Policy Implications of Tax Evasion and the Underground Economy

Jonathan R. Kesselman

Introduction

There is little agreement about the size of the underground economy (UGE) relative to the total economy. No matter how small or how large the UGE might actually be, it has important implications for public policy.  --   This paper concentrates on the public-finance and particularly the taxation implications of tax evasion and the UGE. For earlier analyses of UGE implications for macroeconomic phenomena, including cyclical fluctuations, price stability, economic growth, labour productivity, economic development, and monetary policy, see contributions in Gaertner and Wenig (1985) and Feige (1989).Note   --  These relate to matters of economic efficiency and distributional equity as well as the design, operation, and enforcement of the tax system. Even if the UGE were less than 5 percent of GDP, it could still be vital if its efficiency or distributional effects were significantly adverse. Conversely, a UGE as large as 15 or 20 percent of GDP could be of minimal concern if its efficiency and distributional effets were small or viewed as favourable.

The efficiency and distributional impacts of these overlapping but not congruent fields of tax evasion and UGE activity are of obvious interest in themselves. Yet we also need to consider the tax revenues lost to non-compliance and their economic repercussions. The lost revenues may be recouped through higher taxes on compliant taxpayers, or they may result in lower levels of public spending for various purposes. Each of these compensating policy reactions exerts its own efficiency and distributional impacts.

Since there is so much controversy about the scale of the UGE and tax evasion, we might expect knowledge about their efficiency and distributional effects to be hopeless at a quantitative level. However, economic theory and analysis can be helpful in identifying what features of UGE and evasion activities should be observed to cast light on these policy-relevant issues. We also need guidance about the kinds of evasion, the sectors of the economy, and the types of participants that should be of particular concern for policy purposes. Insights of these kinds should be instructive in guiding society's mobilization of resources to combat evasion and in designing effective strategies for countering these practices.

Such analyses must give due recognition to both the resource costs and the resource savings of alternative approaches to handling the UGE and evasion. Two major types of policy responses may be considered: 1) "ex post" policies related to information reporting, detection methods, and enforcement measures, based on a given mix, structure, and rate of tax; and 2) "ex ante" policies incorporated in the design of the tax system itself with its associated administrative structures and procedures.

In this paper, we draw on insights from an active and growing economic literature on tax evasion and the UGE. We address the policy-relevant issues described above in a relatively non-technical fashion. To provide a more precise context for these phenomena, a distinction will be drawn between UGE activity, which involves tax evasion, and pure tax evasion. We will then offer a discussion of the relationship of the UGE and benefit fraud-the counterpart to tax evasion in the transfer system.

Following a review of some basic insights from the economic analysis of tax evasion and the UGE, we will examine the markets and mechanisms by which the efficiency and distributional effects of these activities operate. Finally, we attempt to extract some key implications of the analysis for the two major public policy aspects: enforcement (the ex post aspect) and tax and transfer design (the ex ante aspect).

UGE versus PTE

When attempting to assess the efficiency and distributional effects of tax evasion, it is useful to distinguish two forms. The first is tax evasion associated with activity in the UGE, which involves the productive supply of goods or services. This entails the use of labour, managerial, or entrepreneurial services, along with the requisite tangible capital, to operate a productive activity. The second is tax evasion not associated with participation in the UGE or the supply of labour or similar services. Such pure tax evasion (PTE) usually involves financial manipulation, the non-disclosure of capital or financial incomes, the overstatement of tax deductions, or misstatements of individual circumstances.

We will now briefly consider these types of tax evasion in greater detail, since each form has different potential effects on economic efficiency. The avoidance of taxes through such legal means as tax shelters and tax incentive provisions does not constitute evasion and therefore lies beyond our present scope.   --  See, for example, Slemrod (1994). The author of the present paper is undertaking an analysis of tax avoidance behaviour using general equilibrium techniques described later.Note  --   Our treatment of the UGE does include the production of legal as well as illegal goods and services. The tax evasion associated with legal activities in the UGE is itself illegal, of course, and may be liable to criminal penalties.  --   A telling commentary on this point was expressed by the chairman of a Vancouver brokerage house, Peter Brown, when a client of the firm was found to have evaded taxes: "There was no money laundering here. I associate money laundering with criminal activity. Lots of people cheat on their taxes." Globe and Mail, July 29, 1994, p. B1.Note  --    Legal activities in the UGE include a wide range of goods and services that are also produced in the legitimate or "above-ground" economy. Common examples are home repair and renovation, food and entertainment, auto repair, gardening, babysitting, and the like.

The illegal portion of the UGE includes such "productive" activities as distributing illicit drugs, smuggling and trading in contraband goods, prostitution, and unauthorized gambling. These items constitute goods and services that are transacted between willing buyers and sellers, even though they are illegal per se. By contrast, purely "appropriative" activities are illegal but are excluded from our notion of UGE activity because they do not add to society's total level of consumer satisfaction. Examples here include extortion, blackmail, and many forms of theft such as robbery, shoplifting, embezzlement, fraud, and forgery. Participants in the illegal UGE evade taxes mainly to conceal their illegal activities; conversely, those who participate in the legal UGE are motivated largely or primarily by the evasion of taxes or benefit fraud.

Pure tax evasion involves non-reporting, understatement, or misreporting of taxable income, profits, or sales, but unlike UGE activity, PTE involves no labour input or no change of mode of business operation relative to the legitimate economy. Some PTE activity is related to extreme financial manipulation that goes beyond the bounds of legal tax avoidance. Other PTE occurs in conventional, legitimate businesses that underreport their receipts or overstate their expenses. Still other PTE stems from misreporting such events as intrafamilial transfers for tax purposes or misrepresenting the type, source, or timing of receipts. Although PTE does not directly affect the allocation of labour or capital resources or the methods of business operation, our later analysis will indicate how it can nevertheless affect the overall allocation and distribution of economic resources.

The distinction between UGE and PTE activities is not always clear-cut. For example, an established jeweller who understates receipts from cash sales but does not significantly alter his mode of business operation to conceal such sales is closer to PTE than the UGE. An individual who repeatedly buys, renovates, and resells homes while claiming the capital gains exemption on principal residences in Canada is probably closer to the UGE than PTE.

Benefit fraud and the UGE

For several reasons, we might expect the size of the UGE to be conditioned by benefit fraud as much as by tax evasion. Public transfer programs-including cash transfers, tax-linked benefits, and in-kind benefits-have grown substantially over the past thirty years. Most transfer benefits are conditional on reported incomes using implicit marginal tax rates (MTRs). When the MTRs of several programs are added together, the total rate can exceed that of the highest rate of personal income tax.

