36
Caleb Ziolkowski What are the principle challenges to the development of deterrence in the cyber domain? Introduction Cyber deterrence, building off of Robert L. Pfaltzgraff, Jr.’s definition of deterrence, is “[t]he ability to prevent someone from doing something they want to do" 1 in the cyber domain. Opinions on the difficulty and means of achieving effective cyber deterrence vary widely. Some, like Martin Libicki, find it painstaking and incredibly different from traditional deterrence. Libicki writes: All this might lead to the belief that the historic constructs of war—force, offense, defense, deterrence— can be applied to cyberspace with little modification. Not so. Instead, cyberspace must be understood in its own terms, and policy decisions being made for these and other new commands must reflect such understanding. Attempts to transfer policy constructs from other forms of warfare will not only fail but also hinder policy planning. 2 1 Robert L. Pfaltzgraff, Jr., “Conflict, War, and Deterrence Theories: Part II,” a lecture delivered at Tufts University, December 12, 2013. 2 Martin Libicki, Cyberdeterrence and Cyberwar, (Santa Monica, CA: RAND, 2009), xiii.

The Challenges of Cyber Deterrence

  • Upload
    tufts

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Caleb Ziolkowski

What are the principle challenges to the development of deterrence in the cyberdomain?

Introduction

Cyber deterrence, building off of Robert L.

Pfaltzgraff, Jr.’s definition of deterrence, is “[t]he

ability to prevent someone from doing something they want to

do"1 in the cyber domain. Opinions on the difficulty and

means of achieving effective cyber deterrence vary widely.

Some, like Martin Libicki, find it painstaking and

incredibly different from traditional deterrence. Libicki

writes:

All this might lead to the belief that the historic constructs of war—force, offense, defense, deterrence—can be applied to cyberspace with little modification. Not so. Instead, cyberspace must be understood in its own terms, and policy decisions being made for these and other new commands must reflect such understanding.Attempts to transfer policy constructs from other formsof warfare will not only fail but also hinder policy planning.2

1 Robert L. Pfaltzgraff, Jr., “Conflict, War, and DeterrenceTheories: Part II,” a lecture delivered at Tufts University,December 12, 2013.2 Martin Libicki, Cyberdeterrence and Cyberwar, (Santa Monica, CA: RAND, 2009), xiii.

On the other hand, some, like Will Goodman, think that a

theoretical excavation of cyber deterrence can lead to an

over exaggeration of the complexity of the task at hand.

Goodman writes, "While deterrence in cyberspace does pose

challenges…[it] remains inextricably linked to the

geopolitics of the physical world. As a consequence, cyber

deterrence turns out to be simpler in real life than it

appears to be in many theoretical models."3 The question

remains. How difficult is it? Certainly, as Goodman

concedes, theoretically cyber deterrence is chock full of

thorny problems.

A list of the main ones includes: identifying the

source of the attack, known as attribution; ensuring

credibility in a domain where guaranteed responses are

difficult; the speed of change in the cyber domain,

especially with an eye to legislation; delineating the

public and private spheres’ respective responsibilities;

public and private cooperation; balancing civil liberties

with public sector control; ambiguity for those attacking 3 Will Goodman, “Cyber Deterrence: Tougher in Theory than inPractice?,” Strategic Quarterly, Fall 2010, 105.

and those attempting to deter or defend; and deficiencies in

the existing domestic and legal frameworks. The traditional

challenges to deterrence, in addition to these more cyber-

specific challenges, must also be considered. These

challenges appear so daunting that some experts recommend

exhausting other options before depending on deterrence.4

Looking in more detail at these challenges will help us

reach some tentative conclusions about the role and

potential of deterrence in the cyber domain.

In analyzing this issue of cyber deterrence it will be

helpful to run through the different levels of analysis to

organize the inquiry. The investigation will begin with the

international system level and then move to the didactic

level—this level focuses “on the nature of pairs of states

(dyads), that is, their mutual or shared characteristics,

and the interaction between these pairs of states.”5 Then 4 "Before contemplating deterrence as its primary response to the threat of state-sponsored cyberattacks, the United States may first want to exhaust other approaches, such as diplomatic, economic, and prosecutorial means." Libicki, Cyberdeterrence and Cyberwar, 176-7.5 Greg Cashman, What Causes War?: An Introduction to Theories of International Conflict (Kindle Edition: Rowman & Littlefield Publishers, 2013), kindle locations 7836-7837.

the unit level will be examined followed by the group and

individual levels.

