Waste reduction and blue skies: assessing the air quality effects of the Zero-Waste City policy with double machine learning

Building zero-waste cities is a central policy effort aimed at cutting solid waste at its source and speeding up the green, low-carbon transition of cities, and it bears heavily on the improvement of urban ecological quality. Treating the Zero-Waste City policy (ZWCP) as a quasi-natural experiment, we exploit a 2010–2024 panel of 271 prefecture-level cities together with a double machine learning approach to gauge how the policy shapes urban air quality and through which channels it operates. Three results stand out. First, launching the ZWCP lowers a city's annual mean PM2.5 concentration by approximately 4.6%, thereby delivering a modest but statistically significant association with improved air quality, and this pattern persists across an extensive battery of robustness checks. Second, the channel analysis yields evidence consistent with two potential routes through which the policy may restrain urban air pollution: industrial structure upgrading and green technology innovation. Third, the estimated abatement appears larger in small and medium-sized cities and in cities outside the old industrial bases, although the absence of formal cross-group tests means that these patterns should be regarded as suggestive. Taken together, the paper broadens the strand of work that links environmental regulation to urban air quality and supplies both a theoretical rationale and preliminary empirical evidence for refining ZWCP design and for considering place-specific approaches to green transition.