AI for Social Good: An Investigation of the Causal Relationship Between Environmental Regulations and Their Effects on Air Pollution in London, UK

2026-06-13

AI Summary

This study addresses the challenge of estimating the causal effect of non-randomly implemented air pollution regulations on PM₂.₅ concentrations, which is confounded by meteorological variability, socioeconomic changes, and multiple concurrent interventions. To tackle this, the authors propose an uncertainty-aware Bayesian deep learning framework that uniquely integrates Bayesian long short-term memory (LSTM) networks, propensity score adjustment, and counterfactual reasoning, augmented with bootstrap resampling to quantify predictive uncertainty. Applied to the evaluation of 32 environmental regulations enacted in London between 2010 and 2020, the method estimates an average reduction in PM₂.₅ of 1.88 μg/m³ (a 12.35% relative decrease; 95% CI: 1.64–2.12), with the most pronounced effects observed during 2018–2019. This approach offers an interpretable and robust framework for causal inference in environmental policy assessment.

Abstract

Air pollution regulation is central to urban public health governance, but estimating its effects is difficult because policies are implemented non-randomly and pollution trajectories are shaped by meteorology, socioeconomic change, temporal trends, and overlapping interventions. This study develops an uncertainty-aware Bayesian deep learning framework to estimate the aggregate effect of air pollution regulations on PM$_{2.5}$ concentrations in London from 2010 to 2020. The framework integrates daily PM$_{2.5}$ observations from Inner London monitoring stations, meteorological covariates, annual socioeconomic indicators, month-of-year and day-of-week indicators, and daily regulation status data for 32 policy measures. A Bayesian LSTM captures temporal dependencies in environmental and socioeconomic covariates, Bayesian embedding layers represent temporal and regulation status inputs, and a regulation status prediction branch supports propensity score-based adjustment for non-random policy implementation. Regulatory effects are estimated by comparing observed PM$_{2.5}$ concentrations with counterfactual predictions under a hypothetical no-regulation scenario, with uncertainty summarized across repeated Bayesian training runs and bootstrap resampling. Results show that London's regulations were associated with an average PM$_{2.5}$ reduction of 1.88 $μ$g/m$^3$, a relative reduction of 12.35%, with a 95% confidence interval of 1.64-2.12 $μ$g/m$^3$. Estimated effects were limited before 2013, became clearer from 2013 to 2017, and were strongest in 2018 and 2019. The findings suggest that sustained and cumulative regulatory interventions contributed to measurable improvements in London's air quality. This study demonstrates how uncertainty-aware causal AI can support environmental accountability, public health protection, and evidence-based governance for environmental decision-making.

Problem

• air pollution regulation
• causal inference
• PM2.5
• environmental policy
• confounding factors

Innovation

• Bayesian deep learning
• causal inference
• air pollution regulation
• uncertainty quantification
• counterfactual prediction