by Peter Byrne
26 Everett and the SIOP
In the councils of government, we must guard against the acquisition of unwarranted influence, whether sought or unsought, by the military industrial complex. The potential for the disastrous rise of misplaced power exists and will persist…. Yet, in holding scientific research and discovery in respect, as we should, we must also be alert to the equal and opposite danger that public policy could itself become the captive of a scientific technological elite.
Farewell address by President Dwight D. Eisenhower,
January 17, 1961.
Calming the Wargasm?
In November 1958, President Eisenhower was briefed on the Air Force plan to launch all of its bombers at the major population centers of the U.S.S.R., China, and their allies the minute that the president (or the Strategic Air Command) believed that the Soviets were preparing to attack U.S. interests, even with conventional forces. Appalled, Eisenhower commented that the number of targets was beyond those needed to destroy the Soviet’s will to fight. The U.S. does not require, “a 100 per cent pulverization of the Soviet Union,” he said. He asked the Pentagon to formulate a way of selectively taking out military targets, while sparing cities. He wanted targeting options because, he observed, there is a “limit to the devastation which human beings can endure.”1
The result of his order was the Single Integrated Operating Plan—a project so secret it was assigned its own security classification, “extremely sensitive information.” It was intended to integrate the duplicative nuclear bomb targeting plans of the Army, Navy, and Air Force into a single plan capable of being reprogrammed—capable of flexibility, of surgical strikes (with hydrogen bombs).
The Joint Chiefs entrusted the development of the SIOP to SAC because it already had a command and control system in place, as weak and undependable as it was, and some understanding of the problems involved in launching and coordinating nuclear attacks. But the chiefs were not naïve. They did not fully trust SAC—nor the other armed services, nor the weapons contractors who benefited from manufacturing instruments of war—to put the interests of the country before their own. So, they ordered WSEG to work with SAC.2
Although many details of this history remained sealed, there is no doubt that Everett, as WSEG’s chief mathematician and resident computer wizard was a leader of the group detailed to design software for the SIOP, as well as the command and control systems of the National Military Command System, which was the Pentagon’s 24-hour-a-day war room. WSEG staff worked with the
actual plans, data, and procedures with a high degree of both national security and political/administrative sensitivity … [The work] was conducted under special access and reporting arrangements…. Results were not published in official WSEG studies but were reported either informally, without a distinctive written product, or in the form of memoranda [transmitted through] a ‘quiet’ reporting channel.3
The very concept of a SIOP was kept secret from the public until 1976, when it was first referred to in news reports, glancingly. It was not until near the end of the Cold War that scholars were granted access to government documents outlining the war plan.4 It is clear that civilian officials had hoped that the first SIOP would provide them with political flexibility by allowing for surgical operations, but, by the time that the military got through beefing it up, it was more sledgehammer than scalpel.5
The center of gravity of SIOP-62 was its first strike capability. It was geared to instantaneously shoot off the entire U.S. nuclear arsenal of 3,423 bombs, obliterating more than a thousand targets and 285 million people in the U.S.S.R., China and Eastern Europe.6 And that damage was to come from blast effects alone; the effects of firestorm and radioactive fallout were not included in the damage calculations. When these were taken into consideration, the potential dead numbered over a billion.7 The SIOP’s retaliatory option alone promised to throw 1,706 bombs against 725 targets. The Wargasm mentality had not been shed. The plan made little or no distinction between communist countries that were at war with the United States and those that were not; it was poised to incinerate them all in a massive nuclear belch regardless of any threat they posed at the time of launch.
Where SIOP-62 principally differed from Wargasm was in attempting to synchronize the uncoordinated strike plans of the Army, Navy, and Air Force. This process was monstrously complex and cumbersome, generating hundreds of new acronyms as it stitched together thousands of tactical instructions into a single strategic plan. Eisenhower was reportedly aghast at the unconstrained violence of the newly integrated attack plan, but he was unable to change it.8 Kennedy was also appalled at the rigidity of the new plan, and he requested more flexibility in the next version, SIOP-63, which did introduce the ability for a commander-in-chief to “withhold” some major target areas. Nonetheless, the five “options” in the first decade of the SIOP were three first strike and two second strike spasms.9
The SIOP was supposed to enhance flexibility, but as it was designed to operate cybernetically, automatically, it provided minimal opportunity for human intervention. Basically, it updated Wargasm to include more advanced weapons systems and greater firepower. Under the minimum option available, at least 80 million people would die.10 And the system was set on a hair-trigger, with tremendous potential for premature launch. According to the National Security Archive,
Policymakers understood the dangers associated with preemptive attacks—the warning of the enemy attack being preempted might be inaccurate and preemptive attack on another nuclear power could not prevent tremendous destruction to the United States.”11
City blasting
Within military circles, thinking differed greatly on “how many and what kind of targets should be destroyed.”12 The chiefs were well aware that Soviet capabilities were overestimated by generals and politicians who sometimes massaged intelligence data to reinforce their own agendas. But the chief’s mandate was to prepare for winning a war, so they strove to maintain the clear superiority of the U.S. in war-fighting forces, even at the risk of nuclear overkill.13
A core question was how best to allocate weapons between cities and military installations, (although striking military targets would certainly impact cities). Those who favored striking directly at population centers argued for that strategy as a necessity because it was impossible to find and destroy enough Soviet missiles and bombers to prevent their retaliatory launching, and the situation would only worsen when the Soviets mirrored the U.S. and armed their submarine fleet with nuclear warheads. So, it was best to threaten the jugular of society: the cities. Adding weight to this argument was that if the Soviets attacked first, it would be impossible to know which of their forces had been left in reserve—especially in the years before powerful reconnaissance satellites were available—so cities were the only stable targets.
