Intervention (Hidden)
For all households, we will show them an informational video on Jakarta’s air pollution levels, health and productivity impacts of pollution, ways to protect yourself, and sources of air pollution in Jakarta. We assume that after viewing the video, all households have sufficient baseline generic knowledge about air pollution for the rest of the experiment. Additionally, we will tell all households about the average efficacy of the air purifiers we use in this experiment, and how opening windows or doors affect the efficacy; this information is meant to standardize beliefs about purifier quality across treatment and control.
Importantly, all the air purifiers (or fans) have built-in non-visible monitors, which allows us to measure indoor PM2.5} levels in the control group as well as the treatment group. We focus on PM2.5 as this is the most dangerous and pervasive pollutant in Jakarta and globally.
Our control group was designed so that the only difference between treatment and control is that the treatment group receives clean air. Based on our pilot work, we believe this sweeps out two confounders. First, one of our main outcomes is WTP for clean air at endline, which we proxy for by giving all households the chance to rent an air purifier for two months at the end of the experiment. If we had not given the control group a fan (that looks just like the purifier), we would have been comparing a treatment bundled with loss aversion to a control group with no loss aversion; the fan rental instead will neutralize any potential for loss aversion to bias the experience wedge measurement upward. Second, our pilot work revealed that households also enjoying the cooling effect of purifiers. Our fan rental will also neutralize this cooling effect confounder.
The main treatment, Comprehensive Experience, is designed to give people the holistic experience of clean indoor air, in order to estimate the experience wedge. Households this treatment group will receive both an air purifier (with filter) and an air quality monitor, to be located next to the air purifier. This bundled intervention is designed to improve indoor air quality and unambiguously show households their exposure levels. Note that because both groups will be shown a theoretical, generic information video about air quality, we are targeting the characteristics of clean indoor air that cannot be conveyed without the experience itself.
Our secondary treatment, the `"sensory treatment'', only includes an air purifier. Households will experience the physical benefits of clean air and accompanying health and productivity benefits, but without the visible measurements provided by the monitor. Within our conceptual framework, we assume that households who only receive the purifier will know that the purifier is providing cleaner air in relative terms, but do not know (without a monitor) (1) the level of air pollution that they are exposed to, and (2) cannot quantify how effective the purifier is. In this way, we are targeting the sensory wedge. We make this assumption based on our piloting work, but will also test the assumptions as a part of our final results.