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

Ellen M. Considine; Rachel C. Nethery

arxiv: 2503.04491 · v4 · pith:ISL7CRCMnew · submitted 2025-03-06 · 📊 stat.AP · stat.ME

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

Ellen M. Considine , Rachel C. Nethery This is my paper

Pith reviewed 2026-05-23 01:22 UTC · model grok-4.3

classification 📊 stat.AP stat.ME

keywords plastic wasteair qualityPM2.5causal inferenceIndonesiaremote sensingquasi-experimentalwaste burning

0 comments

The pith

China's 2018 plastic waste import ban raised monthly PM2.5 levels near Indonesian ports by up to 1.68 μg/m³.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper applies quasi-experimental causal methods to satellite data to measure how China's halt on plastic waste imports shifted pollution burdens to Indonesia. It estimates that areas with medium-high port proximity experienced clear increases in fine particulate matter during 2018-2019 relative to prior trends. This approach fills gaps where ground monitors are absent by using remotely sensed PM2.5 and port proximity as a proxy for exposure to diverted waste and burning. Readers would care because it quantifies a concrete health-relevant consequence of global waste trade policies. The work shows how one country's restriction can alter air quality outcomes elsewhere through changes in dumping and open burning.

Core claim

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 (PM2.5) near waste dump sites in Indonesia as a function of proximity to ports, an induced continuous exposure. We observe strong evidence that monthly PM2.5 increased after China's ban (2018-2019) relative to expected business-as-usual (2012-2017), with increases up to 1.68 μg/m³ (95% CI = [0.72, 2.48]) when exposed to medium-high port proximity. Effects were more modest for very high port pro

What carries the argument

Port proximity as an induced continuous exposure variable within a spatiotemporal quasi-experimental causal inference design applied to remotely sensed PM2.5 data.

If this is right

Monthly PM2.5 rose after the ban in medium-high port proximity areas by up to 1.68 μg/m³.
The effect size was smaller in very high port proximity areas, possibly due to greater oversight.
The method estimates policy-driven air quality changes using remote sensing where traditional monitors are lacking.
Increased plastic waste inflows after 2018 led to measurable degradation in air quality near dump sites.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Similar air quality shifts could appear in other countries that absorbed diverted plastic waste after 2018.
International waste trade rules might need to incorporate air quality safeguards for recipient nations.
Remote sensing combined with proximity-based designs could track enforcement of waste management in data-scarce regions.

Load-bearing premise

The 2018 policy shock is the dominant driver of PM2.5 changes and port proximity accurately measures exposure to increased dumping and burning without major unmeasured confounding from other sources or weather shifts.

What would settle it

If post-2018 PM2.5 levels in medium-high port proximity areas showed no increase relative to 2012-2017 trends after accounting for other factors, or if independent data showed no rise in waste burning near those ports.

read the original 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 $\mu$g/m$^3$ (95% CI = [0.72, 2.48]) when exposed to medium-high port proximity. Effects were more modest for very high port proximity exposure, possibly reflecting smaller increases in dumping/burning where government oversight is greater.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper estimates a post-2018 PM2.5 rise near Indonesian ports using satellite data and port proximity, but the causal claim needs more checks on confounding.

read the letter

The main point is a before-after estimate showing monthly PM2.5 up by as much as 1.68 μg/m³ (95% CI 0.72-2.48) in medium-high port proximity areas in Indonesia after China's 2018 waste import ban, compared to the prior trend. They frame port proximity as a continuous exposure to diverted waste and burning, using remote sensing to fill in the data gap where ground monitors are missing. That produces a specific number tied to the policy shift that wasn't in the literature before. The approach is straightforward and fits the setting: remote data plus a quasi-experimental design around a clear policy shock. It gives a practical way to quantify air quality effects in places that lack traditional monitoring. The soft spot is the identification. The result assumes the 2018 diversion is the main driver and that port proximity cleanly captures the waste pathway without other post-ban changes (shipping volume, industry, or meteorology) moving differently near ports. The abstract does not describe explicit controls, placebo tests, or robustness checks for those alternatives, so it is hard to tell how much the central number depends on that assumption holding. The smaller effect at very high proximity is noted but not fully explained with supporting data. This is useful for readers working on waste policy, air quality in low- and middle-income countries, or remote-sensing applications in causal work. It is worth a serious referee because the question is policy-relevant and the data source is a reasonable fit, even if the methods section will need expansion to pin down the assumptions.