In Canada, the principal transfer programs include provincial welfare, unemployment insurance, and subsidies for child care, housing, and health care benefits. It is not unusual for the total MTR to approach or exceed 100 percent, providing a strong incentive for benefit fraud by working in the UGE.   --  Some observers have applied the term "benefit fraud" solely to multiple welfare claims and similar manoeuvres, arguing that concealment of earnings from the welfare authorities is understandable and non-culpable in view of the low level of benefits and the high penalty on earnings. This view mirrors one cited in the last note that tax evasion is commonplace and not really a criminal activity.Note  --    Benefit fraud has been further fuelled by such phenomena as rising numbers of unmarried couples and other unconventional families and the secular growth of welfare dependency. Higher rates of family fragmentation and child support obligations have driven others, particularly non-custodial fathers, into the UGE.

Echoing the distinction between the UGE and PTE for tax evasion, benefit fraud can involve either participation in the underground economy or simple non-reporting of income from the "above-ground" economy. However, since most candidates for public transfers have little capital other than consumer durables and possibly home equity, their income will be overwhelmingly derived from labour earnings. The visibility of most work in the above-ground economy will tend to drive such persons into the UGE if they hope to conceal their earnings from the transfer authorities. Something analogous to PTE can arise in benefit fraud when a beneficiary conceals assets that would be counted against benefits or fails to report capital or investment income. Such "pure benefit fraud" can also arise in cases where two-parent households present themselves as single-parent families to conceal one partner's earnings. In such cases, the second parent can work in the above-ground economy with less risk that visible earnings will be tracd back to the partner who is claiming the transfers.

Economic framework

An economic framework for explaining tax evasion and UGE participation will be based on rational decisions by individuals. These decisions are affected by preferences-including attitudes towards risk, honesty, and alternative occupations-as well as policy, market, and institutional factors. The simplest form of behaviour to model is PTE, which involves no substantive productive activity or occupational choice. Pure tax evasion hinges on a calculated weighing of gains from the concealment of a taxable receipt, sale, or purchase against perceived risks and penalties if detected.

Attitudes towards risk play a pivotal role in the extent to which various individuals will evade taxes. Persons who are risk-neutral will evade to the point where the taxes saved equal the potential penalties; risk-averse persons will evade to a lesser extent. Views about public morality, influenced by perceptions of the value of public services and the fairness of the tax system, may also condition PTE. It is commonly thought that higher tax rates will augment PTE and the associated revenue losses, and yet economic theory finds that this outcome depends on attitudes towards risk. Under certain conditions-"diminishing absolute risk aversion"-higher tax rates will actually decrease PTE.

UGE activity is more complex to explain than PTE since it involves decisions about occupations (often self-employment) and markets with both producers and consumers. To conceal tax evasion, UGE participants also have to conceal the activity that generates their income. This attempted secrecy affects the scale of underground businesses, their modes of operation, and their production technologies, with key implications for economic efficiency that will be discussed below.

In addition to the risk of being apprehended for tax evasion, self-employed UGE participants further bear the risks of being in business. Since the range of industries and occupations that lend themselves to small-scale covert operation is limited, personal tastes and skills in those lines of work will affect decisions about entering the UGE. The goods and services produced in the UGE must be sold, and the tastes of consumers for them as against the output of the legitimate economy will affect relative pricing and the distribution of UGE gains.

The size of the UGE is conditioned by the balance of the benefits versus the costs of moving into that sector. The benefits of working in the UGE include taxes evaded; the costs of working there include lower gross earnings, the expenses of concealment, and potential tax penalties. The need to conceal activities and occupations in the UGE will also mean psychic strain for many individuals. In equilibrium, UGE participants will tend to be, relative to their counterparts above ground, less averse to risk, less honest, more efficient at concealment, and/or more productive in those occupations or industries that are most amenable to tax evasion. Many who are compliant in equilibrium would actually be evaders were it not for the presence of others who are more predisposed to or efficient at evading. A populous UGE drives marginal returns down to the point where it is not attractive for others to enter. The marginal entrant gains little or nothing from participating in the UGE.

In the context of UGE as against PTE activity, it looks more likely that higher tax rates will raise levels of non-compliance. Tax increases will affect decision-making under risk, and the outcome again hinges on the precise characteristics of individual attitudes to risk. In the UGE, however, tax increases also raise the relative profitability of working covertly rather than in the tax-compliant sector-a situation that will attract more people to move productive activities underground. A new equilibrium will be established when the relative prices of UGE output have fallen to the point where no additional individuals find it advantageous to go underground. Note that this process of adjustment reduces the returns to all previous participants in the UGE as well, so that their incremental savings from evading at a higher rate of tax are offset by lower gross returns from their production.

Efficiency effects

PTE and UGE activities both distort the allocation of economic resources and may thereby reduce economic efficiency. That is, they may produce a lower level of material well-being than would prevail in their absence. However, it is also possible for such activities to enhance economic efficiency, particularly in cases where taxation or regulatory systems are highly coercive or confiscatory. For example, if taxes are imposed on incomes at very high rates, even short of 100 percent, the adverse effects on incentives to produce may be so severe that tax evasion yields higher-valued economic activity than tax-compliant activity would.

Whether the effects on economic efficiency are mainly negative or positive will hinge on the institutional, tax, and regulatory environment of a particular economy. In a rigid, centrally planned economy, we might expect PTE and the UGE to enhance overall economic well-being. At the other end of the spectrum, in a low-tax, relatively unregulated economy, non-compliance is likely to erode economic well-being. In advanced economies with substantial tax and regulatory burdens, the net efficiency impact of non-compliance has not been clearly determined.

Since PTE does not in the first instance involve substantive economic activity, its effects on efficiency are more roundabout than those of evasion within the UGE. If PTE is randomly distributed across the industries and sectors of an economy, it is unlikely to affect resource allocation other than through the need to recoup revenues lost through higher tax rates. Conversely, if PTE is concentrated in particular industries or sectors it will raise net returns from activity in those sectors, and this will in turn tend to expand those sectors and their products as against the efficient pattern arising with uniform compliance. This efficiency cost may be partially offset by the fact that evasion raises net returns on savings and investments, thereby mitigating tax distortions in those activities.

Evasion involving the UGE also involves inefficiencies from channelling excessive resources into the sectors or industries most amenable to evasion. The UGE further entails economic inefficiencies that arise directly from the altered production and sales methods businesses have to use for covert activities. Typical UGE operating styles will include smaller-scale production, fewer workers, less specialization of functions, less subcontracting, fewer tools and equipment, and more time wasted between jobs. For example, an off-the-books plumber will spend more time travelling between jobs than a legitimate plumbing firm with radio-dispatched trucks that can schedule its crews to adjacent jobs. Additionally, UGE operators will have to use less efficient means of advertising and incur more costs of concealment, such as changing phone numbers and addresses more frequently to reduce the risk of apprehension.

More than workers in the legitimate economy, UGE participants may also have their consumption and investment choices biased. They may be inclined to consume a high proportion of their current incomes rather than accumulate savings or large consumer durables that would be visible to the taxman. The composition of their savings and investments will be biased towards underground or offshore activities so as to prevent detection of the original cheating. UGE participants will also invest less in the human capital associated with their underground activity if they perceive greater risk to their business lifespans due to apprehension for taxation, regulatory, or licensing offences. Economic rents garnered by UGE participants may be dissipated through bribes to officials, especially in developing economies. To the extent that this practice diverts officials from their primary duties, it decreases the efficiency of the public sector.

