Prediction and Economic Theory

The role of prediction in economics involves a fundamental tension. On the one hand, much of economics is concerned with prediction. On the other, economic predictions are notoriously unreliable. It is, in fact, tempting to see the economist as the trapeze-performer who tends to miss the cross-bar, or as the jockey who keeps falling off his horse. Whether or not such characterization is fair, there can be no doubt that the nature and genesis of this fundamental tension, and its implications, do call for systematic analysis and assessment. The aim of this paper is to attempt a brief examination of these issues. There is a sequence of questions to be faced. How central is prediction to economics? Why are economic predictions so difficult? What techniques does economic theory use to cope with these difficulties? Are these techniques sound? How does economic theory relate to practice? I shall take up these questions in turn.