Political forecasting: The missing link

Abstract Political forecasting provides the contextuality needed for decision‐making and for forecasting ‘non‐political’ trends. To gear political forecasting to these needs, rather than mimicking approaches in other areas, requires recognition of the distinctive nature of political trends, and realism regarding forecast uses, which generally do not benefit from ‘precise’ probabilities, predictions of only major events, or ‘sophisticated’ methodology that sacrifices comprehensiveness for explicitness. Approaches borrowed from other forecasting disciplines have been counterproductive, although contextual approaches, including cross‐impact analyses and developmental constructs that integrate political and non‐political trends, are promising. Explorations of the consistency of scenario dynamics, taking into account policy responses and non‐formalizable complexity, are also useful. Thus the separation of political forecasting from political analysis should be minimized, calling for a redirection of effort away from developing methodology uniquely geared to forecasting, and towards organizing more comprehensive and systematic analytical efforts.