Out Sight, Out of the Constrained Mind : Testing the Effect of a Cheap Talk in Choice Experiments

Cheap Talk (CT) has been found to be an effective ex ante tool for reducing hypothetical bias in Contingent Valuation Surveys (CVM). However, despite the apparent differences in the preference elicitation formats between CVM and Choice Experiments (CE), particularly in the number of stated preferences tasks, CT has been applied uncritically in CE studies without testing whether the effect of the CT is persistent across all choice sets. Testing the effect of CT in a CE study focusing on visual disamenity costs from off-shore wind farms, an effect of the CT on preferences can only be established in the first three of six choice sets, when tested with reference to preferences based on a sample that did not receive a CT. This suggests that CT alone may not take into account the repeated properties of the choice tasks in CE relative to CVM