A test-retest analysis of stated preferences in uncertain times

In environmental valuation, the issue of the temporal stability of stated preferences to changes in environmental (dis)amenities is important because their results can be employed to inform decision-making. This includes cost–benefit analysis for large infrastructure projects such as coastal protection. A couple of studies have investigated stability of stated preferences over varying time periods. However, less evidence is available for temporal stability of stated preferences for (dis)amenities in uncertain times, i.e., times that are characterised by larger degrees of uncertainty regarding the (near) future. Using a choice experiment on coastal adaptation to climate change, this paper examines the test-retest reliability of individual preferences and resulting welfare estimates over the course of the Covid-19 pandemic. We do so by surveying the same respondents at two points in time five months apart during the ongoing pandemic. Using a latent class model, we find similar preference heterogeneity patterns but different class sizes at the two points in time. While the welfare measure of an adaptation scenario that focuses on safety increases across survey waves, scenarios that centre on recreation or nature have decreasing welfare effects. This suggests that individuals set other priorities in uncertain times.