Straight from the Planning Department of Hell
Last Updated on Monday, 16 April 2012 20:10
STRAIGHT FROM THE PLANNING DEPARTMENT OF HELL...
MAXIMISING DECISION-MAKING TIME: -LOWERING OPERATIONAL READINESS – A MINUTE BY ANY OTHER NAME IS JUST AS SHORT
In writing this, I have always had at the back of my mind a statement made by Ambassador Labbe of Chile, at an IPI seminar in which Steven Starr and I debated former ambassador Chris Ford. Ambassador Labbe said at one point
'Gentlemen, we sound like the planning department of Hell'
He certainly had a point. In writing this material I have been acutely conscious that what I am writing about is an event sequence that comes right from the bowels of hell, and that flip terms like 'vaporised' or 'incinerated', or 'counter-value strike' and the cynicism and black humour with which even disarmament advocates protect ourselves, in fact are a thin cover for what is more properly expressed in Greek tragedy. The event sequences contemplated below ought properly to evoke pity and terror, and the absolute determination that none of this is ever going to take place
Nuclear weapons operational readiness has been a prominent issue in UN and NPT meetings since the Canberra Commission in 1996 recommended that the operational status of nuclear weapon systems should be decreased – recommendations that were reiterated much more clearly in 2006 by the Blix Commission, and with great emphasis and at great length, by Gareth Evans and Yoriko Kawaguchi of the ICNND. Lowering operational readiness was part of the '13 points' of the Year 2000 NPT Review Conference Final Declaration, and figured rather more prominently in the 2010 final declaration, in which the interest of non-nuclear-weapon states in the nuclear posture of nuclear weapon states was strongly asserted. One might have wished for a more prominent, more clear, and more unequivocal call for the NWS to lower operational readiness, particularly the silo-based ICBM forces of the US and Russia, however. (See my memo circulated to delegates to this conference.)
Lowering operational readiness has also been the subject of a number of GA resolutions, notably India's Reducing Nuclear Dangers, which deserves much wider support than it achieves, and of course the widely supported Chile/Malaysia/New Zealand/Nigeria/Switzerland resolution on operational readiness, last adopted by a respectable 157-3. It would be helpful if at the next October in the GA it could do better than that.
Operational Readiness also forms part of the very widely supported Japan-Australia resolution, the NAM resolution, and one or two others.
All of this diplomatic pressure at the UN and NPT conference level has however resulted in surprisingly little willingness in Washington and Moscow to contemplate taking the roughly 1/3rd of their warheads (primarily silo-based ICBMs) that are maintained in a state in which they can be launched in less than 2 minutes after receipt of an order to do so, out of that state into one in which it might take hours or days to launch.
Troublingly, little progress has been made on other measures that might help to 'take the apocalypse off the agenda', such as the establishment of a joint data exchange facility, agreed since 1998 but never implemented. Though presidents Medvedev and Obama issued for the fourth time, a statement committing to it, I am not aware that further progress has been made.
This is in spite of arguments by advocates of de-alerting such as Steven Starr, Col Yarynich, Bruce Blair and myself, that the issue of de-alerting, reducing operational readiness or increasing decision-making time (here it is important that the terminology is NOT important), is literally of human survival significance. I have frequently pointed to the BAS Oct2008 article entitled 'Minimising the Risk of Human Extinction', whose checklist of consequential 'to do's' has at its very top, lowering the operational status of nuclear weapon systems, followed by getting rid of them. This expresses our position perfectly.
Even de-alerting opponent Chris Ford, in a thoughtful article entitled 'Playing for Time on the edge of the Apocalypse' recognises in those very words the true nature of the debate over de-alerting/decision-making time. Ford notes that:
“The tense environment of nuclear command-and-control decision-making is perhaps the closest analogue the real world presents to such apocalyptic scenarios in which a small group of people find themselves with the fate of their country and their world in their hands, and may have only moments in which to make an absurdly momentous decision. Yet these nuclear scenarios are not fiction. Nuclear-armed countries have faced each other in tense standoffs in the midst of broader crises, fingers poised perilously over the proverbial nuclear “button,” and painfully alert for signs that they might have to press it. On other occasions, officials have confronted warnings of incoming enemy attack, wondering if these signals are a false alarm and trying to decide what to do. On one occasion, a commander-in-chief even got to the point of opening his “nuclear briefcase” in preparation for a possible retaliatory launch in response to what initially looked like an inbound missile – but was not.”
Ford has expressed the problem in exactly the apocalyptic terms in which it should be expressed. However, Ford does seem to suggest there are some circumstances in which the button 'should' be pressed. Yet if a large-scale nuclear exchange really does mean the end of civilisation and much more, it is more than arguable that there are NO circumstances including the complete destruction of ones own country in which the button 'should' ever be pressed.
The use of a number of thousands of warheads in an exchange between the US and Russia (why are we still even talking about this?), would be not merely a disaster beyond imagination, that would take not millions but hundreds of millions of lives, essentially destroying both societies utterly, but, according to Toon, Robock and others peer-reviewed research it will have catastrophic global climatic consequences lasting for decades. The use of a sufficient number of warheads against cities could indeed, potentially threaten the continuance both of humans as a species, and most other complex land-based life forms.
Given, from the de-alerting and disarmament advocates point of view, the literally apocalyptic importance of de-alerting as a way to minimise the likelihood of nuclear war by misadventure, miscalculation or malfunction, and given the view of the overwhelming majority of governments worldwide (as expressed in GA votes) that holding thousands of nuclear warheads on high alert is an unacceptable risk, one might ask just what possible arguments there could be, for maintaining over 1000 warheads each, basically silo-based ICBMs, in a state in which they can be fired in less than two minutes?