International System Level

If traditionally the international system level has

been seen primarily as a state of anarchy, the cyber domain

can only be seen as more anarchic. At least partially due

to its novelty, many of the norms and laws established in

other domains are nascent or nonexistent. Unlike

traditional forms of conflict, there are widely diverging

opinions as to what actually constitutes cyber attack, cyber

crime, and cyber warfare.6 The realist assertion of anarchy

certainly is apt for this domain.

Two important caveats, however, merit attention. The

realist assertion—shared by virtually all realist schools—

that states are the primary actor appears, if not invalid,

severely weakened in this domain. Often super-empowered

individuals are cited as representing significant actors:7

6 Oona Hathaway et. al., “The Law of Cyber Attack,” (California Law Review, Vol. 100, No. 4, 2012), 819-22.7 Goodman, “Cyber Deterrence,” 10.

more on this below. Second, while international relations

may be a club from which one cannot withdraw,8 the same

cannot be said of the cyber domain—at least not to the same

degree. Individuals, corporations, and states ultimately

have a lot of control over the size of their cyber

presence.9 These represent two important differences in the

cyber realm.

At the international system level—to bring in something

of more concern to neoliberal institutionalists and

neoclassical realists—the current legal framework deserves

analysis. This framework is fragmentary. There is no

overarching coherence.10 More seriously, there is no agreed

upon definition for cyber aggression, as noted above. The

law of war—jus in bello and jus ad bellum—applies to only a “small

slice of the full range of cyber-attacks”11 and poses

difficulties in implementation.12 As far as legal regimes

8 Robert L. Pfaltzgraff, Jr., “Paradigms, Theories, and Levels of Analysis,” a lecture delivered at Tufts University, September 12, 2013.9 Libicki, Cyberdeterrence and Cyberwar, xiii.10 Hathaway et. al., “The Law of Cyber Attack,” 819.11 Ibid., 844.12 Ibid., 855-6.

that in some way directly address cyber-attacks—the UN,

NATO, the Council of Europe, the Organization of American

States, and the Shanghai Cooperation Organization—while

having showed interest in addressing cyber attacks, “these

efforts have fallen short of establishing a rigorous legal

framework that can effectively govern all cyber-attacks.”13

Another potential framework, that of international regimes

that indirectly regulate cyber attacks—such as the

international law governing telecommunications, aviation,

space, and the law of sea—likewise fails to address all but

a “small number of harmful cyber-attacks.”14 The lack of

effective law in the cyber domain adds to anarchy.

Didactic Level

Moving from the international systemic level to the

didactic level of analysis, where deterrence has

traditionally been analyzed,15 attribution possibly the most

cited difficulty in effective cyber deterrence. Martin

13 Ibid., 867.14 Ibid., 874.15 Cashman, What Causes War, kindle location 11118.

Libicki highlights some of these difficulties when he writes

that identifying an attacker poses tremendous difficulty

because (1) the attacker may simply say, "who, me?,” (2)

"mistaken attribution makes new enemies, and (3)...neutral

observers need to be convinced that retaliation is not

aggression."16 This all stems from the fact that, unlike

most conventional aggression, it can be very challenging to

identify the source of the attack. Routing a cyber attack

through the ip address of some third-party—ignorant of even

the existence of the attack—is not only quite easy for cyber

aggressors to accomplish but also quite difficult for a

would-be deterrer to trace back to its source.

Will Goodman pushes back against the idea that

attribution is as hard as many people intimate.

International cooperation can make this possible. But what

if another state does not cooperate? “In such instances,

victim states can, based on mutual legal aid agreements or

the inherent right to self defense, assign responsibility

for the attack to the non-cooperating state."17 An 16 Libicki, Cyberdeterrence and Cyberwar, xvi.17 Goodman, “Cyber Deterrence,” 10.

