A Spatiotemporal, Quasi-experimental Causal Inference Approach to Characterize the Effects of Global Plastic Waste Export and Burning on Air Quality Using Remotely Sensed Data

TL;DR

This study tackles the challenge of evaluating air quality effects from open burning of plastic waste in settings with limited ground monitoring by leveraging remotely sensed $PM_{2.5}$ data and a spatiotemporal, quasi-experimental framework. It introduces a multiply robust efficient influence function–based estimator to construct a continuous exposure–response curve for a universal intervention, using port proximity as the dose and pre-2018 years as controls. Applied to Indonesia after China’s 2018 plastic waste ban, the method finds a post-ban PM$_{2.5}$ increase up to $1.68~\mu g/m^{3}$ near dump sites with medium-high port proximity, with a lower or negative effect near high-proximity sites due to governance and oversight. The analysis demonstrates the feasibility of policy evaluation in data-scarce LMICs with remote sensing, and provides a generalizable framework for assessing distributed, universal interventions and their spatially varying health effects.

Abstract

Open burning of plastic waste may pose a significant threat to global health by degrading air quality, but quantitative research on this problem -- crucial for policy making -- has been stunted by lack of data. Many low- and middle-income countries, where open burning is most concerning, have little to no air quality monitoring. Here, we leverage remotely sensed data products combined with spatiotemporal causal analytic techniques to evaluate the impact of large-scale plastic waste policies on air quality. Throughout, we study Indonesia before and after 2018, when China halted its import of plastic waste, resulting in diversion of this massive waste stream to other countries. We tailor cutting-edge statistical methods to this setting, estimating effects of increased plastic waste imports on fine particulate matter (PM$_{2.5}$) near waste dump sites in Indonesia as a function of proximity to ports, an induced continuous exposure. We observe strong evidence that monthly PM$_{2.5}$increased after China's ban (2018-2019) relative to expected business-as-usual (2012-2017), with increases up to 1.68 $μ$g/m$^3$ (95\% CI = [0.72, 2.48]) at dump sites with medium-high port proximity. Effects were more modest at sites with very high port proximity, possibly reflecting smaller increases in dumping/burning where government oversight is greater.

Figures (24)

• Figure 1: Median, mean, and 75th percentile of PM$_{2.5}$ across GPW [dump] sites in Indonesia, 2012-2019.
• Figure 2: Quantile of the port proximity index for each GPW site in Indonesia.
• Figure 3: The spatial correlation function (Matérn) fit to the residuals from the LLKR stage.
• Figure 4: Average dose-effect on the treated (ADT) estimates and bootstrap 95% confidence intervals for our case study, estimated using both our proposed multiply robust method and a conventional outcome regression. ADT estimates (y-axis) quantify the average change in PM$_{2.5}$ ($\mu g/m^3$) post-China ban, compared to concentrations expected under business-as-usual, at open dump sites in Indonesia for a given port proximity (x-axis). Note that our port proximity metric is in quantile form, so the observed support of the exposure along the x-axis of these plots is constant. The pointwise confidence intervals are generated using our Spatial Weighted Bootstrap.
• Figure 5: Pre-trends test using 2017 as the "treated" period; using Spatial Weighted Bootstrap with the same Matérn parameters as our main analysis.