Study Deterrence in the 21st century No. 16 | May 2020 The views expressed in Metis Studies are those of the authors. They do not reflect the opinion of the Bundeswehr, the Federal Ministry of Defence, or the Bundeswehr University Munich. The primary target audience of Metis Studies are practitioners. Metis Studies are based on analyses of scholarly literature, reports, press articles and expert interviews with academics, think tank analysts and policy-makers. References are omitted. Inquiries about sources can be directed at the author(s) via email. Metis Study | No. 16 Deterrence in the 21st century Summary At first sight, deterrence reflects a simple idea: however, has raised questions ever since the beginning A credible threat of retaliation convinces of the nuclear age. Against the backdrop of current potential attackers that the cost of an act of security policy challenges, this study examines if and aggression will outweigh its benefit. That is how how the concept remains applicable. deterrence prevents war. A closer look at the concept, The concept of deterrence territory. Mutual deterrence between two actors who are The concept of deterrence can be traced through the both convinced that they are able to destroy their oppo- history of political and military conflict. However, it was nent in a retaliatory strike (i.e. mutual assured destruction) not until the Cold War – when it took on the form of nu- establishes a state of strategic stability. clear deterrence – that the concept gained the academic Deterrence does not just happen on its own. The “del- and practical prominence it has to this day. Deterrence icate balance” must be established by political and military can work in two ways: by threatening retaliation, i.e. means. Its effect is systemic, i.e. it has an impact on inter- deterrence by punishment, or by denying success, i.e. national relations. Most important for its effectiveness, deterrence by denial. The former involves state A aiming however, is its influence on (individual) decision makers. to convincingly signal state B that a certain action or at- tack will cause prompt retaliation by state A, including the Theory and practice of deterrence destruction of essential assets in state B. The latter sees The academic and theoretical examination of the concept state A demonstrating resistance against state B without of deterrence emerged as a first wave of research after threatening retaliation, thus suggesting that state B will World War II because of the need for a political response not be able to achieve its political and military objectives to the nuclear age. The core concepts as outlined above through attack. The two forms may overlap. were developed at that time and clearly influenced by the Deterrence has often been pursued in aid of stra- bipolar order of the Cold War. The first wave emphasised tegic ambiguity. For example, during the Cold War era, the fear aspect of deterrence in that it literally focused on strategic ambiguity was the preferred approach of the frightening opponents. US when it came to Germany. It was impossible to defend The second research wave in the 1950s and 60s dis- isolated West Berlin. Its protection was thus ensured cussed the concept in less emotionally charged terms. through the threat of retaliatory measures elsewhere. References to fear were dropped and replaced with dis- The Soviet Union could never be sure of the place and cussions of rational actors, cost-benefit calculations and scale of such measures. The response from Moscow was modelling based on game theory, all in an attempt to infer to exercise restraint. general conclusions about nuclear strategies. The current This example also illustrates the concept of “extended mainstream of deterrence theory, explicitly understood as deterrence”, which seeks to prevent attacks against third the manipulation of an adversary’s cost-benefit analysis, parties such as allies or partner states. Examples include remains rooted in these influential efforts. The key con- the US nuclear umbrella over Europe or the US security cept of “escalation dominance” – i.e. the ability to always guarantees in Asia. “Direct deterrence”, on the other hand, go one decisive, ultimately deterrent step further – also serves only to prevent attacks against a country’s own dates back to this period. 3 Metis Study | No. 16 Deterrence in the 21st century The third wave, which emerged in the 1970s, called on cognitive psychology and used case studies to examine whether real-world decision makers act in a truly rational manner. As it turned out, the assumptions formed in the second wave were limited in their validity because, in the real world, the mispercep- tions, recklessness, ideology and even drug use of decision makers ran counter to the idea of rational deliberation. Moreover, empirical studies revealed that decision makers sometimes sought out con- flict for domestic reasons (to retain power, for example) de- spite the threat of deterrence. All in all, the empirical analysis of deterrence strategies sug- gested that the concept bears the risk of causing exactly the war it is meant to prevent. In other words, the paradox inherent in deterrence theory – that it requires constant preparation for and the cred- ible threat of that which must actually never happen – had been discovered in practice. After the end of the Cold War, a fourth wave emerged in response to the decline in interstate wars, the increase in intrastate conflicts and the phenomenon of interna- tional terrorism. The focus shifted to asymmetric actor constellations and “rogue states” and, in discussing the Fig. 1 Permanent confrontation: Deterrence uses a credible threat of retaliation to prevent military escalation (Checkpoint Charlie, 1961). motives of suicide attackers and moral values of auto- crats, Western concepts of rationality in the mainstream deterrence theory were called into question once more. There has been talk of a current fifth wave, although it is more a hotchpotch of concept of deterrence. With paradoxes and problematic approaches meant to use diplomatic, economic, political basic assumptions left unaddressed since the Cold War, and military means to address non-kinetic, cyber-spe- the concept has long been at risk of being overstretched cific, terrorist and hybrid risks as they occur. Instead, to such an extent as to render the idea of deterrence it creates even more analytical confusion around the meaningless. 