What Is the True Value of Forecasts
Bibliographic Data
| ID | 22332078 |
|---|---|
| Authors | Antony Millner (University of Oxford, corresponding author) |
| Year | 2009 |
| Volume | 1 |
| Issue | 1 |
| Pages | 22-37 |
| Publication date | 2009-10-01 |
| Peer Reviewed | Yes |
| Open Access | No |
| Type | ARTICLE |
| Venue | Weather, Climate, and Society (JOURNAL) |
| Journal identifiers | ISSN: 1948-8327 • E-ISSN: 1948-8335 |
| Publisher | American Meteorological Society (PUBLISHER • US) |
| DOI | 10.1175/2009wcas1001.1 |
| OpenAlex | W2152100541 |
| Language | EN |
| Citations received | 4 |
| References cited | 41 |
Understanding the economic value of weather and climate forecasts is of tremendous practical importance. Traditional models that have attempted to gauge forecast value have focused on a best-case scenario, in which forecast users are assumed to be statistically sophisticated, hyperrational decision makers with perfect knowledge and understanding of forecast performance. These models provide a normative benchmark for assessing forecast value, but say nothing about the value that actual forecast users realize. Real forecast users are subject to a variety of behavioral effects and informational constraints that violate the assumptions of normative models. In this paper, one of the normative assumptions about user behavior is relaxed—users are no longer assumed to be in possession of a perfect statistical understanding of forecast performance. In the case of a cost–loss decision, it is shown that a model of users’ forecast use choices based on the psychological theory of reinforcement learning leads to a behavioral adjustment factor that lowers the relative value score that the user achieves. The dependence of this factor on the user’s decision parameters (the ratio of costs to losses) and the forecast skill is deduced. Differences between the losses predicted by the behavioral and normative models are greatest for users with intermediate cost–loss ratios, and when forecasts have intermediate skill. The relevance of the model as a tool for directing user education initiatives is briefly discussed, and a direction for future research is proposed
Econometrics · Economics · Machine learning · Normative · Operations research · Value of information · Computer Science · Decision-Making and Behavioral Economics · Economic and Environmental Valuation · Forecasting Techniques and Applications · Mathematics · Artificial Intelligence
Stochastic models for learning.
Choices, Values, and Frames
Learning the value of information in an uncertain world
Verification, Validation, and Confirmation of Numerical Models in the Earth Sciences
Ethnographic and participatory approaches to research on farmers responses to climate predictions
Weather Forecasts are for Wimps
Prospect Theory
“A 30% Chance of Rain Tomorrow”
Choices, values, and frames
Effective seasonal climate forecast applications
Loss aversion, diminishing sensitivity, and the effect of experience on repeated decisions
The Rate of Establishment of a Discrimination
Why the Law of Effect will not Go Away
| Unique citing works | 4 |
|---|---|
| Citations per year | 0.25 |
| Citation span | 2010 - 2024 (15) |
| Citation velocity | recent |
| Highly cited | No |
| Citation types | Neutral: 4 |