Isn't the cold war supposed to have been over for the past 20 years, and can't we now put the apocalypse safely to bed?
If we are to believe some of the opponents to de-alerting, then seemingly the answer is no.
Their arguments are highly counterintuitive, and I believe that they are not only ultimately incorrect, but at times the more extreme ones border on the nonsensical if not the downright lunatic. However there is something in some of the more thoughtful arguments that disarmament and de-alerting advocates could well learn from.
Still, to listen to some of the less thoughtful arguments against de-alerting, it is as if, when two gunmen point loaded pistols or AK's at each other, they should remain locked in that stance for all eternity, there being no 'safe' way to climb down from it that does not invite one to shoot the other, or provoke a 're-alerting race'. In fact it is simply physically not possible for the two to remain that way for anything more than a few hours, and those who are unable to climb down will find the stance becomes terminal – for one or for both parties.
Common arguments against de-alerting are:
1)Specific difficulties and problems with particular forms of de-alerting, especially those involving separation of warheads from delivery vehicles, and SLBMs. These objections largely apply to de-alerting of forces other than land-based, silo-based ICBMs, whose alert status is in fact the most destabilising, that are kept on the highest alert, and to which the most attention should be paid. These objections are to proposals that the advocates of de-alerting would anyway regard as unnecessary – namely the de-alerting of forces that are generally not maintained on the highest alert anyway. (However, Colonel Valery points out that there are in fact no insurmountable difficulties involved.)
2) 'Nuclear Weapons are 'not really on Hair-trigger alert', and 'we can't take them off hair-trigger alert if they are not on hair-trigger alert'. The multiple safeguards in place to make sure an accidental or 'rogue' launch does not take place mean there is no hair-trigger alert. Col. Valery here comments that the mere presence of protection measures and safeguards 'does not mean there is no hair-trigger-alert. It exists really'.
3) Neither the US (maybe) nor Russia (probably incorrect) actually HAS a policy of 'launch on warning' or 'launch under attack'. 'We cannot abandon a policy we do not have'.
In connection with (2) and (3) Ford notes that:
'Debates on the subject of de-alerting have also been complicated by a largely unnecessary conflict over whether the United States and Russia presently have a “launch-on- warning” (LOW) policy. In this regard, critics and defenders of current nuclear force postures often simply talk past each other, the former saying or implying that the nuclear superpowers operate on a LOW basis – being set and likely to launch on warning from a “hair-trigger alert” – and the latter denying it. In fact, both sides are both right and wrong, and much time and energy has been devoted to talking around the real issues'.
According to Col Valery, the reality of LoW and LuA force postures whether officially acknowledged or not, 'is clearly reflected in combat instructions at command posts of nuclear forces of the United States and Russia. But of course, neither the Pentagon nor the General Staff will confirm this publicly.'
4) Most of the forms of de-alerting that are easier and more practical to actually do, and that might stand a chance of approval by military leaderships, are unverifiable or hard to verify. In fact the whole of de-alerting fairly bristles with verification problems.
5) Taking nuclear weapons off high alert (presumably the ones that are not on high alert so we can't take them off high alert, no?) would actually decrease stability by removing the deterrent function of those weapons. (Common-sense would seem to suggest this must be dead wrong. The simulation research done by Col Yarynich, Blair and others and published in Foreign Affairs, Global Science and Security, and elsewhere says this is indeed so.)
6)In the absence of reliable verification, one side or the other might secretly re-alert, and strike the other with impunity.
Even if verification could be ensured, during a crisis one side might re-alert, and immediately there would be a 're-alerting – race', as the other side tried desperately to re-alert first, trying to pre-empt the first side, who would then try to pre-pre-empt, etc etc in a spiral like that leading to WWI, but with ultimately apocalyptic results.
I guess the common-sense response to this is that having a large proportion of forces permanently ON alert does not seem better, and that the further inevitable 'generation' of forces during a crisis would anyway send exactly the same destabilising signals, with much less far to go to reach the brink.
Before we riposte to the all seven of the above, it is worth first asking, to be sure we have clarity, just what we are trying to DO, when we de-alert.
The entire problem arises because silo-based ICBMs are 'sitting' targets for the ICBMs of the 'other' side, which will first of all seek to disarm its opponent by striking those ICBMs. It is taken as given by a state which thinks it might be attacked, that if it does not launch its ICBMs at the side that is attacking it, those ICBMs will be destroyed and it will simply lose almost all of them, and thus be unable to counter – attack.
Consider the following anecdote, told by former Carter national security adviser Zbigniew Brzezinski at the Council of Foreign Relations in April2012, that illustrates, in spite of the mans cool, just how the situation puts people under pressure, and above all the compressed timeframes involved:
“..... I remember being woken up one night at 3:00 a.m. to be told by my military assistant that we are under nuclear attack. It obviously didn't happen, since we're all here. (Laughter.) There would have been... 85 million Americans and Soviets dead six hours later....
"Part of my job was to coordinate the response if something like that happened, to notify the president. I had three minutes in which to notify him. During those three minutes, I had to confirm it in a variety of ways. And then he would have four minutes to decide how to respond. And then 28 minutes later, some of us would be dead and we'd be living in a different age...
I got a message from my military assistant, a general, who simply woke me up at 3:00 a.m. at night on the red phone and said, "Sorry to wake you up. We're under nuclear attack." (Scattered laughter.) That kind of wakes you up.... And he adds 30 seconds ago, 200 Soviet missiles have been fired at the United States...
