社會住宅鄰避效應與迎毗效應—以柯文哲兩次台北市長選舉為例

Abstract

2018年台北市市長選舉，柯文哲以41.1%得票率取得連任。過往研究就「柯文哲現象」分析勝選原因有以下：網路媒體運用成功、親民形象、社群平台經營、年輕族群等。本文旨在以柯文哲任內興建五萬戶「只租不賣」社會住宅政策對於選舉結果影響。 本文使用集體資料分析社會住宅政策效果，依變項為柯文哲兩次台北市市長選舉得票率差。以台北市456個里做為觀察值，依循公共設施鄰避效應與迎毗效應理論，解釋變項為社會住宅政策，並控制經社背景變項、族群變項、政治版圖效果、平均房價、空間自相關等變項。 研究發現「社會住宅政策」與選舉結果有正向顯著相關；各里與社會住宅距離越近者，對於柯文哲選舉得票越有幫助，驗證社會住宅迎毗效應。同時，也得出投票行為具有空間自相關，相鄰近地區互相影響。透過地理加權迴歸分析，得出社會住宅政策正向效果在各里間並非普遍均質現象，在靠近陽明山地區的里、萬華區青年公園周遭里及北投區關渡里等地區最為顯著。交互作用分析得出社會住宅政策有助於柯文哲台北市市長選舉得票之結果，不受所在里平均房價高低調節而有所差異。

In the 2018 Taipei mayoral election, Ko Wen-je was re-elected with 41.1% of the vote. Prior studies on the so-called "Ko Wen-je phenomenon" have identified several factors contributing to his electoral success, including effective utilization of online media, a relatable and approachable public image, active engagement on social media platforms, and substantial support from younger voters. This study focus on the impact of Ko's policy—the construction of 50,000 units of “rental-only” social housing—on electoral outcomes. Using aggregate-level data analysis, this research models the effect of the social housing policy on the change in Ko’s vote share between the two Taipei mayoral elections as the dependent variable. The analysis is conducted with 456 villages in Taipei as observation points. Based on the theoretical frameworks of public facilities NIMBY (Not In My Backyard) and YIMBY (Yes In My Backyard) effects, the main explanatory variable is social housing policy. The model also includes controls for social backgrounds variables, ethnic variables, political territorial effect, average housing prices, and spatial autocorrelation. The results reveal a statistically significant and positive relationship between the social housing policy and electoral performance: neighborhoods in closer proximity to social housing sites were more likely to deliver higher vote shares for Ko Wen-je, supporting the presence of a YIMBY effect. Furthermore, the analysis confirms spatial autocorrelation in voting behavior, indicating that electoral outcomes in adjacent neighborhoods are interdependent. Geographically weighted regression (GWR) further demonstrates that the positive effect of the social housing policy is spatially heterogeneous across villages. The strongest effects are observed in areas near Yangmingshan, around Youth Park in Wanhua District, and in the Guandu area of Beitou District. Interaction analysis shows that the beneficial effect of social housing policy on Ko Wen-je’s mayoral election performance in Taipei City does not vary according to the average housing price in a given neighborhood. In other words, the policy's electoral impact is not moderated by local housing market conditions.