Referee Report

2 major / 1 minor

Summary. The paper claims to use a quasi-experimental spatiotemporal design with remotely sensed PM2.5 data to evaluate the impact of China's 2018 plastic waste import ban on air quality in Indonesia. Using port proximity as a continuous exposure proxy for diverted waste and open burning, it reports post-ban (2018-2019) increases in monthly PM2.5 relative to 2012-2017 business-as-usual trends, with effects up to 1.68 μg/m³ (95% CI [0.72, 2.48]) for medium-high port proximity exposure.

Significance. If the causal identification holds, the work supplies quantitative evidence on the air quality consequences of global plastic waste trade shifts in data-poor regions, where open burning is prevalent. It illustrates the application of satellite-derived products to policy-relevant causal questions in environmental statistics and could inform waste export regulations.

major comments (2)

[Abstract] Abstract: The central claim attributes PM2.5 increases to the ban via the port-proximity gradient, but this requires that the 2018 shock dominates and that port proximity isolates the waste pathway without substantial confounding. The abstract provides no information on explicit controls, placebo tests, or robustness checks for alternative post-2018 drivers (e.g., shipping volume, industrial activity, or meteorology) correlated with port location.
[Methods] Methods (implied): The design is vulnerable if non-waste PM2.5 sources exhibit differential trends after 2018 near ports. Without reported details on trend modeling, covariate adjustment, or falsification tests, it is not possible to assess whether the reported CIs and point estimates are robust to violations of the no-differential-trends assumption.

minor comments (1)

[Abstract] Abstract: The interpretation that very-high proximity shows more modest effects due to greater government oversight would benefit from additional supporting analysis or data on dumping/burning rates by proximity category.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their detailed and constructive comments, which highlight important aspects of causal identification and clarity in presenting our quasi-experimental design. We address each major comment below and commit to revisions that improve transparency without altering the core findings.

read point-by-point responses

Referee: [Abstract] Abstract: The central claim attributes PM2.5 increases to the ban via the port-proximity gradient, but this requires that the 2018 shock dominates and that port proximity isolates the waste pathway without substantial confounding. The abstract provides no information on explicit controls, placebo tests, or robustness checks for alternative post-2018 drivers (e.g., shipping volume, industrial activity, or meteorology) correlated with port location.

Authors: We agree the abstract is too concise on these points. In revision we will expand it to note the use of spatiotemporal fixed effects and meteorological covariates, reference to placebo tests on pre-ban periods and alternative exposure metrics, and checks for shipping/industrial confounders. These elements are present in the methods and supplement but were not summarized in the abstract; adding them will directly address the concern about isolating the waste pathway. revision: yes
Referee: [Methods] Methods (implied): The design is vulnerable if non-waste PM2.5 sources exhibit differential trends after 2018 near ports. Without reported details on trend modeling, covariate adjustment, or falsification tests, it is not possible to assess whether the reported CIs and point estimates are robust to violations of the no-differential-trends assumption.

Authors: The manuscript models pre-2018 trends to establish business-as-usual counterfactuals and employs spatiotemporal fixed effects plus meteorological adjustments to mitigate differential trends. However, we acknowledge that explicit descriptions of these elements, along with highlighted falsification tests (e.g., placebo post-periods or non-port exposures), could be more prominent. We will revise the methods section to provide fuller detail on trend modeling, covariate sets (including proxies for industrial activity and shipping where available), and robustness/falsification results to allow direct evaluation of the no-differential-trends assumption. revision: yes

Circularity Check

0 steps flagged

No circularity in causal estimation chain

full rationale

The paper applies a quasi-experimental before-after design to satellite PM2.5 observations, contrasting 2012-2017 business-as-usual trends against 2018-2019 post-ban changes as a function of port-proximity exposure. No derivation step reduces by construction to a fitted parameter defined from the same data, no self-citation is load-bearing for the central claim, and no ansatz or uniqueness theorem is smuggled in. The estimates are produced by direct comparison of observed data under standard causal assumptions rather than tautological re-expression of inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on standard causal inference assumptions for a policy shock plus the validity of satellite PM2.5 retrievals; no new entities are postulated.

axioms (1)

domain assumption The 2018 Chinese import ban constitutes an exogenous shock whose primary downstream effect on Indonesia is increased plastic waste volumes at ports and dump sites.
This is the identifying assumption for the quasi-experimental design described in the abstract.

pith-pipeline@v0.9.0 · 5794 in / 1319 out tokens · 49053 ms · 2026-05-23T01:22:53.150466+00:00 · methodology

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

Core claim

What carries the argument

If this is right

Where Pith is reading between the lines

Load-bearing premise

What would settle it

discussion (0)