People rarely act against their own interests. Ask an employee how a team or company could be improved and you will hear every suggestion under the sun—better processes, clearer leadership, more resources—except one. No one volunteers the option of their own departure. The reason is not malice but ordinary self-interest. People do not readily vote against themselves.

The same pattern appears, scaled up, in the behaviour of organisations. Institutions almost never conclude that their work is finished and that orderly closure would serve the public best. Instead they expand their remits, defend their budgets and seek to carry on. The people inside them have careers, status and identities tied to the organisation’s survival; the benefits of shutting it down are spread thinly, while the costs are personal and immediate. Survival and growth become the default.

These two observations find their most consequential expression in democratic politics. Voters, like employees and bureaucrats, act on self-interest. Those who receive benefits—pensions, subsidies, transfers, protected status—have every reason to support their continuation and expansion. The costs fall more lightly and more widely, often on taxpayers who are less organised or on generations yet to vote. Over time the share of the electorate with a direct stake in the benefits tends to grow. What begins as a limited safety net can become an expanding claim on the public purse, sustained by the very people who gain from it. Government budgets in Western countries have, over the long twentieth century and into the twenty-first, risen markedly as a share of national income. The expansion has not been perfectly continuous, nor driven by a single cause, yet the overall direction is clear, and the political difficulty of reversing it is plain.

The result is a system that does not readily correct itself. Self-interest is not an occasional flaw but the reliable engine. Individuals protect their positions, organisations protect their existence, and democratic majorities protect the benefits that majorities enjoy. Each level reinforces the others. Hard decisions—removing underperformance, shutting down obsolete programmes, restraining entitlements when demographics or finances demand it—must therefore be imposed by mechanisms that do not rely on the voluntary assent of those who stand to lose.

Reformers have proposed various remedies: constitutional spending rules and debt brakes, independent fiscal councils that raise the reputational cost of profligacy, automatic adjustments to benefit formulas, sunset clauses, and commissions empowered to force up-or-down votes on packages of savings. Each seeks to bind future majorities or to alter the information and incentives under which they operate. Some have worked for a time in particular countries. None has proved permanently self-enforcing, because the underlying pressure of concentrated interests remains.

One more radical possibility follows the logic to its institutional root. If the franchise were confined to those who make a financial contribution to the state, the decisive voters would be those who fund the benefits rather than those who primarily receive them. The asymmetry that allows non-contributors to outvote contributors on pure redistribution would be removed. The idea is not without precedent; earlier liberal constitutions often linked political rights to property or taxpaying for precisely this reason. Its practical difficulties are real—defining contribution fairly, handling temporary non-payers, preserving legitimacy—and its normative implications are profound. Yet as a pure alignment of incentives it addresses the mechanism at its source. In a system otherwise prone to non-correction, it remains a serious, if uncomfortable, candidate for consideration.