The winter-safe deterrence debate

In a recent opinion column for the Bulletin, "Deterrence, without nuclear winter," Seth Baum argued that the biggest danger posed by world nuclear arsenals is a nuclear winter that could be sparked by even a limited exchange of nuclear weapons. Baum's piece went on to suggest that "the world’s biggest nuclear powers [might] meet their deterrence needs without keeping the large nuclear arsenals they maintain today. They could practice a winter-safe deterrence, which would rely on weapons that pose no significant risk of nuclear winter."

In his column, Baum, executive director of the Global Catastrophic Risk Institute, proposed that alternatives to nuclear weapons be explored, in terms of their ability to provide the deterrence now assigned to large nuclear arsenals. "Exploring 'good' options for threatening large destruction is a peculiar and regrettable task in which no civilized person should take any joy," Baum wrote. "But if doing so can save many lives, and indeed save civilization itself, then it should be done. The two weapons that stand out are non-contagious biological weapons and nuclear electromagnetic pulse. The former could work well if deterrence requires threatening large human populations. The latter could work well when deterrence requires threatening large amounts of infrastructure. In both cases these are tentative conclusions, backed only by my limited, preliminary study. Governments should not move forward with either weapon without more careful examination."

Baum's column and the study from which it draws, "Winter-safe Deterrence: The Risk of Nuclear Winter and Its Challenge to Deterrence," published in the journal Contemporary Security Policy, have been vigorously disputed in social media. In this roundtable, security experts Gregory Koblentz, Martin Furmanski, Brett Edwards, Gigi Kwik Gronvall, and Sonia Ben Ouagrham-Gormley and Baum debate his column and winter-safe deterrence ideas in more depth.

Round 1

On winter-safe deterrence and biological weapons

The status quo of large nuclear arsenals risks global catastrophe so severe that human civilization may never recover. If we care about the survival of human civilization, then we should seek solutions to make sure such a catastrophe never occurs. This is my starting point for exploring the potential for winter-safe deterrence. Perhaps, upon closer inspection, winter-safe deterrence will prove infeasible. But I do believe it is worth closer inspection. To that effect, I welcome this roundtable discussion and the broader conversation that my research has sparked.

I have defined winter-safe deterrence as military force capable of meeting the deterrence goals of today’s nuclear-armed states without risking catastrophic nuclear winter. Winter-safe deterrence recognizes two basic issues: first, large nuclear arsenals pose a devastating catastrophic risk; and second, nuclear-armed states may refuse to relinquish almost the entirety of their nuclear arsenals unless their deterrence goals are met through other means. Winter-safe deterrence thus aims to make the world safer given the politics of deterrence.

Personally, I would prefer that today’s nuclear-armed states simply decide that they no longer need deterrence based on the threat of massive destruction. Such destruction is immoral and against the spirit of international humanitarian law. To that effect, we should seek solutions that reduce the demand for this sort of deterrence, for example by improving relations between nuclear-armed states or by stigmatizing this deterrence. The ongoing initiative on the humanitarian impacts of nuclear weapons is enhancing precisely this stigma. I am proud to support the initiative. It is a first-best solution to nuclear war risk.

Winter-safe deterrence is a second-best solution because it avoids global catastrophe while leaving intact deterrence based on massive destruction. It is no surprise that winter-safe deterrence is controversial, because it forces us to consider which devastating weapons we would like militaries to have. This is a repugnant task. I assure you I take no joy in it. But if this task could prevent a major global catastrophe, then it is an important task to pursue.

In my research, I attempted an initial analysis of winter-safe deterrence. I do not claim to have any final answers, and I explicitly discourage governments from pursuing winter-safe deterrence until more thorough analysis has been conducted. The issues are just too complex for one research paper. While I could have done more research on my own before publishing, I decided that progress could be better made through open discussion. That said, I was cognizant of the risks that mere publication could pose, to the point that I even considered censoring myself. I am thus sympathetic to Brett Edwards’s thoughtful comments on the risks of discussing these ideas. He raises important issues on when international security policy research is too risky to publish. In this instance, I believe it would have been too risky not to publish the research, risking the possibility that the status quo would lead to global catastrophe.

My research considered a sizable list of weapons: small nuclear arsenals, conventional military force, conventional prompt global strike, neutron bombs, biological and chemical weapons, cyber weapons, and electromagnetic weapons. Perhaps there are other weapons also worth considering as replacements for deterrence with large nuclear arsenals; I would welcome any suggestions. I further considered these weapons for two types of destruction: destruction to infrastructure and destruction to human bodies. For humanitarian reasons, I hope that deterrence can succeed with threats to infrastructure alone. I do not know if it can, and would welcome any thoughts on that score, also.

Much of the conversation about winter-safe deterrence has focused on a hypothetical decision between deterrence with nuclear weapons and deterrence with non-contagious biological weapons (NCBWs). I regret that the conversation has been framed in this way. Winter-safe deterrence does not require picking one weapon over another. To the contrary, it can be achieved via combinations of weapons, including small nuclear arsenals. But my suggestion of some constructive role for NCBWs has been controversial, and I am glad for this to be debated.

Gregory Koblentz, Martin Furmanski, and Sonia Ben Ouagrham-Gormley all argue that NCBWs are not effective deterrents because their effects are too unpredictable and too easy to defend against, among other reasons. In short, they cannot reliably cause enough harm in counterattack. These points are well taken. The biggest question they raise in my mind is whether these technological shortcomings of NCBWs may be altered by technological advance, making them more suited for deterrence. Given advances in biotechnology, I hesitate to place bounds on what may be possible. I would be interested in further comments on this.

Ben Ouagrham-Gormley says something that catches my eye: “Non-contagious does not mean less lethal.” I am surprised to read this. It is my understanding that NCBWs can only kill people within the area in which they are released, whereas contagious biological weapons can spread, potentially even killing people worldwide. Could NCBWs spread as widely? If not, then how could NCBWs be as lethal?

Another point I am left unclear on is the proliferation risk posed by NCBWs. On one hand, Gigi Kwik Gronvall writes that “biological weapons development is within the technical capability of most countries,” suggesting a large proliferation risk. On the other hand, Ben Ouagrham-Gormley describes how countries and nonstate actors have repeatedly tried and failed to develop biological weapons, suggesting a small proliferation risk, or perhaps even none at all. I am unsure how to resolve this apparent conflict.

Gronvall’s comments on the international bans on biological weapons raise a difficult issue. If NCBW deterrence were technologically feasible, pursuing it could significantly harm the international treaty system. However, it could also help avoid global catastrophe from nuclear war. I hope we never have to make such a difficult choice, but if I had to choose, I would choose to avoid global catastrophe. That is, unless harming the international treaty system could itself cause global catastrophe (and perhaps it could), I would favor accepting this harm to avoid global catastrophe from nuclear war. When it comes to humanity’s long-term survival, the stakes are simply too high not to make hard choices.

