may overspend in one year if it underspends subsequently. Canadian fiscal policy therefore lacks the inertia of its American counterpart. On the other hand, the American law and practice make the trade-offs in proposed spending plans more apparent.

The American deficit limits can be altered during periods of recession whereas the Canadian limits cannot. However, since uncontrolled program spending consists largely of Unemployment Insurance payments, and these are the most cyclically sensitive of program spending, the structure of the program itself effectively responds to recessionary pressures. In addition, the yearly flexibility in the limits set also accommodates recessionary pressures.

Another unique feature of the federal program is its limited duration; the law expires at the end of fiscal year 1995-96. Because of this, any breach of the law's provisions in the last year would not be confirmed by the Auditor General until after the law had expired and would thus generate no consequence other than the public disapproval.

The government chose from the start not to control total spending, leaving debt service charges out of the law's reach. The reason for this was largely practical; the Department of Finance has found it much more difficult to forecast debt servicing costs than to forecast total program spending and set the program to control what seemed inherently controllable; thus the exclusion of debt servicing charges reduced the volatility in the target variable. Nevertheless, as several witnesses pointed out, the control program aimed to limit the impact of spending on deficit and debt growth, but these are affected by spending on interest charges as much as by spending on other programs.

Another issue raised during the public hearings was the size of the caps in relation to trends in government spending. The proposal was popularly described as limiting program spending growth to 3% per year. Controlled program spending had a higher effective growth rate because the uncontrolled part was expected to grow more slowly.

More importantly, though, with inflation at unprecedented low levels, these spending limits represented real growth higher than had taken place prior to their introduction. Moreover, nowhere in the official discussion of these limits was there any rationale for the numbers chosen; for example, there was no set target in terms of percentage of GDP to be devoted to total, program, or controlled program spending.

The spending limits amount to a decline in the ratio of controlled program spending to Gross Domestic Product (GDP) from 14.22% in 1991-92 to 12.5% in 1995-96, based on the forecasts for GDP growth in the 1992 federal budget. Those estimates of GDP growth have now proven to be overly optimistic.

The 1993 federal budget announced that the limits would be further reduced and the program extended for an additional two years. These changes were presented in Bill C-130, An Act to amend the Spending Control Act, which did not go beyond first reading in the House of Commons. Part of the reduction in the limits was done for technical reasons. The restructuring of the child benefit system amounts to a conversion of some program spending into a tax expenditure. Such a change could have represented a loophole in the Act. In addition, legislation has made public sector pension plans fully funded, with the government no longer required to cover shortfalls through separate payments. These reductions are expected to amount to approximately $1,700 million in 1992-93 and $3,500 million thereafter.(25)

But the cuts go beyond such technical changes required to keep within the spirit of the original controls. For 1995-96, for example, the reduction in the cap is $6,450 million and it is $5,600 million for the previous year. The proposed limits in Bill C-130 represent an annual increase in controlled program spending of 2%, down from the 3.4% increase inherent in the original law. Total program spending is expected to grow by about 1.85% per annum. These relatively low rates of growth are more consistent with a low inflation environment than the amounts in the existing law.

But, despite the fact that the growth of spending was reduced, the lowering of the economic growth forecasts means that controlled program spending would represent a higher portion of GDP than was originally hoped for under the existing law. Since the forecasts in the 1993 budget are also likely to be overly optimistic, this problem is even further compounded. Moreover, once we take into account the structural changes in spending resulting from the revision of the child benefits, we see that controlled program spending as a proportion of GDP is higher every year under the C-130 proposals; this increase is equivalent to 0.6% of GDP, a significant amount.

In the 1991 budget, spending limits were promised as part of a comprehensive fiscal plan to reduce the deficit. It was projected at that time that the deficit for 1993-94 would be about $16,600 million. It now looks as if the figure will be more than twice as high, even though there has been no breach of the Spending Control Act, in letter or in spirit.

In May 1993, the provinces of New Brunswick and Alberta put in place legislation establishing a schedule for balancing the budget and mandating zero deficits thereafter.