Goods and services produced in the UGE carry less in the way of guarantees and general consumer protection than their legitimate-sector counterparts. With informed consumers, this factor is presumably reflected in lower prices. The UGE offers opportunities to individuals who might not be able to enter conventional production because of licensing or union barriers. If the conventional markets have monopolistic or non-competitive elements, the presence of the UGE may provide effective competition. Such competitive pressure can actually make the above-ground suppliers more efficient.

Of course, many regulations have to do with safety, health, and environmental standards which may be shortchanged by UGE producers. One might argue that consumers of UGE products are willing to accept lower standards for a lower price. Even if they are fully aware of this, however, there may be external effects on other people who are compromised by UGE activity. For example, a home built below the established standard may be a fire risk for the next-door neighbours, or substandard plumbing will be hidden in walls so that subsequent purchasers are unable to gauge its quality.

Even if the positive and negative efficiency effects of PTE or UGE activity offset each other, there would still be an impact on economic efficiency through the public finances. Public revenues lost because of these activities would have to be either recouped through higher taxes in the compliant economy or reflected in lower levels of public spending. Raising rates will simply exacerbate tax distortions of economic decisions in a variety of areas. In other words, the marginal cost of raising public funds is increased over what it would be without tax evasion.

If, on the other hand, revenues lost through non-compliance are reflected in reduced public spending, this would ordinarily be construed as an inefficiency.   --   The higher cost of raising public funds due to non-compliance should itself induce an efficient public sector to choose a lower level of public spending. Fewer prospective public projects will survive a cost-benefit test when revenues are more costly to raise.Note  --   This view assumes that the outcome of the political and bureaucratic process is to supply the optimal amount of public goods and services relative to the public's preferences and willingness to pay taxes. Yet for those who believe that governments tend, Leviathan-like, to oversupply public goods and services, reduced public spending resulting from PTE and the UGE might be regarded as enhancing efficiency.

Distributional effects  --   Any policy assessment of the distributional effects of tax evasion must contain ethical judgments about the relative weights to be assigned to gains enjoyed by compliant and non-compliant individuals.Note

A common assumption of lay observers is that tax evaders and UGE participants pocket all of their ill-gotten gains. Our review of the economic framework and efficiency effects has showed however that these gains may be dissipated in a variety of ways and that the beneficiaries may include other individuals in the affected markets. In particular, consumers of goods and services that are augmented through PTE and UGE activity stand to gain.

As in our previous analysis, it will be useful here to look at the movement of distributional effects separately for PTE and the UGE. One might expect PTE to be more concentrated at upper income levels, since it typically involves capital rather than labour. However, individuals of modest means can also underreport their receipts from interest, dividends, or rents. Conversely, given the markets and occupations involved, UGE activity might be expected to occur more frequently at lower to middle income levels. Of course, exceptions will arise with high returns to some participants in the criminal sector of the UGE.

The distributional effects of PTE hinge on its dispersion or concentration across the economy. When it is so dispersed as to be almost random across sectors, PTE perpetrators enjoy most if not all of the benefits. No market reaction arises to offset their gains, and the losers are either compliant taxpayers who have to make up the lost revenues or else people who suffer from curtailed public spending. At the other extreme, with PTE highly concentrated in a few sectors, the cost of capital will be depressed in those sectors and the expansion of goods and services produced in them will benefit consumers. Hence, the net gains to PTE practitioners will be less than their apparent savings from evasion, because they will accept lower gross rates of return in order to invest in sectors or industries that are conducive to evasion. This case is similar to the outcome of legal tax avoidance, where tax shelters for particular sectors depress their gross rates of return to the point where marginal investors gainnothing from the preferential tax provisions. With PTE, some of the gains will be dissipated in costs of concealment; in tax avoidance, the gains are partially spent on professional tax advisors.

The UGE has two distributional aspects that interest us here. The first is the distribution of gains between UGE participants and persons in the legitimate sector. This is the issue of UGE producers versus consumers of UGE goods and services, whether or not these buyers work in the UGE or the legitimate sector. If underground output is perfectly substitutable for the counterpart goods and services of the legitimate economy, it will sell at the same price and consumers will not benefit. But if the UGE output is so large as to lower the prices of those goods and services on the legitimate market, consumers can benefit even with homogeneous products. If the UGE output is imperfectly substitutable for consumers, it will command a lower price and consumers will benefit thereby. One would expect typical UGE output to be imperfectly substitutable, if not because of differences in quality or other characteristics, then at least because of the lack of warranties or legal recourse for consumers.

The second distributional aspect of interest is how gains are spread across UGE participants. Here, we apply the concept of "evasion costs" for individuals-costs like risk aversion, the psychic stress of evading or working in the UGE, concealment, and lower productivity. The critical point for distributional effects in the UGE is the degree to which evasion costs vary across individuals. If everyone-at least, everyone in the UGE in equilibrium-bears the same evasion costs, then everyone will benefit to the same degree. Yet in equilibrium, the marginal UGE entrant will derive no benefit from the UGE; fully homogeneous evasion costs mean that no underground workers are better off than they would be working legitimately. However, in the much more realistic case of heterogeneous evasion costs, the people with the lowest evasion costs gain the most from UGE work, with successively smaller gains for those with higher evasion costs.

Any policy assessment of the UGE must be concerned with the characteristics of UGE participants with low evasion costs. Some elements of evasion costs-particularly concealment and productivity losses in the UGE-might be lowest for groups such as youth, recent immigrants, and persons with anonymity in communities set apart by ethnicity or language. Many of these individuals are disadvantaged when it comes to finding work in the legitimate sector or in markets with unions or other entry restrictions. Hence, policies that would effectively address tax evasion in the UGE could have a disproportionate effect on disadvantaged individuals. Still, some industries are so dominated by UGE activity that forcing workers out of the underground would increase opportunities for many of the same individuals in the legitimate sector of the same industries.

The other distributional effects of PTE and UGE activity arise from the government response to lost tax revenues. As already mentioned, a government can respond by raising tax rates or trimming public spending; an increase in the public deficit can be regarded as an increase in future tax rates. We cannot form general conclusions about likely responses or their distributional effects without reference to the type of party in power or fiscal structure. In governments pressing against the political limits of the tax burden and seeking ways to curtail spending and control deficits, the likely response to revenue losses from evasion will probably be more spending curtailment. Similarly, effective policies to reduce evasion would probably reduce constraints on spending sooner than tax rates on compliant taxpayers. To the extent that public spending programs are typically tilted in favour of lower and middle income households, revenue losses from evasion are more than likely regressive in their ultimate effects.

