This paper shows how to bootstrap hypothesis tests in the context of the Parks (Efficient estimation of a system of regression equations when disturbances are both serially and contemporaneously correlated 1967) estimator. It then demonstrates that the bootstrap outperforms Parks's top...