Proposed approach to evaluate the deterrence of limited nuclear attacks

The extensive power and drastic effects associated with nuclear weapons have resulted in posturing these weapons in a manner so as to prevent their use and is articulated as the Theory of Deterrence.

The rational deterrence theory argues that, in order to deter attacks, a state must present to the potential attackers that: 1)

it has an effective military capability;

‘Capable nuclear threats require both the nuclear weapons and the means to deliver them’. (Zarare 2004) The third criteria of rational deterrence hinges on the reasonableness, and hence credibility, of the threatened response. Credible threats are ‘threats that the threatener prefers to execute’ (Zarare 2004).

This is alternatively stated as:

In their Deterrence Operations Joint Operating Concept, the United States Strategic Command (2006) expands the scope beyond the punitive damage by examining the adversary’s decision calculus of costs and benefits for both conducting the undesired action versus being deterred. United States’ President Biden’s administration considers a broader means to achieving deterrence. ‘The Department will employ an integrated deterrence approach that draws on tailored combinations of conventional, cyber space, and information capability together with the unique deterrent effects of nuclear weapons’ (U.S. Department of Defense 2022a, p. 10). The ‘United States’ current policy is ‘integrated deterrence: the seamless combination of capabilities to convince potential adversaries that the costs of their hostile activities outweigh their benefits’ (The White House, 2022). Furthermore, the United States acknowledges the threat of employment of a few nuclear weapons. ‘The capability to deter limited nuclear attacks is critical given that some competitors have developed strategies for warfare that may rely on the threat of nuclear escalation in order to terminate a conflict on advantageous terms’ (U.S. Department of Defense, 2022a, p. 7).

Deterrence depends upon holding at risk what the potential attacking country cherishes. A common construct within the United States military is that countries have instruments of power, categorized as Diplomatic, Information, Military and Economic (DIME), that may impact other countries’ Political, Military, Economic, Social, Information and Infrastructure (PMESII) aspects (Worley 2012; Hartley 2015). Ny (2023) points out that the criteria for superpower status include the ability to exert global influence through economic strength, military forces and soft power. Of these national capabilities, we focus on four aspects of a targeted country that attacking military forces may directly affect: ‣

political control,

The effects on these national attributes by a nuclear strike may be assessed and the dimensions provided by our proposed approach. Other attributes may also be added into our approach.

The massively destructive power of nuclear weapons places them in their own class. Forsyth et al. (2010) consider the political and psychological aspects of using a nation’s nuclear arsenal. ‘Any adversary use of nuclear weapons, regardless of location or yield, would fundamentally alter the nature of a conflict, create the potential for uncontrolled escalation, and have strategic effects’. (U.S. Department of Defense 2022b, p. 9). Despite the case where the most powerful of conventional weapons can have a yield similar to the smallest of nuclear weapons, the employment of nuclear strikes still results in prompt and residual radiation along with the psychological implications of fear and anxiety. Consider that the United States deployed the GBU-43/B Massive Ordnance Air Blast (MOAB), the largest-yield conventional weapon on 13th April 2017, which resulted in minimal international condemnation. We postulate that if that attack had been conducted with a lower-yield nuclear weapon, the world condemnation would have been intense. The extent of damage from nuclear weapons is generally at a significantly more severe level than from any other weapons (Wirtz 2018; Karam 2021). Hence, nuclear weapons are in a distinct category than all other weapons due to the immense associated damage, prompt and fallout radiation and the potential to escalate to much more devastating weapons. Song and Weiss (2023) evaluate deterrence with nuclear and conventional weapons and conclude ‘nuclear deterrence is superior …because it at least prevents nuclear warfare.’ Therefore, we concur with Becker (2020) that a sufficient condition for deterrence of nuclear attacks necessitates threatening retaliatory nuclear strikes.