Implications for enforcement policy

Policies to enforce the provisions of the tax system against PTE and UGE activities face a diverse matrix of factors. These include individual and social psychology, strategic behaviour by evaders, political pressures on public policy, and ethical issues concerning the appropriate balance between law enforcement and personal freedom. Efficiency and distributional outcomes have to be considered when defining an optimal enforcement policy. The economic and psychic costs of enforcement must be weighed against the benefits, and each needs to be evaluated at the margin to determine policy. The marginal benefits of increased enforcement should also include the deterrent effect on other evaders as well as the incremental revenues collected from the ones who are apprehended. The emphasis to be placed on the well-being of tax evaders is equally relevant to the distributional assessment of enforcement policy. At one extreme, evaders would be totally ignored in computing social welfare, while a more individualisti approach would encompass their well-being.

We will first examine the factors affecting enforcement policy for pure tax evasion, many of which will also apply to policy for the UGE. The penalty rate and the probability of detection both raise the anticipated costs of evasion, and an optimal policy will choose an appropriate combination of the two elements.  --   Penalties can be lump-sum fines but are more commonly expressed as fines related to amounts of tax evaded.Note  --   Raising penalties is costless for society, whereas enhancing detection efforts entails expensive audit resources. It might therefore appear desirable to raise penalty rates sky high while reducing resources spent on audits. One observer has called this a policy of "hang tax evaders with detection probability zero."  ---  Benjamini and Maital (1985, p. 254) attribute this statement to Serge-Christophe Kolm.Note

This extreme strategy is limited by several considerations. The acceptable penalties for evasion may be constrained by penalties applied to other criminal offences of comparable severity. Unduly severe penalties are also ruled out by the possibility of, for example, the government or an offender being mistaken as to when an act of evasion took place. The complexity of tax laws increases the likelihood of "honest mistakes" through poor advice, misinformation, or misjudgment. The possibility of erroneous conviction for tax evasion will make Draconian penalties unacceptable.

So how stringent should a detection program be? One measure of stringency is the expected penalty for the act of evasion-the punishment multiplied by the probability of detection. Even if audits or other means of detection were costless, an optimal enforcement policy would stop short of attempting to eliminate evasion altogether. Policy actions to decrease evasions may, beyond some point, reduce the level of society's well-being, because they reduce the utility of risk-averse evaders. In the optimum there should be positive marginal tax revenues from raising either the penalty rate or detection probability. Only if the welfare of evaders does not count should enforcement be pursued so as to generate the maximum total revenues. Considering the resource cost of tax enforcement will also reduce the optimal stringency.

The economic hypothesis of PTE decisions by risk-neutral individuals suggests that much tougher penalties may be needed to deter evasion than are currently in force. For an act of evasion that saves one dollar of tax, a risk-averse person will choose to evade so long as:

                            1 - P > P F,

where P is the probability of apprehension and F is the fine per dollar of tax evaded. In other words, deterrence of evasion requires that fines be levied at a rate of

                           F > (1 - P) / P.

If the probability of detection is as high as 50 percent (P = 0.5), then fines of just over 100 percent (F = 1.00) will suffice. This is roughly in line with penalties for evasion in many countries. However, a more realistic appraisal of the chances of being caught for many kinds of evasion will suggest that penalties at the above rate are woefully inadequate. For example, if the perceived probability is just 5 percent, then the requisite fine should be over 1900 percent, or 19 times the amount evaded.  --   The formula also shows how an unconstrained, optimal strategy is to "hang tax evaders with detection probability zero": P = 0 implies an infinite value for F.Note

Tax enforcement in the UGE involves the same factors as with PTE, but policy makers must also examine the effects on real resources absorbed in underground activity. The old saying that "an economy breathes through its tax loopholes" can be extended from tax avoidance to tax evasion. Excessively stringent enforcement can reduce productivity, output, and work incentives in the UGE and possibly in the economy as a whole. The effects on total economic performance are more likely to be adverse if cracking down on the UGE pushes labour into leisure time or tax-free household production activities rather than into the above-ground economy.

Another dimension of enforcement policy for the UGE has to do with detection methods. If the tax authorities investigate and apprehend only individuals who have filed tax returns or had taxes withheld on regular jobs, some of them will be induced to go completely underground. Such persons have been called "ghosts" since they have no official existence as far as the tax authorities are concerned. UGE participants will have a greater risk of being caught if they also report or engage in regular work. Tightening enforcement for these non-ghosts may propel enough of them into the ghostly state that total tax revenues actually decline. The obvious solution here is to organize an appropriate effort to seek out and apprehend the ghosts, possibly by following up on tips from the public and pursuing underground workers through their classified ads.

Other strategies for tax enforcement also follow from the economic framework reviewed earlier. Publicity campaigns based on evader convictions can raise the perceived probability of detection without increasing public spending for audits. Appeals to morality and public-spiritedness may also play a role, but are less likely to be effective with today's jaded taxpayers. There are high returns from improved knowledge about what signals to use for focussing tax investigation and audit resources. As indicated above, these signals should include elements that will detect ghosts as well as individuals with reported earnings. Moreover, the strategic aspects of evader behaviour need to be better understood by the enforcement authorities. More creative penalty schemes could also play a role in deterring tax evasion. For example, someone with a recently detected offence could be subjected to detailed retrospective investigation for previous years; this would greatly increase the expected penalties for recurent evasion.

Implications for tax and transfer policy

The existence of PTE, benefit fraud, and the UGE has significant implications for the design of tax and transfer policies. Most enforcement problems are primarily the result of inadequacies in the basic design of these policies. Compliance cannot be stronger than the policy design supports in terms of information reporting, source withholding, and related incentives for taxpayers and other parties. Of course, compliance should be just one major concern in the design of tax and transfer policies; it should not be allowed to dominate other key objectives such as efficiency and equity. Still, poor compliance itself poses inequities across taxpayers and raises the costs of operating the tax system.

Compliance will also be affected by broader issues of tax policy and public finance as well as specific tax structures and rates. For example, taxpayers will be better motivated to pay if they perceive that they are getting good value for their tax dollars in public goods and services and that the tax system is basically fair. Replacing general taxes with user charges-at least within the limited range where they can sensibly be applied-also reduces incentives for evasion since payments are attached to the receipt of particular goods or benefit entitlements.   --  User fees are limited to those goods and services provided by governments that do not have any of the following characteristics: "pure public goods," "merit goods," or other social externalities, high costs for excluding non-payers, and distributional goals for their use or consumption.Note

The design of a tax or transfer system contains several elements that have a particular bearing on compliance and enforcement. The first is the base of the tax in both the theoretical and implementation senses. Two taxes can have the same formal base and be equivalent in their economic content and predicted effects but still differ radically in their operational and compliance characteristics. For example, a flat-rate consumption tax can be implemented either as an indirect tax like a value-added tax or as a direct tax, like a personal income tax with a full deduction for net savings. The second element is the intended rate structure and degree of progressivity. The third comprises definitions of the taxable unit, whether it be the individual, family, or some other notion. The fourth includes the methods of tax collection, tax remittances, and source withholding or self-reporting with installment payments. These are essential aspects of tax design for compliance purposes, and they imply systems for iformation reporting and verification. Parallel issues arise in the design of transfer programs, such as defining the measure of need and verifying reported needs.

