🤖 AI Summary
Traditional instrumental variable (IV) methods often suffer from causal bias due to weak or invalid instruments and reliance on external data. To address this, we propose a novel data-driven approach that constructs synthetic instrumental variables (SIVs) solely from observed covariates. Our method introduces the “double-tilting (DT) condition”—a newly established identification criterion that enables valid IV selection without external instruments and further determines the sign of the correlation between the endogenous variable and the structural error. By integrating DT-condition testing with heteroskedasticity-robust estimation, our framework substantially improves causal effect estimation accuracy in both simulations and empirical applications. It effectively mitigates weak instrument and instrument invalidity issues while drastically reducing dependence on exogenous instruments. This work establishes a verifiable, purely observational paradigm for addressing endogeneity, advancing causal inference methodology beyond conventional IV assumptions.
📝 Abstract
Traditional instrumental variable (IV) methods often struggle with weak or invalid instruments and rely heavily on external data. We introduce a Synthetic Instrumental Variable (SIV) approach that constructs valid instruments using only existing data. Our method leverages a data-driven dual tendency (DT) condition to identify valid instruments without requiring external variables. SIV is robust to heteroscedasticity and can determine the true sign of the correlation between endogenous regressors and errors--an assumption typically imposed in empirical work. Through simulations and real-world applications, we show that SIV improves causal inference by mitigating common IV limitations and reducing dependence on scarce instruments. This approach has broad implications for economics, epidemiology, and policy evaluation.