Deterrence of nuclear-weapon use is based on using threatening responses with nuclear weapons. For example, the United States strategy states ‘Nuclear forces [are] the ultimate backstop to deter attacks’ (U.S. Department of Defense 2022b, p. 9), and ‘Nuclear weapons will continue to provide unique deterrence effects that no other element of U.S. military power can replace’ (U.S. Department of Defense 2022b, p. 12). Integrated deterrence augments nuclear responses with other costs and benefits.

An entirely conventional response to a nuclear strike has the challenge of marshalling and massing conventional forces without those military units becoming a target for an adversary that has already escalated to employing nuclear weapons. Our goal is to evaluate nuclear responses that cause a commensurate and just amount of damage, and hence, should deter a potential attacker.

The United States and the Soviet Union initially built more nuclear weapons to enhance their counterforce strategy of targeting their adversary’s nuclear forces (Kelsey 2021). A counterforce strategy is generally deemed a better moral approach and hence more just or credible (Gallagher and Sorice 2014; Gallagher and Cevallos 2021). Targeting the adversary nuclear forces has four significant disadvantages: First, it has considerably fewer civilian casualties than targeting cities, so the threshold of using nuclear weapons is reduced (Avey 2021), however, others morally prefer this strategy due to its potentially fewer causalities. Second, the strategy incentivises a nation to strike first to limit the damage the adversary may inflict in retaliation (Mueller et al. 2007). Third, an arms race results because the side with more nuclear weapons could destroy more of the adversary’s weapons while also making it more difficult for their adversary to destroy their large arsenal. Fourth, countries are incentivised to make their nuclear forces and command infrastructure harder to target (Kelsey 2021).

Over the last several decades, major nuclear powers have enhanced their ability to survive attacks on their nuclear forces through combinations of three aspects: (1) hardening, including building deeply buried bunkers or launch systems in mountains, (2) increased mobility of nuclear-launch systems whether on roads, on train tracks or at sea and (3) regional and global defences to intercept bombers or re-entry vehicles. Even with improved sensing and delivery accuracy, nuclear weapons have limited capability to destroy deeply buried bunkers or silos. Hence, no country can effectively execute a counterforce strategy.

A common counterforce contention is that deterrence exists if potential attackers cannot destroy all the deterring country’s nuclear capability because the attacked nation could respond in a devastating manner. Even worse, if the aggressor does not even attack the opponent’s nuclear capability, their opponent has all of its nuclear arsenal for a retaliatory response. Consider two opponents with hundreds of nuclear weapons. They each deter the other from a counterforce attack because neither can destroy all of their adversary’s arsenal. Even if one side had a small percentage of weapons unavailable due to maintenance, the other side would not launch a counterforce attack. However, similar conditions would exist if one side attacks with only a few nuclear weapons. Would the responder launch an assault against the attacker’s remaining nuclear capability? Probably not since they are unlikely to destroy the attacker’s defended, hardened or mobile systems (Gallagher and Cevallos 2021). The US 2018 Nuclear Posture Review (Office of the Secretary of Defense 2018) acknowledges that Russia has been aggressively enhancing its tactical nuclear arsenal; the use of a tactical nuclear weapon would not change the balance of weapons in the strategic standoff. Yeaw (2023) highlights that both ‘Russia and China determined that limited nuclear employment might be required in any conflict with the United States’. Similarly, countries with fewer nuclear weapons do not intend to target the weapons of countries with a larger nuclear arsenal. We conclude that a country may consider nuclear strikes against targets other than their adversary’s strategic weapon systems. Furthermore, that nation would prefer their attack only consume a few of their weapons so as to retain most of the weapons to prevent or limit retaliation from the targeted country. Additionally, they should consider potential attacks from other adversaries possessing nuclear weapons.

The situation for deterrence of nuclear warfare has significantly changed with the dispersion of weapons to at least eight countries, including China’s growing nuclear arsenal (Mount 2023; Nye 2023). Any nuclear-enabled country needs to consider more than one potential nuclear-capable adversary. A country with large nuclear arsenal that expended most of its weapons may no longer be able to deter another country. Therefore, countries may consider limited nuclear attacks to conserve weapons for future deterrence or use. In this article, we hypothesise about what a country might strike along with its implications, with the goal of assessing how to deter those potential limited nuclear attacks.

