應用空間迴歸分析探討洪汎風險對都市不動產價格之影響–以台北市為例

Applying Spatial Regression Analysis to Explore the Effect of Flood Potential on Urban Real Estate Prices: The Case of Taipei City

徐敏舜(Min-Shun Hsu)

指導教授 ： 江謝令涵

中原大學/工學院/環境工程學系/碩士(2019年)

https://doi.org/10.6840/cycu201900639

摘要

全球天然災害中，洪災的發生頻率、影響人數以及造成的經濟損失為所有天然災害中最高的，所以洪汎風險為全球必須面對的重要議題。面對洪汎風險，很難準確的預測洪災於何時、何地發生，而科學化所模擬出的淹水潛勢圖時常被各國當作預測洪災發生之工具，如美國和日本多使用100-年洪汎區(100-year floodplains)，相當於洪水發生機率為1%的洪水災害地圖，而我國水利署也在民國106年完成淹水潛勢圖更新，製作出「第三代淹水潛勢圖」。以環境經濟的角度，我們將淹水潛勢圖作為風險特徵的來源，以特徵價格理論為基礎量化洪汎風險於住宅交易價格內的隱含價值，並且加入固定效果模型(Fixed Effect Model)以及地理加權迴歸(GWR)，探討洪汎風險於空間中的異質性。
根據固定效果模型的估計結果顯示，淹水潛勢會對住宅交易價格產生負面效應。同時，本研究結果顯示洪汎風險存在空間異質性，除了固定效果估計的結果揭示洪汎風險對住宅價格有負面之影響僅存在於新城區(松山、大安、信義以及內湖)，地理加權迴歸分析亦顯示，各地區存在不同的風險影響程度，舊城區(大同、萬華、中山以及中正)普遍的影響程度低於新城區。除了空間上的差異性，洪汎風險也存在時間的差異性，103年至106年，每一年有不同的交易行情，而交易行情間接的改變洪汎風險對住宅交易價格的影響。
不論是固定效果模型或是地理加權迴歸，皆顯示洪汎風險確實會對住宅價格產生負面效應，然而此現象也代表了民眾對洪汎風險的認知，當結果表示洪汎風險會對住宅價格造成負面效應時，同時也表示風險充分的被居民所認知，為後續的防洪政策提供相關輔助，哪些地區發生洪水災害時是被民眾所關注的，則可針對高影響地區最先進行防洪政策的施行。

關鍵字

住宅價格 ； 特徵價格理論 ； 固定效果模型 ； 地理加權迴歸 ； 洪汎風險 ； 淹水潛勢

並列摘要

Among global natural disasters, flood ranks top in the frequency, the number of people affected, and the economic losses. Therefore, flood risk is an important issue that must be faced globally. It is difficult to accurately predict when and where flood occurs. The flooding potential maps are often used as tools to predict flood occurrences. United States and Japan use 100-year flood plains, which is equivalent to a frequency of 1% flood probability. In Taiwan, Water Resources Agency has produced flood maps under different rainfall conditions since 1999. From the perspective of environmental economics, flood potential maps can be used as a representitive of flood risk. Based on hedonic price theory, the implicit value of residential transaction price of flood risk is quantified. Fixed effects models and geographically weighted regression (GWR) are used to mitigate the problem of endogeneity and heterogeneity in space.
The results of the fixed effect model shows that the flood potential has a negative effect on the price of residential property transactions. The negative impact on the residential price only exists for new urban area (Songshan, Daan, Xinyi, and Neihu) implies the occurance of spatial heterogeneity. The fixed effect model combined with geographically weighted regression analysis further confirms the concern. There are different levels of risk impact in the region, and the general impact of the old urban areas (Datong, Wanhua, Zhongshan, and Zhongzheng) is lower than that of the new urban area. In addition to spatial differences, flood risks also have time differences. From 2014 to 2017, the impact of flood risk on residential transaction prices fluctuates over time.
Flooding risks in fixed-effect models and geographically weighted regressions do have a negative effect on housing prices. However, this phenomenon also represents the public's perception of flooding risks. When the results indicate that flooding risks have a negative effect on housing prices, it also indicates that the risk is fully recognized by residents.It provides relevant assistance for the subsequent flood control policies. In those areas where flood disasters are concerned by the public, the flood control policies can be implemented first in high-impact areas.

並列關鍵字

Residential price ； Hedonic price theory ； Fixed effect model ； Geographically weighted regression ； Flood risk ； Flood potential

參考文獻

Bin, Okmyung, Jamie Brown Kruse, and Craig E Landry. 2008. 'Flood hazards, insurance rates, and amenities: Evidence from the coastal housing market', Journal of Risk and Insurance, 75: 63-82.

Bin, Okmyung, and Craig E Landry. 2013. 'Changes in implicit flood risk premiums: Empirical evidence from the housing market', Journal of Environmental Economics and management, 65: 361-76.

Bin, Okmyung, and Stephen Polasky. 2004. 'Effects of flood hazards on property values: evidence before and after Hurricane Floyd', Land Economics, 80: 490-500.

Brookshire, David S, Mark A Thayer, John Tschirhart, and William D Schulze. 1985. 'A test of the expected utility model: Evidence from earthquake risks', journal of Political Economy, 93: 369-89.

Daniel, Vanessa E, Raymond JGM Florax, and Piet Rietveld. 2009. 'Flooding risk and housing values: An economic assessment of environmental hazard', Ecological Economics, 69: 355-65.