Replication data and code for urban water-system resilience index of 286 Chinese cities, 2011-2023

Published: 23 July 2026| Version 1 | DOI: 10.17632/7ks5fztcf8.1
Contributor:
qiang xin

Description

Water system resilience index, spatial weight matrix, and replication code for 286 prefecture-level cities in China.

Files

Steps to reproduce

The dataset is a balanced panel of 286 Chinese prefecture-level cities over 2011-2023 (3,718 city-year observations). DATA SOURCES. Nine raw indicators of urban water-system resilience were collected from the China City Statistical Yearbook and the China Urban Construction Statistical Yearbook (2012-2024 editions). Per-capita GDP was deflated to constant 2011 prices using province-level GDP deflators. INDEX CONSTRUCTION (01_build_index.do). The nine indicators (x1-x9) were winsorised at the 1st and 99th percentiles, then min-max normalised (indicator x6, per-capita domestic water use, was reverse-coded as negatively oriented). Indicator weights were derived by the entropy method and combined in equal proportion with equal weights (0.5 x entropy + 0.5 x 1/9). The composite index (WRI) is the weighted sum of the normalised indicators. Three dimension sub-indices - Resistance (x1-x4), Adaptability (x5-x7) and Recovery (x8-x9) - were computed using weights renormalised within each dimension. SPATIAL WEIGHT MATRICES (02_spatial_weights_moran.do). Great-circle distances between all city pairs were computed from longitude and latitude using the Haversine formula. Three row-standardised weight matrices were constructed: inverse-distance (Wd), k-nearest-neighbour with k=10 (Wk), and inverse-distance scaled by relative economic mass (Wec). ANALYSIS. Global Moran's I was computed by year under both Wd and Wk with analytical variance. Convergence was estimated by spatial Durbin models (static and bias-corrected dynamic), and local clustering by permutation-based local Moran's I. SOFTWARE. Stata 17 or later; the winsor2 package is required. Set the working directory to the folder containing the data files before running each .do file. Run 01_build_index.do first, then 02_spatial_weights_moran.do.

Institutions

Categories

Regional Economics, Spatial Statistics, Urban Sustainability, Water Resource Management

Funders

  • Shaanxi Provincial Philosophy and Social Sciences Foundation, Youth and Doctoral Program
    Grant ID: 2025QBR009
  • Key Laboratory Project of the Department of Education of Shaanxi Province
    Grant ID: 25JS046

Licence