Beware the Trojan Horse of INF ‘Inspections’

After decades of argument over the possibility of on-site inspections, the superpowers are finally on the threshold of carrying them out. This new experiment in U.S.-Soviet relations will begin 30 days after the treaty eliminating intermediate-range nuclear missiles takes effect. Over a 13-year period, hundreds of inspections will be carried out. But before this novel process has even been tried, critics and supporters of the INF treaty are advocating far more sweeping inspections under a strategic arms reduction treaty.

On-site inspections have obvious political appeal, but their substantive value appears to be widely overestimated. To begin with, inspections will be far less authoritative and valuable than “national technical means” of intelligence gathering--reconnaissance satellites and other collection methods that are under our exclusive control.

Inspections cannot substitute for remote surveillance; at best, they can supplement it. Inspections will provide the most authoritative information when they are based on data already collected by independent means. Manipulation then becomes harder for the cheater and easier to detect for the observer. Conversely, inspections provide the least reliable information when our data base is already quite weak.

Inspections are not a cure-all; they cannot ensure proper compliance. After all, hundreds of inspections and more than 1,000 inspectors could not ensure compliance by a defeated Germany with the provisions of the Versailles Treaty after World War I. How, then, can a few hundred U.S. inspectors enforce superpower compliance in the INF treaty?

While inspections cannot prevent cheating, they can still serve useful purposes if properly designed. Modern nuclear-weapon systems are quite difficult and expensive to build at facilities designed for producing nuclear weapons. They become even more complex and costly to build covertly. Military leaders also lose confidence in the operational effectiveness of these weapons if they are secretly stored under less-than-ideal conditions. For this reason, inspections at plants where weapons are produced and assembled have clear benefits because these inspections raise the cost of cheating. The superpowers have started down this road in the INF treaty.

In contrast, challenge inspections at suspect sites offer greater risks than benefits, despite the prevailing wisdom that we as an open society have more to learn from “anytime/anyplace” inspections than the secretive Soviets. Instead, our openness is likely to make it easier for the Soviets to direct their challenges at especially sensitive installations. As Reagan Administration officials have testified during the INF ratification hearings, the United States still has secrets that need to be kept. A challenge-inspection regime jeopardizes these secrets, without providing guaranteed access to comparable information at Soviet sites.

No one expects the Soviets to hand over new incriminating evidence during challenge inspections, so the argument for anytime/anyplace inspections rests primarily on their public-relations value: If the closed society refuses an inspection, then the open society succeeds in making the case for a violation. But the closed society has many ways of deflecting a challenge, like demanding counterinspections. A closed society also finds it much easier to manipulate evidence than a democratic society does. Inspections can thus provide numerous opportunities for mischief-making and misinformation. Open societies have fewer options because of an aggressive and unrestrained media. Challenges would be even easier to manipulate in START if both sides were permitted exemptions to refuse inspection demands.

The Reagan Administration wisely dropped its demand for challenge inspections at suspect sites in the INF negotiations, acknowledging its own inability to determine the size of Soviet inventories to be eliminated and the Kremlin’s ability to hide missiles from U.S. inspectors. Nevertheless, the multiple risks of challenge inspections at suspect sites could be avoided, especially since the INF treaty mandates the removal of the support equipment for these missiles and prohibits their flight testing and training exercises.

Even though a START agreement will not contain these prohibitions, the same logic applies as in INF. Short-notice challenge inspections of suspect sites will not appreciably increase confidence in compliance. Instead, they will generate a false sense of security because the risks inherent in anytime/anyplace inspection rights are far greater than the benefits.

In the INF treaty the Reagan Administration designed a verification regime that permits detection of militarily significant violations in time to take an effective response. This criterion can best be met in START by placing primary reliance on national technical means of intelligence gathering and by demanding routine inspection rights at the most verifiable points in the production and deployment of new weapons. Dropping demands for short-notice challenge inspections of suspect sites will avoid security problems, reduce opportunities for mischief-making and produce a better START agreement.