The Alberta Deficit Elimination Act set a zero deficit target for fiscal year 1996-97 and every year thereafter. During the transition to a zero deficit, target deficits are $2,500 million for 1993-94, $1,800 million for 1994-95, and $800 million for 1995-96. The deficit is measured on a consolidated basis; it includes net spending on the capital account. The Act applies specifically to the forecast deficit in any budget. The interim years leading up to 1996-97 can have deficits that exceed the target but the subsequent year's target must be reduced by any such excess. There is no provision in the law to allow an excessive deficit in 1995-96 or after. Should the deficit in any year be less than allowed, future deficits cannot be increased. Similarly, should revenues in any one year exceed expectations, the excess must go to paying the debt.

YEAR	C-56		C-130			SPENDING
	in million $	as % of GDP*	in million $	as % of GDP**	as % of GDP, adjusted*	in million $
1991-92	97,200	14.22	97,200			95,671****
1992-93	101,000	14.13	99,300	14.4	14.7	101,538****
1993-94	104,100	13.5	99,300	13.8	14.3	99,300
1994-95	107,400	13.0	101,800	13.2	13.7	101,800
1995-96	111,250	12.5	104,800	12.7	13.1	104,800
1996-97			106,600	12.1	12.5	106,600
1997-98			109,200	11.8	12.1	109,200

*** Based on GDP projections in 1993 budget and accounting for structural changes in spending programs.

Since the law applies to projected deficits, it is important that revenue and spending forecasts be reasonable. Alberta relies heavily upon revenues from non-renewable resources, which can fluctuate widely. The law specifies that budgetary forecasts for such revenues cannot exceed the average of the five most recent years for which data are available.

The law also requires that the provincial Treasurer provide quarterly reports on the government's fiscal position, indicating whether its budgetary targets are on track. If not, the government is to take corrective action. (Such action was taken in October 1993.) In this way, a limit on forecast deficits can almost become a limit on actual deficits.

The law contains no explicit penalties for failure to meet the budgetary targets. An amendment was proposed stating that legislators must take a pay cut and the Premier resign should the target be missed by more than 4%. The proposal was defeated, however.

In meeting these targets, the government of Alberta expects to see a dramatic decline in its operating deficit, from $2,777 million in 1992-93 to a surplus of $510 million in 1996-97. Added to the operating deficit is an amount for net capital investment as well as the deficit (surplus) for other funds. These other funds are in an overall surplus position, largely as a result of the surplus in Workers' Compensation Board. Also deducted is the amount of capital amortization, which represents the amount of depreciation of physical capital and is already accounted for in operating expenditures.(26) (27)

The province of New Brunswick put into law in 1993 An Act respecting the balancing of the ordinary expenditures and ordinary revenues of the province, which requires budgetary balances. The law establishes a series of fiscal periods, the first being from 1993-94 to 1995-96, with subsequent periods of four years, essentially the period between elections. The provincial "ordinary account" must balance within each period. Consequently any deficits must be matched by equivalent surpluses. This law applies to after-the-fact fiscal performance as reported in the Public Accounts; it does not apply only to budgetary forecasts.

The Public Accounts of the province are to track the annual and cumulative record of the government in complying with this law. Exempted from the provisions of the law are fiscal consequences, taking place near the end of the fiscal period, of changing entitlements resulting from revised estimates of the Government of Canada with respect to fiscal transfers or tax collection agreements.

It is possible that, at the end of a period, the law could be breached due to unforseen circumstances. The final accounting does not take place until some time after the fiscal period has ended. There are, however, no provisions for penalties in the event of non-compliance. An excessive deficit is not carried over into another period. The drafters of the legislation see public opinion as the final arbiter of government performance.

What is this law attempting to control? The statement of ordinary revenue and expenditure for 1991-92 showed total revenues of $3,689.1 million, expenditures of $3,906.6 million and a deficit of $217.5 million. Revenues include essentially all tax receipts, royalties, profits from the liquor corporation and other revenue, including transfers from the federal government. Expenditures include virtually all current expenditures, excluding net capital expenditures. Thus the increase in the net debt is not equal to the ordinary deficit; rather, it is two-thirds higher.

