| Session | ||
Thematic Session 1: Air pollution: Information, Interventions, and Behavior (HYBRID)
| ||
| Session Abstract | ||
|
Air pollution remains a persistent problem in low-and-middle income countries (LMICs). The incidence of damages is especially high amongst underprivileged communities due to lower pre-existing inputs into health and the rise of costly defensive investments (e.g. air purifiers) that are only accessible to the rich. Given regulatory failures in improving ambient air quality (Greenstone & Hanna, 2014) there is a need to explore the effectiveness of other defensive interventions in reducing the damages of exposure to high levels of air pollution. The reality of reducing air pollution in LMICs adds urgency to the investigation of the effectiveness of defensive action: even with aggressive regulatory approaches it will be many years before ambient levels are reduced to safe levels in much of the world. This session is focused on examining the effectiveness of providing information to individual households in promoting the adoption of defensive action to reduce exposure to pollution. Each paper in the session examines an intervention that provides information to households about either (a) air quality or (b) ways to take defensive action (or both). They then evaluate the effectiveness of these actions with respect to whether households take action, whether that action has health consequences, and whether the information provided changes their Willingness To Pay for additional reductions in exposure to pollution. All four papers are unified in their examination of these questions in LMICs. | ||
| Presentations | ||
Know thy foe: Information provision and air pollution in Tbilisi 1The World Bank, United States of America; 2National University of Singapore Middle-income countries host the majority of the world’s population exposed to unhealthy levels of air pollution, and the majority of this population lives in urban environments. We investigate the impact of information provision on household behavior in connection with indoor and outdoor air pollution in a middle-income country’s urban center—Tbilisi, Georgia. We implement an RCT to assess whether providing households with different levels of pollution information changes their knowledge of air pollution and avoidance behavior and improves their health outcomes. We find daily text messages significantly enhanced knowledge about pollution, led to increased avoidance behaviors, and improved health outcomes. The Demand for Clean Air: Experimental Evidence from Delhi 1University of British Columbia; 2University of Chicago; 3Pharo Foundation; 4University of Chicago Do hazardous levels of air pollution in developing countries reflect low public demand for air quality or incomplete information about its benefits? This paper reports on a experiment that 1) elicits the demand for clean air by combining pollution mask offers with ambient variation in air quality and 2) tests for demand-limiting market failures in information. We find low-income residents of Delhi, India would pay $1.16 for a 10-unit reduction in PM2.5, but that value increases to $6.33 when health impacts are described beforehand. Notably, the information effects are largest for respondents with less income and education. The Power of Perception: Limitations of Information in Reducing Air Pollution Exposure 1Kennedy School of Government, Harvard University; 2Inter-American Development Bank; 3University of Southern California; 4Amazon We conduct a randomized controlled trial in Mexico City. Households that receive SMS alerts are more likely to report receiving air pollution information via SMS, a high pollution day in the past week, and staying indoors on the most recent perceived high pollution day, but are not more likely to correctly identify days with high pollution. Similarly, households that received an N95 mask are more likely to report utilizing a mask in the past two weeks, but not on high particulate matter days. A theoretical model shows that this seemingly puzzling behavior could be consistent with learning from the alerts. | ||