Operational and environmental disparities in parcel last-mile delivery across spatial contexts: Why an urban-centric perspective falls short

Rising parcel volumes increase operational and environmental pressures in last-mile delivery (LMD) systems. While these impacts are highly visible in dense urban areas, existing research and model-based analyses largely focus on urban contexts and rarely examine suburban and rural delivery structures in a systematic way. As a result, the operational and environmental relevance of non-urban deliveries for system-wide LMD performance remains insufficiently understood.

This study applied the Hanover Agent-based Geospatial Regional Delivery Simulation Model (HAGRID) to analyze parcel delivery performance across the full urban–suburban–rural gradient of the Hanover Region, Germany. HAGRID combined a spatially explicit parcel-demand model with a large-scale agent-based MATSim freight simulation, generating carrier-specific delivery patterns for major logistics service providers.

The results show that delivery density is a key determinant of LMD efficiency, costs, and emissions. Dense urban areas exhibit strong local traffic and emission impacts due to concentrated stopping activity. In contrast, suburban and rural deliveries are markedly less efficient, resulting in substantially higher delivery costs and CO₂ emissions per parcel. Although these impacts are spatially diffuse, they account for a dominant share of system-wide emissions due to long access distances and limited consolidation potential. Disaggregating urban, suburban, and rural areas into eight functional logistics contexts further reveals pronounced heterogeneity within these broad categories, particularly in rural areas, where clustered parcel shipments along major transport corridors can partially offset structural disadvantages.

Our findings demonstrate that the system-level operational and environmental performance of LMD cannot be adequately assessed from an urban-centric perspective alone. By explicitly addressing the under-representation of non-urban contexts in current LMD research, this study provides a scalable, data-driven framework for evaluating context-specific strategies to improve last-mile logistics across diverse spatial contexts.