Net capital expenditures in that year amounted to $293 million; the special purpose fund was $5 million in deficit and offsetting this was the $149 million in earnings from the government's sinking fund. These amounts also affect the level of the net debt, but are not regulated by the operating controls.

The design of the New Brunswick legislation is quite different from its Alberta counterpart. The former is designed to balance actual operating account deficits over a four-year period while the latter is designed to balance projected comprehensive budgets on an annual basis, with a mechanism for prompt action should the projections get derailed. Unlike the Alberta legislation, the New Brunswick law also contains some escape clauses related to changes in federal transfers.

Neither law contains any kind of automatic formula to come into play if the law is breached, neither contains any kind of penalty clause, and neither attempts to set any kind of limit on the spending side. Since provincial governments do not have the power to print money, deficit control might be sufficient for spending control in the provinces.

The Maastricht Treaty of the European monetary union contains several clauses dealing with the fiscal policy of member states of the EC. These clauses set out some preconditions for full monetary union and the creation of a single currency under the auspices of a European central bank. These requirements are thought to be necessary for the effective functioning of a single monetary regime.

There are four economic convergence criteria for each member state: inflation cannot be substantially higher than (i.e., no more than 1.5 percentage points above) the average inflation rate of the three best performing member states; exchange rates should be relatively stable prior to the state's joining the EMU; nominal interest rates should converge with those of the three lowest inflation countries (i.e., they should be no more than 2 percentage points above those of the best performing member states); and public deficits should be below 3% of GDP, with public debt no more than 60% of GDP.(28) The last criterion will be the subject of this section.

As of 1992, the budget balance conditions were met only by Luxembourg, Ireland, Denmark and France. Germany just narrowly missed the mark, while Italy and Greece were not even close. The public debt criteria were met by Luxembourg, Britain, Germany, Spain and France, with Denmark just off the mark and Italy again not coming close. The worst offender in this category was Belgium.(29) The estimates for 1993 are even worse, with only Luxembourg meeting the deficit conditions.(30)

The status of Germany is likely to create little problem. It is comfortably below the debt condition and its deficit status is more the exception than the rule. The real problem lies with countries like Belgium, Italy, Portugal and Greece, all of which are well out of step with respect to both debt and deficit, indicating loose fiscal policies at present as well as a history of such policies.

The fiscal rules apply to the general government sector for each member nation; that is, they apply to the sum of the central, regional and local governments as well as the social security accounts. State-owned commercial enterprises are excluded from this calculation. The deficit is defined on a National Accounts basis and the debt rule refers to gross, not net, debt.

Article 104B specifies that the European Commission is to monitor fiscal developments in member states. These member states must regularly report both planned and actual deficits. The Commission is to examine deficit and debt results to see if they comply with the fiscal convergence requirements. If they do not, the Commission is to determine whether the fiscal variables are moving away from the required levels or whether the deviation appears to be only temporary. If these two criteria are not met, the Commission is to prepare a report, taking into account the economic circumstances in the member state and the deficit in relation to government investment spending.

The Commission first writes and presents a report to the Council of Ministers, which makes recommendations to the non-compliant member state; these could be made public in the event of further non-compliance. If this is deemed to be insufficient, the Council can insist that a timetable for compliance be drafted. If this also fails, the non-compliant nation can be obliged to issue warning statements when issuing new debt.(31)

More severe sanctions can also be applied. The European Investment Bank can restrict its lending to the offending nation. The non-compliant state can be required to place interest-free deposits with the EC until its fiscal position is improved, and could eventually be fined an appropriate amount.(32)

Article 104 and its sub-articles, which contain the stipulation that member states avoid excessive deficits, commences with provisions limiting the role of the European central bank in financing the debt of member states and prohibiting a bailout of excessively indebted states. The article also makes it clear that the indebtedness of one member state is not the responsibility of another state, although nothing prevents members from aiding others should they wish to do so.

