Techniques for dealing with reverse causality between institutions and economic performance

Luciana Cingolani & Denis de Crombrugghe


This article provides a succinct review of the arguments stressing the mutual relationship between institutions and economic performance, and a scholarly account of some of the most popular econometric strategies used to minimize reversed causality problems in impact estimation. Among the techniques revisited we find the instrumental variables (IV) approach, distributed lags and vector autoregressions (VAR), quasi-experiments, and identification by heteroskedasticity (IH). Ultimately, the review is conceived as a methodological aide to researchers seeking to explore causal relationships through the use of the Institutional Profiles Database (IPD) produced by the Agence Francaise de Developpement (AFD).

Keywords: Institutions and growth, endogeneity, instrumental variables, dynamic analysis, identification through heteroskedasticity.

JEL codes: O42, C33, C36, P14

Download the working paper