Household willingness to pay for forest ecosystem restoration under climate risk perception in Khyber Pakhtunkhwa Pakistan

Mountain forest ecosystems in South Asia provide watershed, biodiversity, and livelihood services on which millions of downstream and forest-adjacent households depend, yet because these services are non-marketed, their loss rarely enters formal cost-benefit calculations that guide restoration investment. Forest degradation in Khyber Pakhtunkhwa (KP), Pakistan, poses acute threats to watershed integrity, biodiversity, and the livelihoods of forest-dependent rural communities, and the economic value of ecosystem restoration in the region remains empirically unquantified. This study elicits household willingness to pay (WTP) for forest ecosystem restoration across five attributes Canopy Cover, Carbon Sequestration, Biodiversity, Watershed Protection, and Recreational Quality using a Choice Experiment with attribute levels biophysically calibrated through the InVEST model (full experimental-design details, including D-efficiency statistics, are reported in the Methodology), administered to a stratified sample of 430 forest-dependent households in the Gallies Forest Division, KP. Each respondent completed three choice tasks, each consisting of two forest-restoration alternatives and a status-quo alternative, yielding 1,290 choice decisions and 3,870 alternative-level observations in long format. Preference heterogeneity is modelled using Mixed Logit (ML) and Latent Class Modelling (LCM), with a Yale-adapted climate risk perception scale integrated as a segment-membership covariate. Results confirm that sampled households strongly preferred both restoration alternatives to the no-intervention status quo i.e., to the continuation of the current, ongoing decline in forest condition, with Biodiversity and Watershed Protection emerging as the highest-valued attributes, reflecting direct livelihood dependency on these services. The LCM identifies three interpretable preference segments Livelihood-Focused, Conservation-Focused, and Payment-Resistant with climate risk perception significantly predicting membership in the Conservation-Focused class. Mean household WTP totals PKR 755 per year under ML and PKR 818 per year under the class-share-weighted LCM; aggregating the LCM compensating surplus (PKR 8,411/household/year) across forest-adjacent households yields an aggregate annual restoration benefit of PKR 361.7 million (approximately USD 1.29 million), roughly 7–8% above the ML estimate. A four-panel robustness battery validates the credibility of these welfare estimates. Findings provide an empirically grounded welfare foundation for KP forest restoration policy, including a tiered Community Forest Conservation Fund design, Reducing Emissions from Deforestation and forest Degradation (REDD+) co-financing negotiations, and the development of a Payment for Ecosystem Services scheme across the Indus watershed.