Optimal Taxation under Imperfect Trust

Abstract

We study optimal taxation when citizens are not fully confident that the government will transform tax revenue into useful public goods. In an otherwise standard Ramsey framework, a representative agent values a public good financed by distortionary taxes, but believes that the government is honest only with some given probability and may otherwise divert all revenue. This simple departure from the canonical model delivers two central results. First, there is a sharp trust threshold: if perceived government honesty is too low, any positive tax rate lowers expected welfare and the optimal policy is a zero-tax corner, even though the public good is valued. Second, once trust exceeds this threshold, the usual sufficient-statistics logic of optimal taxation re-emerges, but with a trust-adjusted marginal value of public funds that scales down the benefits of raising revenue. In a simple parametric example we obtain closed-form expressions that map trust into the optimal tax rate and the size of the public sector. The framework provides a compact way to incorporate government credibility into tax design and suggests that in low-trust environments credibility-enhancing reforms should precede attempts to expand the tax base.