Journal of Rural Development
Estimating willingness to pay for attributes to restore forests degraded by forest disasters: A comparison of mixed and conditional logit model results on preference heterogeneity and risk attribute treatment
-
AuthorJeon, Chulhyun
-
Publication Date2026.03.21
-
Original
This study aims to estimate the willingness to pay (WTP) for attributes related to forest restoration and the continuous supply of ecosystem services by forest disasters, such as wildfires and pests. Using a choice experiment approach, WTP estimates were compared across different methods for handling preference heterogeneity and coding of risk attributes. Two coding approaches were employed: continuous variable coding (conditional logit model I) and dummy coding (conditional logit model II). The mixed logit model revealed substantial preference heterogeneity among respondents, with WTP estimates of KRW 10,570 for wildfire risk reduction and KRW 8,360 for pest risk reduction. In conditional logit model I, the corresponding WTP estimates were KRW 13,483 for wildfires and KRW 9,236 for pests. In conditional logit model II, focusing on statistically significant levels, the WTP for the safest wildfire scenario (interest level) was estimated at KRW 6,589, while the WTP for pest risk reduction at the second safest level (caution level) was KRW 6,513. For forest outdoor activity restrictions, representing a preventive management measure targeting human behaviors associated with forest disasters, the highest WTP (KRW 8,014) was observed when restrictions were introduced from a no-restriction baseline. The results from the mixed logit and conditional logit model I were consistent with conventional expectations, whereas conditional logit model II produced negative coefficients for wildfire and pest risk attributes at certain levels. These findings were interpreted in terms of risk aversion, trade-offs between cost and risk, and model-specific characteristics related to cost–benefit balance. Overall, the findings provide economic evidence to support budget allocation decisions in forest disaster response policies and contribute to a understanding of risk attributes in stated preference studies.
- Prev
- Ripple Effects of Content Industry on the Agri-food Industry