Producing Consistent Sectoral Forecasts: A Dynamic Input–Output and Macroeconometric Framework with Application ​to Spain

This paper presents a coherent and operational framework for producing medium‑ and long‑term sectoral projections fully consistent with a macroeconomic forecasting scenario. The methodology integrates econometric techniques, dynamic factor models, and classical input–output (IO) analysis to generate internally consistent projections for 64 productive branches in terms of gross value added, employment, wages, output prices, and production volumes. A dynamic non‑survey procedure is developed to reconstruct complete symmetric IO tables for both historical and forecast periods, ensuring compatibility between sectoral structures, aggregate demand, and price dynamics. The framework requires only statistical information commonly available for most economies, enabling its application beyond data‑rich contexts. An empirical illustration for Spain up to 2050 demonstrates the model’s predictive accuracy, even during periods of exceptional volatility such as the COVID‑19 crisis. Validation exercises using forecasting errors, multiplier analysis, and comparisons with OECD STAN data confirm the robustness of the approach. The resulting sectoral projections offer a valuable basis for further applications, including regional modeling, energy demand forecasting, emissions analysis, and labor market projections by qualification. JEL codes: C67, Input–Output Models, C53 – Forecasting and Prediction Methods; Simulation Methods, E27 – Forecasting and Simulation: Models and Applications