Understanding how public transportation infrastructure investments shape regional and local economic development is essential for planners and policymakers. Traditional approaches, such as input-output based economic modeling, computable general equilibrium (CGE) analysis, econometric estimation, and stakeholder surveys, have yielded valuable insights but face persistent limitations. These include heavy data and resource demands, reliance on specialized expertise, and restrictive assumptions of linearity between investment and economic growth. This paper introduces a machine learning–based framework that addresses these challenges by capturing both linear and non-linear relationships between transportation investments and economic outcomes. Using detailed data on project costs, investment types, and regional socioeconomic conditions, the framework estimates impact multipliers for key economic indicators, including business establishments, employment, income, and GDP. These multipliers can be integrated into a GIS-enabled decision-support tool, allowing agencies to conduct rapid, scenario-based assessments from both ex-ante and ex-post perspectives. By leveraging extensive historical data and advanced modeling techniques, the proposed framework improves predictive accuracy, quantifies uncertainty, and enhances the transparency of economic impact analysis. The results demonstrate how AI-assisted methods can provide transportation agencies with more robust, timely, and policy-relevant evaluations of infrastructure investments, contributing to more informed and equitable decision-making.
Chen et al. (Sun,) studied this question.