But there were subsequent confirmations and clearly within – well, within actually almost two minutes prior to me calling him on the third minute, it was clear that this was a false alarm. So I did nothing. I went back to sleep. (Laughter)"
But then came the real punch line. The interviewer asked, "And if the confirmation had been a little late, could we have had a problem?" Brzezinski's answer: "We might have had."
Thus, the inexorable 'logic' of deterrence goes, that if at 3am some morning my early warning system picks up evidence that the 'other' as launched, I must launch my missiles as soon as, (or sooner than), I can, thereby ensuring that when his missiles reach my missile silos on what was to be a disarming first strike, MY missiles are no longer there to be destroyed because they are now incinerating HIS cities, darkening the sky with their burning, and ensuring a body count in the tens to hundreds of millions and global climatic catastrophe….If that confirmation (disconfirmation, really) had come a little late.....
However this compulsion to launch quickly slides past a number of major considerations, aside from the need to be sure that an attack is what is really taking place, not just another of many operational glitches.
--Why on earth is a crisis like this happening anyway? Are threats of nuclear annihilation in any way a meaningful response to the realities of an admittedly sometimes difficult relationship between the US and Russia? Were they EVER, a meaningful response, even during the height of the cold war?
--Two thirds of my (and his) warheads are not in fact in silo-based ICBMs at all, but are in submarines or aircraft, or, in Russia, on mobile land-based launchers that are hard or impossible for the other side to target. These warheads are NOT maintained on high alert, but are entirely capable of reducing an opponent to rubble. If I really want to destroy the other half of civilisation and complete the job of inducing catastrophic climate change, then even without my silo-based ICBMs I can most probably do it. I do not actually 'NEED' to launch silo-based ICBMs immediately at all, though the pressures to do so will be considerable.
Col Valery comments apropos of this, that:
'Here I would like to remind why in our de-alerted scheme the first echelon consists of only silo single-warhead ICBMs. This is only possible variant when at the “first combat line” the opposing forces of the parties are completely symmetrical. All other variants of the first echelon (using subs and mobile ICBMs) will inevitably cause a lot of questions and speculations. The American submarine is harder to detect by opponent than Russian one. However, the destruction of one such sub immediately eliminates a large number of missiles and warheads at its board. At the same time, Americans will be forced to allocate a certain number of warheads to destroy every Russian mobile missile in the forest. There is a large asymmetry, which increases the uncertainty.
But it's not just that. Our symmetric scheme of the first echelon of only silo single-warhead ICBMs provides relatively low efficiency of the first attack between them: under this scheme, to destroy one silo you can take only one warhead (no additional ones in “hot reserve” at all). It is known that the probability of survival of the existing silo missile under impact by one warhead is equal to 0.85 - 0.90. This means that if the first echelon has, say, 100-150 silo missiles, much of them will survive first attack and participate in retaliation strike.
In short, even with a sudden attack of the enemy, first echelon provides some deterrence, not to mention the fact that there is the second echelon, which will also contribute.
I think that we should not exclude consideration of all possible schemes of construction of nuclear forces in the de-alerted status. However, if our scheme with silo ICBMs on “the front line of defense” will be effective, then why look for something even better?
--(To return to our postulated 3am alarm) It is more likely than not that this alarm, like all the other alarms so far is a false alarm, just like Brzezinski's wakeup call. It could be vertical high clouds over North Dakota that looked to a then state of the art Soviet surveillance satellite in 1983 exactly like a series of launches, or a faulty chip in Colorado in 1980 and 1981 that had the entire US nuclear forces ready to launch and the 'doomsday plane' launched thrice before the chip was traced. Such events are actually MORE likely than the real apocalypse. However the paradox is that if they are taken to be the said apocalypse, they will indeed become the said apocalypse.
So far the bacon of the whole planet has been saved by Russian Colonels, security advisers, and others down to lowly Minuteman silo commanders who kept their cool amidst wailing sirens and flashing lights. At some point, however, we may run out of heroes.
What if Brzezinski had been just slightly too quick on the draw?
The point to be noted here is that the problem of high alert and the highly compressed decision-making time it brings with it (and Brzezinski's account of his 3am call confirms just how compressed in practical reality that timeframe is, with the President being given just 4minutes to decide how to respond) -the said problem of high alert applies preeminently to land-based ICBMs, that may cease to exist 20-30 minutes after the other side has launched, unless they are immediately themselves launched. No such consideration applies to submarines or mobile ICBMs. Indeed, one possible approach to the problem is to go from a 'triad' to a 'dyad' of submarine and aircraft only forces, as do the UK and France, who have also de-alerted their SLBM forces (and seem not to have had any particular difficulty with doing so) from 'minutes' notice to fire, to 'days'.
All this is important not only because of the light it casts on the de-alerting question more broadly, but because of the light it casts on some of the specific difficulties postulated with de-alerting SLBM forces – forces that hardly need to be de-alerted anyway, and that can be kept in a 'relaxed' state of alert in the simplest way, without recourse to complex verification, and can, in an absolute emergency, be re-alerted without anyone noticing it has been done. Indeed, they could be randomly returned to alert for brief periods without notice, apart from an expectation that this will happen from time to time, as militaries like to practice this kind of thing.
These considerations also apply to the whole idea of de-alerting, which is above all an attempt to give senior decision-makers TIME to make rational decisions about, potentially, the fate of what we call civilisation, and to make the use of nuclear weapons as unlikely as possible.