Information reporting and source withholding are particularly important for achieving high rates of tax compliance. The income types that are subject to withholding-wages and salaries-have higher rates of voluntary compliance than types subject only to reporting, such as interest and dividends. The lowest rates of voluntary compliance are found in income types that require neither withholding nor reporting by an agent independent of the taxpayer-rent receipts, capital gains, and self-employment income. Yet requirements to withhold and report must offer the parties genuine incentives. For example, most firms have a strong interest in withholding and reporting the correct amounts of personal tax from their employees since they can claim deductions for payrolls from their own taxable incomes. Most firms face comparable or higher rates of tax than their employees, so there are few incentives for collusion between employer and employee to underreport payroll.

The weak link for compliance with most taxes arises at the final stage-purchases by households and sales of goods and services directly to households. Company-to-company transactions usually have the built-in safeguard that they are tax-deductible by the purchasing firm, so that a revenue paper trail leads back to the vendor firm. This point has often been cited in support of the credit-invoice method of value-added taxation exemplified by the Canadian goods and services tax. Households, however, do not ordinarily need receipts for tax purposes. This is one reason, along with the small scale of production for many services purchased by households, for the high incidence of evasion related to the sale of services to households. An interesting exception is the requirement that Canadian households claiming the child care tax deduction must report the names, addresses, and social insurance numbers of persons paid for child-minding services.

It is useful to examine four issues of tax design to see how they are affected by the presence of PTE and the UGE. First, we consider optimal progressivity and levels for personal income tax. Choices here hinge on the value of public spending at the margin and the trade-off between efficiency costs and distributional effects of taxes. Since evasion itself reduces the distortions of high tax rates, its presence does not necessarily imply that a less progressive rate structure is desirable. The existence of the UGE also does not necessarily mean that marginal tax rates should be reduced-the reason here being that the UGE is distorted by penalties imposed on evaders, which can mean a less than optimal supply of labour and other resources for this sector. Hence, an increase in tax rates on the legitimate economy can in some circumstances improve efficiency by driving a more optimal level of resources to the UGE. While the presence of evasion can raise the levels and progressivity of tax rates, it unambiguously reuces the optimal total level of tax revenues for program spending. However, the optimal level of gross tax revenues including those needed to finance tax enforcement can be higher with tax non-compliance.Note This is because PTE and the UGE raise the marginal efficiency cost of taxes. Their presence also clearly reduces the total income and spending of a government that acts like a Leviathan.

A second issue involving tax design has to do with relief for individuals at lower income levels. The main methods in use are personal exemptions for an initial amount of income and taxing all incomes from the first dollar while providing an offset through either payments or refundable tax credits. The latter method allows the extension of source withholding to many additional types of income, including interest and dividends, thereby improving compliance rates. The personal exemption method makes it more difficult to withhold taxes at a uniform rate from such payments, since many persons with low and moderate incomes, including retirees, would have tax withheld on interest and dividend receipts and need refunds. The use of a flat rate of tax, at least over lower to middle income levels, further facilitates the use of withholding at source. A flatter rate schedule would support the extension of withholding to purchases from independent contractors; a penalty could then be applied to purchasers who failed to wthhold. Australia has required homeowners buying home repairs or renovations over a threshold level to withhold and remit tax.

A third tax design issue is the appropriate revenue mix of direct and indirect taxes. It has commonly been asserted that the presence of evasion should prompt greater reliance on indirect taxes, particularly the value-added type. While direct personal taxes can be evaded in numerous ways, indirect taxes of both the retail and value-added types are also amenable to a variety of evasion manoeuvres. It is said that shifting the mix from direct to indirect taxes would allow taxes to be collected on the consumer spending of people who evade income tax. Yet the authorities find it as difficult to apply an indirect tax on UGE output as to collect direct tax on the earnings of UGE participants. This implies that the effects of such additional indirect taxes will be felt mostly or entirely by workers in the legitimate sector. The economy will reach a new equilibrium in which the competitive prices of goods and services from the legitimate sector fall relative to UGE prices. UGE workers would pay the indirect taxes on heir purchases from the legitimate sector but be fully compensated by a fall in the prices of those items.

Empirical studies of the Canadian economy support the theoretical prediction that a shift in the mix from direct taxes toward greater use of indirect taxes will not improve tax compliance or reduce the size of the underground sector.   --   The results cited here are based on Hill and Kabir (1996) and Spiro (1993, 1994).Note  --   Changes in indirect taxes have been found to account for most growth of the UGE since 1964. Moreover, the change in form of the federal sales tax in 1991, from a single-stage tax on manufacturers to a multi-stage value-added tax, has been associated with a large increase in the UGE even without a shift in the direct-indirect tax mix. Estimates of the resultant impact on the Canadian UGE range form 1 percent to nearly 4 percent of GDP.

A fourth issue around tax design is the appropriate use of excise and property taxes. The presence of PTE and the UGE suggests that greater reliance on particular forms of those taxes could improve overall compliance. By raising taxes on the purchase or use of items that are large, visible, and registered, governments can be relatively assured of collecting their money. The shift allows them to reduce rates on taxes that apply to other items such as earnings, which can more easily be concealed. In the optimum, this strategy should be pushed to the point where the social gains from improved tax compliance, measured in efficiency and distributional terms, are just offset by the marginal inefficiencies and distributional effects from the distorted consumption patterns of taxing a few selected commodities more heavily. Two natural targets for this kind of tax policy are automobiles and homes or real estate; both are highly visible, with ownership registered and valuations either already available or readily obtaied.   --  Also relevant to the formulation of this kind of policy is the correlation of consumption of these items with the incidence of evasion of other taxes.Note  --   The US pursued a tax strategy of this kind in 1991 by imposing excise taxes on purchases of higher-valued cars, boats, airplanes, jewellery, and furs.

Benefit fraud has major implications for the design of transfer policies. Here we will focus on fraud in the form of undeclared or underreported earnings. In contrast with tax policies, one can design transfer policies in a way that provides incentives for full reporting of earnings.

Traditional welfare programs provide net benefits that are negatively conditioned on reported earnings. For this reason, beneficiaries have a strong incentive to conceal or underreport earnings, a practice that is facilitated by working underground. For employable individuals, the transfer scheme can be structured as a wage rate subsidy with benefits that rise proportionately with hours worked but fall with hourly wage rates. An individual who has no earnings receives no benefits. It is far harder for the public authorities to detect unreported earnings than to verify the existence and level of reported earnings.

Another way of structuring transfer policies that discourage work in the UGE for the purpose of underreporting earnings is to attach a work requirement to the payment of benefits. This approach has a variety of forms that range from special public employment to community work. By taking up the beneficiary's regular working hours, such a policy sharply raises the cost of taking UGE employment. Of course, workfare or work-for-welfare policies can offer other benefits such as socially useful output, the maintenance of skills and work habits, skills training, and the dignity of beneficiaries. Careful design of the content and operation of the work program will be required to achieve all these goals and avoid stigmatizing recipients.

