연구제목: Effects of particulate matter (PM10) on tourism sales revenue: A generalized additive modeling approach

일정: 2018년 12월 13일

장소: Washington D.C., U.S.A.

연구배경

The impact of PM on other economic sectors, especially those involving human behaviors and activities, are less well studied than its human health effects. Tourism is one sector in which such research has been more frequently conducted, and it has become evident that tourists perceive PM as posing potential health risk or reduced utility and alter their travel plans. From the relevant literature, we can easily deduce that tourists and residents might alter their detailed plans and schedules and refrain from going out on a day of high PM concentration. This behavioral pattern would affect their consumption activities and perhaps lead to a significant economic disadvantage for retailers. While health problems occur occasionally and for a specific portion of the population, behavioral changes are more frequent and more sensitive reactions to PM among the general population. In spite of this magnitude, however, questions remain about how PM levels influence consumption patterns in the tourism and recreational sectors in an urban context.

연구목적 

We investigate the relationship between retail revenues and the level of PM10, in the context of Seoul, South Korea for about two years from April 2015 to February 2017.

연구방법

We adopted a twofold GAM strategy; in the first exploratory modeling, we specified the PM-related variables as smooth functions and derived inflection points of the trajectories. In the second model, we parametrized the line segments with interaction terms to estimate the effects' size and direction.

연구결과

The study confirms that the level of PM10 and sales revenues in the tourism industry exhibit a statistically significant relationship in a non-linear fashion. People begin to alter their consumption related to dining and recreation activities only after the level becomes worse than the “Bad” level. When the PM10 level is “Good” or “Normal” up to some low value of the “Bad” category, its concentration has a positive relationship with sales revenue. When the level exceeds that point, the relationship becomes negative. The turning points are at 109 μg/m³ and 121 μg/m³ for the PM10 level of today and the day before yesterday, respectively. In the case of today's PM10 level, every 10 μg/m³ increase is associated with an incremental increase of sales revenue by 3.8% up to 41 μg/m³ and by 0.3% up to 109 μg/m³. Thereafter, the same increase is associated with an incremental decrease of sales revenue by −1.2%. In the case of a two-day lag in the PM10 level, every 10 μg/m³ increase is associated with an incremental increase of sales revenue by 0.5% up to 121 μg/m³. Thereafter, the same increase is associated with an incremental decrease of sales revenue by −1.2%.