The Power to Tax: Analytical Foundations of a Fiscal Constitution
Toward Authentic Tax Reform
Prospects and Prescriptions
No doubt the raising of a very exorbitant tax, as the raising as much in peace as in war, or the half or even the fifth of the wealth of the nation, would, as well as any gross abuse of power, justify resistance in the people.
The analytical setting in this book is dramatically different from that which informs most of the literature of tax policy. One of our central purposes is simply to shift the grounds for discussion and debate, quite apart from any specific policy prescriptions that might subsequently emerge. Tax reform deserves to be examined in a constitutional perspective and not as some prize to be captured in a partisan political struggle over relative shares nor as some abstracted exercise within the political naiveté of the economist's study.
In order to discuss tax reform constitutionally, however, an analytical dimension must be introduced over and beyond that within which orthodox analysis has been conducted. Comparatively speaking, it is relatively easy to discuss alternative tax instruments in a normative manner when the operation of the political-governmental process can be assumed away or put to one side, either by the partisan advocate who seeks to capture political power for his own ends or by the economist who blindly assumes government to be benevolent as well as despotic.
The constitutionalist, quite independently of his own ideological or normative predispositions, cannot even begin argument until and unless he first models the operation of government in the postconstitutional sequence. A "theory of politics," defined as the theory of the working properties of the political process under alternative sets of rules, is logically and necessarily prior to any responsible discussion of constitutional alternatives themselves. An acknowledgment of this methodological principle confronts us with difficulties. In a very real sense, all of the public-choice theory and analysis that has been painstakingly developed over the course of three decades becomes prolegomena to analysis of constitutional design, including design for tax reform. But public-choice theory itself remains a long way from its own long-term equilibrium, potentially described by established and widely accepted paradigms.
By necessity we are required to restrict our scope and range. To do so, we have modeled politics in an admittedly extreme, and indeed simplistic, framework, one that we have called "Leviathan." The analysis of preceding chapters has demonstrated that, given this model of government, many of the standard norms for tax reform, for idealized tax arrangements, are totally unacceptable. And by "unacceptable" we mean that the arrangements would never be selected by the rational person in a genuine constitutional choice setting where he is assumed to be empowered to choose as among alternatives for constitutional policy. It is perhaps worth noting that this conclusion of our analysis emerges directly from our positive examination of the individual's constitutional calculus, given the model of politics that is plugged in. The conclusion does not stem from any ideological mind-set that we presume the individual to possess and certainly not from any presuppositions of our own.
Objections to the normative implications, assuming that the technical analysis holds up against possible criticisms, may properly be directed at the Leviathan model of politics. In such a sense, these objections may be at least similar in kind to those that we have ourselves advanced, in this book and elsewhere, against the benevolent despotism model that has for too long informed the orthodox analysis of tax reform. We have argued in Chapter 2 that "natural government" embodies Leviathan-like properties. There is, however, an important difference between the application and use of our model and the application and use of that model implicit in the orthodox policy discussion. In the latter, government is modeled as a benevolent despot, as an imaginary entity that can listen to, accept, and act upon the policy advice proffered by the economist. By contrast, our use of the Leviathan model of politics does not involve any offer of advice or counsel to governments. We use this model to generate predictions of a "worst-possible" sequence of outcomes, predictions that facilitate our analysis of ways and means of ensuring that such "worst-possible" results will not, in fact, be realized. In a very real sense, our whole effort is in close affinity with a Rawls-like minimax strategy.*110 We should also note, at least in passing, that our whole exercise is within the spirit of the classical political economists and the American Founding Fathers, some of whom are cited at the beginning of our chapters.
Even if, operationally, our Leviathan model of politics should be wholly rejected, however, there seems to be no reasonably legitimate basis for jumping to another extreme and acquiescing in the benevolent despotism presumption. Once the latter is called into question, a constitutional approach is almost necessarily required. For example, a model of politics described by the domination of the median voter under majoritarian rules might be introduced as a substitute for our Leviathan. In this case, some of the implications for tax reform would, of course, be different from those that emerge from our analysis. But they would also, and perhaps more sharply, diverge from those that are implied from the standard treatment.
Tax reform must be analyzed constitutionally. But we need to go beyond suggestions for an approach; we need to look somewhat more directly at tax-policy alternatives as these have entered into the political discussions of the later 1970s and 1980s. Our critique, whether direct or implied, of the proposals for tax "reform" that have emerged from orthodox tax-policy discussion is sufficiently contained in earlier chapters. What these chapters have not contained is any treatment of the various proposals for constitutional change that have now surfaced to command positions of reckoned importance in current public-policy debates. Constitutional changes have been made, and many more changes are being quite actively discussed, within a realm of popular-public-political discourse that has remained isolated from the "in-house" talk of tax economists and tax lawyers. Professionals in these groups find themselves unable to articulate their own positions largely because constitutionalism is alien to their thought patterns. The so-called "taxpayers' revolt" of the late 1970s, brought into focus by California's Proposition 13 in June 1978, has been populist and constitutional rather than elitist and legislative in its origins. And it has taken substantial form in actual or proposed changes in constitutional rules that impose constraints on the taxing powers rather than changes in legislated tax-rate levels and tax-structure arrangements.
