Testing bias in professional forecasts

Professional forecasters can rely on econometric models, their personal expertise or both. To accommodate for adjustments to model forecasts, this paper proposes use two stage least squares (TSLS) (and not ordinary [OLS]) the familiar Mincer–Zarnowitz regression when examining bias in professional where instrumental variable is consensus forecast. An illustration 15 with quotes real gross domestic product (GDP) growth, inflation and unemployment United States documents usefulness of new estimation method. It also shows that TSLS suggests less than OLS does.