These fiscal rules were designed to enhance the monetary credibility of the EMU and the proposed European central bank. A new central bank and a new currency without any history lack the credibility that other currencies, especially the Deutschmark, now enjoy. This is one reason why the treaty provides the new bank with an explicit instruction to pursue a policy of price stability.

There are, then, several reasons why excessive fiscal policies in any one member state are to be avoided. The possibility of bailout by other members could put them in a more precarious fiscal position and heighten pressures on the central bank for monetary accommodation. It could also provide a perverse incentive for members to allow their own irresponsible fiscal actions to lead to transfers of resources from other countries.

Secondly, excessive fiscal deficits and debt in one member state, if not bailed out, could spread to other countries, again putting pressure on the central bank to resolve the crisis via money creation. Finally, a deficit crisis and subsequent higher interest rates in one part of the union might spread to the rest of the community because of the monetary union. Such higher interest rates lead to higher deficits both directly, given no offsetting spending or revenue changes, and indirectly, because of their dampening effects on economic performance.

On the other hand it is instructive to recall that there is scope for some laxity the operation of these rules, which are really more like guidelines. There are good reasons for this. Temporary economic shocks do not hit all Europe equally. That is why Europe as a whole does not constitute an optimal currency area. Monetary adjustment is one way in which such shocks have been accommodated in the past. The single currency arrangement takes away this option. Consequently, fiscal policy is the only remaining tool for regional stabilization.(33) Without fiscal rules, it is feared that member states might be tempted to export the costs of their own adjustment to other states by means of lax fiscal policies.

If we think of Europe under Maastricht as a federal state, with the member nations as its "provinces," it is easy to compare it with Canada. In the late 1980s many provinces were concerned that the fiscal policies of Ontario were damaging their own economies, both directly and indirectly as a result of the actions of the Bank of Canada. In addition, early in 1993, there was short-lived speculation that the Government of Canada was looking at ways to intervene should a province be shut out of international capital markets and thus find itself unable to refinance its debt. These are the kinds of difficulties foreseen by the drafters of the Maastricht rules.

While the speculation about the second issue proved to be incorrect, it did heighten public awareness of debt problems and prompted the C.D. Howe Institute to publish a short memo,(34) arguing that the federal government should make it clear it would not bail out provinces in the event of default. Such a strategy, while it might increase the borrowing costs of some provinces, could also reduce the borrowing costs of the federal government and indeed all Canadians. The provinces would have to take the steps necessary to get their fiscal houses in order and fears that Canadian debts might eventually be monetized would be alleviated. Such fears are important because the Bank of Canada's balance sheet indicates that its holdings of government debt represent only a small part of its overall asset holdings and the amount held is small in relation to the annual amount of provincial net borrowing. Thus any purchase of provincial issues would substantially increase the size of the monetary base, unless offsetting actions were undertaken. Moreover, such a bailout would penalize taxpayers in provinces that had run more prudent fiscal policies.

These concerns are very similar to those that have been used to justify the fiscal convergence rules of the Maastricht Treaty. The need for such rules is less obvious in Canada than it is in Europe. In the first place, Canada has a well established and reputable currency and central banking system. Secondly, the provinces are not the sole agents for engaging in regional stabilization policies; the federal government, through its tax and spending policies, also provides a large degree of regional stabilization. This is in sharp contrast to the European monetary union, where virtually all fiscal stabilization would take place at the regional (i.e. national) level. Finally, there is likely to be a stronger sense of community in Canada than in Europe, where the community is only now being built; consequently it is less likely that one region would attempt to pass on its adjustment costs to others.

Recently, a senior official of the Ontario government has suggested that the federal government take over responsibility for virtually all provincial government debt, in exchange for a constitutional requirement that the provinces balance their budgets in the future.(35) This represents a very strict version of the Maastricht rules, one that places virtually all fiscal responsibility at the federal level.

