氣候變遷引發洪險增加對風險規避者之影響

When Households of Risk-Aversion Face Increased Flood Risks from Climate Change

The effects of climate change on future flood risks have been found in many studies to be a more intense hydrological cycle with increases in probability and magnitude of precipitation that increases peak runoff. Households living within the flood boundary may therefore face increased flood risks resulting from climate change. This study investigates household’s responses to climate change-induced changes in flood risks through residential location choices. A Random Utility Model (RUM) is used to incorporate the frequency and severity of flooding as housing characteristics. Choices of flooding variables are investigated and used to estimate household willingness-to-pay (WTP) to reduce these risks. For risk-averse households, the utility function is nonlinear in income and traditional RUM estimates of WTP through simple linear aggregation are biased. This study uses a simulation process to approximate mean WTP given three nonlinear utility functions. Estimation results suggest that for risk-averse households, economic costs of increases in flood risks measured by WTP will be affected by the degrees of risk-aversion from utility function used. Estimated WTPs for reducing flood frequency and severity by exponential and logarithmic utility functions are similar, but are much higher than estimated by the quadratic utility function. The amount of WTP for reducing frequency is much greater than that for reducing severity, suggesting households are more risk-averse toward experiencing flood events rather than suffering from flood damage. 許多研究顯示，氣候變遷對於未來洪水風險之影響在於增加降雨之機率及幅度，導致更密集的水文循環，進而增加洪峰發生之可能性；洪區住戶也因而必須面對氣候變遷下所增加之洪險。本研究藉由洪區住戶對於居住地點之選擇作為探索住戶對於氣候變遷所導致洪險變化之反應；本研究採用隨機效用模型（Random Utility Model），並將洪水之發生頻率及平均每次發生洪水之幅度納為模型中之住宅變數。研究洪區住戶對這些變數之選擇可用以估算住戶對於降低洪險之願付價格（Willingness-To-Pay）。對避風險之住戶而言，其效用曲線是收入的非線性方程式；傳統隨機效用模型以簡單線性加總求出之願付價格將導致偏誤。本研究採用三種非線性方程式，以模擬方式估算出願付價格之平均值。研究結果顯示，以願付價格估算出洪險增加之經濟成本會受採用之非線性方程式所導致避風險程度之影響；其中採用指數方程式和對數方程式對降低頻率及幅度之估計結果相近，但比採用二次方程式高出許多。而採用三種非線性方程式之估計結果都顯示降低頻率之願付價格遠高於降低頻率之願付價格，此結果顯示洪區住戶趨避洪險頻率之偏向大於趨避洪險災害之偏向。