「疫」卡「通」–新冠肺炎、天氣因子對捷運人流之影響

Abstract

本研究建立台北捷運人流的估計模型，著重於載客量最多的板南線、松山新店線和淡水信義線。被解釋變數是捷運站的進站人數，解釋變數則有氣溫、雨量、捷運站周遭人口數。對於這些變數，捷運站的真實數值未知，我們透過地理資訊系統協助推算每個捷運站的雨量、溫度以及周遭人口數。根據固定效果模型的估計結果，上述變數皆顯著影響捷運人流。 我們假定，新冠肺炎疫情則會經由「每日新增確診人數」與和「入境人數的變化」兩種方式影響捷運載客量。稱前者為心理效果，後者為政策效果。我們從基本模型逐一引入新變數，可以估計這兩種效果對捷運人流的影響。本研究顯示，這兩種效果造成2019年至2020年捷運人流降幅的65%。

This study establishes an estimation model of the Taipei MRT passenger volume, focusing on the Bannan Line, Songshan-Xindian Line, and Tamsui-Xinyi Line, which carry most passengers. The dependent variable is the number of people arriving at the MRT station, and the explanatory variables are temperature, rainfall, and the population living around the MRT station. For these variables, the true values of the MRT stations are unknown. We use the geographic information system to help estimate the rainfall, temperature, and surrounding population of each MRT station. According to the estimated results of the fixed-effects model, all of the aforementioned variables have a statistically significant impact on the MRT passenger volume. We assume that the COVID-19 pandemic will affect the passenger volume of the MRT through two channels: “daily new cases” and “the change in the number of arrivals.” The former is called the “psychological effect”, and the latter is the “policy effect”. Starting from the basic fixed-effect model and introducing new variables one by one, the impact of these two effects on the MRT volume can be estimated. This study shows that these two effects will jointly explain 65% of the decrease in MRT passenger volume from 2019 to 2020.