Structural Analysis

Overview of Structural Estimation Uses and Methods

Overview/Intro

Reduced Form Estimation via linear regression is most common way to estimate impact or effectiveness of programs in data science. Most attention is put on achieving "significant results", fixing standard errors, best designing panel data to control for unobservable factors and assure that control variables are exogenous.

Linear regression is well-known, easy to interpret, and simple to communicate. However, there is a limit to what can be achieved as far as causal interpretation.

OLS regression has high variance-- there's no guarantee that findings will hold under new data in the "real world"! This can obviously lead to costly mistakes in business strategy. If your customer base comes from a very specific cluster of the market, reduced form estimates and even Machine Learning predictions won't be very helpful at forecasting your revenue in the case a different segment was captured.

Structural estimation is a well-studied technique that makes it possible to obtain robust conclusions and predictions from your data.