Public Attention Elasticity to Air Pollution in Toronto

Air pollution is a major urban health risk, but public attention to it is uneven. This thesis asks whether active public attention in Toronto responds proportionally to changes in PM2.5, or whether it becomes much more responsive only when pollution is visible or publicly interpretable. Using monthly Toronto data from 2015 to 2024, the analysis combines PM2.5 concentrations with Google Trends search activity to estimate attention elasticity: the model-implied responsiveness of active search behaviour to changes in pollution. Public attention is treated as a theoretical construct, but measured pragmatically through a Public Attention Index built from multiple pollution-related search terms. Exploratory factor analysis shows meaningful common movement but mixed evidence for a settled latent trait, and therefore PCA is used to construct a transparent composite rather than a validated scale. The results point away from a stable proportional-attention process, as routine pollution shows weak and specification-sensitive associations with active attention, while AQHI7 months produce a much stronger pollution-attention relationship. The thesis therefore argues that air pollution is not only an exposure problem but also an attention problem, as routine pollution may remain publicly muted unless it becomes visible and socially interpreted as actionable risk.