It should be evident that our whole analysis is directly relevant to the policy options that have been discussed in the context of the taxpayer revolution, whether real or presumed. It is, however, also important to recognize that the proposals for explicit constitutional changes in tax rules that have emerged in the course of the shift in attitudes characteristic of the late 1970s have not represented the outcome of analysis like that we have attempted in this book. As noted in the Preface, our book is two or three years late in serving this function. The taxpayer revolution has indeed been born without an analytical blueprint or even an analytical map. Nonetheless, the potential usefulness of our effort seems evident. We try to assess critically some of the proposals under popular discussion from a more sophisticated, but still constitutional, perspective than that offered by more orthodox public-finance specialists. From this assessment, our own prescriptions for authentic tax reform follow straightforwardly.
10.1. Taxation in Constitutional Perspective
We stated above that tax reform deserves to be discussed constitutionally. To justify our argument, the differences between the nonconstitutional and the constitutional setting must be emphasized. What does it mean to say that taxes and tax reform are sometimes treated nonconstitutionally? In such a context, tax rules and institutions are considered to be subject to period-by-period changes; they are not treated as permanent or quasi-permanent features of the political structure. In the limiting case, the allocation of tax shares among individuals and groups in the economy and the choice of tax instruments that generate the imputations of such shares are considered "up for grabs" during each and every new budgetary period. In such a nonconstitutional setting, the prospective taxpayer is, of course, vulnerable to exploitation by government to the maximum limits of his taxpaying capacity. Under anything resembling our Leviathan assumptions about governmental processes, the prospective or potential taxpayer will clearly have some interest in imposing constitutional constraints on the taxing power in advance of the budgetary period, constraints that will act to bind the exercise of fiscal authority over the whole postconstitutional sequence.
Even if we relax the Leviathan assumptions, however, and allow for some genuine electoral in-period controls to be exercised by a majority coalition of taxpayers, the single individual must face the prospect that, within any given budgetary-fiscal period, he may find his tax obligations arbitrarily settled by the dominant political coalition which may act contrary to his own interests. Under almost any set of projections or predictions about governmental processes, even those that model politics as ideally benevolent in some respects, the individual would prefer that basic tax rules be considered to be constitutional. There is positive value in predictability. And, indeed, a constitutional interpretation of tax rules implicitly informs much of the tax practice in Western countries. Tax arrangements, once in being, tend to be relatively long-lived, and the adage that "an old tax is a good tax" is a part of a more general attitude toward tax practice even if it does not seem to have been recognized widely in the economic analysis.*111
In a sense, our argument becomes a plea for a more explicit constitutional attitude toward tax reform. Tax rules should be considered, analyzed, and discussed as a set of quasi-permanent arrangements within which persons can anticipate making appropriate behavioral adjustments, including those that require a long planning horizon. As earlier analysis has indicated, one means of constraining Leviathan lies in the restriction of its ability to modify tax rules with such frequency that taxpayers must face continued uncertainty as to just what the rules will be period by period. If government can succeed in keeping taxpayers off balance in their planning, additional revenue potential may be available for exploitation, provided, of course, that the fiscal uncertainty itself is not pushed beyond revenue-maximizing limits.
The categorical distinction between tax rules as elements in the basic constitutional contract and tax rules as pawns in in-period conflicts over tax shares emerges only in a non-Leviathan context. If electoral politics is assigned the task of imputing or reimputing relative shares in total tax liability as among differing groups of taxpayers in each budgetary period, or even if such reimputation is treated as being available for consideration, the analytical model becomes that of a never-ending, negative-sum, n-person game. Continual manipulation of the basic tax-share distribution can be judged to be an undesired attribute of "tax reform," even in the most "democratic" electoral-political process, even when political outcomes sensitively mirror the true demands of members of effective majority coalitions among voters and, hence, among taxpayers. In such a setting, "tax-reform" advocacy largely takes on the pattern of mutually offsetting attempts to shift tax shares among groups. Little or nothing is gained by any group in the process and, over a sequence of periods, all groups would reasonably expect to lose. In playing off groups against each other, the opportunities available to any revenue-seeking authority can be effectively exploited.