The United States has now had quite an extensive experience with fiscal rules at the federal, state and local levels. This experience, especially at the federal level, makes it clear that rules by themselves are not sufficient to ensure balanced budgets. This might, however, simply reflect the quality of the rules put in place. State governments, on the other hand, have experienced a much better fiscal performance in recent years. They have much more extensive fiscal rules, and, although these also leave room for avoiding the constraints, the evidence suggests that these states have taken dramatic steps in recent years to avoid breaches of the law. (At the federal level, on the other hand, the response has been to change the law in the face of potential violations.) When the state rules are coupled with the strong veto powers of the executive branch (something that is lacking at the federal level), they appear to be effective tools.

The American municipal experience is also relevant. In the 1970s, New York and Cleveland defaulted on their debt while Philadelphia came precariously close to doing so.(36) All these cities were obliged by law to balance their books, yet their respective fiscal crises show clearly that they violated the spirit, if not the letter, of the law for quite some time. This was possible because the relevant laws applied to budget proposals not budgetary results (much like the GRH law), bookkeeping "innovations" masked the true state of government finances, and local governments could draw upon other sources of funds, for example by increasing the extent of unfunded pension liabilities.

When judging the relevance of the American experience to Canada there are two things to keep in mind. First, there is a different division of powers between the federal and state governments in the two countries. Second, the political pressures in a Congressional system of government differ from those in a parliamentary system, even though they both appear to give similar types of fiscal results.

The evidence indicates that there are strong political pressures on governments to provide spending programs without imposing taxes on the beneficiaries of those programs. Fiscal rules do not remove those pressures, they merely constrain the response. But there are many ways to achieve the same result and rules that are not comprehensive are likely to fail.

What does experience teach us about the proper design of fiscal rules? We know that in their absence the federal government has not had great success in curtailing the growth in its net debt. The apparent success of the Spending Control Act could be as much due to its generous limits as to the government's ability to control its own spending.

A fiscal rule must do more than have legislators go through the motions of fiscal control. In the United States, new management techniques such as zero base budgeting and sunset laws are widely used, yet their performance has been very disappointing.

It is also apparent that controls must be applied ex post, not just ex ante, and that the results be monitored in some systematic and consistent fashion. Constitutional amendments could be more effective because they would make it more difficult to alter the rules; however, constitutional change is difficult to achieve. A legislative rule with oversight by, for example, the Auditor General could prevent some of the more blatant abuses.

Finally, to be effective a rule must imply action in the event of non-compliance. The provincial laws ultimately appeal to the "court of public opinion"; however, public opinion did not prevent the financial problems that led to the demand for fiscal rules. The GRH experience indicates that the non-compliance sequesters should apply broadly, and should not exempt favoured programs. While cuts need not be draconian, they would likely be more effective if they generated results that politicians, bureaucrats and voters found undersirable, for example an across-the-board reduction in government salaries and personal transfers.

According to the Public Choice view, existing political constraints are not sufficient to curtail excess government and excess deficits and the court of public opinion does not work. Fiscal rules of one sort or another are needed, but poorly designed or insufficiently comprehensive rules might even be counterproductive, leading to more subterfuge on the part of legislators and complacency among the public.

(1) United States General Accounting Office, Balanced Budget Requirements: State Experiences and Implications for the Federal Government, Washington D.C., March 1993, GAO/AFMD-93-58BR, p. 1 of letter of submission.

(2) Congress of the United States, Congressional Budget Office, Reducing the Deficit: Spending and Revenue Options, A Report to the Senate and House Committees on the Budget, Washington, D.C., February 1993, p. 7.

(3) A. Wildavsky, The New Politics of the Budgetary Process, Second Edition, Harper Collins Publishers, New York, 1992, p. 241-247.

(4) A. Wildavsky, How to Limit Government Spending, University of California Press, Berkeley Calif., 1979, p. 44-47.

(5) H.G. Grubel, "Constitutional Limits on Government Spending Deficits and Levels in Canada," in H.G. Grubel, D.D. Purvis, and W.M. Scarth, Limits to Government – Controlling Deficits and Debt in Canada, The Canada Round: A Series on the Economics of Constitutional Renewal – No. 13, J. McCallum, Series Editor, C.D. Howe Institute, Toronto, 1992, p. 25-27.

