Long-term forecasts of military technologies for a 20–30 year horizon: An empirical assessment of accuracy

• Published in: Technological forecasting & social change Vol. 137; pp. 272 - 279
• Main Authors: Kott, Alexander, Perconti, Philip
• Format: Journal Article
• Language: English
• Published: New York Elsevier Inc 01-12-2018; Elsevier Science Ltd
• Subjects: Accuracy; Armed forces; Assessments; Data processing; Electronic warfare; Evaluation; Forecast assessment; Forecasts; Impact analysis; Long-range forecast; Military technology; Missile systems; Missiles; Prediction accuracy; R&D; Research & development; Statistical significance; Technology; Technology forecast; Unmanned aerial vehicles; Union of Soviet Socialist Republics > USSR; Prediction accuracy; Military technology; Technology forecast; Forecast assessment; Long-range forecast
• ISSN: 0040-1625; 1873-5509
• DOI: 10.1016/j.techfore.2018.08.001

During the 1990s, while exploring the impact of the collapse of the Soviet Union on developments in future warfare, a number of authors offered forecasts of military technology appearing by the year 2020. The forecasts covered diverse systems ranging from unmanned systems to guns to missiles to electronic warfare. This paper offers a quantitative assessment of the accuracy of this group of forecasts. The overall accuracy — by several measures — was assessed as quite high. In particular, when measured by the expert-assessed accuracy of the forecast statements, the average accuracy is 0.76. This points to the potential value of such forecasts in managing investments in long-term research and development. Major differences in accuracy, with strong statistical significance, were found between forecasts pertaining primarily to information acquisition and processing technologies, as opposed to technologies that aim primarily at physical effects. This paper also proposes several recommendations regarding methodological aspects of forecast accuracy assessments. Although the assessments were restricted to information available in open literature, the expert assessors did not find this constraint a significant detriment to the assessment process.
•Accuracy of forecasts made in the 1990s for 2020 is high: average accuracy of 0.76.•This is broadly comparable to two other known technology forecasts with horizons over 20 years.•89% of forecasts name directions of R&D that materialized or remain promising 20 years later.•Forecasts for “informational” technologies are significantly more accurate than for “physical” technologies.•Long-range forecasting of military technology should use readiness as a key measure for assessment.