Association for Public Policy Analysis & Management

Panel: The Use Of Evidence-Based Simulation Modeling Tools For Improving Public Policy Decisions

(Tools of Analysis: Methods, Data, Informatics and Research Design)

The rapid expansion and availability of government administrative data is challenging researchers’ ongoing efforts to effectively and appropriately use this data to inform public policy decisions. Researchers must keep pace with developing evidence-based analytical tools to handle this expanding data. In the public health and education sectors, these administrative data are especially useful when combined with government survey data to simulate the effects of various policy choices. Simulation modeling provides a useful empirical framework for forecasting the outcomes of various policies, allowing for the often unforeseen effects of complex systems. For example, agent based modeling (ABM) approaches, including spatial microsimulation, allow consideration of individual level heterogeneity. Combining ABM with other methods, including system dynamics modeling, yields even greater insights into nonlinear dynamics and emergent properties of complex systems. This panel’s theme focuses on the advantages of employing computer simulations in policy and management research to maximize the potential benefits of government data.  Our purpose is to provide both theoretical and practical perspectives on how government data can be used to inform public policy decisions.  Panelist #1 will present an ABM developed to evaluate food availability of households and to test the potential impacts of the policy interventions to decrease food insecurity. This study exemplifies the integration of both survey and census data in analysis of food security and the use of interdisciplinary data analytics methods. Panelist #2 will present a data driven Early Warning System prediction model that determines an individual student’s risk of dropping out of high school. This tool is designed to alert educators and parents to the risk of dropping out at an early stage to allow more effective interventions to occur.  Panelist #3 will present a hybrid modeling framework for evaluating policy effects on the scientific researcher workforce. This uses empirical data to model career path decisions of Ph.D. recipients. Finally, Panelist #4 will demonstrate an interactive system dynamics model that simulates the relative effects of policy interventions designed to reduce statewide infant mortality rates. The model was calibrated using Medicaid claims data with results that provided public health officials and other key stakeholders with unexpected and valuable insights. Together, all four presentations represent both the academic and practitioner perspectives, as well as policy and management perspectives, on computer simulation methods.