How Useful Are Oil Projections?

Edward Tufte:

n 1974 the Federal Energy Administration asked 4 statisticians to provide independent estimates of the amount of oil still underground. The four groups worked completely independently; I found out the names of the other three groups long after the reports were filed. The results, as I recall, were one very low forecast, one very high forecast, and two skeptical reports (including mine) in effect saying the error bound around the forecasts covered all reasonable policy alternatives. Thus the collective result confirmed the views of the two skeptical reports!

Here is my analysis. Perhaps this should have appeared as a short case study in Beautiful Evidence, but the idea never occurred to me.

What about now, 31 years later? My skepticism about resource forecasts might be confirmed or might not by a fresh analysis, which would reveal what a fresh analysis of the evidence would reveal. In policy relevant studies of evidence, there is too often a rage to conclude.