Conclusion

Pure tax evasion and the underground economy pose fundamental questions about the effects of taxation. Benefit fraud poses analogous questions about the transfer system. The importance of these phenomena for public policy cannot be judged by their scale alone; it stems from the magnitude and nature of their associated efficiency and distributional effects. Even a relatively small underground sector could generate serious inefficiencies and inequities. However, our analysis suggests that most of the benefits associated with the UGE are in fact more widely dispersed to consumers as well as underground producers.

With pure tax evasion, which is not highly concentrated in particular sectors of the economy, greater inequities may arise between compliant and non-compliant individuals. The efficiency costs of non- compliance include distortion of overall resource allocation, low productivity from petty producers trying to conceal their operations, public enforcement costs, and reduced public provision of otherwise desirable goods and services. Offset against these inefficiencies are the potential efficiency gains from economic activity in a sector free of tax distortions and the possible reduction of excessive activities by a Leviathan-like government.

PTE and the UGE also raise important questions about the design, operation, and enforcement of tax and transfer systems. Many policy decisions can be made at the enforcement stage in the form of increased audit or penalty rates and changes in reporting or withholding requirements. Yet our analysis stresses the importance of considering compliance factors at the initial design or major reform stages of tax and transfer policies. The feasibility and cost of source withholding depend very much on tax structure, and the basic design of a tax determines what kinds of information need to be reported to the authorities. Furthermore, it is vital to consider the incentives for third parties such as employers, payers, or purchasers to comply with withholding or reporting requirements.

No matter how many resources a society spends on enforcing tax and transfer provisions, it will ultimately be limited by the integrity of its policy design. And there are economic and social considerations that should restrain a society from devoting excessive resources to enforcement. These factors include the efficiency and distributional impacts of enforcement as well as concerns about individual privacy and Draconian justice.

References  --  This bibliography provides key references for the economic literature on pure tax evasion but focusses on the smaller body of studies exploring the linkages between evasion and the underground economy.Note

Alm, James, "The Welfare Cost of the Underground Economy," in Economic Inquiry 24 (April 1985), pp. 243-63.

Benjamini, Yael and Shlomo Maital, "Optimal Tax Evasion and Optimal Tax Evasion Policy: Behavioral Aspects," in Gaertner and Wenig (1985), pp. 245-64.

Besley, Timothy and Stephen Coate, "Workfare versus Welfare: Incentive Arguments for Work Requirements in Poverty-Alleviation Programs," in American Economic Review 82 (March 1992), pp. 249-61.

Clotfelter, Charles T., "Tax Evasion and Tax Rates," in Review of Economics and Statistics 65 (August 1983), pp. 363-73.

Cowell, Frank A., "Public Policy and Tax Evasion: Some Problems," in Gaertner and Wenig (1985), pp. 273-84.

--- , "Tax Evasion with Labour Income," in Journal of Public Economics 26 (February 1985), pp. 19-34.

--- , Cheating the Government: The Economics of Evasion: Cambridge, MIT Press, 1990.

Erard, Brian and Jonathan S. Feinstein, "Honesty and Evasion in the Tax Compliance Game," in Rand Journal of Economics 25 (Spring 1994), pp. 1-19.

Feige, Edgar L., ed., The Underground Economies: Tax Evasion and Information Distortion: Cambridge, Cambridge University Press, 1989.

--- and Robert T. McGee, "Sweden's Laffer Curve: Taxation and the Unobserved Economy," in Scandinavian Journal of Economics 85 (1983), pp. 499-519.

Feinstein, Jonathan, "An Econometric Analysis of Income Tax Evasion and Its Detection," in Rand Journal of Economics 22 (Spring 1991), pp. 14-35.

Fortin, Bernard and Nguyen M. Hung, "Poverty Trap and the Hidden Labor Market," in Economics Letters 25 (1987), pp. 183-89.

--- and Guy Lacroix, "Labour Supply, Tax Evasion and the Marginal Cost of Public Funds: An Empirical Investigation," in Journal of Public Economics 55 (November 1994), pp. 407-31.

Frey, Bruno, "How Large (or Small) Should the Underground Economy Be?" in Feige (1989), pp. 111-26.

Gaertner, Wulf and Alois Wenig, eds, The Economics of the Shadow Economy: Berlin: Springer-Verlag, 1985.

Gordon, James P. F., "Evading Taxes by Selling for Cash," in Oxford Economic Papers 42 (January 1990), pp. 244-55.

Hansson, Ingemar, "Tax Evasion and Government Policy," in Gaertner and Wenig (1985), pp. 285-300.

Hill, Roderick, and Muhammed Kabir, "Tax Rates, the Tax Mis, and the Growth of the Underground economy in Canada: What Can We Infer?" in Canadian Tax Journal 44 (no. 6,1996), in press.

Jung, Young H., Arthur Snow, and Gregory A. Trandel, "Tax Evasion and the Size of the Underground Economy," in Journal of Public Economics 54 (July 1994), pp. 391-402.

Kaplow, Louis, "Optimal Taxation with Costly Enforcement and Evasion," in Journal of Public Economics 43 (November 1990), pp. 221-36.

Kesselman, Jonathan R., "Income Tax Evasion: An Intersectoral Analysis," in Journal of Public Economics 38 (March 1989), pp. 137-82.

--- , Rate Structure and Personal Taxation: Flat Rate or Dual Rate? Wellington, New Zealand, Victoria University Press for the Institute of Policy Studies, 1990.

--- , comment on "Taxation and the Service Sector" (by John Whalley), in Richard M. Bird and Jack M. Mintz, eds, Taxation to 2000 and Beyond: Toronto, Canadian Tax Foundation, 1992, pp. 286-94.

--- , "Evasion Effects of Changing the Tax Mix," in Economic Record 69 (June 1993), pp. 131-48.

--- , "Compliance, Enforcement, and Administrative Factors in Improving Tax Fairness," in Allan M. Maslove, ed., Issues in the Taxation of Individuals, a volume of research studies for the Ontario Fair Tax Commission: Toronto, University of Toronto Press, 1994, pp. 62-84.

Lemieux, Thomas, Bernard Fortin, and Pierre Fréchette, "The Effect of Taxes on Labor Supply in the Underground Economy," in American Economic Review 84 (March 1994), pp. 231-54.

Persson, Mats and Pehr Wissén, "Redistributional Aspects of Tax Evasion," in Scandinavian Journal of Economics 86 (1984), pp. 131-49.

Pyle, D.J., "The Economics of Taxpayer Compliance," in Journal of Economic Surveys 5 (no. 2, 1991), pp. 163-98.

Sandmo, Agnar, "Income Tax Evasion, Labour Supply, and the Equity-Efficiency Tradeoff," in Journal of Public Economics 16 (December 1981), pp. 265-88.

Schweitzer, Urs, "Welfare Analysis of Excise Tax Evasion," in Journal of Institutional and Theoretical Economics 140 (June 1984), pp. 247-58.

