A Quantitative Assessment of AI-Driven Predictive Analytics for Economic Development Decision Support in U.S. Public Policy Centers

Abstract

This study examined the problem of limited empirical evidence on whether AI-driven predictive analytics meaningfully improves economic development decision support in U.S. public policy centers. Although these institutions increasingly use data integration, forecasting systems, analytical automation, and real-time insight tools, the measurable value of these capabilities for policy decision quality, efficiency, forecasting accuracy, and responsiveness has remained insufficiently established. The purpose of the study was to quantitatively assess how AI-driven predictive analytics capability influences economic development decision-support effectiveness in selected U.S. public policy centers. A quantitative, cross-sectional, case-based research design was adopted, using cloud-enabled and enterprise-oriented public policy cases involving policy analysts, economists, research officers, program managers, strategic planners, and administrative professionals. From 260 distributed questionnaires, 228 were returned and 220 valid responses were analyzed, producing a usable response rate of 84.6%. The key independent variables were data integration capability, forecasting capability, analytical automation, and real-time insight generation, while the dependent variables were decision quality, decision-making efficiency, forecasting accuracy, policy responsiveness, and overall decision-support effectiveness. The analysis plan included descriptive statistics, Pearson correlation analysis, multiple regression modeling, and hypothesis testing using SPSS. The findings showed a high overall level of AI-driven predictive analytics capability, with a mean of 3.96, and high decision-support effectiveness, with a mean of 4.01. Predictive analytics was significantly correlated with forecasting accuracy, r = .760, decision quality, r = .710, policy responsiveness, r = .680, and decision-making efficiency, r = .640, all at p < .001. The regression model was statistically significant, R = .794, R² = .630, F = 91.84, p < .001, indicating that analytics capability explained 63.0% of the variance in decision-support effectiveness. Forecasting capability was the strongest predictor, β = .340, followed by data integration, β = .280. The study implies that U.S. public policy centers should strengthen predictive forecasting, integrated data infrastructures, and real-time insight systems to improve evidence-based economic development planning.