interesting case study that is often brought up in

discussions of attribution is Estonia. Many, Libicki

included, highlight the attack in 2007 as evidence of the

impossibility of attribution in the cyber domain.18

Goodman, rightly, says it is actually an example of how

attribution is not as difficult as many imply. He points

out that Russia failed to cooperate in helping Estonia even

though the two countries had a legal aid agreement.19

Therefore “[i]nternational law provides a basis for

assigning the culpability of the attacks to Russia even if

Russia did not officially direct them.”20

The problem, then, is not attribution in the case of

Estonia. It is a problem of geopolitical asymmetry. What

can Estonia do with aggression from Russia? Its most

feasible option, and the one it pursued, was to seek help

from NATO and establish the NATO Cooperative Cyber Defense

Centre of Excellence.21 The case of Estonia is, contrary to

what some posit, actually a good example of how attribution 18 Libicki, Cyberdeterrence and Cyberwar, 2-3.19 Goodman, “Cyber Deterrence,” 113.20 Ibid.21 Ibid., 114.

is often not the roadblock to deterrence in the cyber

domain. Geopolitical concerns can trump it. Some

attribution cases may be more difficult, though.22

An important point regarding international cooperation,

one that Goodman himself mentions, is that quid pro quos are

involved. If, say, the United States wants other countries

to aid cyber related investigations by sharing information,

then U.S. security agencies must be willing to help others

by opening up their systems as well. One may question

whether the FBI or other agencies are ready for such a

thing.23

There is one more part of attribution that must be

discussed: super-empowerment. Goodman admits that super-

empowered individuals appear impossible to deter, but

insists that since geopolitics and the cyber domain are

inextricably connected, and that the state is “supreme,” it

is less daunting. Putting pressure on the state from which

individuals base their attacks should reduce the problem,

22 Ralph Langner, “Stuxnet's Secret Twin,” Foreign Policy, November 19, 2013.23 Goodman, Will, “Cyber Deterrence,” 122.

provided a favorable balance of geopolitical power. 24 The

validity of this assertion may be questioned. First, there

are times when other countries have been blamed only to find

out later that the attack was actually domestic in origin.

An example comes from February 1998 when two California

teens along with an Israeli hacked unclassified Department

of Defense networks; initially Iraqis were suspected. The

notion that states are completely “supreme” in their

domestic cyber domain seems at least partially at odds with

the initial false attribution.

Another example that could cast doubts on state control

of the cyber domain comes from China. China, frequently

thought of as one of the main culprits of cyber attacks, not

only suffers a great deal of foreign-originated cybercrime

itself, but also “data triangulation from multiple sources

indicates domestically originated attacks are no less

severe.”25 China seems to have problems controlling its

domestic cyber domain. Are there states that are simply not

24 Ibid., 112.25 Nir Kshetri, “Cyber-Victimization and Cybersecurity in China,” Communications of the ACM, Vol. 56 No. 4, 35-37.

“supreme” in their cyber domains? In defining a state Max

Weber says it is a "human community that (successfully)

claims the monopoly of the legitimate use of physical force within a

given territory."26 When considering cyber, have all states

“successfully” made this claim in the cyber domain? Is it

possible to have failed states in a single domain, where

super-empowered individuals easily evade the state’s wishes

and use force against domestic and international targets,

including foreign states? If this is the case, is it

realistic to expect a country like China—obsessed with

portraying the image of controlling its domestic affairs—to

admit that the state is not in complete control? Do all

countries control their cyber domain in the same way they

monopolize force, as Weber’s definition requires? Failed

states have proved fertile ground for terrorism. This could

be true for states failing in the cyber domain. These are

serious questions representing grave challenges in the cyber

domain.

26 Max Weber, “Politics as a Vocation,” Essays in Sociology, pp. 77-128 (New York: Oxford University Press, 1946), 78.

Yet, in spite of these concerns, Will Goodman’s

argument that geopolitics and the power of states—which will

likely grow in the cyber domain as time goes by to

compliment their dominance in other domains—to control

groups and individuals within their boarders certainly makes

issues of attribution less daunting than many portray them

to be. Many cases with super-empowered individuals will

also involve states that do exercise enough control to

warrant some of Goodman’s optimism. It is important to

remember, conversely, that in Goodman’s analysis—while

attribution becomes less of a challenge—achieving

geopolitical symmetry becomes essential. 27 This is not

feasible for many states.

