Demand for urban tree cover: A two-stage hedonic price analysis in California

- We conduct a two-stage hedonic price model to identify demand for tree cover.
- We use SAR model to identify the effects of urban tree cover on house price.
- We use segmented markets and instrument variable technique to estimate the demand.
- We find that own price elasticity of demand for tree cover is − 0.075.

This research seeks to use extensive home sale and socioeconomic data, coupled with urban tree cover data to identify the influence of urban trees on house values, estimate the demand for urban trees and ultimately calculate the welfare change of forest loss in California, with a two-stage hedonic price model. In the first stage model, we detected spatial dependence using the Lagrange-Multiplier Robust tests and found significant spatial correlations in house prices for each of five California counties. Our identification strategy relies on flexibly controlling for unobserved spatial effects by using the Spatial Lag Model (SAR) to get consistent estimates of urban tree cover. The SAR model solves the problem by establishing the spatial weight matrix to incorporate the neighboring house price. In the second stage, we then used market segmentation to identify the demand parameters by collecting data from five geographic markets, assuming that residents in each market share common preference structures. Consequently, the first-stage analysis provides evidence of positive effect of tree cover on home values, which is robust with different specifications. We further found that the estimated own price elasticity of demand for tree cover within each parcel was − 0.075, suggesting an inelastic demand curve.

102