Slemrod, Joel, ed., Why People Pay Taxes: Tax Compliance and Enforcement: Ann Arbor, University of Michigan Press, 1992.

---"Fixing the Leak in Okun's Bucket: Optimal Tax Progressivity When Avoidance Can Be Controlled," in Journal of Public Economics 55 (September 1994), pp. 41-51.

Spiro, Peter S., "Evidence of a Post-GST Increase in the Underground Economy," in Canadian Tax Journal 41 (no. 2, 1993), pp. 247-58.

----"Estimating the Underground Economy: A Critical Evaluation of the Monetary Approach," in Canadian tax Journal 42 (o. 4, 1994), pp.1059-81

Usher, Dan, "Tax Evasion and the Marginal Cost of Public Funds," in Economic Inquiry 24 (October 1986), pp. 563-86.

Watson, Harry, "Tax Evasion and Labor Markets," in Journal of Public Economics 27 (July 1985), pp. 231-46.

Yaniv, Gideon, "Withholding and Non-withheld Tax Evasion," in Journal of Public Economics 35 (March 1988), pp. 183-204.

Benefit Principle and Taxation: Possible User Taxes and Fees in Canada

François Vaillancourt

Introduction

The purpose of this paper is to examine the possibility of increasing the role of user taxes and fees as revenue sources for governments in Canada. This is of interest in the context of an examination of the underground economy, since it is argued that the current tax mix contributes to the growth of that economy. We begin by reviewing the conceptual framework used for assessing revenue sources and then look at the revenues and expenditures of governments in Canada, ending with some ideas for a few possible user taxes.

Conceptual framework

Governments finance themselves in three major ways: seigniorage income, also referred to as the inflation tax; deficits, covered by domestic or foreign borrowing; and general revenues encompassing taxes, fees, returns on investment, and other sources of current income. With inflation running at under 2 percent, the first method is not significant in Canada at present.

Deficits and general revenues are the two major sources of income. Deficits emerged as an important source of financing in the 1970s with the oil shocks of 1973 and 1979 and came into full bloom in the '80s with the recession of l981-1982. They now, depending on our exact definition of the public sector and the year considered, account for about 20 to 25 percent of government spending. The general revenue source is over 80 percent made up of taxes.

Taxation has been founded for generations on two major principles: ability to pay and benefits received. Early on, the main emphasis was put on benefits, with income sometimes called on as a measure of benefits, not of ability to pay. In the early part of this century, the generalization of the progressive income tax, justified in part by arguments that equal sacrifice required increasing proportions of income because of the decreasing marginal utility of income, was accompanied by a greater and indeed almost exclusive emphasis on ability to pay as a basis for taxation. An example of this is a recent Quebec government paper on taxes and expenditures where the benefit principle is not even mentioned in the list of tax principles.

From our perspective, the three interesting questions to be asked with respect to taxation are:

     •     What are user taxes/fees and when should they be used?

     •     How are they currently being used? and

     •     What can be done?

We address the first question below and the other two in our next two sections.

"What are user taxes/fees?" is more easily answered for fees than for taxes. User fees are payments for specific and clearly identified units of goods or services such as so many cubic metres of water delivered to a specific outlet, parking permits valid for given streets for limited time periods, fishing or hunting licences for particular seasons, and so on.

User taxes, on the other hand, are not conceptually well defined. They must be earmarked insofar as they are linked to a well-defined expenditure. However, not all earmarked taxes are user taxes, since by our definition user taxes are taxes paid directly by users of specific public programs, or on their behalf by others such as parents or employers, for access to the benefits-goods, services, transfers-provided by the programs.

User taxes can be solely entry-restrictive: they can also be benefit-defining. Entry-restrictive user taxes must be paid to gain access to benefits, but once payment has been made the value of benefits does not depend on the amount paid. With benefit-defining user taxes, the amount of benefits is tied in some way to the payment by rules about benefit duration and so on.

"When should taxes/fees be used?" is at least partly a normative question and thus difficult for economists to answer. Nonetheless, we can argue that:

     •     User fees/taxes are an appropriate way of charging for private goods and services provided by government when their provision cannot be privatized. Why? User fees/taxes act as pricing mechanisms and thus help to allocate resources optimally over various goods and services. This does not mean that publicly provided goods and services will not be produced at too high a cost: what is does mean is that the quantity produced may not be too great. Proper pricing would set the fees/taxes to cover the costs of producing these goods and services privately. Any excess cost, associated with the inefficiency subsidy arising from public production, should be raised from general revenues.

     •     User fees/taxes must take into account, not only the efficiency gains of using them rather than general revenue sources, but also their associated administrative and compliance costs. Thus, it may be appropriate to offer individuals and small businesses a bundle of services for a flat inclusive fee.

     •     User fees/taxes should not be used for redistributive purposes. Redistribution should be explicit: in some cases, it will result in individuals or families being fully compensated for user payments.

Revenues and expenditures of governments in Canada

Table 1 presents the most recent data on expenditures and revenues for all levels including local governments. The most important expenditures are social services (22 percent of all governments) followed by debt charges (20 percent), health (14 percent), and education (13 percent): these four spending areas account for about 70 percent of the total. The revenue field is dominated by personal income taxes (36 percent for all governments) followed by general sales taxes (l3 percent) and health and social insurance levies.

Click here to view Table 1: Government Own-Source Revenues and Expenditures Canada, 1993-1994 ($ millions)

The information in the table does not enable us to comment on the relative importance of user fees/taxes. An interesting paper by Thirsk and Bird (1993) attempted to measure the importance of earmarked taxes in Canada, some of which are user fees, but they were unable to come up with a precise figure. They concluded that user fees were more important at the local level than at other levels of government. If we examine the four main areas of spending identified above, it can be seen that debt charges, health benefits and elementary and secondary education are financed from general revenues. User fees are often present in postsecondary education (the exception being the CEGEPs in Quebec) but they account for only a small part of total expenditures.

With respect to social services and user taxes, there is a link between payroll taxes and benefits received in terms of eligibility for worker's compensation benefits (WCB), unemployment insurance (UI) and Canada/Quebec Pension Plan (CPP/QPP) benefits. Employees of uninsured employers cannot receive WCB benefits and non-contributors cannot collect UI or CPP/QPP benefits. In these last two cases, individual benefits are weakly (UI) or strongly (CPP/QPP) linked to contributions. In 1994, about $40 billion in user taxes, representing about 15 percent of all government revenues, covered a part of social services expenditures in Canada-UI=$18 billion, CPP/QPP=$15 billion, and WCB=$7 billion.

Potential user fees/taxes for Canada

Before addressing the question of potential new fees and taxes for Canadian jurisdictions, we want to mention a few points to be kept in mind when introducing levies of this kind:

     •     Borrowing should be the method used when governments are buying multiyear capital assets. In a steady-state economy we could see total borrowing increase while per-capita debt remained the same as the population increased. Depending on technological change, real debt and assets could rise in the economy. Thus, introducing user fees/taxes does not imply a balanced budget.

