Conference Agenda

Leisure and recreation are an important and temperature-sensitive source of welfare, yet the welfare consequences of extreme temperatures across diverse recreational activities have rarely been quantified in a unified and comparable empirical framework. Using tens of millions of Weibo check-ins matched with Dianping point-of-interest (POI) information over 400 days in China during 2023–2024, we estimate the temperature response of visits and expressed sentiment across 21 leisure and recreation categories. We find that visits–temperature relationships fall into three dominant shapes: inverted-U, monotonic increasing, and flat. Patterns of offsetting changes in visits across categories, together with opposite-signed responses of sentiment and visits in some indoor activities, are consistent with cross-category substitution under extreme heat. We further identify two channels of adaptation: acclimatization, meaning reduced sensitivity to locally common extremes, and substitution across activity categories. To obtain a welfare measure, we combine category-specific average spending and demand elasticities to translate temperature-induced changes in visits into the price compensation required to maintain baseline visitation. Aggregating across activities, we estimate that if each city experiences one additional hour of extreme heat/cold on a given day, nationwide recreational welfare losses amount to approximately 182/714 million RMB. Under SSP5-8.5 projections, net effects become positive by the end of the century (about +58 RMB per person per year), primarily because the aggregated gains from fewer extreme-cold exposures outweigh added losses from increased extreme heat. Our study provides a scalable welfare accounting framework for the leisure and recreation domain that incorporates substitution and offers quantitative evidence to inform urban heat and cold adaptation planning.