Climate-air-health impacts of traditional to online shopping transition in China

Abstract

The rapid shift from traditional “in-store” shopping to online “pure-player” retail has transformed the retail sector and consumer behavior, raising concerns about its climate, air pollution, and health impacts. Previous studies offer conflicting conclusions due to differences in the scope, context, and time frame, complicating comparisons between the two shopping modes. Critical determinants of impact vary substantially, including transportation methods (e.g., air and road freight), packaging-related lifecycle impacts, and complex consumer behaviors. The urgency for dynamic, comprehensive assessment is further amplified by the rapid evolution of retail supply chains, driven by technological advancements, efficiency gains, and shifting consumer dynamics. Without resolving these knowledge gaps, identifying critical supply chain hotspots and developing effective strategies for sustainable retail development remains challenging. This study quantifies the climate, air pollution, and health impacts of the transition from traditional to online shopping in China—the world’s largest e-commerce market. An emission inventory spanning 1990–2023 was developed for China’s retail sector, covering 16 merchandise categories across six supply chain phases: wholesaler transport, long-distance transport, last-mile delivery/shopping trips, device use/store operation, and storage. It includes both greenhouse gases (CO₂, CH4, N₂O) and air pollutants (primary PM2.5 and PM10, BC, OC, NOx, SO₂, NMVOCs, NH₃, and CO). Average emission intensities (EIs), measured as emissions per item purchased, and annual shopping-related emissions were calculated to compare the environmental impacts of online and traditional shopping. We adopted the EI to account for the primary differences in product features, including weights and basket sizes (the number of items per purchasing order) that drive emissions, particularly from long-distance transport, last-mile delivery/shopping trips, and packaging use, across 16 merchandise categories. This indicator has been used in previous research to precisely compare the emissions intensity of shopping methods. Air pollutant emissions were integrated with city-level Multi-Region Input-Output modeling, logistics data, and transport networks to generate high-resolution (0.1°× 0.1°) spatial emission distributions. These were coupled with the Community Multiscale Air Quality adjoint model and the Global Exposure Mortality Model to estimate ambient PM2.5-associated premature deaths associated with shopping. Economic costs associated with climate change impacts associated with greenhouse gases emissions and health impacts from ambient PM2.5 pollution were evaluated to provide a comprehensive impact assessment. Based on these findings, key supply chain hotspots were identified, and targeted mitigation strategies were proposed. CO₂ emission assessments show that approximately 13.4% of the total shopping-related emissions can be attributed to the shift from traditional to online shopping in China for 2023. The average EI in online shopping (0.344 kg item⁻¹) was four times higher than that in traditional shopping (0.068 kg item⁻¹) during the early 2000s. This level decreased by 34% by 2023 due to improved operational efficiencies, while that of traditional shopping tripled largely due to increased car usage, lead to a convergence in current EIs: 0.227 kg item⁻¹ for online versus 0.182 kg item⁻¹ for traditional shopping. Mitigation strategies targeting hotspots, including packaging use and air freight in online shopping, show potential of reducing shopping-related CO₂ emissions by up to 85% (or 162 million tonne year⁻¹) by 2060, underscoring the role of online shopping in advancing low-carbon retail development. Air pollutant assessments showed that online shopping transition reduces primary PM2.5 emissions by 25% (EIs: 0.017 g item⁻¹ vs. 0.057 g item⁻¹), primarily due to more efficient courier deliveries compared to individual store visits. However, emissions of secondary PM2.5 precursors (NOx, SO₂, and NMVOCs) increase substantially—by 23%, 270%, and 30%, respectively—driven by air freight use and overpackaging. Targeted strategies—including shifting from private vehicles to walking/cycling, replacing air freight with high-speed rail, and adopting reusable packaging—can mitigate these effects up to 80% (11 Gg year⁻¹), 59% (180 Gg year⁻¹), 97% (220 Gg year⁻¹), and 96% (430 Gg year⁻¹), respectively by 2060. Implementing such measures is essential for promoting sustainable retail practices amid e-commerce growth. Health assessments indicate that per-billion-item premature deaths associated with shopping were 36% higher in online shopping than in traditional shopping (34 vs. 25 deaths billion item⁻¹), primarily due to extensive packaging use and long-distance transport. Spatially, the transition to online shopping resulted in both increases and decreases in premature mortality (−26 to +69 deaths billion item⁻¹), due to disparities in packaging production, freight mode use, and last-mile delivery efficiency across regions. Mitigation strategies targeting key emission sources—including promoting active travel, adopting reusable packaging, and electrifying freight—collectively reduce total premature mortality of shopping by up to 80% (or 5,800 deaths avoided). Economic cost assessments show that per-item climate–air–health costs in online shopping were 46% higher than in traditional shopping (0.071 vs. 0.049 USD item⁻¹). This difference includes a 57% increase in climate change costs (0.026 vs. 0.016 USD item⁻¹) and a 40% increase in health loss costs (0.046 vs. 0.033 USD item⁻¹). These disparities are primarily driven by the greater use of packaging, air freight, and road freight in online shopping. Mitigation strategies—such as promoting active travel, adopting reusable packaging, and electrifying freight—have the potential to reduce these costs by up to 77% or 12 billion USD year⁻¹. In conclusion, this study demonstrates that online shopping generally results in greater climate, air pollution, and health impacts than traditional shopping, primarily due to overpackaging and increased air and road freight use. Despite this disparity, the EIs for CO₂ and air pollutants in online shopping have decreased during 2000–2023, compared to the increases in those for traditional shopping. Moreover, future predictions highlight that online shopping is expected to dominate in the retail sector (87% of total sales), underscoring the urgent need to mitigate the online shopping supply chain. Taken together, the observed reduction in EIs from 2000 to 2023 and the projected expansion of online shopping supports that online shopping has a larger mitigation potential for CO₂ and air pollutant emissions in the retail sector compared to traditional methods. For long term, shifts of online shopping operation strategies, including shifts from air to HSR freight and from disposable to reusable packaging, can substantially reduce CO₂ and air pollutant emissions by 2060 while in the short term, adopting active travel reduces those for traditional shopping. For both shopping methods, electrifying freight is a primary strategy for reducing health impacts of shopping. Future research should focus on more comprehensive and large-scale assessments that integrate temporal, spatial, and product-specific variations in climate and health impacts. Such analyses are essential for developing targeted mitigation measures and supporting the sustainable development of the retail sector amid the rapid expansion of e-commerce.