It is essentially in the context of a presumed "democratic" model of politics that F. A. Hayek has advanced two distinct proposals for fiscal reform worthy of brief discussion here. In his treatise, The Constitution of Liberty,*112 Hayek strongly argued against progressivity in rate structures of income taxation. His argument for proportionality in rates was related directly to his more general argument for the "rule of law." Proportional taxation was classified as falling within his normative requirement that all rules or laws be general in the sense that all persons in the community be equally subjected to their impact and effect. By contrast, progression in tax rates was held to violate this basic precept of generality.
As our earlier analysis in this book has indicated, a requirement for tax-rate proportionality may not seriously constrain a governmental Leviathan bent on maximizing surplus, either because the decision makers are effectively isolated from the taxpaying citizenry or because the public-spending structure can be manipulated to achieve any result desired. If, however, we shift out of the extreme Leviathan model, as we move toward more "realism" with models that contain Leviathan-like elements but which also require that decision makers (politicians and bureaucrats) remain as members of the taxpaying citizenry and which, further, restrict the range for overt transfers, the proportionality restriction might well serve to constrain discrimination among different groups of taxpayers. One effect would surely be to reduce, or even substantially to eliminate, the debate-conflict over relative tax shares (over the "degree of progressivity") that characterizes much modern discussion of alleged "tax reform."
In his more recent work, Law, Legislation and Liberty, vol. 3,*113 Hayek proposes a different, and more structural, reform of political process that relates directly to the taxing power. He proposes that the structure of taxation, the distribution of relative tax shares among persons and groups, be chosen in the deliberations of a new and differently elected and differently organized assembly, an upper house, whose sole function is limited to the enactment of general laws or rules, which would presumably, once enacted, remain in force over long periods of time. Hayek's "general laws" seem equivalent to what we should here call "constitutional rules." He would then allow the other assembly, the ordinary legislature or parliament, to choose levels of taxation and, of course, levels of outlay along with the allocation among uses.*114 As with his earlier proposal for tax proportionality, the structural reform suggested by Hayek is designed primarily to reduce or to eliminate the in-period political squabbles over the distribution of relative tax shares. In a sense, both of Hayek's proposals have as their objective some insurance that tax rules be treated constitutionally rather than nonconstitutionally.
10.2. Tax Reform as Tax Limits
I would add, however, that there are certain tax "reforms" under discussion that threaten to contract rather than to augment net tax revenues and therefore tend in the wrong direction. [Italics added.]
The intellectual integrity of E. S. Phelps, as indicated in the citation above, is to be applauded. But in his candor, he has really let the cat out of the bag of orthodox tax-reform advocacy. As our earlier analysis has revealed in more technical detail, many of the orthodox proposals for "tax reform," especially those that are alleged to be directed toward the achievement of enhanced economic efficiency, may be reinterpreted as directives that suggest how Leviathan may secure additional revenues. "Minimization of excess burden" and "maximization of net tax revenues" become, in many instances, the two sides of the same coin. Income-tax-reform advocacy provides the simplest, and most important, illustration. The dominant emphasis is on making the tax base more comprehensive rather than less, a suggested "reform" that would, if enacted, tend to ensure a larger total tax take under political assumptions that allow any scope at all for Leviathan proclivities. Similarly, proposals for commodity taxation involve the introduction of differential rates that are related to the different degrees of substitutability between taxed and nontaxed outlays.
These, and other, suggestions for "tax reform" are justified within the strict confines of the economists' ivory tower by the equi-revenue framework imposed on analysis. The suggested changes are made to appear almost wertfrei, and they seem to allow the "oughts" of tax reform to be derived from the minimal ethical postulate required for the acceptance of the Pareto criterion. However, the very usage of the equi-revenue constraint depends upon the acceptance of the benevolent despotism model of politics, a proviso that is rarely made explicit and may not even be realized by many of those who participate in the discussion. Such an acceptance is not merely something that is convenient for analysis and without substantive importance. It has implications for the real fiscal world of taxing and spending.
We may illustrate by analogy. "It is costly to build a fence or to purchase a chain. It is possible to prove that the no-fence, no-chain solution is more efficient than either, provided that we model the behavior of our dog in such a way that he respects the boundaries of our property." As we have put this example from personal experience, the exercise seems, and is, absurd. But is it really very different from that procedure which argues that tax structure X is more "efficient" than tax structure Y provided that we model the behavior of government in such a way that it seeks only to further efficiency in revenue collection?
Once the benevolent despotism model of governmental behavior is abandoned, the orthodox suggestions for tax reform that tend to emerge from the equi-revenue analytical framework cannot stand alone. Alternative tax institutions and rules must be evaluated on criteria other than efficiency and equity, although, of course, these standard objectives remain relevant. If, for example, the comprehensiveness in income-tax base suggested by the efficiency criterion is predicted to offer Leviathan the opportunity to extract more revenues from taxpayers, this instrumental target for reform may well be discarded in favor of some alternative that produces more effective predicted results consistent with the political model adopted. As the analysis of earlier chapters has indicated, a change in the political model may turn many of the orthodox precepts for tax change on their heads.
