Construct Overlap Inflates Apparent Economic-Capacity Alignment: Evidence from a Hebei City Panel, 2014–2023

Index-based urban sustainability studies often ask whether urban resilience and economic development evolve harmoniously, but the answer is unreliable when the two constructs contain the same economic indicators. A targeted audit of recent Chinese coupling studies shows that economic variables are commonly embedded in urban-resilience systems without a reported construct-overlap weight check. Using Hebei, China, as a measurement-validity test rather than as another provincial economic-growth case, this study analyzes a balanced 2014–2023 panel of 11 prefecture-level cities. Duplicated economic concepts account for 44.4% of the baseline resilience-index entropy weight and 88.2% of the baseline economic-index weight, indicating substantial mechanical exposure in the resilience-development correlation (Pearson r = 0.953). We, therefore, construct a measured non-economic service-infrastructure-ecological resilience-capacity index and a separate economic-development index. The redesigned index is not a comprehensive urban-resilience measure because it excludes governance, social capital, emergency response, hazard exposure, social vulnerability, and institutional adaptation. Removing only the shared resilience-side components lowers the all-year correlation to 0.783; after full redesign, it is 0.562. In constant 2014 prices, economic development rises sharply from 2014 to 2023, whereas measured non-economic resilience capacity improves modestly. Robustness tests separate stable city findings from measurement-sensitive ones: Langfang and Shijiazhuang remain economy-ahead across all 13 selected specifications, while Cangzhou does so in 11 of 13, Tangshan is economy-ahead in 11 of 13 selected specifications but sensitive to the gas-supply proxy, and Qinhuangdao and Xingtai remain measured-capacity-ahead across all specifications. The measured non-economic resilience capacity shows a detectable global autocorrelation under queen contiguity, whereas a mismatch shows no statistically detectable global autocorrelation under the tested specifications. The primary contribution is methodological: a construct-overlap audit and directional mismatch diagnosis that identifies relative differences between economic development and measured public-service, infrastructure, utility, and ecological resilience capacity. These diagnostics do not establish a causal conversion failure.