International humanitarian law (IHL), also known as the Law of Armed Conflict (LOAC), restricts the use of military force.

Various views exist on how LOAC applies to the use of nuclear weapons. We categorise and discuss three schools of thought: strict interpretation, a relaxed interpretation and belligerent reprisals.

A strict interpretation of the LOAC allows only attacks on military targets with minimal civilian casualties.

Some contend, based on the widespread effects of a nuclear detonation, that nuclear weapons should never be used. For example, Krasny and Kawano (2019) assert that the use of nuclear weapons would not necessarily violate the LOAC; however, they conclude it would ‘violate the customary rule of prohibition of employment of weapons of a nature to cause superfluous injury or unnecessary suffering’. While tolerating their use in strategic counterforce attacks, Richardson (2021) argues that any use of tactical nuclear weapons violates the LOAC.

Generally, strict interpretation leaves only the counterforce strategy of targeting strategic and conventional military units and facilities along with national command and control installations (Fetter and Glaser 2022). The viable targets may not be adequate for deterrence. A nation possessing only relatively small military force and few nuclear weapons may not have many counterforce targets. In contrast, a major nuclear power may have only a few of these targets that are not hardened, defended or mobile. ‘A counterforce doctrine is not the United States’ best option for dealing with opposing states that have deployed large capable arsenals, which make significant damage-limitation essentially infeasible’ (Fetter and Glaser 2022).

One relaxed interpretation of the LOAC would expand to include war-supporting facilities as potential targets. Fetter and Glasser (2022) suggest civilian industrial and infrastructure targets that are essential for military operations, such as oil refineries, electrical grids and communication networks. Some contend this interpretation is a slippery slope leading to the targeting of cities. For example, while the bombing of Hiroshima was justified as destroying military targets and the war industry, the actual attack was focused on the city (McKinney et al. 2020). Jamison (2022) contends that the United States should implement ‘tailored targeting’, which includes conventional and cyber attacks on ‘critical economic functions’ under the guise of counterforce deterrence.

A third school of thought is belligerent reprisals, which invoke conditions where the LOAC constraints do not apply.

Past US interpretations did allow belligerent reprisal with US attacks against civilians if the adversary had violated the LOAC by attacking US civilians, if the response is proportional, and if the response has the purpose of deterring future attacks. Sagan and Weiner (2021) acknowledge this view, however they reject these arguments. Similarly, Goodman and Jinks (2004) contend that the LOAC restrictions apply in combating terrorists, even when the terrorists are not complying with the LOAC. Sagan (2023) rejects the arguments for belligerent targeting along with arguing against retaining nuclear weapons. In contrast, Fetter and Glaser (2022) contend that eliminating reprisals presents situations without ‘adequate retaliatory options’. Our evaluation in the last section indicates a lack of deterring responses depending on the interpretation of the LOAC applied.

One may conjecture that since nations are not bound to follow their stated policies, any country initiating a nuclear attack should consider the possibility of a reprisal response from their adversary. World leaders and experts continue to debate the legality of reprisals (McCarron and Holt 2001; Ford et al. 2021; Kang and Kugler 2023a,b,c).

Alternatively,

In this article, we are not advocating for any particular view regarding the LOAC; however, we present various nuclear attack options and note which would be allowed under various LOAC interpretations.

Zagare and Kilgour (2000) propose the Perfect Deterrence Theory. Capability is a necessary condition, however, not a sufficient condition for deterrence (Zarare 2004). ‘The logic of perfect deterrence theory is consistent with a policy of minimum deterrence, which rests on a threat that is costly enough to deter an opponent but that is not so costly that the threat itself is rendered incredible (Zarare 2004)’.

A more recently proposed nuclear strategy is proportional deterrence. A proportional deterrence policy deters opponents by inflicting a cost that exceeds the potential gain that an opponent could acquire by using nuclear weapons (Wiitala 2016).

