- Time series data involves observations across time that are related as part of a stochastic or random process. We can only observe one realization of this process. - Static regression models assume the dependent variable is related only to independent variables in the same time period. Finite distributed lag (FDL) models allow for lags where past values of independent variables may affect current dependent values. - In FDL models, coefficients measure the impact of temporary or permanent changes in independent variables on dependent variables over time, showing the "memory" or lag effects. Coefficients of permanent changes sum to indicate total long-run impact.