Whither less is more? Understanding the contextual and configurational conditions of polycentricity to improve urban agglomeration efficiency

Polycentricity is at the core of the urban policy and planning debate on both city size and spatial configuration, but there is no conclusive evidence for how and when polycentricity can improve urban agglomeration efficiency. This study examines the impacts of polycentricity, as well as features of its spatial configurational, on urban agglomeration efficiency. An improved relative threshold method based on local isolines (IRT-LI) is developed to identify urban centers and sub-centers, and this indicator is regressed against agglomeration economic efficiency acquired through an input-output method—super-slack-based measure (Super-SBM). Taking 252 prefectural-level cities in China as examples, this study finds that polycentricity does not demonstrate any linear or U-Shape impact on the agglomeration economy efficiency of all cities, while the impacts become positive only when urban population and central city density surpass certain thresholds (population over 6 million and density over 6000 people/km2). Compactness in density and balance in the size of multiple centers are key spatial configurational features to improve economic efficiency, while the number of centers should only be increased for large size and high-density cities. Policymakers and planners are encouraged to limit polycentric planning only to megacities/mega-city-regions and combine compact, multi-functional, mixed land-use spatial configurations with polycentricity.