But we should hope that this is a false choice. We should hope that other options exist such that we can avoid global catastrophe from nuclear war without harm to the international treaty system.

In closing, I thank the contributors to this roundtable for their thoughtful and constructive comments and look forward to continuing the discussion.

Bioweapons not an alternative to nuclear weapons

In his article “Deterrence, without nuclear winter,” Seth Baum argues that the greatest danger associated with current nuclear stockpiles is that nuclear winter will inevitably follow a nuclear strike and “threaten the whole of humanity.” To avoid such a catastrophic outcome, and thus promote drastic reductions in nuclear weapons, Baum proposes that nuclear states consider alternative weapons that would at once meet their deterrence objectives and preserve humanity from the devastating effects of a nuclear winter. Baum identifies five qualities that such ideal alternative weapons should possess: They should be “politically acceptable,” “not pose a significant proliferation risk,” not “destabilize the international system,” and possess a deterrent value while being “affordable and technologically feasible.” Baum concludes that “non-contagious biological weapons” possess such qualities and would constitute a “good” alternative to nuclear weapons.

I suspect that most nonproliferation analysts—particularly bioweapons experts— have found this proposition to be shortsighted. Baum’s suggestion simply redounds to advocating the violation of the Biological Weapons Convention (BWC)—a nonproliferation treaty that bans the development, production, and stockpiling of biological weapons. The article also ignores historical realities that are important to highlight because they demonstrate that biological weapons possess none of the attributes that Baum lists as essential for an alternative to nuclear weapons.

Among the nine nuclear weapons states—the United States, the United Kingdom, Russia, France, China, India, Pakistan, North Korea, and Israel—all but one (Israel, which has not openly declared its nuclear state status) have signed the BWC. Under the treaty, states agreed to destroy their bioweapons arsenals and refrain from developing, producing, and stockpiling bioweapons. Therefore Baum’s suggestions to promote nuclear disarmament by nudging nuclear states toward bioweapons development is akin to erasing over 40 years of international negotiations and achievements, the greatest of which was the development of an international norm against the development and use of biological weapons. In addition, if powerful states such as Russia and the United States were to resume their past bioweapons programs, they would not only place other nations at a disadvantage due to their lack of accumulated expertise in the field, but this would also have a chilling effect on all types of nonproliferation efforts, including nuclear disarmament. Why would North Korea renounce nuclear weapons when they could use them to counter the nuclear or bioweapons threat from more powerful states? Baum’s proposal is neither politically acceptable nor strategically advisable, while it also carries significant proliferation risks.

In his article, Baum puts the emphasis on “non-contagious biological weapons,” presumably because if these weapons cause non-contagious diseases, their effects would be more limited, making them somehow more acceptable. But the BWC makes no distinction between contagious and non-contagious bioweapons. All bioweapons are banned under the treaty. None is more acceptable than the other. Besides, historically most state and non-state bioweapons programs have focused on agents that do not cause contagious diseases, such as Bacillus anthracis (anthrax) and botulinum toxin. Very few programs have actually worked on contagious agents such as the smallpox virus. Non-contagious does not mean less lethal.

Further, biological weapons are poor deterrents. In addition to the fact that states can protect their populations beforehand, and in some cases even after an attack, depending on the agent used and the length of the incubation period, biological weapons do not guarantee a certain level of predictable destruction as nuclear weapons do. They are sensitive to light and weather conditions, making their use less effective during the day, when ultraviolet light can destroy the agents, or in windy conditions, when agents can inadvertently disseminate over unintended populations. Changes in air currents within cities and over variable terrain make it difficult to achieve accurate targeting. Their effects can also be mitigated by individual immune responses. Because of their sensitivity to environmental conditions, biological weapons are also difficult to deploy, making their use for a second strike problematic.

Historically, because of their unpredictability and the uncertainty of their outcomes, biological weapons were never fully integrated in US military doctrine. In the Soviet Union, the military also viewed bioweapons as unreliable, due to the difficulty of controlling their dispersion and the possible risks of infecting Soviet troops. Attempts by Soviet scientists to develop new biological agents, unknown in nature, which the enemy could not protect against, lasted about 20 years and did not go beyond the exploratory and research and development phases. Furthermore, both the United States and Soviet Union struggled to design effective missile systems for the delivery of these weapons. The US program did not produce a dedicated bioweapons cruise missile, and the Soviets could not solve technical difficulties associated with the development of a ballistic missile warhead that could effectively protect fragile bioagents against the stress of re-entry.

These challenges underscore my final point: Bioweapons are technically complex and can be very expensive. Their complexity derives from their reliance on living microorganisms that are highly sensitive to their environmental and handling conditions. Their behavior throughout the development process is unpredictable, which—as past state and terrorist programs show—causes many failures and delays in program advancement. Furthermore, the stages of a bioweapons lifecycle—research, development, production, scale-up, weaponization, and testing—are highly interdependent, and the successful passage from one stage to the next requires organizational and managerial conditions that promote coordination, cooperation, and information exchange among the various teams of experts involved. Such conditions are particularly difficult to achieve. As a result, most past state and non-state bioweapons programs have been unsuccessful at reaching their goals. The Soviet Union, which had the longest-running and largest program, spent an estimated $35 billion over the last 20 years of the program’s life but was unable to develop the new agents that it aimed to produce. The US program cost about $700 million over 27 years without producing a weapon that satisfied military requirements. Iraq spent over $80 million during the last five years of its program but was only able to produce ineffective weapons that would have destroyed most of the agents they contained upon impact. Finally, South Africa and the Japanese terrorist group Aum Shinrikyo spent about $30 million and $10 million respectively, but failed at all stages in a futile quest for bioweapons. Thus, bioweapons are hardly affordable, and their technical feasibility is constrained by a host of factors, of which I have mentioned only a few.

In short, bioweapons are not an alternative to nuclear weapons. Proposing to sacrifice the biological nonproliferation regime to ensure nuclear disarmament is simply shortsighted.

Bringing a knife to a gunfight: biological weapons as deterrents in a nuclear-armed world

Seth Baum’s recent column, which called for legitimizing non-contagious biological weapons to replace the deterrent function of nuclear weapons, is problematic for me on a number of levels, but in this essay I will address only one: Would biological weapons be effective as weapons of deterrence, as potential replacements for some of the existing or future nuclear weapons?

The terrible destructive power of nuclear weapons is obvious: Hiroshima and Nagasaki demonstrated the power of first-generation fission warheads, and then, in the mid-20th century, atmospheric testing showed the even more awesome potential of augmented fission and true thermonuclear warheads, which can be tens and hundreds of times more powerful. Nuclear weapons can be delivered by a variety of systems, primitive or sophisticated, and are effective in any environment or weather, against essentially any target: cities, military bases, hardened underground bunkers, and mobile military assets such as naval battle groups.