'Nuclear Weapons are not really on 'Hair-trigger Alert' '(2)
No-one seriously denies that extensive safeguards exist to ensure that deranged commanders a la General Ripley cannot launch their own personal armageddon, or that 'no-one goes to nuclear war lightly'. Indeed so.
However, the sheer existence of the OPTION to launch silo-based ICBMs in minutes, coupled with the 'use them or lose them' mantra, means that not only do decisions as to their use have to be made also in minutes, but, absent a conscious, planned-for, policy decision always to 'ride out' a possible incoming attack, (remembering that it is most likely a false alarm), then the pressure will be very much on decision-makers to launch those ICBMs, whether or not the US or Russia has an 'official' policy of LoW or LuA.
Podvig notes that:
As long as the strategic forces keep the launch-on-warning option, we cannot completely rule out a sequence of events that would lead the military and political leadership to conclude that an attack is under way and to exercise this option. Even though an accident of this kind would be extremely unlikely, its probability is not zero even during peacetime. Given the truly catastrophic potential consequences of a nuclear strike, this probability should not be ignored or tolerated.
In this context it is interesting to note that the Russian/Soviet 'Perimitr' or 'Dead Hand' device that my colleague Col Yarynich had a hand in, was a second-strike device, designed to assure retaliation after a successful first strike by the US, and after an attack has not only been ridden out but after that attack has successfully incinerated military and civil leadership.
Ford concedes that some nuclear forces are indeed kept on high alert, as indeed the US Nuclear Posture Review itself does, in saying that current alert status will be maintained.
According to Chris Ford:
'It is certainly true that the United States and Russia appear to devote much energy and expense to maintaining some nuclear forces on an alert level that would permit launch in a very short period of time. In the late Cold War, some 90 percent of the U.S. land-based Minuteman missile ICBM force was said to be launch-able within three minutes of receipt of an authenticated order, while the Soviets were by some accounts estimated to be able to launch up to 80 percent of their force in an equivalent time period. The Soviets first demonstrated the ability to launch an ICBM on tactical warning in 1982, had fully deployed a satellite-based early-warning system by 1987, and by 1988 possessed (and had exercised) the technical option of launching sizeable nuclear forces on warning of an incoming attack. By 1990, according to U.S. intelligence officials, “[m]ost, if not all, Soviet ICBMs could be launched within minutes of a valid launch order.”12 Years after the head of the Soviet Strategic Rocket Forces first publicly mused about LOW in 1967, there are “strong indications” that Moscow has ensured this option.' (Yarynich comments that this is indeed so.)
The entire text of an email exchange between the Arms Control Association and the Pentagon on the subject of hair-trigger alert is given below. In the same breath, essentially, the Pentagon denies, and then reaffirms, that it keeps its ICBM forces on high alert. At one point it expressed an outright rejection of de-alerting based on precisely the flawed arguments I am now refuting.
Q1) If "hair-trigger alert" and "launch-on-warning" are incorrect terms for describing the status of US forces on alert, what is the proper term?
A1) U.S. nuclear forces are not on “hair trigger” alert. The term “hair trigger” ignores the safeguards, deliberate actions, and procedures required in order to employ nuclear weapons. The U.S. nuclear force posture has evolved since the end of the Cold War. Only a portion of the operationally deployed U.S. nuclear forces is maintained on a ready alert status. No strategic bombers, 450 Minuteman III IBCMs, and a small number of SSBNs at sea are on alert at any given time.
U.S. policy is not to rely on a "launch on warning" strategy. U.S. strategic forces are postured to provide maximum flexibility so the U.S. is not faced with a “use or lose” dilemma. A major strike on the U.S. would be required to eliminate the responsive ICBM capability. The ICBM force could be launched prior to impact, but only if the President were to direct such an action. In addition, should the ICBM force not be able to respond, the U.S. SSBNs at sea could deliver an overwhelming response if directed by the President.
Should the international security situation call for it, the U.S. could bring its nuclear forces to a higher state of readiness (i.e., “generated alert”), putting a larger portion of its submarines to sea and returning heavy bombers to alert, to increase their survivability.
Q2) The US government has made statements to the effect of as long as nuclear weapons exist it is necessary for us to keep some portion of our forces at some level of alert. What is the proper description or term for that "level of alert?"
A2) See answer 1. A portion of U.S. nuclear forces are on day-to-day alert. Since the end of the Cold War, however, the U.S. has reduced dramatically both the overall number of nuclear weapons and nuclear systems maintained on day-to-day alert. To ensure deterrence, U.S. nuclear forces must be postured such that, under any credible scenario, a sufficient number of nuclear weapons would survive to respond to an aggressor’s attempt to carry out a disabling attack on the U.S. The proper term would be “on day-to-day alert”.
Q3) The US government statement also noted that US forces have evolved away from "rapid reaction high alert levels." Is that the proper term to describe the alert status of some US nuclear weapons systems today?
A3) The U.S. nuclear force posture has evolved since the end of the Cold War. Only a small portion of our SSBNs and the 450 Minuteman III ICBMs remain on day-to-day alert. Nuclear capable bombers have been removed from alert status, but could be re-postured in a national crisis and additional SSBNs could be generated to alert status.
Q4) Has the US completely stopped this previous practice of "rapid reaction high alert?"
A4) See answer 3.
Q5) What measures marked this shift? What steps were taken that no longer classifies or makes US weapons as on "rapid reaction high alert?"