In almost any nonbenevolent model of political process, which need not, of course, be so severe as the specific Leviathan constructions we have introduced, the choice of the desired "tax constitution" arises, and this constitution is defined by the constraints or limits that it places on government's power to tax. The tax-reform exercise becomes categorically different from that which posits the giving of normative advice to the benevolent and all-wise government. In the latter, and standard, version, any constitutional-legal constraint on the ability of the fiscal authority to respond directly to the advice so proffered can only be negatively valued. "Good" government can only be limited from doing "good." The desire to limit government constitutionally, to define in advance the range and scope for the subsequent implementation of the taxing authority, arises only from a presumption-prediction that government may, at least on some occasions, act in ways that are not within the interests of taxpayers. In this setting, the tax-reform exercise becomes that of choosing among alternative sets of limits.
We shall, in subsequent sections of this chapter, examine some of the fiscal limits that have been discussed in the "tax revolt" of the late 1970s. Before launching into such an institutional array, however, it is useful to look briefly at the possible objectives for "constitutional tax policy," considered in their more general sense. On the assumption that the potential taxpayer, at the stage of constitutional choice, does not model government in idealized terms, what protections will he try to secure through constitutional limits on the power to tax? We may distinguish between relative and absolute guarantees. The individual may be interested in his own position in the postconstitutional budgetary sequence relative to other persons in the political community. He also may be interested in his position vis-à-vis the fiscal authority defined or described in absolute terms.
Some insurance that the fisc will not arbitrarily discriminate against the individual, any individual, may be a desired feature of almost any acceptable constitutional-policy set. This objective is, of course, the same as that which appears in different guise under the horizontal equity norm of the traditional tax literature and also under the more inclusive "legal equality" norm familiar in the discourse of jurisprudence. The methodological advantage of the veil-of-ignorance-constitutional analysis is that it allows us to derive the logical basis for such a norm from the choices of the individual rather than from some presumed external ethical standard. As we have noted in earlier chapters, existing fiscal or tax constitutions in Western nations incorporate some of the limits against arbitrary discrimination in the distribution of tax shares that almost any person's "efficient" constitution would embody.
Existing constitutions do not, however, embody protections to the potential taxpayer in the absolute dimension. In the United States, existing constitutional law would presumably prevent the levy of a confiscatory tax on Mr. A while allowing Mr. B to remain free of tax, assuming that both persons are, somehow, "equally situated" as deemed relevant for tax purposes. By contrast, there is nothing in existing constitutional law or its interpretation that would deter government from levying a confiscatory tax on both Mr. A and Mr. B. Much of the support for new and additional constitutional constraints on government's fiscal powers stems from a general, if vague, sense of the existence of this anomaly in the fiscal constitution.
Our Leviathan construction is helpful in isolating and identifying the potential value of constitutional protection against absolute exploitation of the individual by the state. The requirement for uniformity in treatment, or insurance against relative tax deprivation, has been shown to complement constraints aimed at limiting absolute revenue potential in many cases. The significant exception here lies in the possible revenue productivity of progressive and proportional taxation. If the uniformity or legal equality precept is interpreted to require rate proportionality, as it is by Hayek and as it was by the U.S. judiciary prior to the passage of the Sixteenth Amendment in 1913, the shift to nonuniformity in the direction of rate progression may well tend to reduce rather than to enhance the maximum revenue potential available to the fiscal authority. This possible relationship tends to be obscured by the American historical experience in which the twentieth-century federal government revenue explosion was facilitated by the introduction of progressivity in rate structures. Rate progression serves to ensure that revenues increase disproportionately with real economic growth and with inflation, but this relationship would also operate under proportionality if rates should be set and adjusted continuously to revenue-maximizing levels.
It is inappropriate in this summary chapter to expand and elaborate analysis that has been developed earlier. But it is useful to remind those who seek to move toward tax-rate uniformity as a means of constraining overall revenue growth that the relationship may sometimes be reversed. The aggregate revenue potential of a broad-based proportional tax (e.g., a value-added tax) tends always to be greater than the aggregate revenue potential of a highly progressive tax on a roughly comparable base. The argument here is not to suggest that movement toward more restrictive constitutional guarantees of tax-rate uniformity may be undesirable. Our point is only that, in some cases, the guarantee of equality of tax treatment may be secured at the partial expense of tighter guarantees against absolute revenue limits.
One further qualification should be made before we discuss specific proposals for constitutional constraints on the taxing power, as these have variously emerged in the 1970s. For the most part, and almost necessarily, the debates over constitutional tax policy have proceeded in a setting where the economic positions of individuals and taxpaying groups are well identified, at least over short-term planning horizons. In such a setting, self-interest dictates, of course, that individuals and groups favor policy actions that promise to yield the most advantageous results. That is, the conflict over the assignment of tax shares implicit in all the orthodox tax-reform advocacy is not absent from the conflict over constitutional tax policy.
