Estimating the effect of central bank independence on inflation using longitudinal targeted maximum likelihood estimation

Abstract The notion that an independent central bank reduces a country’s inflation is controversial hypothesis. To date, it has not been possible to satisfactorily answer this question because the complex macroeconomic structure gives rise data adequately incorporated into statistical analyses. We develop causal model summarizes economic process of inflation. Based on and recent data, we discuss identify assumptions under which effect independence can be identified estimated. Given these alternative assumptions, estimate using modern doubly robust estimators, i.e., longitudinal targeted maximum likelihood estimators. estimation procedure incorporates machine learning algorithms tailored address challenges associated with data. do find strong support for hypothesis having long period time necessarily lowers Simulation studies evaluate sensitivity proposed methods in settings when certain are violated highlight importance working appropriate estimation.