Spatial locking of rural incomes: terrain gradients and spillovers across 150 counties in Sichuan, China
Description
This deposit contains the two supplementary-data workbooks referenced in the Data Availability Statement of the manuscript submitted to Spatial Economic Analysis (RSEA-2026-0452), supporting a spatial analysis of farmers' income across a terrain gradient in Sichuan, China. File 1 — farmers_income_panel_Sichuan.xlsx (Supplementary Data S1) County-level balanced panel: 150 counties × three benchmark years (2000, 2010, 2020) = 450 county-year observations, no missing values. Cross-section covers 18 prefecture-level cities and autonomous prefectures of Sichuan. 43 variables including the dependent variable RuralPerCapitaDisposableIncome (current yuan), six mutually exclusive geomorphic-gradient dummies (Non-mountainous omitted), and a control set. Sheets: README, Variable_dictionary, Panel_data, Descriptive_statistics (live formulas), Data_quality_flags, Known_issues_corrections. Supports the Spatial Durbin model and Moran's I analysis (manuscript Table 6), reproducible with row-standardised k-nearest-neighbour weights (k = 5), inverse-distance and contiguity alternatives, and LeSage–Pace direct/indirect/total effects. Known issues are documented transparently: a corrected county-centroid duplication (F1), a spreadsheet-formula error in the tertiary-industry and GDP-per-capita controls (both published and corrected columns retained), electricity-consumption magnitude anomalies (F2), and rural-population zeros (F3); the geomorphic-gradient coefficients are unaffected. File 2 — policy_effect_DiD_rural_revitalization.xlsx (Supplementary Data S2) Difference-in-differences panel derived from S1 (same 150 counties, 450 records; county_id join key), 12 columns. Supports Section 5.5.1, Table 7 and Figure 6(a): the effect of China's targeted poverty-alleviation campaign (2013–2020) on rural income, interacted with deep-mountain (prominent-/entirely-mountainous) priority zones — 88 treated vs 62 control, 2010 pre-policy base, 2000 interaction as a pre-trend placebo. Sheets: README, Variable_dictionary, DiD_panel, Model_specification, Estimation_results, Known_issues. Reproduces Table 7 end to end; known-issue notes cover treated-group definition sensitivity and cumulative-effect timing. Both files ship Chinese administrative names alongside romanised forms. Data are released for verification and replication; underlying raw yearbook volumes are not redistributed where publishers restrict it.