The Importance of Listening to Your Own Forecast

Tim Lister: It's easy to say, look, I'm a statistician. I'll just pick whatever forecast better in the past. That's a bad thing to do because first of all, if only wanted to forecast, then just look at the central banks forecast. Don't bother even looking at households, firms or someone else. Second, it's a general principle forecasting that you should combine different measures, never pick one. More important than whether a forecast performs very well is whether the person making the forecast is responding to its own forecast. Is its forecast being actually important for the decisions that it makes?