Nonetheless, this conflict is necessarily reduced in the constitutional choice setting. When tax rules are treated as quasi-permanent features of the legal structure, individuals are naturally more uncertain about the effects of differing rules on their own positions.*115 Further, the conflict need not emerge at all, whereas it must emerge in the alternative tax-share allocation context.
10.3. Tax-Rate Limits: The Logic of Proposition 13
The most significant fiscal event of the late 1970s was the approval of Proposition 13 in California in June 1978. In a state-wide referendum, voters by a two-to-one margin approved a constitutional limit to restrict the tax on real property to 1 percent of market value. There were many features of Proposition 13 peculiar to the California fiscal-political environment in 1978 that need not be examined here. It is instructive, however, to try to look at this event within the analytical framework of this book. From a truly constitutional perspective, is there any logical basis for imposing maximum limits on rates for specific taxes? Are there plausible reasons for predicting that maximum rate limits will be imposed on real property taxes but not on other taxes? What are the predicted consequences of constitutional rate constraints on some but not all of the tax instruments within a government's fiscal bag? Can rate limits alone restrict total spending? Can rate limits offer any guarantee of equal relative treatment?
It is first of all necessary to distinguish clearly between constitutionally imposed limits on rates for specific taxes and constitutionally imposed limits on overall revenue collections, defined as some share of product or income and, hence, as some inclusive "rate" of tax. We shall postpone discussion of the latter type of limits until Section 10.5. In this section, we shall discuss rate limits for specific taxes.
If the potential taxpayer in some constitutional choice setting conceives of government in Leviathan terms, he will recognize that the imposition of maximum rates for any particular tax will result in a diversion of fiscal pressures toward those taxes that may not fall under the rate-limit constraint. However, any rate limit on one tax from among the allowable set available to government must reduce the total revenue potential collectible by government from the whole set. Whether or not the introduction of specific rate limits offers a desirable or efficient means of achieving the overall absolute constraint on revenues that may be desired is another issue.
There may be desired objectives of constitutional tax policy that are at least partially independent of the aggregate results defined in terms of total revenue potential. Rate or base limits may be aimed at some of these peripheral objectives, even if the taxpayer is not primarily concerned with the absolute levels of revenue potentially extractable from the citizenry. In the analysis of Chapter 5, the announcement effects of taxes were discussed with special reference to taxes on wealth. Emphasis was placed on the temporal dimension, which separates the taxpayer's decisions on saving and capital accumulation from the potential tax levy by government. The extreme vulnerability of the individual to fiscal exploitation in this setting relative to that in which he can make behavioral adjustments after taxes are announced will be recognized by the individual at the constitutional state of deliberation. The analysis of Chapter 5 suggested that the individual would be reluctant to grant government the constitutional authority to levy taxes on wealth and capital.
The most important tax on wealth in most fiscal systems is the tax on real property. We observe that this tax is largely concentrated within the fiscal authority of localized units of government. And there is a sound, logical reason for such a constitutional assignment of tax base, as our analysis makes clear. The prospect for interunit resource mobility, human and nonhuman, tends to restrict the potential exploitation of the real property tax. If, however, local units are not sufficiently competitive, one with another, and if strictly locational rents are significant, localized assignment alone may not provide effective guarantees. In this case, individuals may reasonably demand some imposition of maximum rate limits in order to allow the making of long-term decisions concerning investments in real assets.
Individuals save in order to accumulate wealth, but the rate of saving in any one period is small relative to the total value of wealth potentially accumulated. A rational saving plan, projected forward through many periods, requires that the individual possess some reasonably accurate expectations concerning the tax liabilities that asset ownership will involve. He cannot readily adjust his desired portfolio to changing tax burdens. And for most taxpayers, the purchase of real property represents a commitment to a long-term saving-investment profile, and one that is not readily adjustable at the margins of value. Some constitutional guarantee that the tax claims against such assets will not exceed a specific share of market value may be desired by substantially all potential taxpayers, quite independently of personal persuasion concerning the proper size of government or the proper place of a property tax within the overall tax mix.
It should not, therefore, be surprising that Proposition 13 embodied rate limits on the tax on property values and that other tax-protest emphases have been somewhat differently orchestrated. The assessment here does not, of course, enable us to say anything at all about the appropriateness of the particular rate limit adopted. The argument suggests only that there may be a logical basis for imposing maximum rate limits on taxes on wealth. We should predict, therefore, that this approach toward fiscal constraints, generally, will be more likely to surface with respect to wealth and capital taxes, along with specific limits on public-debt issue, which involves somewhat comparable logical support.
