Forecasts are Either Wrong or They are Lucky
An important part of forecasting is evaluating past forecasts for errors.
The following two mini-studies illustrate the level of accuracy for the economic forecast produced by the CU Leeds School.
Accuracy of the CU Leeds School by SIC Sector 1972-2001. Key findings from this brief analysis are:
- There is greater accuracy foretelling employment changes for large sectors or service producing sectors. Most service sectors have lower volatility.
- It is easier to project job gains than job losses. Forecasters don’t do well projecting job losses.
- Most employment forecasts tend to be too conservative.
Measurement of Forecast Accuracy for the Leeds School of Business Colorado Business Economic Outlook – 1972-2010. This brief analysis looks at forecast accuracy from three perspectives: jobs added vs. jobs lost, over/under (aggressive vs. conservative), and turns.
The following is recommended if you are interested in learning more about forecasting.
Macroeconomic Forecasts Microeconomic Forecasters, by Owen Lamont.
Abstract (from NBER): In the presence of principal-agent problems, published macroeconomic forecasts by professional economists may not measure expectations. Forecasters may use their forecasts in order to manipulate beliefs about their ability. As forecasters become older and more established, they produce more radical forecasts. Since these more radical forecasts are in general less accurate, ex post forecast accuracy grows significantly worse as forecasters become older and more established. Available for download at various websites, including NBER.