As far as attribution is concerned, at this point in

time there are certainly some serious difficulties for

deterrence, including issues of super-empowered individuals

and a lack of a desire for cooperation between national

security agencies. It must be remembered, however, that in

the cyber domain major technological change can come

27 Goodman, “Cyber Deterrence,” 109-10.

quickly. Current technological difficulties in attribution

may be of minor significance in the near future.28

Another topic that bridges the international system

level and the didactic level is the ambiguity that pervades

cyber aggression. Martin Libicki notes, "No one knows how

destructive any one strategic cyber war attack would be."29

One of the things that theoretically made nuclear deterrence

so effective was “assurance;” in the case of nuclear weapons

and U.S.-Soviet relations, this meant the guarantee of

mutual destruction. In the cyber domain, however, attackers

can only guess as to how their attack will work. Imagine,

for instance, one state simply wants to test another state’s

cyber defenses but ends up bringing down critical systems.

It is highly likely the attacked state will assume that

inflicting severe damage was the original intent and respond

accordingly. If they attempt to respond in the cyber-

28 Stuart Baker, “The Attribution Revolution: A five-point plan to cripple foreign cyberattacks on the United States,” http://www.foreignpolicy.com/articles/2013/06/17/the_attribution_revolution_plan_to_stop_cyber_attacks#sthash.egfHQvkT.dpuf29 Libicki, Cyberdeterrence and Cyberwar, xv.

domain, however, they too will have to try to account for

the unpredictability of cyber. A calculation of

effectiveness for a cyber attack is far more difficult than

for a conventional attack. Another related issue is that

cyber attacks quickly lose efficacy. “Attacking and

retaliation in cyber both have the problem of subsequent

attacks (using similar methods) being less effective.”30

It is clear, with all this ambiguity, that there is

plenty of room for misunderstanding. It is not clear,

however, if this difficulty of predicting the results of an

attack makes deterrence more difficult. For one thing, it

likely makes cyber at best a supplement for conventional

attacks—it is not dependable enough to be an integral part

of a vital operation.31 There are arguments that even an

operation like Stuxnet, which captured the imagination of

analysts because of the amazing reach of cyber power, was

really the result of a failure of the real program and

represented only a temporary setback for Iran.32 This

30 Libicki, Cyberdeterrence and Cyberwar, xvii.31 Ibid., xx.32 Langner, “Stuxnet's Secret Twin.”

uncertainty in cyber efficacy brings to mind Robert Osgood,

who argued, “up to a point, the element of uncertainty in

nuclear deterrence...may contribute to caution and

restraint.”33 Like Osgood warns, however, too much hope

should not be put in this uncertainty.

Another challenge at the didactic level is credibility.

Greg Cashman writes, “[T]he central problem in deterrence

theory is how to make threats credible.” 34 This can be very

challenging in the cyber domain. Thomas Schelling's method

of solving the credibility problem was the fear of things

getting out of hand.35 This seems difficult to apply to the

cyber domain. The idea that a rational state would be

willing to go to full-scale war over a relatively mild

cyber-attack would leave many potential cyber aggressors

incredulous.

Goodman says that, traditionally, credible deterrent

declarations depend on certainty, severity, and immediacy.36

33 James E. Dougherty and Robert L. Pfaltzgraff, Jr., Contending Theories of International Relations, 356.34 Cashman, What Causes War?, kindle locations 11339-11341.35 Dougherty and Pfaltzgraff, Contending Theories of International Relations, 359.36 Goodman, “Cyber Deterrence,” 107.

The immediacy component seems especially difficult—even if

attribution can be accomplished, doing so requires time.

Likewise, as pointed out above, an overly severe threat may

actually render the threat less credible in cyber space.

Goodman, however, holds that immediacy and severity are not

as important in this domain. Cyber attacks are usually not

nearly as grave as nuclear attacks or even most conventional

attacks; the response to a cyber attack should be

commensurate.37 Likewise, the necessity of immediacy was

based on the premise that in a nuclear war the counterattack

must happen before all the capabilities to counter are

destroyed or disabled. In the cyber domain, this level of

destruction—such that it wipes out all counter-attacking

capabilities—is highly unlikely. Therefore, the requirement

of credibility for the cyber domain is only based on

certainty, according to Goodman.38 This brings us back to

the centrality of attribution. If adequate time will result

correct attribution, the credibility of a deterrent

declaration seems more tenable. Libicki, as mentioned 37 Ibid., 108.38 Ibid.

above, does note that, if attribution drags on too long, it

could be difficult to convince neutral observers “that

retaliation is not aggression.”39 This, however, seems

manageable.