“Counterforce” proponents, on the other hand, believed that selectively destroying a few hundred military targets would decapitate the Soviet response capabilities:
If we were to strike certain urban and control centers, the Soviets would be incapable of prosecuting the war, and the United States would emerge on top. Therefore, … it is a waste of money to build a strategic delivery system capable of attacking more than a few hundred targets.14
Counterforce was a surprise attack aimed at preventing a Soviet strike. Of course, the United State’s offensive warheads could equally be destroyed by a Soviet counterforce strike. So, in a counterforce war, the winner would be the nation that shot first—which undermined the logic of deterrence by assured destruction, as there would be no mutual destruction if one side’s forces were decapitated before they could retaliate in a meaningful manner.
Critiquing counterforce, the Navy noted,
In a surprise attack it is not inevitable that enemy missiles will land before our weapons are launched, but the chances are great that they will. Under these conditions, it would be injudicious to launch the remainder of our greatly depleted forces against a primary target sys
tem of empty bases and missile sites, even though we should know their locations, which we probably won’t.15
Counterforce, no-cities was out.
On paper, if not in cruel reality, the SIOP was designed to operate after a surprise attack due to the sheer number of warheads deployed, especially on submarines. But it was meant to function most effectively as a first strike mechanism, whether preemptive or preventive. And that was about the limit of SIOP-62’s flexibility.16
Using Everett’s algorithm
WSEG’s main task for the SIOP project was to write programs that minimized the expenditure of weapon resources while maximizing damage to the enemy. Everett’s λ multipliers were undoubtedly used to calculate the trade-offs of deploying specific weapons systems as a function of “kill ratios,” or destruction probabilities. Optimizing destruction required predicting the amount of damage caused by various explosive yields.
The “damage function” was also constrained by a “circular area of probability,” i.e. the radius around a target in which a bomb was likely to fall if its “delivery vehicle” managed to penetrate enemy air defenses. Expected destruction was a function of the height at which the bomb exploded. Surface bursts delivered more fallout and localized destruction than air bursts, but required more bombs to ensure a high probability of kill over a wide area. The amount of toleration allowed for “collateral damage” to friendly forces or untargeted populations provided another constraint.17
Differing values for these variables could be plugged into Everett’s multiplier formulae and the “price” for maximizing desired results would pop out as a probability measure, i.e. according to λ, X amount of dollars will buy Y amount of damage on target T with probability P.
Air Force v. Navy
The SIOP was based on a list of targets drawn from the ultra secret National Strategic Target List. Intelligence data for each Desired Ground Zero was fed into a sealed folder to be opened by its keeper only during time of war. The effectiveness of the SIOP was premised on the data for the ground zeros being automatically updated as new intelligence became available, or when political objectives changed, or when budget constraints required tweaking the optimal mix of weapons and targets. But the Joint Chiefs could not easily agree on the who, what, where, when and why of targeting.
The Air Force favored configuring the SIOP to surprise and obliterate the Russians. But as Navy forces were more likely to survive a surprise attack on the United States than were the SAC bombers, the Navy favored configuring the SIOP to employ its aircraft carriers and submarines in a more graduated response.
The heads of these services disagreed, almost violently, on how to set standards for destruction:
It can be stated that the Army and Navy favor a lower level of destruction while the Air Force favors a higher level of destruction … because of [its] experience that it is almost always cheaper to destroy a target in the initial attack, even if it requires more force, than to have to reattack the same target.18
Adjudicating that particular question, said Air Force Major General Nathan F. Twining, the chairman of the joint chiefs, “can be better handled by analytical and mathematical techniques than can other aspects of the problem.”19 Twining was eager to reduce friction caused by inter-service rivalry by tasking WSEG to demonstrate choices using war games.20 Everett invented a war game for the SIOP, as well as a RSIOP, the Pentagon’s mirror of the Russian war plan; and his simulations were used by the joint chiefs to monitor SAC.
