The Pigouvian Bias: Distributional Weights and Discrete Choice Valuation

Discrete choice experiments (DCEs) have become the standard tool for environmental policy evaluation, applying a willingness-to-pay (WTP) criterion to assess whether environmental interventions generate net social benefits. This practice assumes distributional neutrality, a marginal cost of public funds equal to one, and behavioural responses limited to consumer-side substitution. These conditions are inherited from the marketing origins of DCEs and are misaligned with the institutional and distributive structures shaping environmental taxation.Drawing on the sufficient-statistics framework of optimal taxation, this paper shows that the standard WTP-based criterion is a restrictive special case of a broader welfare-consistent environmental tax formula incorporating distributional weights, avoidance behaviour, and firm-side rent-seeking.

Using calibrated simulations based on European income distributions, we quantify what we term the Pigouvian Bias as the systematic divergence between the aggregate-WTP benchmark and the socially optimal environmental tax. The magnitude of this bias ranges from 14.6 to 36 per cent across parameterisations, with a baseline component of 14.6% driven solely by the marginal cost of public funds and independent of any normative assumptions about inequality aversion. The direction and magnitude of the remaining component depend on the income distribution and the degree of inequality aversion, varying systematically across European economies.We then propose the DCE-Saez, a redesigned stated-preference instrument that recovers the distributional, behavioural, and firm-response parameters required for welfare-consistent evaluation.

The design follows Johnston et al. (2017) on incentive compatibility and consequentiality, extending current stated-preference practice to recover the parameters that social welfare evaluation requires.