We use the term structure of disagreement of professional forecasters to document a novel set of facts: (1) forecasters disagree at all horizons, including the long run; (2) the term structure of disagreement differs markedly across variables: it is downward sloping for real output growth,...