Abstract (may include machine translation)
We address the problem of optimally forecasting a binary variable for a heterogeneous group of decision makers facing various (binary) decision problems that are tied together only by the unknown outcome. A typical example is a weather forecaster who needs to estimate the probability of rain tomorrow and then report it to the public. Given a conditional probability model for the outcome of interest (e.g., logit or pro-bit), we introduce the idea of maximum welfare estimation and derive conditions under which traditional estimators, such as maximum likelihood or (nonlinear) least squares, are asymptotically socially optimal even when the underlying model is misspecified.
Original language | English |
---|---|
Pages (from-to) | 308-319 |
Number of pages | 12 |
Journal | Journal of Business and Economic Statistics |
Volume | 28 |
Issue number | 2 |
DOIs | |
State | Published - Apr 2010 |
Externally published | Yes |
Keywords
- Decision-based forecasting
- Maximum likelihood
- Maximum welfare estimator
- Misspecification
- Multiple decision makers
- Nonlinear least squares