Panel Dataset and Empirical Codes for Research on Digital Inclusive Finance and Rural New-Quality Productivity in China's Counties (2014-2021)

This dataset is the supporting research data and empirical analysis code for the paper *Digital Inclusive Finance and New-Quality Productivity Forces in Rural Counties: Causal Mechanisms and Empirical Evidence*. It focuses on the impact mechanism and effect test of digital inclusive finance (DIF) on the new-quality productivity (NPF) in rural counties of China, and provides standardized panel data and reproducible empirical analysis tools for research on high-quality development of rural county economy and rural revitalization.The sample of this dataset covers 242 county-level administrative regions in China (including rural counties, non-rural counties, county-level cities, autonomous banners, etc.), with a time span from 2014 to 2021. All data are from official, public and authoritative data sources, mainly including: *China County Statistical Yearbook*, *China Regional Economic Statistical Yearbook*, patent public data from the State Intellectual Property Office of China, *Peking University Digital Inclusive Finance Index* released by the Institute of Digital Finance of Peking University, Digital Rural Index jointly released by the Institute of New Rural Development of Peking University and Alibaba Research Institute, county-level nighttime light data from Harvard Dataverse, and PM2.5 emission data from the Atmospheric Composition Analysis Group.The dataset mainly includes three parts:1. Raw and preprocessed panel data: including full-sample datasets and sub-sample datasets (rural counties / non-rural counties / county-level cities / autonomous banners) in .dta format (directly readable by Stata) and .xlsx format, as well as datasets for core variable measurement. It covers full-dimensional indicators of explained variable (county-level new-quality productivity, NPF), core explanatory variable (digital inclusive finance, DIF), mediating variables and control variables. Standardized preprocessing such as linear interpolation for missing values, winsorization, and outlier elimination has been completed.2. Full-process empirical analysis codes: including executable .do code files in Stata format, covering the full empirical process of the paper, including data preprocessing, core indicator measurement via entropy weight-TOPSIS method, benchmark regression, threshold effect regression, mediating effect test, heterogeneity analysis, robustness test (Bootstrap sampling, system GMM dynamic panel regression). All codes can be run directly to reproduce all empirical results of the paper.3. Auxiliary analysis files: including auxiliary files such as descriptive statistics of core variables and output tables of regression results, which fully restore the data processing and analysis process of the paper.This dataset can be used for academic research and policy analysis in the fields of digital economy, rural finance, county economy, new-quality productivity, etc. It can support empirical research such as mechanism test, heterogeneity analysis, spatio-temporal evolution of digital finance and high-quality economic development at the county level. All data and codes have passed empirical tests with strong reproducibility.