But Wait, There's More! Maximizing Substantive Inferences from TSCS Models

Political scientists rarely take full advantage of the substantive inferences that they can draw from time-series cross-section data. Most studies have emphasized statistical significance and other standard inferences that can be drawn from single coefficients over one time period. We show that by simulating the quantities of interest over longer periods of time and across theoretically interesting scenarios, we can draw much richer inferences. In this article, we present a technique that produces graphs of dynamic simulations of relationships over time. Graphical simulations are useful because they represent long-term relationships between key variables and allow for examination of the impact of exogenous and/or endogenous shocks. We demonstrate the technique's utility by graphically representing key relationships from two different works. We also present a preliminary version of the dynsim command, which we have designed to extend the Clarify commands in order to produce dynamic simulations.