Defenses against nuclear weapons are ineffective. The heat, blast, and ionizing radiation of nuclear explosions can only be survived in hardened underground facilities, and then only if those facilities are not specifically targeted. Mediation of radioactive fallout damage is localized and feeble at best. There is no way of reversing or ameliorating the biological and physical damage after the attack has occurred. The sole defense against nuclear weapon attack is total interdiction of its delivery systems, a goal that may be unattainable when a single failure results in unacceptable death and destruction.

In contrast, the effects of biological weapons are speculative and unproven. The alarming, well publicized models of the dire consequences of bioweapon attacks, resulting in mass casualties, are based on a number of important but undisclosed technical assumptions and typically portray a “most damaging case” scenario. While this type of estimation is prudent in making risk assessments, it has also been, historically, self-serving to those persons and institutions making the estimates, because careers and appropriations are dependent on convincing scientifically unsophisticated military and political authorities of the potential danger of biological weapons. When one examines the actual constraints of field use of biological weapons, as discovered during the US offensive bioweapons program, which lasted from 1942 to 1969, one discovers fundamental limitations.

Biological weapons potentially useful as strategic deterrents would be delivered by wide-area aerosols. The effectiveness of bioweapon aerosols is profoundly influenced not only by the agent preparation and delivery system, but also by a large number of independent factors, such as time of day, temperature, humidity, local landforms and vegetation, degree and type of urban development, atmospheric temperature gradients, and wind speed and direction. Each of these compounding variables can influence the ultimate effectiveness by several orders of magnitude, so that unfavorability of even a few variables will render an aerosol attack ineffective. A deterrent weapon must offer the assurance of an immediate, effective retaliatory attack under conditions not of the defending nation’s choosing. Biological weapons cannot give this assurance.

Real-world biological targets offer considerable resistance to bioweapon effects, when compared to field tests on animals in cages. The modern filtration in air-handling systems in typical late 20th- and 21st-century buildings and vehicles significantly reduces the effectiveness of biological aerosols. Moreover, bioweapon agents are, after all, biological, and show biological frailty. They are exquisitely vulnerable to industrial type air pollution, an effect noted by United Kingdom investigators at Porton Down, when mid-20th century England was still burning soft coal. Much of China and many Asian and other cities in the developing world today would be similarly resistant to an aerosol bioweapon attack. These fundamental limitations on the practical employment of biological weapons were well known on an operational level during the US offensive bioweapon program, and are responsible for the distinct lack of enthusiasm for biological weapons by any branch of the US armed forces, except for the sponsoring Chemical Corps. Soldiers require weapons that will function as expected in the environment in which they find themselves fighting, when they are wanted and needed.  Biological weapons cannot meet this requirement.

Biological weapons are vulnerable to both pre-attack and post-attack defenses. Preemptive immunization would nullify the deterrent effect of a biological agent, and this has already been done by the US military in the case of anthrax and smallpox. In a true threat scenario, or to nullify another nation’s deterrent, mass immunization of the entire population could be undertaken: The US Strategic National Stockpile was created with this in mind. Such emergency mass immunization programs have been undertaken in non bioweapon-threat scenarios: The ill-fated 1976 swine flu immunization program immunized 40 million people in three months.

Biological weapon effects can be nullified or mitigated after the attack with post-exposure treatment. This had been evaluated experimentally in humans during the US offensive bioweapon program, but was most dramatically demonstrated in the 2001 postal anthrax attacks. At the Senate office building the threat was immediately recognized, and exposed personnel were given post-exposure antibiotics immediately. No cases occurred in that cohort. At two postal centers, the exposure was not immediately appreciated, and seven unexpected symptomatic inhalation cases developed. These first cases appeared in personnel who worked closest to the automated letter-handling machines that generated the aerosols, and who therefore received the highest dose. Upon recognition of the postal center exposures, all potentially exposed employees were put on post-exposure antibiotics. Only two of the seven symptomatic postal cases died, far less than the 90 percent fatality rate predicted before the experience with post-exposure antimicrobial therapy. Quite remarkably, no further cases appeared among postal personnel after the institution of post-exposure antimicrobial therapy, even though the 1979 Sverdlovsk anthrax accident in the Soviet Union and animal studies predicted new cases would continue to appear for 60 to 90 days.  Post-exposure therapy truncated the attack effects as soon as it was recognized.

To be a credible strategic deterrent, a weapon needs to offer assured unacceptable damage when used in retaliation against an enemy’s attack. Nuclear weapons are credible. Biological weapons, not so much.

The “false allure” of biological weapons deterrence

In reading the initial Twitter response of academics and nongovernmental organizations to Seth Baum’s recent article in Contemporary Security Policy, "Winter-safe deterrence: The risk of nuclear winter and its challenge to deterrence," one would have been forgiven for assuming that the paper was advocating the use of non-contagious biological weapons as an alternative to nuclear deterrence in a way that was entirely uncritical. However, this is simply not the case. While the article does not provide in-depth analysis of the broader implications of Baum's argument, several of the themes outlined in recent critiques are at least partially reflected in his analysis. This includes for example the idea that the "Biological and Chemical Weapons Conventions must be considered among the greatest successes of the international community," and the caveat that "[m]any people would probably rather not even consider rescinding these conventions to re-proliferate these weapons, as well as the potential proliferation risks of such weapons." The paper is then, a provocative thought experiment that considers a hypothetical approach to reducing the size of nuclear arsenals, rather than a polemic against biological weapon prohibition. It involves describing a situation of such devastating human consequence that even the most morally repugnant of acts might become justifiable.

While it is tempting to engage in a more technical critique of assertions that are made about biological weapons in this article, I will leave this to colleagues with greater relevant expertise in this area. Instead, I want to draw attention to the almost allergic response of some commentators to Baum’s article, and consider its meaning and potential implications. The first key question one must ask is why such criticism has been forthcoming. A key reason: The paper is incredibly alien to those currently working in biological weapon control. This is because the paper is not motivated by the goal of ensuring the continued prohibition of biological weapons. Instead, the article provides a potential justification for developing and using such weapons. The article also places little emphasis on the benefits of the comprehensiveness of the international prohibition laid out within the 1972 Biological Weapons Convention. This convention prohibits the development, production, retention, stockpiling and use of such weapons under any circumstances. Any argument for "roll-back" within this treaty could also have implications for those working on sister treaty systems associated with the Conference on Disarmament

The unusual nature of the argument should not automatically invalidate the findings of the paper, but it does mean that the author would have to work hard to support such tentative assertions for them to be generally convincing. Such work would also likely have to go well beyond the scope of a single paper. If the author were to take up such a challenge, however, it would be entirely antithetical to the majority of academic and nongovernmental projects I am familiar with. From my perspective, the only potential value of such a project would be that it continued to cause other experts to recount the arguments against biological weapons and to explain why bioweapons cannot effectively replace nuclear weapons for strategic deterrence purposes.

