The Restraint

The power that collects the tax is itself part of the problem of taxation. Earlier plates asked what a levy does to buyers, sellers, workers, and owners. They could hold the state's purpose steady while tracing the consequences. Public choice opens that closed box. The people who administer public revenue have purposes of their own; the institutions that restrain a private monopoly cannot simply be assumed to restrain the government that writes them. A state can supply the courts and roads on which prosperity depends and still possess an appetite for resources beyond what its citizens would choose to surrender. The constitutional question asks what powers people would entrust to that state before knowing who will exercise them, against whom, and with what freedom to increase the claim.

Brennan and Buchanan sharpened the question with a deliberately demanding model: a government seeking to maximize the revenue available to it. Their Leviathan is an institutional test, with electoral restraints assumed insufficient for the exercise. Its usefulness lies in asking what a fiscal arrangement permits when the officeholder's interest diverges from the taxpayer's. It is neither a finding that every official pursues revenue alone nor a reason to forget useful public services. Imagine designing a treasury for an official whose intentions are unknown. A rule that works only when that official is unusually restrained leaves the public dependent on a personality. A rule that binds an ambitious successor gives the institution a life beyond the virtues of its first occupant.¹

Under that assumption, the limit of revenue is a different boundary from the limit of a justified burden. Take an invented tax base B = 100(1 − t), where t is a tax rate between zero and one. Revenue is 100t(1 − t). At a rate of one-quarter, it is 18.75; at one-half, it is 25; at three-quarters, it returns to 18.75. The revenue-seeking authority chooses the peak, one-half. That calculation says nothing about the value of the services financed, the distribution of sacrifice, or whether raising the last unit is worth its cost. Nor does the falling side of this invented curve tell us where a real economy stands. The model exposes a distinction: a restraint that stops government only when additional extraction becomes self-defeating may stop it long after the citizen's preferred limit.²

A constitutional restraint must specify what is constrained and who can enforce the boundary. A cap on a named rate can be bypassed if the base is widened without the same scrutiny. A ceiling on taxes can move the claim into fees, mandatory private payments, or borrowing. A spending rule can invite expenditure to be placed outside the account it covers. These are possibilities to test in a rule's design, not descriptions of every government facing one. The arithmetic is plain: a ten-percent levy on a base of one hundred raises ten; retain the rate while doubling the base and the claim becomes twenty. If citizens meant to restrain the total claim, they have restrained only one of its components. Effective limits require a defined perimeter, visible accounts, and a decision maker able to apply them.³

The capacity to say no also needs a means of saying yes when the public has a good reason. An inflexible nominal ceiling may become tighter as prices and population change; a rule that allows every difficulty to be called an emergency may cease to constrain anything. Suppose an ordinary spending limit is one hundred and a genuine disaster creates a temporary need for twenty more. A public procedure can require the additional amount to be identified, its purpose stated, its duration limited, and its renewal decided openly. The value lies in making the departure contestable and temporary rather than silently changing the baseline. The design must also account for recessions, when receipts can fall without a new tax decision and an abrupt spending cut may deepen the shortfall.⁴

The test must be applied to the restraint's beneficiaries as carefully as to the treasury's. A minority empowered to block every change may protect itself from an unfair levy, or preserve a privilege while others fund the common services. A local government's competition for mobile taxpayers may discipline its spending, while shifting burdens toward residents with fewer choices. A constitutional promise is credible only through institutions that sustain it: transparent budgets, review, representation, and consequences for evasion of the rule. These safeguards cannot be reduced to a single low number in a statute. The relevant comparison is between workable arrangements, including what each permits, what each prevents, and who bears the cost when it works exactly as designed.⁵

The restraint asks the political community to judge fiscal power before the next attractive use of it arrives. The public may authorize a substantial common enterprise and still demand a tightly governed treasury; it may choose a smaller enterprise and discover that an opaque rule protects waste inside it. The size of government and the discipline of its powers are related questions with distinct answers. Public choice contributes the refusal to treat the collector as disinterested by definition. Constitutional design supplies the means to bind a future collector whose interests may differ from today's promises. The next plate turns from those limits to the authority within them: what makes a compulsory contribution a legitimate obligation, and what kind of consent can a whole society meaningfully give?