Currently, the commonly employed instrumental variables strategy relies on the knife-edge assumption of perfect exogeneity for valid inference. To make reliable inferences on the structural parameters under violations of exogeneity one must know the true correlation between the structural error...