A5) Under the 1991 Presidential Nuclear Initiatives, all nuclear bombers were removed from alert. Only the Minuteman III ICBM and a small portion of our SSBN force remain on day-to-day alert and neither force is targeted against any country. All 50 Peacekeeper ICBMs, 50 Minutemen III ICBMs, the B-1 Bomber fleet, and four SSBNs have been deactivated or removed from strategic service.
Q6) The USG statement further said that "few of the operationally deployed US nuclear forces are maintained on a ready alert status." What is meant by the term few?
A6) Only the Minuteman III ICBM force and a small number of SSBNs are on day-to-day alert.
Q7) Independent nongovernmental analysts say that regardless of what the alert status is called, the reality is that some US nuclear weapons are capable of being fired in "minutes." Is that assertion accurate?
A7) The United States maintains the ability to launch its nuclear weapons in a timely basis as directed by the President. Minuteman III ICBMs are designed to be capable of delivering a rapid response prior to being struck by an adversary’s ballistic missile force. This is an important aspect of our deterrent because it complicates an opponents’ pre-emptive strike planning. However, the fundamental fact is that U.S. forces are postured such that the President is not confronted with a “use or lose” situation in that other strategic forces could be directed to respond to an attack. See answer 1.
Q8) In a Nov. 6 paper, nongovernmental analyst Bruce Blair wrote, "the fact remains that the US posture is still geared for firing thousands of weapons with a few minutes." Is that an accurate statement?
A8) No, this is not true. Under the Moscow Treaty, the U.S. will have only 1700-2200 operationally deployed nuclear weapons. The U.S. is well on its way to achieving this limit. Only a portion of these are on day-to-day alert.
Q9) Now that the Soviet Union is gone and the United States says Russia is no longer an enemy, why is it necessary for the US to keep some of its forces on alert for possible launch in minutes?
A9) The security environment of the 21st century is dramatically different from the East-West rivalry of the Cold War era, but the goals of U.S. security policy remain much the same: to strengthen deterrence and limit risks that could result in serious -even catastrophic- damage to the United States, its allies, and friends. Nuclear capabilities continue to play an important role by providing options to deter a wide range of threats, including the use of WMD by a variety of adversaries. These capabilities also contribute to our non-proliferation goals by assuring allies and friends that the U.S. will be able to fulfil its security commitments, thereby negating any need to develop their own nuclear weapons.
“De-alerting” strategic forces raises other unique concerns, related to the safety and the credibility of the deterrent force. Additionally, the generation of nuclear forces during a crisis, when none had been on alert, could cause an already tense situation to be come unstable.
Q10) What measures does the US have in place to prevent nuclear weapons from unintentional or accidental use?
A10) There are multiple, rigorous technical and procedural safeguards to protect against accidental or unauthorised launch.
These safeguards include positive measures such as weapon design features, safety rules, procedures (including two-man rule), accident prevention or mitigation measures, and other controls. Such controls include physical security and coded control systems, which are used collectively or individually to enhance safety and to reduce the likelihood, severity, or consequences of an accident, unauthorised actions, or deliberate destructive actions.
Col Yarynich comments on this email exchange:
'Yes, these measures [safeguards] are very reliable. But it absolutely does not give the right to ignore the possibility (albeit very unlikely), the breakthrough of these protective barriers, which would have catastrophic consequences. These two things are so disparate that they can not be compared on the scales, in principle. One scale (severe consequences of a war) will always outweigh another one (a small probability of such event).'
The bottom line of all of this email exchange for all the Pentagons beating around the bush, is that, precisely as Bruce Blair says, the ICBM force is on 'day-to-day alert', which Blair and Col Valery characterise as 'hair-trigger'. And this does amount to over 1000 warheads or did until very recently. The response to Q(8) is thus at best evasive or misleading, (and according to Valery just plain wrong). However it is possible to see an admission of the truth – that the ICBM force and a 'small portion' of the SLBM force are on 'day-to-day alert'. And this has been the case for over a decade. Nothing is new here.
Statements of this kind that nuclear forces 'are not on hair-trigger alert' simply slide past the fact that in fact, both in the US and Russia, land-based ICBM forces are indeed maintained on high alert, able to be launched within roughly two minutes of an order to do so, as indeed confirmed by the very same email Q&A.
Given that these very forces are the likely targets of any incoming strike, that very fact means that decisions as to their use MUST be taken in ridiculously compressed timeframes, and that massive pressure will exist to launch them, whether they are said to be on 'hair-trigger' alert or on 'day-to-day' alert. Brzezinski's account of his 3am wakeup call indicates clearly that in fact decision-making times are exactly as compressed as Blair says they are, and is fully consistent with Blairs account of the sequence following a missile alert. The following sentence from the email exchange between ACA and the Pentagon IS consistent with Brzezinski's 3am wakeup call:
'The United States maintains the ability to launch its nuclear weapons in a timely basis as directed by the President. Minuteman III ICBMs are designed to be capable of delivering a rapid response prior to being struck by an adversary’s ballistic missile force'.
However, the following sentence does NOT seem consistent with Brzezinski's statement that President Carter would have had four minutes to decide what to do:
'….the fundamental fact is that U.S. forces are postured such that the President is not confronted with a “use or lose” situation in that other strategic forces could be directed to respond to an attack.'
This would seem to be a statement of how forces SHOULD be postured, but would not appear from Brzezinski's account, to have been the operational reality. If it were, the President would have had more than four minutes to decide what to do. Furthermore there is no indication that force postures or combat instructions to missile forces have fundamentally changed over the 30 years or so since the incident. The Pentagon statements that forces are postured to AVOID the situation Brzezinski found himself in thus lack substance.