I don't think Baum’s original motivation was to encourage a discussion about the allure of biological deterrents. Instead, his treatment of biological weapon norm was just collateral damage in his attempt to address the issue of winter-safe deterrence.  As stated above, while this is perhaps a questionable approach academically, it does not mean that the paper should not have been published in a peer reviewed academic paper such as Contemporary Security Policy. However, I do draw the line in at the appearance of these arguments, in a largely unqualified form, in a feature piece for The Bulletin. This publication has a respected, public-facing, and humanitarian profile and addresses biological weapons as a core issue area. The organization has a responsibility to its readership to ensure that controversial arguments such as Baum’s are sign posted as such, at the point of publication.

It seems likely that this is somewhat of a one-off paper. However, if it becomes the norm that biological weapon concerns are viewed predominantly and uncritically through a nuclear security lens within this publication, then this will likely have implications in terms of who contributes as well as the readership in the future. It would behoove The Bulletin to review its policy in relation to future publications of this type, to ensure concerns are flagged before the article goes to press. This will allow for decisions to be taken on the best way to present the article, if it is to be published. Controversial thought exercises, which go against the prevailing wisdom in relevant arms control communities, should be labelled just that. This is not to discredit such work, but to provide non-experts with a sense of the articles’ broader context.

The second key question is the extent to which further discussion of Baum’s paper is of broader value. I do not believe Baum’s academic paper presents a convincing enough thesis to be a genuine challenge to the international status quo, nor do I expect to see his arguments made publicly elsewhere by other experts. However, the paper has motivated some scholars and nongovernmental organisations to reflect on the norm against biological weapons. This is important because the question of "why" biological weapons should be prohibited has often taken a back seat to the question of "how" biological weapons are prohibited. This is understandable, considering the recent history of biological weapons. For over a decade, concerns about sub-state and pariah-state terrorism have contributed to the further stigmatization of biological weapons in press and media coverage. However, as we have seen in Baum's article, the very same things which make such weapons abominable, may also contribute to their appeal.

To sum up, I think the following key issues are worth further discussion. First, in the absence of any prospect of scientifically sound risk-benefit analysis, what sense does it make to advocate pursuing alternative modes of deterrence? Second, does Baum’s line of reasoning appeal to those working in the field of nuclear security? And finally, how can we ensure that the "false allure" of biological weapon deterrence does not take hold in the minds of the public and military strategists?

The biological weapons ban increases US security

A recent Bulletin column has suggested that non-contagious biological weapons may be a useful alternative to nuclear weapons as a deterrent, and could reduce the threat of nuclear devastation. While the United States may consider a variety of mechanisms toward a more stable deterrence strategy with fewer nuclear risks, biological weapons development will not be one of them. Doing so would violate US and international law and would be morally reprehensible. It would also leave the United States less secure.

There are multiple legal restrictions. The Biological Weapons Convention (BWC, formally titled the Convention on the Prohibition of the Development, Production, and Stockpiling of Bacteriological [Biological] and Toxin Weapons and on Their Destruction) is the first agreement among nations that declared an entire category of weapons to be off limits. It has been in force since 1975, and there are currently 170 states parties that have agreed not “to develop, produce, stockpile or otherwise acquire or retain:  Microbial or other biological agents, or toxins whatever their origin or method of production, of types and in quantities that have no justification for prophylactic, protective or other peaceful purposes; Weapons, equipment or means of delivery designed to use such agents or toxins for hostile purposes or in armed conflict.” The 1925 Geneva Protocol for the Prohibition of the Use in War of Asphyxiating, Poisonous or Other Gases, and of Bacteriological Methods of Warfare also bans the use of biological weapons in warfare. Taking the ban on bioweapons one step further, UN Resolution 1540 (2004) prohibits nations from supporting non-state actors toward developing, acquiring, manufacturing, possessing, transporting, transferring or using nuclear, chemical, or biological weapons or their delivery systems.

Biological weapons development is within the technical capability of most countries, so a revocation of the bioweapons ban could lead to a rapid proliferation of highly dangerous weapons. Even non-contagious biological weapons have enormous lethal potential that could rival the use of nuclear weapons, and could, in the case of anthrax, make the attack area unusable for a long time. The results of nation-state sponsored tests—back in the days before the BWC, and before the US renounced the development and use of biological weapons—demonstrate this. In 1968, a US military exercise took place 1,000 miles south of Hawaii, where hundreds of rhesus monkeys were loaded onto barges and exposed to aerosolized anthrax released from a Marine Phantom jet. Monkeys 50 miles downwind died of anthrax. In 1942, the United Kingdom tested bombs filled with anthrax spores on Gruinard Island, off the Scotland mainland, which left it uninhabitable for 48 years, until it was decontaminated at great expense.  

Instead of pursuing a weapons class that the international community has decided is off the table—a decision that has saved lives—the US government should continue what it is doing: working towards strengthening the BWC and rallying international partners to diminish the threat of sub-state actors or even individuals from pursuing biological weapons. While it would be a mistake to discount the technical skill, biological knowledge, and manufacturing technologies that were utilized in offensive biological weapons development when it was legal in the 1960s, there are many more technologies and shortcuts that would be available to a modern-day bioweaponeer. One can easily imagine how modern biological advances could be used to make a biological weapon more lethal, or difficult to detect or treat. 

In short, the ban on biological weapons continues to be the best policy. It has worked well to reduce proliferation—though there have been violations, no nation state openly declares that it is developing a biological weapons capacity—and it promotes US security.

The myth of biological weapons as the poor man’s atomic bomb

In his recent column, “Deterrence, without nuclear winter,” Seth Baum concludes that non-contagious biological weapons are one of two viable alternatives to replacing nuclear weapons in order to achieve what he calls “winter-safe deterrence.” He writes that non-contagious biological weapons “could work well if deterrence required threatening large human populations” without posing the risk of a global catastrophe like nuclear winter or a pandemic. Leaving aside the disturbing normative and legal implications of Baum’s proposal to start a global biological arms race, I will focus on the strategic logic underpinning his proposal to replace nuclear weapons with biological weapons. Baum’s conclusion is based on an uncritical acceptance of the long-standing myth that biological weapons are “the poor man’s atomic bomb.” This myth is based on the simplistic notion that because biological weapons could potentially cause mass casualties on par with those caused by nuclear weapons, these weapons should have similar political effects and implications for international security. Although biological and nuclear weapons are both considered weapons of mass destruction, biological weapons differ from nuclear weapons in three important ways that undermine the utility of biological weapons for deterrence: uncertainty of effects, availability of defenses, and the need for secrecy and surprise.

The first significant difference involves the level of uncertainty associated with the employment of these weapons. Nuclear weapons deliver instantaneous, overwhelming, and predictable levels of destruction. The effects of biological weapons, on the other hand, are delayed, variable, and difficult to predict due to their sensitivity to environmental conditions and the importance of pathogen-host interactions. In addition, the lack of operational experience with these weapons and the inability to simulate realistically their effects (short of massive human experimentation) impedes the ability of states to substantially reduce this level of uncertainty.

