THE DECLASSIFIED
COLLECTION
← All articles

Nuclear age

How close had a NATO exercise brought the world to a nuclear crisis?

A 1990 review argued that American intelligence had underestimated Soviet fear in 1983. Its sharpest warning concerns how analysts handle a dangerous alternative explanation.

An exercise seen through another country's fears

The NATO exercise Able Archer belongs to a wider 1983 story of deteriorating relations and Soviet concern about a surprise attack. The 1990 President's Foreign Intelligence Advisory Board report, The Soviet War Scare, revisited how American intelligence interpreted that concern. Its central criticism was that reassuring judgments had been expressed with more confidence than the evidence justified.

The board argued that information available then, supplemented by later reporting, supported genuine fear among important Soviet leaders. It challenged assessments that had treated alarming Soviet statements and behaviour principally as propaganda. That is a finding about perception and intelligence analysis. It should not be inflated into a documented nuclear launch countdown.

Two explanations for the same signal

A military alert can be interpreted as strategic theatre, prudent precaution or preparation for imminent conflict. Analysts rarely receive an unambiguous label identifying which explanation is correct. The report describes how an established interpretation can become resistant to evidence: an observation inconsistent with it is explained away, while confirming observations receive greater weight.

The danger is not simply choosing the wrong prediction. Decision-makers also need to understand the consequence of the less likely alternative. If a counterpart truly fears attack, an action intended as a routine signal of resolve might be read very differently. A low-probability interpretation can deserve attention because the cost of dismissing it is unusually high.

Hindsight is useful and unfair

The board acknowledged that some of its confidence came from information unavailable to analysts at the time. That admission matters. Retrospective reviews can reveal failures while still recognizing the constraints under which earlier decisions were made. An archive should preserve that qualification beside the criticism.

The surviving public report is also declassified in part. Redactions are evidence of limits on what readers can inspect, not blank spaces into which a preferred dramatic narrative can be inserted. Claims about specific orders, weapons or timing should be attached to the documents that support them rather than inferred from the report's overall severity.

The recommendation was methodological

The board urged better understanding of Soviet decision-making and fuller assessment of plausible alternatives. Its discussion did not advocate a debate in which competing offices simply defended their institutional positions more aggressively. The aim was to give policymakers a clearer account of uncertainty and risk.

That is the enduring value of the report. A warning system fails if it reports only what its analysts consider most likely while concealing what they cannot safely exclude. Reading Able Archer through this document therefore means asking two questions together: what did intelligence believe was happening, and what could have happened if that belief was wrong? The report makes the second question harder to ignore without pretending that the historical record answers every detail.

Sources and further reading

  1. The Soviet War Scare ↗

    PFIAB, 15 February 1990; executive summary pp vi–xii and recommendations; 109-page declassified-in-part file

Continue reading

The fallout study that catalogued babies' bones around the world ↗The Air Force commissioned a study of a nuclear explosion on the Moon ↗Two nuclear bombs fell on North Carolina. Why didn't they explode? ↗