Verification is indeed a potential difficulty, but it is one that cuts both ways. Pavel Podvig argues forcefully that de-alerting measures are actually best done WITHOUT verification and that verification is in fact destabilising.
According to Podvig:
'To address the problem of an accidental launch in the short term, the United States and Russia, while continuing to work toward deep reductions of their strategic nuclear forces, should develop and implement measures that would keep their entire forces at low levels of readiness without revealing their actual alert status.'
Steve Starr and myself would not go quite as far as Pavel, but he does have a very strong point. Podvig argues that especially in the face of fears and suspicions about potential re-alerting races, verification, far from being confidence-building, is entirely unhelpful. Incorrectly interpreted verification could in fact LEAD to destabilising re-alerting races, in which one side believes the other side has re-alerted and themselves re-alerts. In any case, many of the most practical measures of de-alerting, such as flipping 'safing' switches in silos, are not practically verifiable. Still LESS verifiable would be the deep changes in war plans that are really needed to ensure adequate decision-making time.
Col Valery comments with respect to Pavel that:
'Difficulties in the verification, of course, exist, but you can not abandon them. Pavel would have been right if the verification was the only way to solve the problem. But we offer a second way: open assessment of mutual threats and risks through the joint modeling. If you take this way, then the verification will be superfluous addition.'
Podvig notes that:
'One example of the contradiction between risk reduction and transparency is de-alerting. It is generally assumed that if weapons are taken off alert, it should be done in a transparent and verifiable manner. This, however, creates a possibility of a dangerous re-alerting race during a crisis, which in some circumstances can create an instability that might be worse that the one de-alerting was supposed to stave off.'
'It is important to note, however, that neither transparency nor verifiability plays any role in reducing the risk of an unintended launch. If one side takes its missiles off alert to prevent them from being launched by an accident, it will not be able to launch them regardless of whether the other side can see status of the missiles or verify the fact that they have been de-alerted. In short, the task of reducing the risk of an accident is quite separate from the one of confidence-building and it should be treated as such. Decoupling of these two tasks would remove the most serious objections to de-alerting—that it creates additional instabilities and that it is cumbersome and hard to implement.'
We see that on the matter of verification, that different points of view amongst de-alerters exist. This very fact however, problematises arguments that difficulties in verification are therefore arguments as to the 'impracticality' of de-alerting. In the light of Podvig's arguments that verification is anyway destabilising rather than stabilising, arguments against particular de-alerting measures on the grounds that they are unverifiable or not easily verifiable, should cut no ice whatsoever. Finally, as Col Valery points out, given a modicum (just very little) of mutual good will and trust, many if not all of these purportedly insurmountable difficulties dissolve.
Re-Alerting Races and Surprise Attacks – Devious malice versus chaos
(5)(6)and(7) all canvass the possibility that it might at some point seem rational for either the US or Russia, to launch a surprise attack on the other side, having either secretly re-alerted, failed to truly de-alert, or believing they can re-alert so much more quickly than the other that they could mount a successful disarming first strike.
The essence of this argument is that it is paradoxically de-stabilising to de-alert, because, - during a crisis-, (one might fairly ask, in the current US Russia context, just what kind of crisis might promote such a response) - one side or another might secretly re-alert, prompting either a mutual re-alerting race, or possibly, a (pre-emptive??) first strike by one against the other.
(Again, one wants to ask just what on earth would prompt such a move, from either side, in current, real-world politics? Is there any cause or interest at all that is worth risking the destruction of the US and Russia, and possibly of everyone and everything, for? What possible rational calculus of interests would lead that way? Surely, none.)
In fact, just as we might expect, and notwithstanding the extremely dangerous arguments of Lieber and Press that the US has some kind of 'nuclear primacy', and could get away with a first strike, it is NEVER rational for either side to attempt a so-called 'splendid first strike' against the other whose weapons are not on alert, simply because the likelihood of completely disarming the other is so low that there are just no circumstances where it can be worth it: It can be effectively discounted as a possibility, even when the attacking state has an 'effective' ABM system.
The number of warheads needed to destroy either the US or Russia as functioning societies is relatively low – even a single warhead exploded in space above the continental US is enough to destroy all communication, electronic, and even electrical systems, thereby taking it to a pre-industrial age. If even a tiny fraction of the strategic forces of either side survives a first strike, it can STILL reduce the other side either to rubble, or to a non-organised, preindustrial society. In the perverse discourse of deterrence theory, this means that 'deterrence' will always be preserved, in that it will never be rational for one side to try to destroy the other even if one side has successfully and secretly re-alerted and the other side is completely unaware of this.
In even articulating a scenario in which both sides de-alert and one side then secretly re-alerts, and attacks the other out of as it were, nowhere, I am acutely aware of the utter unreality of such a scenario. I would challenge the rationality of any strategic planner who said they could honestly envisage any such event sequence.
However it is clear enough even from these preliminary considerations, that the thesis that re-alerting races and/or surprise attacks out of nowhere are not arguments against taking precisely the measures that are needed to make the use of nuclear weapons less, not more, likely. Whether we call that de-alerting, lowering operational readiness or giving decision makers more time to deliberate is purely semantic.
There is a still more profound reason as to why arguments from devious malice are implausible. It is simply that the most powerful motivation for one side attacking the other is that one side believes, probably falsely, that the other side is about to attack them.
By far the greatest risk of itchy fingers on real or metaphorical nuclear triggers, is the belief that the other side has itchy fingers, so I pre-empt them, and they pre-pre-empt me, and so on.
