The understanding of applied modeling in causal effects is of particular importance in econometrics, according to recent developments and research in causal inference applications. We also provide an outline of econometrics’ use of causal inference. The majority of economists would agree that the randomized controlled experiment is the gold standard for drawing conclusions, but actually, a significant portion of empirical work in econometrics relies on observational data, where, among other things, the possibility of confounding or loss of exogeneity must be taken into account. We focus in particular on two types of contemporary research: randomized experiments and observational studies. Our review of the dynamic causality study approach, the linear method, which includes LP and VAR, and nonlinear statistical modeling which includes BART, and their use in econometrics, are all reviewed in this paper. Modeling dynamic systems with linear parametric models usually suffer limitation which affects forecasting performance and policy implications. On the nonparametric framework, BART specifications can produce more precise tail forecasts than the VAR structure. Finally, BART has the lowest RMSE in linear and non-linear data generation processes, and also the performance of BART important variables in a set of macroeconomic data has an optimal performance than other regression estimators. Keywords