(6) J.M. Buchanan and R.E. Wagner, DEMOCRACY IN DEFICIT – The Political Legacy of Lord Keynes, Academic Press, New York, 1977.

(8) All references to American budgetary numbers are in U.S. dollars. The fiscal year 1986 refers to the 12-month period ending 30 September 1986.

(10) D. Altig, "Some Fiscal Advice for the New Government: Don't Let the Sun Go Down on BEA," Economic Commentary, Federal Reserve Bank of Cleveland, 1 February 1993.

(13) S.E. Collender, The Guide to the Federal Budget, Fiscal 1992, The Urban Institute Press, Washington, D.C., 1991, p. 25.

(15) Much of the material in this section comes from the following publications:
United States General Accounting Office (GAO), Balanced Budget Requirements: State Experiences and Implications for the Federal Government, Washington D.C., March 1993, GAO/AFMD-93-58BR; and
Advisory Commission on Intergovernmental Relations (ACIR), Significant Features of Fiscal Federalism: Volume 1, Budget Processes and Tax Systems, Washington, D.C., February 1993.

(16) U.S. Department of Commerce, State Government Finances, 1991, GF/91-3, September 1992.

(17) M. Goldstein and G. Woglom, "Market-Based Fiscal Discipline in Monetary Unions: Evidence from the US Municipal Bond Market," in: M.B. Canzoneri, V. Grilli, and P.R. Masson eds., Establishing a Central Bank: Issues in Europe and Lessons from the US, Cambridge University Press, Cambridge 1992, p. 228-260.

(18) J.G. Stotsky, "Coping with State Budget Deficits," Federal Reserve Bank of Philadelphia Business Review, January-February 1991, p. 16.

(19) R.P. Inman, "Anatomy of a Fiscal Crisis," Federal Reserve Bank of Philadelphia Business Review, September/October 1983, p. 15-22.

(23) D.A. Kenyon and K.M. Benker, "Fiscal Discipline: Lessons from the State Experience," National Tax Journal, Vol. XXXVII, No. 3, September 1984, p. 433-446.

(24) D.G. Bails, "The Effectiveness of Tax-Expenditure Limitations: A Re-Evaluation," American Journal of Economics and Sociology, Vol.49, No. 2, April 1990, p. 223-238.

(26) Government of Alberta, Budget '93 – Update, Edmonton, 8 September, 1993, p. 39.

(27) The reported fiscal position of the government has not been affected by the $273-million gain from sale of shares in the Alberta Energy Company. This amount went directly to reducing the net debt of the province.

(28) W. Buiter, G. Corsetti, and N. Roubini, "Excessive Deficits: Sense and Nonsense in the Treaty of Maastrict," Economic Policy, Vol. 16, April 1993, p. 58-100; and Amendments to the EEC Treaty – Economic and Monetary Union – as Agreed in the European Council of Maastrict on 10 December 1991.

(29) M. Butler, "Europe's Currency Tangle," The Economist, 30 January 1993, p. 21-23.

(30) "A Rude Awakening," The Economist Survey of the European Community, 3 July 1993, p. 14.

(31) M. Fratianni, J. von Hagen, and C. Waller, The Maastricht Way to EMU, Essays in International Finance, No. 187, Princeton University, Princeton N.J. June 1992, p. 39-40.

(32) K. Habermeier and H. Ungerer, "A Single Currency for the European Community," Finance & Development, September 1992, p. 26-29; and
R. Owens and M. Dynes, Tuttle Guide to The Single European Market, Charles E. Tuttle Company, Inc., Boston, 1992.

(33) M. Artis, "The Maastricht Road to Monetary Union," Journal of Common Market Studies, Vol. XXX, No. 3, September 1992, p. 299-309.

(34) T.E. Kierans, D.E.W. Laidler, and W.B.P. Robson, The Hazards of a Federal-Provincial Debt Bailout, C.D. Howe Institute Backgrounder, March 16, 1993.

(35) G. Yonk, "Bold Plan Would Give Provinces Debt Relief," The Globe and Mail (Toronto), 29 November 1993, p. A1, A3.