To use Chris Fords terminology it is precisely the risks posed by INADVERTENCE and miscalculation that are the real risks. The risks posed by advertence and rational calculation simply make no sense at all, for the reasons outlined above: An attack is simply never rational.
So if we want to make an apocalypse less likely, we must make the risks of an accidental apocalypse less likely, because there is no other credible apocalypse. In Chris Fords terms I am saying that what he calls 'typeB risks' (inadvertence) are real, and that type A risks (advertence) are not credible. However, the type B risks are all TOO credible. Even in a time of high tension over, say, NATO missile defence installations next to Russia's and border, and in the light of Medvedev's press-released orders to Russian military units to target those installations, I do not find a pre-planned surprise attack by anyone remotely credible.(Not even under a Romney presidency, ghastly as I find that to contemplate.) I do find a stumbling into hostilities a la August1914 possible. (And the big lesson of 1914 which Ford cites is that 1914 was war by inadvertence, not by advertence.) .
The final question that resolves everything is this: Even if one can, with some effort, envisage situations in which de-alerting nuclear forces could be awkward or inconvenient, and even might entail some risks, are we, on the whole, more likely, or less likely, to experience the actual use of US or Russian nuclear forces, if we take them off high alert, or take other measures to positively ensure that presidents and senior military do not have to make decisions about the use of silo-based ICBMS under intense time pressure? Are we more likely, or less likely, to see the use of nuclear weapons that can be launched in minutes, or in days?
The answer would seem, to Blind Freddie's common-sense, to be pretty obvious. Even if potential nasties like re-alerting races still lurk on the sidelines, the decision-making time that a lower alert rate and a considered, deliberate, refusal of LoW and/or LuA buys us, and the 'firebreak' it creates, are literally world-saving.
In fact, the statistical analysis of this done by Colonel Yarynich, Generals Esin and Zolotarev, and Blair and Mc Kinzie shows this is indeed so, with quantitative rigour.
In their analysis, entitled '100 nuclear wars', Yarynich, Blair, Esin, Zolotarev and McKinzie assume that nuclear forces are de-alerted in two 'echelons', one that can be launched in hours,(silo-based ICBMs), and one that can be launched in days.(mobile ICBMs, SLBMs, and bombers).
In fact the important part of the analysis is that which concerns the land-based ICBMs.
'100 Nuclear Wars', in which the entire project is outlined (Science and Global Security Aug 2011), notes that:
'Two fundamental criticisms of nuclear de-alerting have been advanced: (1) strategic nuclear forces off alert are vulnerable to a disarming first strike; and (2) a future crisis between the United States and Russia would grow dangerously unstable as missile forces race to return to launch-ready status. De-alerting could, the argument goes, create exploitable advantages from breaking out and re-alerting. This criticism slides past the fact that the current nuclear postures are fully geared to generate two-thirds of their arsenals as rapidly as possible during a U.S.-Russian confrontation, and to launch them preemptively or on warning. Our de-alerting scheme in fact suppresses such re-alerting impulses. The solution to a stable nuclear deterrent with all forces off alert is to divide the nuclear forces of both countries into distinct groups, termed Echelons, with different degrees of reduced combat readiness (i.e., different generation times to launch-ready status). By “echeloning” the forces, our model constructs a stable nuclear deterrent whole from more vulnerable, de-alerted parts. The partitioning of nuclear forces into a First and Second Echelon serves both as a barrier to surprise nuclear attack and to a re-alerting race.'
As Col Yarynich (to whom I suspect this entire analysis owes its existence) is sitting next to me, I won't try to duplicate his detailed understanding of the quantitative aspects of this. My own grasp of that mathematical side is tentative indeed.
What is important from the de-alerting and decision-time question is that even if it is assumed for reasons that truly defy comprehension and have no foundation in current political reality, that, having de-alerted, one side successfully 'cheats', re-alerts its forces, and carries out a first strike with its silo-based ICBMs on the silo-based ICBMs of the other side, it will STILL always suffer unacceptable devastation and the destruction of its society, and will know this before it starts. This picture is not changed even if the attacking side has a 'highly capable' missile defence system, technologically implausible as that is.
This in turn means that it is NOT the case that de-alerted forces would cease to deter. In fact, the real deterrent forces properly speaking even as things now stand, are forces that are NOT kept on high alert – largely SLBMs, and in Russia, mobile ICBMs. These relatively invulnerable forces are still able to inflict utterly apocalyptic devastation on an adversary if that must really be ones objective. Indeed it will be THEIR use – the use of these second-strike forces – that administers the coup de grace to civilisation and much else. Whether deterrence of this kind SHOULD ever be a rational objective, or whether it even works at all, (it will not work for war by inadvertence because inadvertence is by definition non-rational), is questionable to put it mildly. Our very concern with war by inadvertence, miscalculation, malfunction or misadventure should say that, somewhere, sometime it will fail not because some crazed national leader with lunatic objectives was undeterrable, but because amidst wailing sirens, and flashing lights, with complete panic all around, and false data, someone simply made an unfortunate choice.
As Brzezinski noted, 'if the confirmation had been a little late....we could have had a problem.'
Finally, a note on the issue of maximising decision-making time.
Both the excellent EWI analysis of operational readiness (Reframing De-Alert), and Chris Ford in his thoughtful paper for the Hudson Institute, point to the centrality of actual decision-making time in the de-alerting debate.