4 Metis Study | No. 16 Deterrence in the 21st century regulates the interaction of nuclear weapons states, although how much of it is due to rational deliberation versus sheer fear remains un- clear. When it comes to newly emerging security challenges, however, it is obvious that some of the fundamentals re- quired for deterrence to work simply no longer apply. In the information space, cyberattacks raise the ques- tion of whether they may be deterred at all. The key prob- lem when applying deterrence to cyberspace is the so-called attribution problem, i.e. the inability to clearly identify the originator of a cyberattack. When a state is able to locate an attacker, for example in a cybercafé or private home in Asia, that information is only of limited use for identifying the actor who actually carried out the attack. Even if the attack is traced back to a computer centre of a local military force, the risk remains that elements of the cyber architecture of the apparent perpetrator state have actually been compromised by third parties. The issue is made all the more complex by some countries tasking non-state actors with operations in cyberspace. As a result, the apparent perpetra- tor state maintains plausible deniability, which the attacked state struggles to disprove even with advanced cyber intelligence and time-con- suming forensic investigation. A threat of prompt retaliation does nothing in terms of de- terrence by punishment if it cannot be directed at anyone in particular. Deterrence in new domains Thanks to the efforts of state actors to protect cyber- The “classical” deterrence theory of the Cold War and space and critical infrastructures against cyberattacks, its practical application have certainly had some effect, successfully mounting such an attack has become much although it may have not been as reliable and generalis- more difficult in recent years. Adversary network opera- able as was hoped during the second wave. It undeniably tions have to invest more resources, energy, personnel 5 Metis Study | No. 16 Deterrence in the 21st century and time to successfully attack a state’s key capacities. The underlying idea would still be to convince a Attacking military structures from a laptop in a cybercafé potential attacker that their plan is useless because it obvi- is plausible only in Hollywood movies. We may therefore ously has no chance of succeeding. But while the classical assume that at least deterrence by denial can be used theory of deterrence by denial would involve signalling successfully against many actors with few resources. The in the military context, which bears the risk of being concept of resilience plays a major part in this respect, as misperceived, deterrence by resilience would simply work this study will go on to show. as a result of robustness and the demonstrated ability to For one thing, some space scenarios call into ques- absorb attacks. It is therefore not a question of defence tion the applicability of classical deterrence concepts. For and subsequent counterattack in response to hybrid at- another, the military use of space creates new scenarios tacks, for example, but rather of establishing “absorption that might affect the “delicate balance” on the ground. dominance” and thus the ability to control how damage In the Cold War era, limited space capabilities meant unfolds and how long it takes to return to the previous there was only ever a very small group of “usual suspects” status quo after an attack. For example, if state A remains in orbit. Following advances by private actors, however, largely unharmed and unimpressed (i.e. unaffected by we now face an attribution problem when it comes to ki- major cost) by a hybrid intervention of state B involving netic effects much like the one for non-kinetic operations fake news or cyberattacks, the probability of a second, as previously described above. Once again, deterrence as similar attack will be drastically reduced. To implement commonly practiced has no clear target. this approach, states would have to focus on developing Moreover, new space capabilities such as anti-satellite an all-state capacity for resistance and absorption which systems or other weapons 1 may jeopardise, for example, comprises all critical areas in order to become more resil- early detection of a nuclear first strike or even the sec- ient to attacks and disruption. ond-strike capability and thus erode strategic stability. In Such scenarios are not without analytical and prac- such a case, while the logic of deterrence would remain tical pitfalls because, as in the above example, an attack intact, new risks would emerge which could limit its effec- may barely be registered as such and if it is, it may be tiveness as a stable guarantor for the continued non-use difficult to attribute to a certain actor. The attribution of nuclear weapons. For example, a nuclear weapon state problem thus persists, although it is somewhat less rel- – in a case of “use them or lose them” – might respond to evant because a strategy of resilience does not imply a an attack against its space-based capabilities with nuclear threat of retaliation and the lack of attribution would thus retaliation in order to forestall being deprived of its sec- not undermine its credibility. The identity and intention of ond-strike capability. an attacker would initially be irrelevant – as long as they are thoroughly discouraged and deterred from renewed Deterrence as resilience attempts. In other words, a strategy of resilience is itself Deterrence has thus accompanied us into the 21st century, more resilient than a deterrence strategy, at least when it with old and new questions raised along the way. A prom- comes to cyberspace. ising, forward-looking approach is to consider deterrence The concept of deterrence, especially its nuclear in radically simplified terms of resilience, especially in dimension, entails risks but remains an integral part of new domains of application. This would also put to rest strategies throughout the world. Not every new aspect the deterrence paradox and other legacies of deterrence of security policy can – or should – be addressed with theory such as the lingering problem of credibility. Such deterrence. But deterrence theory and the scientific ob- an approach would use the logic behind deterrence by servation of its practical use may stimulate new strategic denial as its starting point. considerations that reflect the complexity of today’s secu- rity challenges. 1 See “Space Security”, Metis Study No. 13 (August 2019). 6 IMPRINT Publisher Metis Institute for Strategy and Foresight Bundeswehr University Munich metis.unibw.de Author Dr. Konstantinos Tsetsos [email protected] Creative Director Christoph Ph. Nick, M. A. c-studios.net Images Cover: Graphic by ndul on 123RF p. 4 / 5: Photo by USAMHI from wikimedia Original title Abschreckung im 21. Jahrhundert Translation Federal Office of Languages ISSN-2627 – 0609 This work is licensed under the Creative Commons Attribution 4.0 International License. Metis | 2020