Abstract This paper describes and implements a simple partial solution to the most common problem in applied microeconometrics: estimating a linear causal effect with a potentially endogenous explanatory variable and no suitable instrumental variables. Empirical researchers faced with this...