10.4. Tax-Base Constraints
Proposals to introduce new constitutional restrictions on the bases for taxation made available to government were not within the set of suggestions for practical constitutional reform in the late 1970s. However, as our analysis, particularly that developed in Chapters 3 and 4, suggests, base constraints warrant more serious consideration. Moreover, one important reason for directing attention to tax-base questions is that orthodox tax advocacy is preoccupied with this aspect of tax reform. The policy stance that emerges from the conventional treatment, and that is now taken for granted in virtually all professional discussion of tax policy, leads inexorably to broader tax bases and correspondingly larger potential tax revenues. Much of our discussion is designed to indicate the extent to which such policies depend on heroic political assumptions. Apart from our intellectual and academic purposes, however, we believe that tax-base limitations may have a practical role to play in "tax-limit" policy.
One way of looking at tax-base assignment is in terms of "governmental property rights." As we noted, the constitutional delegation of that base amounts to the assignment of a monopoly franchise in the exploitation of that base. The behavior of government, in possession of such a franchise, becomes predictable in a manner analogous to that of the profit-seeking monopoly firm. And with an ability to make such predictions, individuals can anticipate and plan for their own adjustment responses. If the tax-base assignment is made with care, individuals will be protected against undue fiscal exploitation without more complex constraints on the fiscal authority.
The individual will not, of course, want to allow government access to taxable bases that are sufficiently general (inclusive, comprehensive) to allow the generation of revenues far in excess of estimated public-goods financing requirements, given some complementary estimates for the share of collected revenues likely to be expended on goods provision. To ensure that base restrictions will indeed be constraining on governmental action, the individual must try to ensure that government will limit its tax-rate imposition because of the predicted behavioral responses of taxpayers, responses that will, beyond certain rate limits, reduce rather than increase total revenues.
If we translate this relationship into the tax-reform jargon, we can say that the individual at the constitutional stage will seek deliberately to build certain "loopholes" or "escape routes" into the tax structure. These provide the protection or guarantee against undue fiscal exploitation that the individual wants the constitution to embody. This argument in favor of loopholes and against comprehensiveness in tax base runs directly counter to the norm or principle that is central to much of the orthodox tax-reform advocacy.
In this summary section, it may be useful to specify precisely what sort of "defense of loopholes" our analysis implies, and to discuss this defense in the context of existing tax systems. The argument for leaving open avenues for flexibility in behavioral response to tax rates suggests the rationality of constitutional loopholes. But what about a possible postconstitutional opening up of loopholes or maneuvering of an agreed-on tax base? Is this action a desirable or undesirable characteristic of in-period political "reform"? The answer to this question must be ambiguous. On the one hand, unless the rate structure is effectively constrained, a Leviathan government could, through a combination of high nominal tax rates and a set of tax preferences, move toward the discriminating monopoly level of revenue analyzed in Chapter 4. On the other hand, governments may seek to further particular objectives other than straightforward revenue maximization by manipulation of tax rules. "Tax preferences" may be granted to citizens in "exchange" for certain modifications in their behavior. To the extent that affected taxpayers respond, they are clearly better off than they would be without the tax preference. Other taxpayers need be no worse off; their tax treatment by a Leviathan government would be unaffected; and the activities of others encouraged by the tax preference will, presumably, exert favorable spillover effects at least equal to any reduction in public-goods financing necessitated by the reduction in tax revenues. Our analysis suggests, therefore, that there may be an important normative distinction between constitutionally and postconstitutionally imposed tax loopholes.
10.5. Aggregate Revenue and Outlay Limits: Ratio-Type Proposals for Constitutional Constraint
Most economists who support some constitutional constraints on governmental fiscal powers beyond those in being tend to favor that set of proposals that attempts somehow to relate aggregate revenues or outlays to some aggregate economic base, such as total income or product. In 1973, one of the first of this type of proposals, Proposition 1 in California, was soundly defeated in a referendum. In 1978, proposals of this nature were made parts of the state constitutions in Tennessee, by convention, and in Michigan, by a referendum vote. In early 1979, the National Tax Limitation Committee sponsored a proposed amendment of this type to the U.S. Constitution. As drafted, and as introduced for consideration in the U.S. Senate, this amendment would restrict percentage increases in rates of federal government outlays to percentage increases in gross national product over preceding years, with a tempering adjustment designed to penalize government for generating inflation. The state proposals, somewhat more simply, tie state tax revenues or state outlays directly to state income, in proportionate terms, more or less at the status quo of the date of enactment.
