THE LOGIC OF TAX LIMITS: ALTERNATIVE CONSTITUTIONAL CONSTRAINTS ON THE POWER TO TAX

Abstract

entities, total budgets and total product or income. But setting these problems of electoral psychology aside, are there any purely technical problems that ratio limits seem likely to pose? There are two that seem to us particularly important one relating to whether the limits are effective, and the other relating to whether those effective limits are achieved in the most desirable way. In the simple Leviathan model outlined above, government seeks to maximize its own surplus or income, measured by revenues, R, minus the necessary outlay on public goods and services, G. If it is constrained as R, either in absolute or share-of-income terms, it may seek to provide the necessary G indirectly rather than directly, and thereby increase its surplus, Y. One means of accomplishing this result would be to manipulate the structure of taxes while remaining within the overall uneven constraint. Consider, for example, some publicly provided good such as education. Clearly, the government can achieve a required level, either by direct provision of education (i.e. it can individuals and provide the service free of direct charge) or by subsidizing private provision. And the subsidy in question may take the form of vouchers (or direct subsidy of some other type) or concessions. In principle, each of these alternatives could produce precisely the same outcome, but they clearly involve quite different amounts of nominal revenue. The subsidizing of private activity will typically achieve a larger response per dollar of revenue raised than direct provision; obversely, for any given level of activity, subsidy will typically involve less revenue than direct provision. Subsidies given as direct payments will likewise involve more nominal revenue and a higher nominal government budget than concessions which induce an identical private response. At the very least therefore we should require that such tax expenditures be incorporated into the relevant measure of total revenue for 'share limiť purposes. But this in turn raises the question as to whether we ought also distinguish between subsidies and direct provision; for, as we have observed, an identical amount of revenue can involve a different level of government influence depending on the form of government action. Even if the implicit subsidy embodied in taxexpenditures is made explicit, therefore, the problem associated with translating any given revenue figure into some measure of government output remains. And there are important conceptual issues at stake in this. Is it clear, for example, that we do wish to minimize government output or influence, independently of its revenue cost? Is there a distinction to be drawn between policies which involve revenues passing through the hands of government agencies, and those which do not? We do not attempt to answer these questions here. But it does seem clear that any share limit which accurately reflected what it is we wish to measure would be extremely complex and itself somewhat open to debate, and the problems of electoral ignorance would be correspondingly aggravated. Equally, the extent to which an aggregate revenue limit would succeed in achieving the objective of placing limits on government activity seems open to some doubt. Suppose, however, that in the presence This content downloaded from 157.55.39.171 on Sat, 25 Jun 2016 05:31:38 UTC All use subject to http://about.jstor.org/terms 16 SUPPLEMENT NATIONAL TAX JOURNAL of this maximum share (appropriately defined), government is genuinely constrained. Nevertheless, government can achieve its maximum share in a variety of ways between which the citizentaxpayer behind the veil of ignorance could not be expected to be indifferent. For example, different arrangements will induce different excess burdens: in the extreme case government could exploit instruments considerably beyond their maximum revenue limits and the excess burden induced by the could be many times greater than the revenue raised. There is nothing inherent in the nature of share limits which would give government the incentive to seek out broadly based, minimally distorting taxes. As we have already seen, the government, on the contrary, has the incentive to acquire its revenue limit by the application of very high rates to a base made narrow by strategic concessions.7 Whether these high rates raise much or indeed any revenue is largely beyond the point: the associated excess burdens seem likely to be very substantial indeed.