It should be noted that central as decision-making time is, de-alerting is not solely about that issue: De-Alerting is also about decreasing the role played by nuclear forces in security planning, and thus very much about paving the way to global zero. Disarmament advocates have no need to be apologetic about seeing both global zero and lowered operational readiness as human survival-related issues, with one leading to the other. (and with global zero as an oft reaffirmed ArtVI NPT legal obligation.)
But it is certainly the case that increased decision-making time is utterly central to what de-alerting is about, and vital to the aim of preventing an accidental apocalypse. The problem however, is that it is almost impossible even logically, to disentangle 'increasing decision-making time' and 'de-alerting'. There is a deep sense in which the two are not so much exactly one and the same, as different aspects of each other.(in spite of what was said above).
The utter centrality of decision-making time, and the limited time there in fact is under LoW or LuA to actually decide things (including to decide whether we are going to DO LoW/LuA at all), is shown by Bruce Blairs account of what is actually supposed to take place when incoming missiles are detected or are thought to be detected. Everything is governed by the ticking clock of a 30 minute flight time, two to five minutes for detection, a threat assessment conference, waking the President at whatever unearthly hour it may be, and eight to zero minutes to make a decision – even if that decision were to be 'wait for detonations' or 'ride it out'. There is little time to reflect on the meaning of life or to do much more than go very hastily through a checklist. The diabolical paradox is that unless I decide for rideout, there is insufficient time to determine for sure if the incoming missiles are real at all, and therefore to determine if I even 'NEED' to launch my silo-based ICBMs.
The catch is that if the missiles are not real – and so far they have never been real – then I never needed to launch. If they ARE real I cannot know that until AFTER it is too late to launch silo-based ICBMs.
(All of which suggests to me, that silo-based ICBMs are really not very useful unless I am prepared to risk the accidental destruction of pretty much everything.) The only way to be absolutely sure is to do as Colonel Stan Petrov did, and wait for an impact – or for there NOT to be an impact. Colonel Stan found it a nervous wait, all the more so because, had he been wrong, he would have been incinerated. Colonel Stan TOOK his decision-making time, knowing as he did so that he traded the risk of being wrong and being vaporised for the certainty that he was not about to initiate an (accidental) apocalypse.
However we slice it, increased decision-making time and increased certainty as to whether or not we have a false alarm, and a decreased likelihood of an accidental apocalypse, MUST mean decreased responsiveness.(i.e. a lower alert level). And to demand a more and more timely response (= a higher alert level) MUST mean decreased decision-making time, and increased risk of an accidental apocalypse. There is just no way around these mathematics, unless I explicitly abandon even thinking about LoW or LuA – which means they must cease to exist as options in my OPLAN, and must not be anyplace on my nuclear briefcase. Once they exist as options at all I must decide on them in that compressed timeframe governed by a missiles flight time.
The argument from the disarmament side is simply that large-scale use of nuclear weapons is likely to terminate civilisation and much else and that this risk is THEREFORE unacceptable and that its avoidance rightly trumps absolutely all other considerations including the highest and most dire considerations of national security, above which it must be ranked. It is potentially a human survival consideration, and to put any consideration whatsoever in front of human survival can never be rational.
That said – having gotten 'mere' human survival out of the way as it were – in terms of the relationship between alert status and decision-making time, we have simply gone in a circle. To increase decision-making time is precisely to lower alert status, because that is simply central to what lowering alert status is about. And to increase decision-making time is to FORGO certain options, namely to launch before the other fellows missiles reach my ICBMs – if indeed those missiles exist at all, and are not a glitch somewhere.
What is also true is that a decision that I need to take more time, come what may, to take decisions on retaliation (is retaliation ever rational? So-called Deterrence has now already failed if my EWS is indicating incoming missiles.) - does indeed require re-designing many things ranging from basing modes to everyday procedures, to the OPLAN. Ford and the NPR are quite right to point to a need for more survivable basing modes (which might mean abandoning silo-based ICBMs altogether) as well as more survivable C3I.
There is a ray of hope in that the NPR said there was a need to investigate or study these issues. One must hope above all however, that this isn't just a way of fudging the issue of operational readiness/decision-making time. The NPR Seemed to recognise a real need to move on the issue of decision-making time. It is indeed vital that ways are found, to give decision-makers in the US and Russia more time to make decisions that really do, at least potentially, bear on short-term human survival.
A Kind of Conclusion
--Accidental nuclear war (nuclear war by inadvertence) remains possibly the single most important short term threat to human species survival, and to what we call 'civilisation', and much else including most complex – land-based living things.
--War by inadvertence is considerably more probable than war by advertence. Miscalculation, malfunction and misadventure are far more likely to undo us as a species than devious malice.
--Most of the measures that address war by misadventure and miscalculation do in fact also feed back into longer-term relationships between the parties, making war by advertence also less likely.
--The measures that do most to make war by inadvertence less likely are those that increase decision-making time. However, the nexus between increasing decision-making time and lowering operational readiness is tighter than it looks at first, and meaningfully disentangling the two is not really possible.
--Increasing decision-making time will involve deep changes to overall operating procedures and may also require changes (such as increased survivability) in C3I infrastructure. It may also require moving away from silo-based ICBMs, or other changes in basing modes to allow greater survivability and remove the pressure for a quick response.
--A variety of other measures will also be helpful and these include the establishment of the Joint Data Exchange Centre (JDEC). However, far too little progress has been made on these simple and commonsense measures.
I wish to acknowledge the editorial help of Professor Peter King of Sydney University Centre for Peace and Conflict Studies, and Col Valery Yarynich, formerly of the Soviet Missile Forces.
People for Nuclear Disarmament NSW