REPLICATION DATA FOR: A Scenario-Based Approach to Post-War Sectoral Recovery Prioritization in Ukraine

This dataset is the third part of a research trilogy on the structural and dynamic analysis of intersectoral linkages in the Ukrainian economy, following Part 1, Sectoral Multipliers and System Linkages in Ukrainian Industry Markets (10.5281/zenodo.19348878), and Part 2, Macrodynamic Interrelationships and Intersectoral Shock Transmission Effects in the Economy of Ukraine (10.5281/zenodo.19645520). It combines the network indicators from Parts 1–2 with external war-damage assessments — the Kyiv School of Economics (KSE) Institute and the World Bank's Fifth Rapid Damage and Needs Assessment (RDNA5) — to construct an objectively-weighted Recovery Priority Index (RPI) for 42 NACE Rev.2 sectors of the Ukrainian economy. The RPI integrates three dimensions for each sector: (1) its systemic role within the pre-war (2021) production network, measured by a composite Systemic Impact Index and PageRank centrality; (2) its shock vulnerability, revealed by the statistically significant structural break in the network confirmed in Part 2 (Chow test, F = 3.50, p = 0.016, year 2022), captured through the change in Rasmussen quadrant and the magnitude of change in the Systemic Impact Index; and (3) the scale of direct physical war damage it has sustained, allocated from aggregate KSE/World Bank categories to the 42-sector NACE classification in proportion to each subsection's 2021 value-added share, and rescaled to match the RDNA5 total. Component weights are derived objectively via the Shannon entropy method rather than assigned by a priori expert judgement, yielding w_S = 0.1505, w_V = 0.2276, and w_D = 0.6219. The robustness of the resulting sector ranking is verified through three independent procedures, all included in the accompanying R script: cross-scheme Spearman and Kendall rank correlation across four alternative weighting specifications, bootstrap stability analysis (1,000 iterations simulating measurement error), and one-at-a-time weight-sensitivity analysis (±20% perturbation). The index is then applied to simulate three investment-allocation scenarios — a proportional baseline, a structural-transformation scenario weighted by Rasmussen forward linkage (operationalizing the “Build Back Better” principle), and a network-resilience scenario weighted by systemic centrality — through a Leontief inverse matrix built from Ukraine's 2023 input-output table. The structural-transformation scenario yields the highest aggregate output multiplier (3.043) compared to the proportional baseline (2.914). Contents: consolidated input data (42 sectors × 20 variables combining 2021 network indicators, 2021→2022 shock deltas, and mapped war damages); full computed RPI results under four weighting schemes with corresponding ranks and three-scenario simulation outputs; the 42×42 Leontief inverse and technical-coefficients matrices built from the 2023 input-output table; a fully documented, reproducible R script covering entropy weighting, all robustness checks, and the scenario simulation; a variable-by-variable codebook; and a methodological note explaining the dataset's deliberate dual base-year design — 2021 for RPI components (the last undistorted pre-war year) and 2023 for the Leontief matrix (the most recent available technological structure, required for forward-looking scenario simulation). External data sources: Kyiv School of Economics Institute, war damage assessment as of November 2024; World Bank Group, Government of Ukraine, European Commission and United Nations, Fifth Rapid Damage and Needs Assessment (RDNA5), February 2026; State Statistics Service of Ukraine, Input-Output Table for 2023, at basic prices. License: Creative Commons Attribution 4.0 International (CC BY 4.0).