Fiscal Forecast Errors in Jamaica, 1968 - 2016 : Are Government’s Budgetary Projections Systematically Biased?

AbstractAccurate forecasting is critical for an efficient fiscal policy framework since it impacts consumer sentiments, business confidence, domestic and international institutions. Forecasts will be evaluated for accuracy, bias and efficiency, using ME, MAE, RMSE, Theil’s U statistic and the weak efficiency test. Biased forecasts usually result in a budget deficit, leading to financing and debt issues, which influenced the country’s economic growth rate. Such biases have implications for macroeconomic projections in the formulation of short-run and medium-term plans. Efficiency tests focus on the use of available information at the design stage. The aim is to estimate the deviation between fiscal forecasts and actual outcomes for Jamaica for the period 1968 to 2016, focusing on government expenditure, revenue and the fiscal deficit. It derives estimates of the size and trend of fiscal forecasting errors, at the aggregated and disaggregated levels. The period 2013 to 2016 will be used to test the impact of fiscal rules on the size of forecasting errors. Results show that expenditure and revenue errors were on the optimistic side, 0.38 per cent of GDP. Capital and current expenditure and tax revenue were also over-estimated, but at lower ratios. Fiscal and primary balances were at higher levels, 0.44 and 0.41 per cent of GDP. The adoption of fiscal rules reduced these ratios marginally. The results from both the F test and t-test suggest that all revenue variables were biased and inefficient. For the expenditure categories, total and current expenditure were unbiased and efficient, while capital expenditure was unbiased but inefficient