Proportionality of a declared deterrent response engenders credibility, which the US Joint Chiefs of Staff (2011) define as ‘maintaining a level of believability that the proposed actions may actual be employed’.

One concern could be that the proportional response may result in continued similar responses and proceed into a slow escalation. Each side should be considering the possibility of further exchanges. The United States Department of Defense’s (2022b, p. 8) declaration is ‘The United States would seek to end any conflict at the lowest level of damage possible on the best achievable terms for the United States and its Allies and partners’. Kang and Kugler (2023a,b,c) examine conditions for nuclear stability. For deterrence to work, a defender must be able to deter its opponent not only from initiating a low-level conflict but also from escalating if and when the defender chooses to respond-in-kind’ (Zarare 2004).

The nuclear capabilities required to implement proportional deterrence differ significantly from other strategies. According to proportional deterrence, it contends that deterring any nuclear weapon use requires the capability to respond appropriately, which may necessitate capabilities that scale from small-yield precision strikes to large-yield strategic attacks (Perkovich abd Vaddi 2021). Hence, a much more robust nuclear arsenal, capable of a broader spectrum of yields and outputs, may be tailored depending on the opponent and the situation (Lieber & Press 2009). Hersman et al. (2020) note the challenges of initiating the design of a new nuclear weapon in the United States, and Hippel (2023) questions if making nuclear weapons more useful increases their likelihood of use. Sisson (2023) explores the United States’ potential vulnerability to manipulation and coercion due to the lack of a credible response to near-peers’ limited use of low-yield nuclear weapons.

While a damaging limiting strategy of counterforce is often characterised with a simple metric of the opponents remaining nuclear weapons, proportional deterrence needs to be evaluated in several aspects. We include political control, military strength and economic and civilian casualties in our proposed approach. Other criteria may further be added.

Deterrence depends on holding at risk what the country cherishes. In our earlier discussion such as the national basis for power that would be affected by a nuclear attack, we selected four national attributes of political control, conventional military strength, economic capability and population. In this section, we define those and assess categorical measures, which Table 1 summarises.

Political control characterises the ability of the targeted government to maintain control and exercise authority over its population and territory. We propose a graduated scale from ‘no change’ through protests, civil unrest, an insurgency and ongoing revolt to end of regime. While a nuclear detonation would certainly invoke a major crisis to address, the proposed limited nuclear attacks do not directly affect the government’s ability to control their country. Conventional or general-purpose forces assess their ability to project power. Our levels expand from local to theatre and then national with the degrades progressing from losing offensive capability to both offensive and defensive effectiveness. Economic impacts may be affected by destroying raw materials, transportation capabilities or production facilities. Similarly, economic impacts increase in geographic scope and extent of production impacted. Countries value their people; hence, population, assessed by the number of civilian casualties, is critical to evaluating a potential nuclear attack. Table 1 proposes damage levels within each category. Our approach can be applied with different attributes or with some other attributes in addition to the four considered. Different nations may vary in their priorities of these value categories; therefore, as we are not evaluating any particular country, we do not equate these levels or combine across the categories.

Additional diplomatic and political repercussions, probably in the form of sanctions, may be imposed on a country launching a nuclear strike. Whether initial or retaliatory, world opinion may perceive a nuclear strike as unjust. As such, the world community may add additional costs to this action. Sanctions may affect travel, education, arts and sports; restrict participation in international forums; limit trade or other economic aspects or restrict weapons trade and other military activities. National leaders may be tried for war crimes. However, these political consequences are significantly less severe than the damage produced directly by nuclear weapons. Economic and diplomatic sanctions, while very impactful, are not commensurate with the damage inflicted by a nuclear weapon. Therefore, sanctions alone are very unlikely to deter a nuclear strike. Hence, a proportional response does require nuclear counterstrikes. Since the application of these diplomatic actions varies in types and degrees, we do not attempt to rate their impact; however, we do, given a round of nuclear exchanges, indicate if one country is more likely to be perceived as acting inappropriately. Perceived compliance with LOAC is our primary determinant of whether the attacker or responder is sanctioned.