Aggregate revenue and outlay constraints, defined in relative terms, appeal to economists because they seem to be directed at the central objective, which is one of limiting the overall size of government, of keeping the Leviathan-like proclivities that seem to have surfaced only in recent years from further encroachment on the private sector of the economy. By contrast with these apparently direct approaches, rate and base limits necessarily seem to leave more scope for governmental manipulation and evasion, still within the allowable fiscal powers of the constitution. With some effective aggregate revenue or outlay limits, governments would be forced to nonfiscal alternatives in order to evade the constitutional requirements.
On the other hand, ratio-type proposals have major disadvantages which may well prevent their effective implementation. The proposals seek to establish some constitutionally enforceable relationship between revenue or outlay totals (R and/or G) and the total income or product in the economy (NI or GNP). Both elements in this sort of relationship are aggregates that require, first, abstract definitions, and, second, expert measurement by specific criteria. Neither element in the relationship carries direct meaning for the taxpayer other than as an abstract idea. Neither element means much in personally measurable value. By sharp contrast, a rate or base limit is readily translatable into personal terms. The taxpayer knows roughly what the Jarvis-Gann ceiling of 1 percent of market value implies for her tax bill next year. And a taxpayer also would know what it means when the government is not allowed to tax interest on municipal bonds or on the rental value of owned homes. Generalized taxpayer support for base and rate constraints falls within the realm of the possible in the constitutional tax-policy debates. Comparable support for ratio-type proposals to limit government's fiscal powers may be exceedingly difficult to organize.
Another difficulty with ratio-type proposals lies in the relationship between the specific ratios or shares that are normally suggested and the currently existing status quo. There is little other than historical accident that determines the government's current share in aggregate product. Why should this share necessarily be frozen at its 1980 limits? Arguments may be advanced concerning the appropriate share, quite independently of the present limits, but the constitutional-policy discussion has not, to this point, taken this form. It becomes difficult, therefore, to assess the whole ratio-type approach to fiscal constraint on analytical grounds that are comparable to our earlier discussion of rate and base limits.*116
10.6. Procedural Limits: Qualified Majorities and Budget Balance
All of the proposals for constraining the fiscal powers of government that have been examined in Sections 10.3 through 10.5 can be classified as result- or end-state-directed, in one sense or another. They are designed to set specific limits on what government can and cannot do. They involve the setting of maximum rates on specific tax rates, the defining of the bases upon which government is to be allowed to levy taxes, and the maximum shares in total economic income or product that government is to be allowed to take and to spend. None of these proposals is directed at the governmental decision structure as such. None of the proposals is aimed at processes or procedures through which end-state results are produced. An alternative, and conceptually quite distinct, approach to constraining government's overall fiscal powers is to modify and to limit the structure within which governmental outcomes emerge.
Mention was made earlier in this chapter of Hayek's proposal to separate the power for setting the tax-share distribution among individuals and groups and the power of setting the level of tax rates, given the tax-share distribution. This proposal falls within what we shall call here the set of procedural constraints on fiscal powers. Earlier in the book we have, on several occasions, made reference to Knut Wicksell's proposals for constitutional change which took the form of requiring qualified majority approval of spending legislation by members of legislative assemblies. Wicksell moved from his idealized process requiring unanimous approval to the qualified majority process which involves the approval of as much as five-sixths of the members. Some participants in the discussion of the 1970s have called for constitutional requirements that dictate three-fifths, or two-thirds, approval of spending legislation in legislatures. Proposition 13 in California, in one of its less familiar clauses, requires a two-thirds majority in the state legislature for the enactment of new taxes. Further, almost all of the specific proposals previously discussed are framed in such a way as to include escape clauses, for overriding restrictive limits in times of war or national emergency. These escape clauses are almost all stated in terms of qualified majority approval in state or national legislative bodies.
These procedural reform proposals seek to constrain fiscal outcomes indirectly by modifying the process through which governments are allowed to reach fiscal decisions. One particular proposal that falls within this set has received widespread support and warrants brief discussion here. In early 1979, thirty state legislatures had approved resolutions calling for a constitutional convention for the purpose of adding a budget-balance amendment to the U.S. Constitution. Some thirty members of the U.S. Senate had sponsored resolutions that would have proposed such an amendment to the constitution but without the convention route. In effect, the proposal for requiring the federal government to balance its budget is designed only to ensure that government cover its outlays with tax revenues rather than with public-debt issue or with new money creation. It does not aim directly at the level of revenues or outlays. In a sense, this proposal may also be interpreted as falling within the base-limit set; this would be the case if we should treat public-debt issue and money creation as forms of taxes. The amendment would effectively deny these "tax bases" to government. In a more fundamental sense, however, the amendment for budget balance may be interpreted in procedural terms. It seeks to modify the governmental decision process by requiring that decision makers, whoever these may be, balance off costs against benefits.