The second major difference between nuclear and biological weapons concerns the availability of defenses. There are no effective defenses against the effects of a nuclear attack. There are, however, a number of countermeasures that can be taken before, during, and after a biological attack that can mitigate the consequences of such an attack. Masks and filters can prevent exposure to biological agents. Biological weapons are also unique in that vaccines can be used to protect soldiers and civilians before an actual attack occurs. Because diseases have an incubation period of days to weeks, defenders have a window of opportunity to detect an attack using sensors and biosurveillance systems. Early detection can trigger the distribution of medical countermeasures to treat the victims of an attack and there are already vaccines and /or treatments available for the most lethal diseases such as anthrax, plague, smallpox, and tularemia. As a result, the effects of a biological attack are not absolute and incontestable; they can be mitigated and limited by a well-prepared defender. This possibility is likely to reduce the confidence of states in their ability to reliably inflict unacceptable damage against an adversary in a retaliatory strike. The full panoply of defenses need not be deployed constantly at full readiness because the very availability of these defenses may be sufficient to dissuade a state from calculating that it can inflict unacceptable damage. Although civilian populations will remain more vulnerable to biological weapons than will military forces, damage limitation remains a viable option for larger, more advanced states facing less sophisticated adversaries.

Third, biological weapons have limited value as strategic deterrents due to the need for states to shroud their biological weapons programs in strict secrecy. This need for secrecy is driven by normative, legal, and strategic considerations. In the strategic context, the availability of defenses against biological weapons places a premium on the attacker achieving surprise. This undermines the ability of a state to use biological weapons as a deterrent in two ways. First, the secrecy required to retain the element of surprise in a biological attack reduces a state’s ability to issue credible threats to inflict unacceptable damage against an adversary. To make a deterrent threat credible, a state would not only have to admit that it was violating international norms and laws but it would also have to reveal details about its offensive biological warfare capabilities such as the types of agents it has developed and their means of delivery. These revelations could reduce the effectiveness of these weapons by allowing the defender to mobilize appropriate countermeasures. In contrast, the superpowers flaunted their nuclear forces during the Cold War for deterrent purposes. They were able to do this because these demonstrations of their nuclear capabilities did not provide the other side with an improved means of defending against them. Second, secrecy is a flimsy means of protecting strategic forces designed for deterrence. Strategic forces that depend on secrecy for their protection are vulnerable to intelligence breakthroughs by an adversary. If a defender gained inside information about an attacker’s capabilities, it would be possible to develop and stockpile new pharmaceuticals, immunize the at-risk population, distribute protective masks and treatments, enhance public health surveillance, and take other precautions that could substantially mitigate the impact of a first-strike or retaliatory attack with biological weapons. Although such information is difficult to acquire, the cases of Soviet biologist Vladimir Pasechnik, former Soviet bioweapons program official Ken Alibek, and Iraqi weapons official Hussein Kamal attest to the risk posed by the defection of high-level government officials knowledgeable about their nation’s biological warfare programs.

A careful analysis of the technical and strategic aspects of biological weapons reveals that while biological weapons have the potential to inflict unacceptable damage against an adversary, they are unable to offer states an “assured” capability for doing so. This shortfall significantly undermines the suitability of biological weapons to serve as a strategic deterrent. Whatever the merits may be of pursuing “winter-safe deterrence,” promoting the discredited concept of biological weapons as a “poor man’s atomic bomb” is not an analytically defensible means of achieving that objective.

Round 2

On winter-safe deterrence and interdisciplinary research

I thank the roundtable participants for their continued discussion of winter-safe deterrence and biological weapons. With an eye toward the third and final round, my one request to them is that they suggest references in which interested readers could find further detail on these topics. Sonia Ben Ouagrham-Gormley has already suggested her book, Barriers to Bioweapons. I would appreciate any additional suggestions. I am especially interested in references for Martin Furmanski’s insightful comments on biological weapons testing and delivery systems.

In this post, I will reply to one comment from Gigi Kwik Gronvall and then offer a more general discussion of winter-safe deterrence as an interdisciplinary research project. I am not going into further detail on biological deterrence because I do not have much more to say at this time. Instead, I will be taking the points made into consideration as I continue my research.

Gronvall concludes her second round post by stating that “[a]dding a biological or other deterrent threat that can wreak more death and destruction would be going down the mountain, not up.” This is (I believe) to say that alternative deterrents should not be pursued in the context of nuclear disarmament. But some nuclear-armed states are already shifting their deterrence from nuclear weapons to weapons that cause less harm. The clearest example is the United States, which is actively shifting its deterrence towards conventional weapons. The 2010 Nuclear Posture Review states: “The United States will continue to strengthen conventional capabilities and reduce the role of nuclear weapons in deterring non-nuclear attacks.” I support the US effort to reduce the role of its nuclear weapons, and I do not believe that this is a particularly controversial position. All this is to say that some aspects of winter-safe deterrence are already mainstream military policy that is making a constructive contribution to climbing up that great mountain to a world without nuclear weapons.

With that, I will now make some more general comments about the winter-safe deterrence research and other research of this nature. My starting point is Gregory Koblentz’s second round post, which describes several flaws he sees in my winter-safe deterrence research. Clearly, he does not like this research or interpret it in the same way I do, and I do not think there is anything I can say in this space to change that. But his comments prompt a reflection on what it means to conduct good research.

Winter-safe deterrence is an inherently interdisciplinary topic. It begins with an analysis of winter-safe nuclear arsenal limits, which depends on a range of complex and poorly understood environmental and social phenomena as well as a careful treatment of risk and ethics. My suggested winter-safe limit of 50 total nuclear weapons worldwide is no doubt debatable. The research then analyzes a long list of candidate deterrents. Each of these weapons could have its own community of experts able to poke holes in my analysis—although, for one counterexample, I do not think neutron bombs have any active community of experts out there to chime in.

Ideally, winter-safe deterrence would be studied by a sizable team of experts on nuclear winter, risk analysis, deterrence theory, and each of the candidate weapons. Unfortunately, I did not have the luxury of assembling such a team. What I could have done is reach out to experts in all these fields for comment on my draft prior to its submission to the journal. I did get some expert comment, mainly on the nuclear winter risk analysis. I could have done more, but instead I judged, given my own constraints, that the paper was ready for submission. The paper then went through a constructive peer review that resulted in significant improvements to the paper. But peer review cannot catch every detail, nor should it be expected to. Peer review is especially difficult for this sort of interdisciplinary research, as it can be difficult to recruit reviewers with expertise across all the relevant areas. I know this issue firsthand as co-editor of a special issue on catastrophic threats forthcoming in the interdisciplinary journal Futures.

But peer review does not end at publication. A core virtue of the open publication of ideas is that they can be critiqued and improved by any interested parties. This is why I have gone out of my way to label my research as preliminary and to welcome the debate that my research has prompted. Because of this debate, progress on winter-safe deterrence is being made unusually rapidly. My only regret is that the debate has focused so exclusively on biological weapons. I commend the biological weapons community for its energetic response; I just wish other communities with expertise on other aspects of winter-safe deterrence would speak up, too. Surely other aspects of this research also have room for improvement.

