The NERC Fan: A Retrospective Analysis of the NERC Summary Forecasts
对比北美电力可靠性委员会1974年起发布的年度电力需求增长预测与简单模型（指数平滑/Box-Jenkins方法及动态需求方程）的基准预测，发现委员会预测相对指数平滑法并不过于乐观，但早期预测相对需求方程存在乐观偏差，该偏差与对能源价格持久性和需求价格敏感性的学习过程一致。
Summary forecasts of the growth in demand for electricity, published annually since 1974 by the North American Electric Reliability Council, have proved to be excessively optimistic, although the projected growth rate has been revised downward each year. Should forecasters have been able to do a better job of predicting the slowdown in demand growth since the early 1970s? Should they have responded more quickly as actual demand growth fell short of projections? Our objective in this article is to provide some partial answers to these questions by comparing the published NERC Summary Forecasts (NSF) with benchmark forecasts provided by simple models representing well-established techniques. One alternative forecast is provided by the exponential smoothing/Box-Jenkins method (ES/BJ) and another by a simple dynamic demand equation that uses real price and real income as explanatory variables. We conclude that relative to ES/BJ, the NSF were not excessively optimistic. Relative to the simple demand equation, the early NSF were optimistic; sensitivity analysis indicates that the optimism is consistent with learning about either the permanence of higher energy prices in the mid-1970s or the full extent of demand's sensitivity to price.