The didactic level can, theoretically, be seen as

containing cyber deterrence challenges so intractable as to

be prohibitive. Yet, some of the thorniest issues, such as

attribution and credibility, when looked at from the lens of

geopolitics, appear less daunting—as long as a state enjoys

at least geopolitical symmetry with those states it hopes to

deter.

Unit Level

The unit level, and differences between states, has the

potential to greatly affect the way different states respond

to the demands of the cyber domain. For the purpose of this

paper focus will mostly be given to the U.S. domestic

situation, but a comprehensive study of myriad countries

would surely be fruitful.

39 Libicki, Cyberdeterrence and Cyberwar, xvi.

Moving to the unit level of analysis but returning to

the topic of law as a potential framework for dealing with

cyber attacks and cyber crime, U.S. domestic law provides

some insight into how the unit level affects states’

behavior in the cyber domain. Domestic law, given the

limited applicability of international frameworks, often

represents the best option for such a framework. Indeed,

domestic criminal law is an “important tool for combating

cyber attacks, including those that cross international

borders.”40 The main drawbacks, however, include the lack

of updates to meet the modern challenges of cyber attack and

the severe limits of due to a dearth of extraterritorial

reach.41 At the unit level, like the international level,

many states, like the United States, do not have a suitable

domestic legal framework to address most foreign-based cyber

attacks; domestic legal reform is an important challenge in

cyber deterrence.

The rate of innovation was mentioned earlier as a

potential boon to attribution efforts; it can also pose 40 Hathaway et. al., “The Law of Cyber Attack,” 883.41 Ibid.

unique challenges. There are serious questions about the

ability of legislation and Washington to keep pace with

these changes. Paul Rosenzweig notes that some people

“argue that the pace of threats on the Internet is so swift

that the designation and demand for compliance structure

adopted by the legislative proposal would, inevitably, be

too far behind in addressing actual threats."42 Beyond

simply getting the legal machine to move quickly enough,

there are legitimate concerns over the ability of national

leaders to anticipate important changes and trends in this

domain. As the Center for Strategic and International

Studies Commission on Cyber Security said, "Devising a

national strategy [for cyber] has proved to be difficult for

many reasons, the most important of which has been the pace

and direction of change in the international environment

exceeded our expectations and our ability to predict the

direction that change would take."43 Any talk of effective

42 Paul Rosenzweig, “The Internet ‘Kill Switch’ Debate,” Lawfare, February 2, 2012.43 Center for Strategic and International Studies Commissionon Cybersecurity, “Rebuilding Public Private Partnerships,” 12.

cyber deterrence must come to grips with the adaptability

and agility that such a policy must exhibit.

Delineating the responsibilities of the public sector

and private sector at the unit level also merits

consideration. Some clamor for more government guidance and

involvement. However, hasty reactions should be avoided.

The private sector is just as interested in eliminating

cyber attacks and cyber crime as the government is; power

companies do not want to see the grid down any more than the

government does.44 Only when two conditions are met should

presidential direction be necessary, according to

Rosenzweig. First, when the government has information the

private sector lacks and, second, when “the private sector

won’t voluntarily act on the government’s knowledge if the

government shares it.”45 The Center for Strategic and

International Studies sees the need for government

involvement and potential direction limited to only critical

infrastructures. They write,

44 Rosenzweig, “The Internet ‘Kill Switch’ Debate.”45 Ibid.

To focus the defense of cyberspace, we have identified four critical cyber infrastructures: energy, finance, the converging information technology and communications sectors, and government services...thesecritical cyber sectors are large, interconnected national networks that are the most vulnerable to broaddisruption by cyber attack.46

A melding of these two recommendations seems to provide a

reasonable way forward. Government intervention should be

limited to simply sharing information lacking in the private

sector. More active direction is needed only when there is

a threat to critical infrastructure and there are concerns

that the private sector cannot or will not take appropriate

action even if they have the pertinent information.

Along the same vein, the traditional relationship and

cooperation between the public and private sector matters

greatly.

Securing cyberspace requires government and the privatesector to work together. The private sector designs, deploys, and maintains much of the nation's critical infrastructure...cyberspace cannot be secured by the government alone. There is a bifurcation of responsibility."47

46 Center for Strategic and International Studies Commissionon Cybersecurity, “Rebuilding Public Private Partnerships,” 44-5.47 Ibid., 43.

