This paper develops and empirically supports, using 3 and 6 month interest rates, a theory that political risk can explain the shifting term premia found in U.S. data. We find that incorporating these political regime shifts yield results that support the expectations hypothesis.