Mitigating cost vector effects in stated choice experiments using cheap talk and opt-out reminders

The magnitude and range of the cost attribute levels in stated choice experiments have been found to affect willingness to pay (WTP) estimates. Such cost vector effects are of concern for the validity of derived welfare estimates. This paper employs a treatment design to investigate whether using a cheap talk and opt-out reminder device, which has previously been found to be effective in reducing hypothetical bias, mitigates potential cost vector effects. In the context of a stated choice experiment to value outcomes of marine and coastal management in Germany, results show considerable cost vector effects in a set of four control treatments that do not apply cheap talk and repeated opt-out reminders: WTP estimates are significantly higher as the magnitude of the cost attribute levels increases. However, cost vector effects are reduced substantially in another treatment arm where the combined cheap talk and opt-out reminder device is used, and model fit increases relative to control reatments. Ancillary results demonstrate how the device increases the validity of stated preference responses and resulting welfare measures. The findings suggest that the inexpensive use of this device in stated choice experiments may be an effective way to mitigate potential cost vector effects at the design stage.