Much of what we know about the SIOP comes from memos written by Navy admirals who were incensed that control of the targeting and, consequently, the coming world war, had been handed over to SAC. The Navy argued that SIOP-62 was inflexible because it was a “pre-conceived plan (seldom is such a plan executed).” Consequently, argued the Navy,
We would forfeit the flexibility that is inherent in the decentralized execution of the strike plans by several unified commanders. The military logic of retaining this flexibility is overwhelming. In preparation for World War II France had a single pre-conceived plan that she thought was foolproof, but it was virtually worthless.21
As planning was held hostage to a struggle for control of billions of dollars in weapon procurement funds, WSEG’s analytical capabilities were sought by SAC’s opponents. Trying to curtail SAC’s influence on targeting strategy, the Navy suggested that,
We should subject the target lists and damage criteria to analysis by machine and mathematical techniques. A major objective of this analysis would be to arrive at an estimate of ‘how much is enough.’22
Recognizing that excessive fallout could harm its own forces, the Navy suggested that operations researchers determine whether,
Carefully planned utilization of radioactive fall-out [can] be exploited where practicable to contribute it the destruction of enemy government and military controls and to the general deterioration of the Sino-Soviet bloc.23
In other words, if fallout from bombing Russia wafted over to China, which was not at war with the U.S., so be it!, as long as easterly winds didn’t shower the U.S. Pacific fleet with radioactive dust.
Mirror, mirror
In August 1960, Admiral Arleigh Burke had a long talk about the SIOP with Secretary of the Navy William Franke. There is a transcript of the supercharged conversation, in which Burke threatened to quit the Navy if SAC was kept in charge of the war plan. He was afraid that his carriers would be assigned “insignificant targets.”24
Burke noted that the group putting together the SIOP under SAC control had some Navy officers in it, “But it is just like putting a little bug in a piece of plastic. The bug does not control the plastic. The plastic encases the bug. In other words, these people could be absorbed.”25 Burke’s concerns provides us with a very important insight into how the military works:
If SAC gets control of this thing, the number of atomic weapons will be tremendous and they will be the wrong kind of atomic weapons. The number of horses [bomber planes] will be tremendous. There will be thousands and thousands of Minutemen. They will control the budget. They will control everything, and they will wreck … everything…. And the Joint Chiefs of Staff and the Secretary of Defense and nobody else can stop it because they [SAC] are the ones that have the figures [target data] … In a year of this stuff—you can never undig it. Grave harm. And the President won’t have the guts any more than the past Presidents have had the guts—because these people will be entrenched. The systems will be laid. The grooves will be dug. And the power will be there because the money will be there. The electronic industry and all of those things. We will wreck this country.26
In a remarkably forthright condemnation of SIOP-62, Burke wrote a series of top secret memos accusing SAC of ignoring Eisenhower’s instruction to prioritize retaliation over preemption. He said SAC was overkilling—assigning many more hydrogen bombs to Moscow, Stalingrad, and Kalingrad than Eisenhower had requested. Burke suggested that the high percentage of surface bursts included in the SIOP could result in “world-wide contamination” from radioactive fallout. He questioned the policy of casually obliterating cities whose only sin was that they lay under air corridors leading to more high value targets. He commented that, “misses will kill a lot of Russians and Chinese even if the specified objective is missed, therefore misses should not be given a zero in the box score of damage achieved.” He asserted that the number of targets was “unnecessarily large,” and that the “damage criteria are excessively high,” and that war gaming independent of SAC should be used to “assess validity of the plan.”27
He then damned the doctrine of first strike:
Preemptive, preventative or initiative strikes will not prevent serious damage on United States because first, we do not know where their land-based missile sites are and never will know all of them. And, second, we will not be able to destroy all enemy seagoing missile forces simultaneously before they can get off their missiles. In addition, manned aircraft are of no use in a pre-emptive war. If we were to
try to coordinate missiles and bombers so that the strikes arrive on targets at the same time such tremendous universal effort by United States would be known by Russia several hours before bombers could arrive and Russia would launch their strikes against us. If we use missiles first, the landing of the first missile would cause all bombers and missiles in Russia to be launched and our bombers arriving hours later would be bombing empty sites and empty airfields. It is surprising that the Air Force is sponsoring a preemptive strike …28
Burke was worried that the size of the target list “will determine force levels in the future and also will have a great deal of impact on the types of weapons systems which will be procured.” He complained that raw intelligence was being “cranked” into the plan without having been evaluated by experts. As the Navy viewed SAC as not properly using its computers, Burke, echoing Twining, called for war gaming by WSEG to offset SAC’s technical incompetence. He signed that memo with a handwritten note, “None of the above has anything to do with ‘Merry Christmas’ which I hope you all have. Arleigh.”29
In the last days of Eisenhower’s administration, his science advisor reiterated many of Burke’s criticisms of SIOP-62. George Kistiakowsky concluded that it “follow[s] rather closely the earlier War Plans of SAC,” rather than limiting the potential for nuclear holocaust. He observed that better mathematical procedures were needed to optimize targeting, because the planners were prioritizing targets based more upon human judgment, than computer output and, therefore, erring on the side of overkill (and larger budgets). Kistiakowsky noted,