It may be argued that budget balance was a part of the existing fiscal constitution of the United States prior to the Keynesian revolution in the theory of economic policy. Even if the constitution did not contain a formal, written requirement for budget balance, governmental decision makers acted as if such a constraint did limit their fiscal behavior. The effect of the Keynesian revolution was to repeal this part of the fiscal constitution.*117 Much of the support for the introduction of an explicit constitutional provision for budget balance in the 1970s stems from a widening recognition that, in the absence of such a constraint, governments will revert to their quite natural tendency to generate budget deficits almost continually.
In comparison with the more restrictive proposals for constitutional change, the budget-balance amendment, if approved, would not seriously constrain Leviathan's fiscal appetites. On the other hand, proponents of the amendment predict that the electoral checks on governmental process will work much more effectively if governments are prevented from concealing the genuine costs of governmental outlays through deficit creation. In one sense, budget balance might be considered as a first step toward more comprehensive constitutional constraints on the fiscal powers of modern governments.
10.7. Toward Authentic Tax Reform
Our basic purpose in this summary chapter is not to support or to reject any one or any set of the various proposals for constitutional reform that have been advanced to impose constraints on government's fiscal power. Our whole analysis, however, may properly be interpreted as offering analytical argument in support of some appropriately designed set of limits. Our extreme Leviathan model of politics is not critical for this conclusion. This model, which we introduced to allow us to develop our analysis with some logical rigor, may be substantially modified in the direction of more "realistic" political assumptions without undermining the general conclusion in support of constitutional constraints.
To the extent that our analysis bears on "tax reform" at all, it does so with reference to "constitutional tax reform," and, as noted, with particular reference to "tax limits." We have little or no interest, here or elsewhere, in proffering our own private and personal advice to existing governments about taxes or indeed about anything else.
If we can succeed in shifting the grounds for debate in tax reform, we shall have opened the way toward the only authentic tax reform worthy of serious consideration.
Given this somewhat restrictive methodological stance, it would be inconsistent for us, at this stage, to lay down our own pet prejudices as to "ideal tax arrangements," whether these be supported on efficiency, equity, or some other grounds. We hope to have been able to demonstrate that the individual, as potential taxpayer, placed in a hypothetical position where he is allowed to select among alternative sets of constraints on the taxing power of government, will rationally choose to exercise his option and impose some such constraints. In a broadly defined perspective, therefore, our whole analysis may be interpreted as providing positive argument in support of almost any one of the currently discussed proposals for constitutional fiscal limits. To go further than this, and to try to isolate supporting arguments for any one of the proposed set of prospects that have been suggested, would require more analysis than we have been able to muster. We do not want to make the mistake of suggesting that a unique constitutional solution will necessarily emerge even from the most idealized modeling of constitutional choice.
As we noted in the Preface as well as earlier in this chapter, this book is both badly and excellently timed. Perhaps a two- or three-year advance date in writing and in publication might have allowed some of our analysis to inform more fully some of the current discussions of constitutional change. But a ten-year advance would have surely meant a neglect of our whole argument. The "taxpayer revolution" of the 1970s has been politically exciting regardless of the final tally in terms of ultimate change in the political framework of society. And this political event has offered intellectual challenges that too few of our fellow economists and other scholars have chosen to accept. To the extent that our analysis in this book prompts others, regardless of ideological persuasion, to take up the gauntlet, tax-reform discussion will have moved beyond the realm of partisan advocacy toward authenticity in the desired debate over constitutional alternatives within which we should allow government's fiscal powers to be exercised.
Notes for this chapter
Our use of minimax, however, seems more justifiable than that of Rawls because we are explicitly modeling the behavior of the "adversary player," the Leviathan government, rather than "nature," which presumably distributes talents and capacities with no malign intent.
For a specific discussion of this adage, see James M. Buchanan, Public Finance in Democratic Process (Chapel Hill: University of North Carolina Press, 1967), chap. 5.
F. A. Hayek, The Constitution of Liberty (Chicago: University of Chicago Press, 1960).
F. A. Hayek, Law, Legislation and Liberty, vol. 3, The Political Order of a Free People (Chicago: University of Chicago Press).
We should note that the political structure that Hayek is implicitly suggesting be modified is one of parliamentary government, akin to the British structure. His suggestions for change fit less readily into the U.S. context.
For a comprehensive analysis of the effects of extensions in the time sequence of tax rules on the prospects for securing agreement among persons, see Antonio Pinto Barbosa, "The Constitutional Approach to the Fiscal Process: An Inquiry into Some Logical Foundations," Ph.D. Dissertation, Virginia Polytechnic Institute and State University, 1978.
For a general discussion of tax limits in the context of the 1978 issues, see Geoffrey Brennan and James M. Buchanan, "The Logic of Tax Limits," National Tax Journal, 32 (June 1979), 11-22.
This theme is developed at length in James M. Buchanan and Richard E. Wagner, Democracy in Deficit (New York: Academic Press, 1977).
End of Notes
Return to top