NCJ Number
50231
Date Published
1978
Length
252 pages
Annotation
THE GENERAL PERFORMANCE OF FORECASTING FROM 1930 TO 1980 AND THE BROAD PRINCIPLES WHICH HAVE PRODUCED RELATIVELY GOOD FORECASTS ARE STUDIED. TECHNIQUES FOR EVALUATING THE QUALITY OF A FORECAST ARE DESCRIBED.
Abstract
FOLLOWING AN OVERVIEW OF THE USE OF FORECASTING IN SOCIAL AND ECONOMIC PLANNING, THE ACCURACY OF A NUMBER OF EXPERT FORECASTS MADE OVER A 50-YEAR PERIOD IN THE FIELDS OF POPULATION, ECONOMICS, ENERGY USE, TRANSPORTATION, AND TECHNOLOGY ARE EXAMINED. THESE FORECASTS INCLUDE BOTH THOSE MADE BY GOVERNMENT AGENCIES AND THOSE DEVELOPED BY INDUSTRY AND PRIVATE STUDY GROUPS. TO IDENTIFY THOSE FACTORS ASSOCIATED WITH GREATER ACCURACY AS WELL AS THOSE FACTORS LEADING TO BIAS, THE METHODS USED, THE SOURCES OF DATA, AND THE ULTIMATE ACCURACY ARE EXAMINED FOR EACH FORECAST. IT IS FOUND THAT METHODOLOGICAL SOPHISTICATION, SUCH AS USE OF A COMPUTER MODEL, IS RELATIVELY UNIMPORTANT. THE ACCURACY OF A FORECAST DEPENDS ON THE BASIC ASSUMPTIONS, THE QUALITY OF THE DATA, AND THE LENGTH OF TIME FOR WHICH THE FORECAST IS MADE. INTERVENING FACTORS WHICH NO FORECASTER CAN FORESEE ARE SUMMARIZED. TECHNIQUES AND CURVES ARE THEN GIVEN WHICH PLANNERS CAN USE TO EVALUATE THE ACCURACY OF A FORECAST. IN SOME CASES POLICYMAKERS CAN PREDICT THE LIKELY ACCURACY AND BIASES OF FORECASTS JUST BY LOOKING AT THEIR DATA SOURCES AND METHODOLOGY. SEVERAL CASE STUDIES TRACE THE IMPACT OF EXPERT FORECASTING ON GOVERNMENTAL AND PRIVATE DECISIONMAKING. THERE ARE EXTENSIVE FIGURES AND TABLES ILLUSTRATING VARIOUS DATA CURVES, BIASES, AND ERRORS. EACH CHAPTER IS FOOTNOTED. (GLR)