Dataset for Forecasting U.S. Manufacturing Capacity Utilization Using Operational Variables and External Risk Indicators (2001–2024)

This dataset accompanies the manuscript "Forecasting Manufacturing Capacity Utilization: The Relative Value of Operational Variables and External Risk Indicators." The dataset contains monthly observations of U.S. manufacturing capacity utilization and related operational and macroeconomic variables covering the period from January 2001 through December 2024. It was compiled by integrating multiple publicly available data sources into a single analysis-ready dataset for forecasting applications. The dataset contains the following variables: Observation_Date (YYYY-MM-DD), Capacity_Utilization (percent), Inventory (million U.S. dollars), Production_Hours (average weekly hours), Employment (thousands of employees), Labor_Productivity (output per hour), Economic_Policy_Uncertainty_Index, Global_Supply_Chain_Pressure_Index, Work_Stoppages (percent), Natural_Gas_Price_Index, and Industrial_Electricity_Price_Index. The dataset was used to compare the forecasting performance of AutoRegressive Integrated Moving Average (ARIMA), Random Forest (RF), and Gradient Boosting Machine (GBM) models under two predictor specifications: (1) a Base model using operational variables and (2) an Extended model that additionally incorporates external macroeconomic risk indicators. The original data were obtained from publicly available sources, including the Federal Reserve Economic Data (FRED), the U.S. Bureau of Labor Statistics (BLS), the U.S. Census Bureau, the Economic Policy Uncertainty database, and the Federal Reserve Bank of New York. The variables were aligned by observation month and merged into a single dataset suitable for forecasting analysis.The original sources for each variable are provided in the references section.