     •     A checklist financing system would seem appropriate for advocacy programs and specific semi-public services. At income tax time, the government would attach to the tax return a list of programs to be publicly financed. The initial list would reflect the status quo, while subsequent lists would evolve in part according to funding received. Individuals would use this list to determine a level of support for each program. For example, the Consumer Association of Canada and the Canadian Broadcasting Corporation would receive financial support equal to the sum of all amounts contributed with a one-year lag.

     •     User fees should be introduced as widely as possible. For instance, all municipalities should be required to sell metered water and, when operating transit systems, levy charges to reflect distances travelled. Similarly, users of downtown access roads should be charged for downtown access permits or, ideally, by electronic monitoring of actual use.

Let us now turn to possible user funding in the fields of social services, debt charges, health, and education. In suggesting user taxes, we will assume that:

     •     Existing programs remain more or less unchanged, although user financing may reduce demand for some services.

     •     As the term "taxes" (rather than "fees") indicates, participation is compulsory. The freedom to choose not to be covered by public health insurance or pension plans does not exist. This is important, as we would otherwise move away from taxes to premiums and from a public monopoly to a private market.

     •     User taxes can pay the full costs of each plan examined. This might be achieved in part through a redistribution of income from the rich to the poor to cover the relevant user taxes.

Income security / social services

In general, we wish to distinguish between three types of transfers: those to unemployables, those to employables, and those to retirees (once employables).

The key point with unemployables will be to identify them as precisely as possible to avoid free rides by other groups. We would argue that individuals should be presumed employable and thus need to prove that some serious physical or psychological problems prevent them from working. The basic needs of unemployables should be met out of general revenues, not user taxes.

In the case of employables, premiums paid by workers should be used to fund an employment income insurance system that integrates the various existing schemes such as UI, WCB, and welfare for employables (Vaillancourt, 1994). Benefits would be set at a basic minimum with some topping off to reflect premiums accumulated in the fund. Unfunded benefits would then be recovered, possibly from future premiums. This would be a user tax with a strong link between payments and benefits.

In the case of retirement income, CPP/QPP should become fully funded from individual contributions at a given point in time for everyone then aged 45 or less. To begin with, people would receive statements of their accumulated and projected benefits: they would have the right to make extra contributions over the ensuing 25 years to make up for the shortfall in expected pensions resulting from this change in the rules.

Also from that changeover date, individuals aged 25 or less would be told that they no longer had access to OAS/GIS: they must contribute more to CPP/QPP. If people are free to contribute, then society must be willing to let them suffer, if not die, from lack of resources. To ease the transition, CPP/QPP funds should be invested to earn the market rate of return (Prince, 1993). Again we have a strong link between taxes paid and benefits received.

The use of individual premiums to finance both employment income insurance and retirement income would make it more advantageous to declare income for tax purposes and thus reduce the size of the unreported economy.

Debt charges

At first blush, it may seem strange to make a connection between debt charges and user fees/taxes. But in Canada, debt has been incurred to finance either infrastructures-and should therefore be paid by appropriate user fees and taxes-or current spending. If the latter was the case in the 1980s and through into the 1990s, and we would argue it was (Trahan, 1993), then it is incorrect to raise the taxes of the young or cut the services available to them in order to finance the spending of the old: the young could not vote on these borrowings.

In other words the older generation, having failed as trustees of the nation and misused their borrowing power, must pay an ex-post user tax. The following three possibilities are offered in increasing order of feasibility:

     •     sending each Canadian aged 35 or more a debt bill to be paid now or at the time of departure from Canada or death (applied against his/her estate). Life/departure insurance would have to be paid by anyone electing not to pay the bill on receipt;

     •     imposing an age-specific (say, 35 to 70) income surtax for a limited time period (10 to 20 years) to pay off this consumption-related borrowing.

     •     taxing pension fund earnings at a flat rate (say, 25 to 40 percent) on the grounds that people with access to pension funds are the better-off 35- to 60-year-olds. They will thus see their retirement savings and income reduced.

Age 35 was chosen as the minimum age to make those who voted repeatedly for free-spending governments in the 1980s and '90s pay up for the misuse of their borrowing power.

Health

In this area we need to distinguish the three classes of illnesses and injuries: genetic, natural, and self-inflicted.

•Genetic illnesses are passed on by parents. Some are detectable before conception, some early after conception so that abortion is possible, some are detectable right after birth, and some are detectable only when they materialize. Assuming that society does not wish to coerce women into having their foetuses tested, let alone having abortions, genetic illnesses should then be covered by per-capita health premiums.

•Natural illnesses are complaints such as colds, ear infections, and so on that occur in the natural course of our lives. Little if anything can be done to prevent them, although their severity may vary among individuals. Again, these conditions should be covered by per-capita health premiums.

•Self-inflicted illnesses or injuries should be covered by premiums set to reflect high-risk behaviour. For example, smokers and those who share living quarters with them would pay a Tobacco Consumption Health Premium (Vaillancourt, 1994a). The premium would be collected on the basis of an annual declaration: falsification would lead to ex-post collection of the real premium and, if this could not be paid, denial of services.

Education

Here we should distinguish between elementary/secondary schooling on the one hand and postsecondary education on the other. In the first case, we can argue that basic abilities are a public good, and while user fees should be levied to finance, say, one fourth to one third of the costs of that educational level, general revenues should also be tapped. In the second case, private returns from postsecondary schooling at the bachelor's and master's levels are such that full fees should be charged in combination with an income-contingent repayment scheme. When it comes to PhDs, an argument cay be made for R&D externalities to justify subsidies from general revenues for at least some fields of study.

Conclusion

We have argued here that it is possible to make greater use of user fees and user taxes to finance governments in Canada. This should somewhat slake the thirst for general tax revenues and income taxes in particular, opening the way for reduced tax rates that would diminish incentives for tax evasion and thus the size of the underground economy.

References

Prince, M. (1993), "Reforming the Public Pension System in Canada, Retrospect and Prospect," Centre for Public Sector Studies, University of Victoria.

Thirsk, W. and R. Bird (1993), "Earmarked Taxes in Ontario: Solution or Problem?" in A. Maslove, ed., Taxing and Spending Issues of Process: Toronto, University of Toronto Press,.

Trahan, F. (1993), "Réflexion sur le contrôle des dépenses et la comptabilité générationnelle," M.Sc. term paper.

Vaillancourt, F. (1994), "Income Distribution, Income Security and Fiscal Federalism in Canada," in Future of Fiscal Federalism, R. Banting, ed., Queen's University School of Policy Studies.

Vaillancourt, F. (1994a), "Cigarette: No thanks, I haven't paid my premium," Toronto Globe and Mail, February 17, 1994, p. Al9.

info@fraserinstitute.ca You can contact us at the above email address for any comments or information requests. Please report any dead links or technical problems.

If you know someone who would be interested in this web page, please enter their email address below, and we will forward this URL to them: Email Address:

Last Modified: Wednesday, October 20, 1999.