Abstract
It is argued that ex antewillingness to pay is appropriately measured as a function of subjective supply and demand probabilities. An exploratory case study of a loggerhead sea turtle protection programme is used to show how estimates of these probabilities can be obtained from a sample survey and how a change in probabilities determines contingent choice. Empirical results confirm that subjective probabilities are important when explaining ex antewillingness to pay. Attempts to estimate the determinants of ex antewillingness to pay without controlling for subjective probabilities will result in specification bias. It is important that subjective probability information is gleaned from contingent valuation survey respondents and included in empirical models of contingent choice.