The aim is to evaluate the efficiency of the Congressional Budget Office’s (CBO) budgetary and economic forecasts and to investigate the causes of the inefficiency in its forecasts. While the efficiency is accepted for the CBO’s forecasts for outlays, it is strongly rejected for its revenue and economic forecasts. The conditional forecast evaluation suggests that efficiency is accepted once we control the differences in the underlying economic forecasts. By replacing the CBO’s underlying economic forecasts with another subjective forecast in an out-of-sample forecasting exercise, we find that this adjustment could achieve modest improvements in the accuracy of the CBO’s budgetary forecasts. Though the CBO tends to make positive forecast errors for the budget deficit, there is no definitive evidence for the asymmetry in its loss function.
The 1st International Conference on Econometrics and Statistics (HKUST), Hong Kong University of Science and Technology (HKUST) Business School EcoSta 2017, Parallel Session G, Friday 16.06.2017 10:20 - 12:25, EO0436