天氣對遊客量的衝擊：以臺灣作實證

旅遊業被歸類於天氣依賴型的產業，受天氣變化的衝擊不可小覷，英國的研究中提到特別是國內的旅客在資訊的蒐集上比國外旅客更加方便，進行旅遊決策時更容易受到天氣的影響，本研究的目的在於量化天氣因子所造成的遊客量變化趨勢，研究結果能提供旅遊業者具有價值的參考，藉由預期天氣來減少他們的損失，或是宣傳當地宜人氣候來達到收益增加。樣本蒐集了中央氣象局的天氣因子數據以及觀光局的景點遊客量數據，總共25856筆資料，根據Taylor (2009)調查英國國內旅遊業的研究為主軸，採用隨機效果模型估計天氣因子的影響，另外考慮到天氣訊息對旅遊行為產生的滯後效果，採用動態模型中的Arellano–Bond dynamic panel data estimator重新調查天氣的效果。　　實證的結果顯示溫度、降雨、風速的上升會造成遊客量的下降，而日照時數的增加則可以促使大眾到戶外旅遊，然而天氣的影響在連續假期時並不明顯，春節假期的模型中天氣變數都不顯著，除了暑假的模型中降雨量有顯著負向的效果，因此旅遊的相關產業應該在一般時期注意天氣的變化，在春節假期及暑假期間則聚焦於降雨的機率上。

關鍵字

遊客量；天氣因子；隨機效果模型；動態模型；連續假期；季節性

並列摘要

Tourism is classified as a weather-dependent industry , so the impact of weather changing should not be ignored. The research in UK mentions that especially the domestic visitors in collecting weather information are more convenient than the foreign visitors, while making their travel decisions can be easily affected by the weather. This study aims to quantify the trend of visitor`s change caused by weather factors. Our findings can provide a valuable reference for the tourism industries to reduce their loss by anticipating weather and increase their revenue by promoting the pleasant climate. Samples collected weather-factor data from Central-Weather-Bureau , visitor data of feature spots from Tourism-Bureau (a total of 25856 observations). Following the research of Taylor (2009) with the UK domestic tourism survey, I adopted random effects panel data regression model to estimate the impact of weather factors. Furthermore, accounting for the lagged influence of weather information on travel behaviors, I employ Arellano–Bond dynamic panel data estimator in dynamic model to re-investigate the weather effects. 　　Empirical result indicates that the temperature, rainfall, and wind speed increase will lead to a decline in visitors, while the hour of sunshine increases outdoor traveling. However, the weather influence is not obvious in the continuous holidays. The weather variables in the model of Spring-break are all not significant except for the significantly negative rainfall effects in Summer-vacation. Therefore tourism-related industries should pay more attentions to weather changes in regular days, while focus on probability of rainfall during Spring-break and Summer-vacation.