# Predicting Inflation: Professional Experts Versus No-Change Forecasts

**Authors:** Tilmann Gneiting, Thordis L. Thorarinsdottir  
**arXiv:** [1010.2318](https://arxiv.org/abs/1010.2318)  
**Date:** 2010-10-12  
**Summary page:** https://www.summarizepaper.com/en/paper/1010.2318v1/

## Key points

- Study compares forecasts of United States inflation from the Survey of Professional Forecasters (SPF) to predictions made by simple statistical techniques
- Economic expertise is persuasive in nowcasting
- Novel yet simplistic probabilistic no-change forecasts are equally competitive when projecting beyond the current quarter
- Surveys can be interpreted as ensembles of forecasts, similar to ensemble prediction systems in weather forecasting
- Statistical techniques borrowed from weather forecasting enhance the accuracy and reliability of the SPF forecast
- Findings hold true regardless of adjustments made to the SPF forecast through statistical techniques
- Both economic expertise and simplistic probabilistic no-change forecasts are effective in predicting inflation
- Ensemble approaches and postprocessing techniques can enhance forecast accuracy in economic forecasting, drawing parallels with advancements in weather forecasting methods.

## Summary

In their study titled "Predicting Inflation: Professional Experts Versus No-Change Forecasts," Tilmann Gneiting and Thordis L. Thorarinsdottir compare forecasts of United States inflation from the Survey of Professional Forecasters (SPF) to predictions made by simple statistical techniques. They find that in nowcasting, economic expertise is persuasive. However, when projecting beyond the current quarter, they discover that novel yet simplistic probabilistic no-change forecasts are equally competitive. The authors also interpret surveys as ensembles of forecasts and demonstrate that they can be used in a similar manner to ensemble prediction systems in weather forecasting. This approach allows for a more comprehensive understanding of inflation predictions. Drawing inspiration from weather forecasting practices, Gneiting and Thorarinsdottir apply statistical techniques to postprocess the SPF forecast based on recent past experiences. This borrowing of ideas enhances the accuracy and reliability of the SPF forecast. Importantly, even after survey postprocessing, the authors find that their earlier conclusions remain unchanged. This suggests that their findings hold true regardless of any adjustments made to the SPF forecast through statistical techniques. Overall, this study highlights the effectiveness of both economic expertise and simplistic probabilistic no-change forecasts in predicting inflation. It also emphasizes the potential benefits of using ensemble approaches and postprocessing techniques to enhance forecast accuracy in economic forecasting; drawing parallels with advancements in weather forecasting methods.