It is important that interdisciplinary research on subjects like winter-safe deterrence be conducted. Too many major global issues do not fit neatly into established research fields. Progress on these issues often requires that researchers push outside their comfort zones in order to put all the pieces together and see the whole picture. I expect such research to be critiqued, including by those with expertise on specific pieces of the picture. But I do not think the existence of such criticism ought be taken as evidence that the research should not have been conducted in the first place. Progress is made by publishing and discussing, not by holding back.

Technological advance, proliferation potential, and the unsuitability of bioweapons as a deterrent

I have two responses to Seth Baum’s questions: the first relates to the impact of new technologies on increasing the deterrence value of bioweapons, and the second to the proliferation risk associated with bioweapons development.

Baum asks whether technological advances might alter bioagents, so bioweapons become better deterrents. The short answer is that it’s unlikely due to the nature of the microorganisms used in bioweapons work. Unlike nuclear materials, which have physically predictable properties, living microorganisms are prone to change. As living agents, they mutate, and they are sensitive to environmental and handling conditions. In the US and Soviet programs, for example, scientists found that some strains lost their virulence during manipulation in the laboratory or during dispersion, which compromised their use as a weapon. Laboratory experiments and the production of bioagents were also often subject to unexplained failures, even when highly experienced scientists were involved. The mercurial character of bioagents has not been reduced by new technologies. Currently, gene synthesis companies, which use highly automated processes to produce strands of DNA, routinely yield faulty materials due to errors that can naturally occur in the synthetic process.The pharmaceutical and biotech industries, as well as bioengineering projects, face similar difficulties due to the complexity of living systems. The truth is that in spite of the progress made in understanding the functions and composition of microorganisms, there is still a lot that we don’t know about them, which creates a stiff challenge in the development of viable and stable organisms that can be used predictably an reliably as weapons.

On the proliferation front, biological weapons are indeed complex and difficult to produce, but that does not imply little or no proliferation threat. Nuclear weapons are also very complex and difficult to produce—and yet they pose a proliferation threat. What distinguishes biological weapons is that unlike nuclear weapons, they are not constrained by material acquisition but by the acquisition of specialized skills required to handle and manipulate fragile microorganisms, while maintaining their desirable properties throughout the development process. Yet, most bio-nonproliferation initiatives today are modeled on nuclear nonproliferation efforts, i.e., a primary focus on material acquisition by imposing controls on material and equipment procurement. It is not that such an approach is unimportant but rather that very little is done to target the factors that allow the spread of expertise and the acquisition of specialized knowledge. Hence a major source of proliferation remains under-addressed.

The virtue of the Biological Weapons Convention is that by banning the development of bioweapons, it not only allows the knowledge acquired within past bioweapons programs to erode over time; it also makes it more difficult to create and develop that specialized knowledge. Countries bent on developing bioweapons will have to do so covertly, which requires an organizational and managerial model that is the exact opposite of what is needed for success. The need for covertness has indeed been a key reason for failure in all but the US program. This is why it is important to continue supporting the implementation of the convention: It is indeed one of the few initiatives that target, albeit indirectly, the key barrier to bioweapons development—the acquisition of expertise. (A more detailed analysis of where past programs succeed and where they failed can be found in my book, Barriers to Bioweapons: The Challenges of Expertise and Organization for Weapons Development.)

Testing Bioweapons: the Catch-22

In his posting for this roundtable, Seth Baum asked for a further discussion of several issues: one was the impact of technological advances, another how dismantling the Biological Weapons Convention might threaten global catastrophe.

No matter what technical advances may develop, biological weapons are fundamentally different from conventional, chemical, and nuclear weapons, because they depend upon species-specific biological characteristics, rather than broad phylum-specific biological processes affected by chemical weapons or physical and radiological effects. Bioweapon agents are fragile compared to other weapons, subject to rapid degradation during dispersion. This means candidate bioweapons must be tested on humans, and in the open air.

Scientists, quite naturally, are accustomed to drawing conclusions based on laboratory tests performed in defined, highly constrained systems where variables are controlled or excluded; this is, after all, the scientific experimental method. It is a mistake, however, to assume that laboratory tests alone can insure the development of effective bioweapons.

The US and UK offensive bioweapons program found repeatedly that laboratory tests were fraught with pitfalls and were insufficient alone to predict the behavior of bioweapons.

Bioweapon aerosols must traverse the open air. In 1946, at the remote Suffield proving ground in Canada, bioweapon scientists were chagrined to find that after relatively short downwind travel, bioweapon agents, though viable in laboratory cultures collected downwind, were unable to cause infection in exposed animals. Dealing with this “infectivity/viability dissociation” became a central focus of the US bioweapons program. Major technical installations, including the million-liter “8-ball” sphere at Camp Detrick, were constructed to allow the release of BW agents in controlled conditions and to sample the results at various post-dispersal intervals. Side-by-side exposures of human subjects (see below), non-human primates and other laboratory animals were made. Eventually open-air “proof tests” of biological agents produced downwind infections at the isolated Dugway Proving Ground in Utah and in the open Pacific Ocean off Johnston Island.

The utility of these proof tests was largely invalidated when in the late 1960s workers at Porton Down in the UK found that bacteria and viruses, including bioweapon pathogens, when exposed to outdoor air in a bio-secure apparatus sometimes showed episodes of dramatic loss of viability compared to their survival in closed containers. This was termed the “open air factor” and was linked temporally to conditions favoring importation of air from distant metropolitan areas. Studies in urban areas showed the open air factor to be ubiquitous and characterized as a synergistic combination of ozone and hydrocarbon fragments (olefins) typical of motor-vehicle associated photochemical smog. Its effect disappeared in confined spaces such as buildings or laboratory apparatuses such as the 8-ball. At concentrations of a few parts per billion, the open air factor was capable of rapidly reducing bacterial and viral viability by several orders of magnitude.

Since weapons of strategic deterrence would target urban areas, the existence of the open air factor would seem to deal a fatal blow to most bioweapon agents as candidate deterrent weapons. It also emphasizes that even elaborate, expensive laboratory investigations and field tests in isolated areas cannot model effectiveness in combat conditions.

A valid human dose-response curve with a “downwind” infectivity profile is essential to calculating design and targeting protocols for the strategic deterrent bioweapons Baum proposes. The US program acquired this data by exposing humans to bioweapon agents (with the 8-ball, for example), but with an important Catch-22. The agents so used had to be amenable to rapid and reliable cure with specific therapy. This meant that these agents were of limited lethality: tularemia, typically characterized as “potentially lethal” by bioweapon advocates, had only a two to seven percent fatality rate if untreated, and nil if treated with common antibiotics.  

Experimentation on humans is fraught with ethical pitfalls and has a disgraceful history. Bioweapon programs are no exception.  Most egregious was the Japanese program of 1935 to 1945, which used human captives almost to the exclusion of laboratory animals. At least 3,000 captives, and likely several fold more, perished in ultimately fatal experiments.