In this section, we postulate various general categories of nuclear strikes and provide a general indication of their consequences. We point out that all these are types of potential attacks that need to be deterred. What might a country attack with a limited nuclear strike? We consider the instruments of national power (DIME) and their impacts (PMESII). Reflecting on the capabilities of nuclear weapons, we postulate various limited nuclear strikes with different objectives. Since these nuclear options only expend a few nuclear weapons, we contend these strikes do not significantly affect a counterforce nuclear balance between major nuclear powers.

Although other strikes are possible, we discuss seven types of targets for limited-nuclear strikes. We present these categories of attacks from completely compliant with the LOAC to the extreme of an attack on population that clearly violates the LOAC. The following is a preliminary assessment of hypothetical targets. An actual application of this approach requires subsequent detailed evaluation of actual potential targets. We examine all these categories also as potential responses, mainly looking for adequate and credible responses to deter the attacker from each type of attack.

Strategic military targets include nuclear forces, national intelligence means and national leaders. We include a limited attack against strategic nuclear forces because some military strategists are of the mindset that nuclear war must be counterforce. Our premise is that counterforce attacks against strategic nuclear-delivery systems cannot disable an adversary’s nuclear arsenal due to hardening, defences and mobility. Hence, a limited attack with a few nuclear weapons would have an insignificant impact and not make sense for the attacker. Similarly, we do not include a decapitation strike on the country’s leadership since nations have also protected their leaders through hardened bunkers or mobility and defences.

A limited counterforce attack would be compliant with a strict interpretation of the LOAC. Since our military assessment is on the impact to conventional military forces, a notional strike against the nuclear forces would not significantly impact any of our four damage categories.Hence, we assess no damage in the four categories, as shown in the first row of Table 2, and we do not consider counterforce strikes further. Similar to targeting nuclear capabilities, modern hardening, defences and mobility make targeting national leaders and their control challenging.

A country may conduct a nuclear demonstration that does not result in any casualties or damage property. For example, nuclear demonstrations over the ocean would not directly injure anyone or any system. A country’s incentive to conduct a nuclear demonstration may be to remove any doubt about their possessing nuclear capability and indicate a warning of their willingness to escalate to employing nuclear weapons. Since there is no targeted nation, no damage would be inflicted regarding military forces, economic assets or political control. A nuclear demonstration would not violate any interpretation of the LOAC. However, the world community may impose costs through sanctions or other actions on the demonstrating nation.

We evaluate a hypothetical limited-nuclear strike on conventional or general military forces. Many contend that this very real potential use of a theatre nuclear weapon is against conventional forces (Beck 2021; Redmon 2021). Lieber and Press (2017) state that having nuclear weapons may not deter relatively weak states from using nuclear weapons in a limited capacity, such as denying access to enemies’ allied bases. Russian scholar Fedorov (2002) states that limited nuclear use can deter and de-escalate conflicts against a superior force in favour of a weaker force without risking escalation. The Russians have planned controlled escalation of conflict, which includes the use of nuclear weapons (Kofman et al. 2020). ‘[Russia] would use these forces to … avoid defeat if it was in danger of losing a conventional war’ (U.S. Department of Defense 2022b). Wiitala (2016) contends that countries may resort to using nuclear weapons if they perceive the employment is in their interest and do not believe there is a credible response to their actions. The most likely application would be a tactical nuclear weapon employed on a battlefield. This type of attack may be difficult unless the adversary was mobilised with targetable forces in the field. A strike against military forces complies with the strict interpretation of the LOAC.

We assess that a hypothetical nuclear strike of this type would have devastating level 3 damage on the adversary’s military units reflecting their ability to offensive project force. Such an attack would not necessarily directly affect the nation’s political control, economy or civilian population.

