Recency and projection biases in air quality valuation by Chinese residents

We combine survey responses to subjective well-being (SWB) questions with air pollution data to recover Chinese residents' valuation of air quality improvements. Motivated by theoretical models of 'projection bias' and 'recency bias', we posit that one's SWB (and valuation) is affected disproportionately by more recent experiences with air pollution, even though long-term air pollution is more detrimental to one's actual well-being. Towards this end, we find that valuation for a unit improvement in PM_2.5 is twice as large when air quality on the day of survey is used as the explanatory variable compared to air quality averaged over a year. Our findings have far-reaching research and policy implications as they call into question the appropriate temporal scale of air quality conditions when conducting valuation studies or policy evaluations. Furthermore, our results imply that policymakers could conceivably exploit this behavioral bias to introduce more stringent air quality management policies when air quality is extremely poor.