The US program’s human experiments in the 1950s and 1960s would not be ethically acceptable today. In 1957, a university hospital was contracted to determine the infective dose of tularemia, and it intentionally infected “members of our staff or suitable volunteer patients from the wards of the medical service.” An Army mobile lab determined the inhalation dose/response curves of bioweapon agents by using “volunteer” prisoners. Intensive studies were run on approximately 2,300 Operation Whitecoat volunteers, who were conscripted Seventh-day Adventist conscientious objectors. These vulnerable populations are excluded from human studies today.

So if non-contagious lethal biological weapons are legitimized, how are the essential human tests to be accomplished? Intentional exposure to untreatable lethal bioweapon agents would clearly fail ethical standards and would hardly attract many volunteers, unless participation were considered religious martyrdom. Covert testing on unwilling “disappeared” humans at “dark” sites is a chilling possibility and tragically not very farfetched, even for democracies, given the recent experience with torture.

Baum asks, apparently seriously, how the exit by nuclear powers from the Biological Weapons Convention by could result in proliferation of bioweapons. I will allow another participant to address the political aspects of this rather obvious consequence of his proposal. Of course many non-nuclear nations would be motivated to at least try to develop bioweapons.

The trying is more of an issue than the chances of success. The risks of escape from laboratories of dangerous pathogens and the risks of developing new strains with “gain of function” capabilities is currently of great concern. (See
"Threatened pandemics and laboratory escapes: Self-fulfilling prophecies.") These risks would only increase as classified bioweapon labs proliferated without inspection or accountability. There is no guarantee that any nation, once committed to develop a bioweapon capability, would necessarily comply with Baum’s arbitrary restriction of avoiding contagious pathogens. And as Baum acknowledges, release of contagious bioweapon agents, intentional or accidental, would be a grave threat indeed.

One must remember the failure of the elaborate technical apparatus of the US bioweapons program to predict the real world effects of candidate bioweapons was the result of stringent constraints on release of agents into the environment and a real (if flawed in retrospect) commitment to the protection of the human test subjects, bystanding civilians, and military personnel.

A pariah state or well-funded extra-national organization that felt no such constraints could avoid the trouble and expense of laboratory studies and just fashion “best guess” candidate bioweapons, release them on “enemy” territory (or less provocatively, on some neutral territory), and watch the results. The Assad regime’s use of improvised chemical weapons on its own rebellious citizens exemplifies this risk. Even nominally responsible nations could perform deniable, untraceable field tests using anonymous releases. And if a nuclear-armed nation were to receive such an attack, it might well consider it a threat to its national survival (or the sentinel wave of an emerging bioweapons attack that would) and mount a nuclear response against the presumed perpetrator.

Hello nuclear winter.

Deterring conflict, getting to zero

In the first round of commentary in response to the Bulletin column that suggested that other weapons, such as non-contagious biological weapons, could be a useful alternative to nuclear weapons as a deterrent, I focused on why biological weapons would not be suitable for this purpose. Doing so would be illegal and lead to proliferation of biological weapons that are much more accessible to nations than nuclear weapons; any use of those weapons would be devastating. An additional challenge for using biological weapons as a deterrent: Biological weapons are not easily placed into a verification regime, as they can easily be disguised as legitimate pharmaceutical or biotech research facilities because they do not have the expansive footprint of nuclear weapons facilities or require difficult to obtain fissionable material.  

For this second round of commentary, I focus on the grave risks of relying on nuclear weapons as a deterrent, and what is actually being done—and could be done—about those risks. While it can be imagined that a “safer” weapon could be switched out for nuclear weapons in our strategy of deterrence, in practice this will not happen. The myriad challenges of reducing nuclear weapons numbers across the globe would remain. In any case, governments need to do something to diminish the potential for armed nuclear conflict, for accidental launch, and for the increasing likelihood of nuclear-armed terrorist groups, which threatens to upend the strategic balance between nuclear armed and protected nations in ways that cannot be deterred. Adding another weapon of mass destruction to the mix does nothing to decrease this strategic risk.

The dangers of relying on nuclear weapons for deterrence, the potential for accidental or unauthorized launch, and the challenges of verification of the reduction or elimination of weapons are well known. And there are many current and former world leaders who wish to see nuclear weapons reduced to zero, especially President Barack Obama, former President George H.W. Bush, and former government officials around the world, including those from Russia and South Korea. Other supporters of a "global zero" future include humanitarian-focused networks and groups, the UN Security Council (via a 2009 resolution), and the nonpartisan Nuclear Security Project, led by former Secretary of State George P. Shultz, former Defense Secretary William J. Perry, former Secretary of State Henry A. Kissinger, and former Senator Sam Nunn. Taking on a global zero vision requires global momentum, not only from the nations that have nuclear arms, but also from those nations that do not, but are protected through extended deterrence provided by nuclear powers.

Reduction of nuclear forces in nations that already have them is a primary step toward global zero—and may be the realpolitik end state of affairs. But there are many other actions that need to be taken, and that will require international action, pressure, and technical advances. The Comprehensive Test Ban Treaty, which calls on states to ban all nuclear explosions for whatever purpose, must be brought into effect. Currently, there are states that have signed but not ratified the treaty—including the United States—and others that have not signed, including India, Pakistan, and the Democratic People’s Republic of Korea. The risk of accident needs to be reduced, and one measure that could be taken involves the separation of delivery systems from their nuclear warheads. Such a move would not reduce the numbers of nuclear weapons, but it would make the world more secure if the risk of accidental, mistaken, or unauthorized launch could be drastically lowered; a global norm that eliminates the hair-trigger from nuclear launch would reduce the risks of a nuclear winter. The risks of nations cheating on their stockpile numbers, or of terrorist groups acquiring nuclear weapons or weapons materials, also need to be reduced, and such a reduction requires coordinated international work to secure to the highest standards the nuclear weapons that already exist, to stop production of plutonium and highly enriched uranium for weapons, to phase out use of weapons-grade material in civil or research facilities, to lock down fissile materials, and to implement strong verification and enforcement regimes for nuclear materials. Attribution of the perpetrators of a nuclear attack through nuclear forensics is a less certain process than many might expect; to be able to hold violators responsible, and to deter nations from selling nuclear materials, the international community must create a mechanism to detect such illegal actions and to bring violators to swift justice.

None of these steps—or others required on the road to global zero—will be accomplished immediately. Deep nuclear weapons reductions will almost certainly take years—how many years, of course, is open to debate and will be a function of the commitment of the nuclear-armed states. Even if another class of weapon could inspire a level of deterrence comparable to nuclear weapons, it would do nothing to diminish the risk of the current global nuclear deterrent stockpiles. There are limits to deterrence itself: Neither traditional deterrence nor a “winter-safe” version can be effective against terrorists with no return address.