Possible war-supporting infrastructure targets include manufacturing facility, depot repair, energy production or distribution, transportation hub or communication node. For example, damaging a satellite control facility could affect navigation systems. The military impact may be delayed, such as reduced oil production. Tucker (2017) states that civilians lose their LOAC protection when they directly participate in hostilities. Almost all of these infrastructure targets support both military and civil uses; therefore, they are permissible only under a relaxed interpretation of the LOAC. We assess a hypothetical nuclear attack on a war-supporting infrastructure target would have level 2 impact on the conventional military, level 1 impact on economic and civilian casualties and no impact on political control.

A country can execute a HAB that would affect the electronics in a large geographic region or deploy an attack in space against satellites. Smith (2023) contends that a HAB detonation is a real threat. Since HAB effects are widely dispersed, the targeted nation needs to have a large geographic region; regardless, HAB effects may affect other nations. For example, a HAB might affect the world through disruption or damage to satellites. While a HAB would not cause direct civilian casualties that LOAC prohibits, Green (1988) contends that a HAB could be restricted based on international laws not allowing poisoning. Since a HAB would likely affect the electrical power grid, it would have significant economic consequences. We rate a hypothetical HAB as having level 1 effects on conventional military and level 2 effects on the economy. Responding with a HAB is problematic because the effects cannot be constrained to the targeted nation.

A country might conduct a limited economic attack. ‘Economic warfare has been a feature of armed conflict since antiquity, and has commonly been employed as an adjunct to armed force since that time’ (Lowe and Tzanakopoulos 2012).

Clemens (2013) states that ‘Many geopolitical tactics are, thus, quite explicitly economic, as they seek either to deter an adversary’s worst intentions or reduce its capacity to inflict damage.’ Gray (1979) reports, in the 1970s, the United States selected targets that would slow the post-war economic recovery of the Soviet Union. While the military conducted considerable economic warfare during the two great wars, the Additional Protocols of 1977 to the Geneva Conventions of 1949 restrict attacks to military objectives (Oermann and Wolff 2022). Interesting economic sanctions, such as an embargo, can lead to hundreds of thousands of civilian deaths, potentially more than a military action (Oermann and Wolff 2022). In contrast, Førland (1993) assesses economic targeting as generally ineffective; however, he contends that ‘belligerents ignored jurists’ rules of acceptable conduct in economic warfare’. The trends towards cyber and hybrid warfare may involve more economic impacts. Economic targets, while in violation of the LOAC, are a threat that needs to be deterred. Furthermore, an economic response may be considered a belligerent reprisal to an economic or population attack.

Since the intent is not to cause extensive civilian casualties, our hypothetical economic attack would have to be constrained to a transportation hub, such as a port with container loading, a railway transfer station or major bridges, preferably in rural areas. These economic attacks would require a relative small yield, perhaps as from a tactical nuclear weapon. The attack might be conducted on a day and at a time when few people are present in the area. The attack detonation should be executed so that the fallout would not blow over the populated areas, for example, a coast harbour with prevailing off-shore winds.

We assess that a hypothetical economic attack might result in level 1 damage for civilian casualties and level 2 for economic damage. No direct damage would occur to military forces.

As history demonstrates, nuclear attacks can be conducted on cities. Even with a few weapons, the impact would be severe in terms of civilian casualties. Additionally, an attack on the population would have huge economic implications, where we assess economic damage level 2 for a hypothetical attack. Payne et al. (2023) rejects targeting of civilian populations. Attacks on cities violate the LOAC; however, they may be justified as belligerent reprisal in response to a nuclear strike against cities.

Table 2 summarises the damage levels that we assess for each type of hypothetical limited-nuclear strike. None of these limited attacks are intended to directly disrupt government control; hence, we did not assess the political damage for any of them. Therefore, we do not list the political control to the target nation in our subsequent analysis; however, we do assess potential diplomatic responses in the form of sanctions. Furthermore, we are of the opinion that limited attacks against strategic nuclear systems or command do not have a significant impact so we do not include that option further.