The prospect of a nuclear winter is indeed unthinkably horrific and could threaten all life on earth. It is the duty of all governments to work hard to take this scenario off the table, however long it takes. As Sam Nunn wrote, “The way I view it is that if you view the goal of getting to zero as the top of the mountain… I think we have an obligation to our children and to our grandchildren to build paths up the mountain.” Adding a biological or other deterrent threat that can wreak more death and destruction would be going down the mountain, not up.

Physics favors deterrence; biology doesn’t

In his contribution to the first round of this debate on the desirability and feasibility of developing non-contagious biological weapons as “safer” alternatives to nuclear weapons, Seth Baum defends the intellectual justification for his project on moral and humanitarian grounds. He concedes, however, that he “could have done more research on my own before publishing.” Baum’s research does leave a lot to be desired. For example, in his article in Contemporary Security Policy on which his original Bulletin of the Atomic Scientists column was based, Baum cites the toxin ricin as a suitable candidate for a strategic biological deterrent. While on paper ricin is attractive because of its high toxicity and ease of production, the toxin has failed miserably as a weapon. Several countries, including the United States, United Kingdom, and Iraq have explored the possibility of using ricin as a weapon before abandoning the toxin. According to the Organization for the Prohibition of Chemical Weapons, “Ricin is not known to have been used as a military weapon so its effectiveness as a chemical/biological warfare agent is unproven.” Ricin then is a strange choice to replace the most powerful weapon ever invented. The fundamental flaw in Baum’s column and article, however, is not so much his poor research, but the absence of analytical rigor in assessing the deterrent utility of biological weapons compared to nuclear weapons.

In his column and article, Baum lays out nine technical and political criteria for a weapon that could provide “winter-safe deterrence.” Baum’s ideal deterrent must be able to inflict a level of destruction to people or infrastructure equivalent to that of nuclear weapons, be able to survive a first strike, cause only local effects, be public, not be susceptible to significant countermeasures, not pose a significant proliferation risk, not undermine the geopolitical standing of the nuclear weapon states, not destabilize international security, and be politically acceptable. Non-contagious biological weapons fail to meet most of the criteria he described for an "ideal" deterrent. The only criterion that biological weapons easily meet is being able to survive a first-strike, since this depends more on the type, deployment mode, and alert status of the delivery system than the nature of the payload itself. On all of Baum’s other criteria, biological weapons are highly problematic, if not outright failures.

Could biological weapons cause mass casualties on par with those of nuclear weapons? Sonia Ben Ouagrham-Gormley and I already explained in our previous contributions to this debate that the answer is no, due to the high degree of uncertainty associated with the effectiveness of biological weapons and the availability of defenses against them. As Martin Furmanski pointed out, models used to demonstrate the potential of biological weapons to cause casualties at the same scale as nuclear weapons are “based on a number of important but undisclosed technical assumptions and typically portray a 'most damaging case' scenario.” Far from being an “absolute” weapon capable of causing predictably massive levels of damage in a reliable way, the effects of biological weapons are highly contingent on the specific characteristics of the agent being disseminated, meteorological factors, the health status of the targeted population, and the level of preparedness of the defender.

Could biological weapons provide a visible deterrent against which there are no significant defenses? We all know from Dr. Strangelove that “the whole point of the doomsday machine is lost if you keep it a secret.” Baum does not analyze the ability of biological weapons to meet this criterion, but as I explained in my previous contribution to this debate, states have strong normative, legal, and strategic reasons to keep their biological weapons programs secret. Since Baum is postulating a world where these normative and international legal restraints are no longer applicable, it is important to understand that the strategic motivations for secrecy will remain. The crux of the matter is that since there are effective defenses available against biological weapons, a potential attacker must keep its capabilities secret in order to retain the element of surprise and catch a defender unprepared. In his article, Baum acknowledges that the availability of defenses such as medical countermeasures and air-filtering systems on buildings are a “serious drawback” to the use of biological weapons as a deterrent. He also notes that this drawback could be reduced by a state remaining ambiguous about the types of biological weapons it has, but this would “also reduce the credibility of the deterrent.” Baum’s solution to overcoming the interconnected obstacles of defenses and secrecy is that unspecified advances in biotechnology might allow the development of biological weapons for which there are no defenses. There is certainly the potential for states to develop more lethal biological weapons, but advances in biotechnology also enable improvements in diagnostic tests, sensors, biosurveillance systems, and medical countermeasures which could improve a defender’s chances against even a genetically-engineered threat. Physics favors deterrence; biology doesn’t.

What effects would biological weapons have on geopolitics, international security, and proliferation? On all of these criteria, Baum’s own article provides evidence that the development of biological weapons by the world’s major powers would be detrimental to peace and stability. Since Baum acknowledges that biological weapons are better-suited for first-strike attacks than for retaliation, “the net effect is to destabilize the geopolitical order” and increase the risk of war. While Gigi Kwik Gronvall and Sonia Ben Ouagrham-Gormley disagree on exactly how easy it is for states to develop biological weapons, both would no doubt agree with Baum that biological weapons “present a much greater proliferation risk than nuclear weapons.” Baum does not reconcile the contradiction between the nuclear weapon states wanting to preserve their privileged position within the international community and shifting their deterrent strategies to a more widely available technology that could be used by weaker states to deter stronger ones. Furthermore, in a world where biological weapons are the foundation of deterrence, states will engage in a perpetual biotechnology arms race, constantly seeking an offensive or defensive advantage over their rivals. Such a situation would not only be destabilizing, but also have a major adverse impact on research in the life sciences and the peaceful applications of biotechnology.

Finally, do biological weapons represent a politically acceptable weapon? Again, Baum’s own article argues they are not. He writes, “The Biological and Chemical Weapons Conventions must be considered among the greatest successes of the international community.” Since I have already noted that Baum recognizes the severe limitations that plague biological weapons in a deterrent role, it is difficult to understand how he judges the purported benefits of launching a new biological weapons arms race outweighs the costs of overturning 40 years of work to prevent the proliferation of these weapons.

Perhaps the most disturbing aspect of Baum’s work is that his own analysis does not support his conclusion that non-contagious biological weapons could be a suitable replacement for nuclear weapons. Not only does Baum not provide any evidence that biological weapons possess the necessary characteristics to serve as strategic deterrents, he cites several features that would undercut their ability to function in that role. In the end, the potential lethality of biological weapons appears to be the only factor that Baum considers necessary and sufficient for these weapons to serve as “the poor man’s atomic bomb.” In defending his recommendation to replace nuclear weapons with biological weapons as a means of avoiding nuclear winter, Baum claims “the stakes are simply too high not to make hard choices.” Given the radical nature of Baum’s policy recommendations, the stakes are too high for scholars and publishers not to ensure that their research is backed up by solid evidence and rigorous analysis. 

Round 3

Sorry, we couldn't find any posts. Please try a different search.

Share: [addthis tool="addthis_inline_share_toolbox_w1sw"]


Receive Email