Given the gravity of this relationship, differences between

states at the unit level—otherwise known as intervening

variables—will matter, as a neoclassical realist would

readily assert. As noted by President Toomas Hendrik Ilves

of Estonia speaking at Fletcher in September, some countries

have very “robust” public-private partnerships.48 These

countries do not run into nearly as many difficulties with

cooperation between the public and private sector—a fact

that can be leveraged to nefarious ends if desired.49

Others have a tradition of separation where private firms

often prefer to simply absorb losses than risk damage to

their reputation and share prices.50 This reticence results

in a lack of communication and obfuscation of the severity

of cyber-aggression. It is quite clear that a very close

private-public relationship has its benefits in cyber space

and states with a tradition of separation are finding their

current systems in need of adapting. 48 Toomas Hendrik Ilves, “What Keeps Me Awake at Night: Worries and Challenges for a Small European Ally,” speech delivered at Tufts University, October 9, 2013.49 Ibid.50 Amitai Etzioni, “Cybersecurity in the Private Sector,” Issues in Science and Technology, Fall 2011, 58-62.

This adaptation is not without its difficulties. The

frequent changing and reshuffling of personnel on the

government side of the equation wears away at the trust

between these two sectors. "Large, diffuse groups with a

floating population are not conducive to building trust.

Trust is also damaged when senior officials from government

agencies do not cultivate it and when government plans and

processes are opaque or inadequate."51 Further, there is a

lack of a clear vision for what must be accomplished and no

“articulated strategic initiative to guide private-sector

efforts.”52 To round off the difficulties countries like

the United States face in coordinating the private and

public sectors, the government feels it must share

information with everyone or no one—there is no middle

ground—and “there are the usual issues related to a

fragmented government that does not speak with a single

voice or act as a unified entity."53 Will Goodman makes the

51 Center for Strategic and International Studies Commissionon Cybersecurity, “Rebuilding Public Private Partnerships,” 45.52 Ibid.53 Ibid.

argument that tying cyber security to the pragmatism of

geopolitics can resolve many theoretical problems. When

examining public-private cooperation, however, it becomes

clear that sometimes practice has serious problems that most

theories fail to predict.

Continuing the unit level analysis and delving further

into the public-private relationship, concerns over civil

liberties adds more difficulty. In the United States,

companies are afraid to cooperate, in some cases, because of

the questioned legality of sharing private information.

When laws have been proposed to give them indemnity in such

situations, they have been killed in the legislature.

Whenever the topic [of cybersecurity] is raised, alarmssound from both sides of the political aisle. On one side, the intelligence community stresses that protection from cybersecurity threats is essential to national security, even if some personal liberties are sacrificed. On the other side, adamant proponents of personal privacy online seem to view themselves as someincarnation of Cassandra, announcing prophecies of the demise of privacy that fall on deaf ears.54

54 Carol M. Hayes and Kesan, Jay P., “At War Over CISPA: Towards a Reasonable Balance between Privacy and Security” (August 1, 2012). Illinois Public Law Research Paper No. 13-03; Illinois Program in Law, Behavior and Social Science Paper No. LBSS13-04. Available at SSRN: http://ssrn.com/abstract=2135618 or

Neo-classical realists would be quick to point out this

intervening variable. The foreign policy that the U.S.

government can pursue is highly dependent on a domestic

debate about values.

After assessing the unit level, we see there are a lot

of real challenges, especially for countries with relatively

cumbersome legal systems, complicated domestic politics,

concerns over civil liberties, and a tradition of a

meaningful separation between the public and private

sectors. Simply from the point of view of attempting to

implement practical, effective cyber deterrence measures,

these issues represent serious challenges for which answers

are not immediately forthcoming.

Group and Individual Level

Though many things could be examined at the group and

individual level, attention should be given to something

that they both share in common: the rational actor model.55

http://dx.doi.org/10.2139/ssrn.2135618, 1.55 Cashman, What Causes War?, kindle locations 1597, 3768.

Given that deterrence—cyber or otherwise—depends on

predicting or at least anticipating how potential aggressors

will make their choices, the rational actor model has often

been at the center of this calculus. Greg Cashman says,

“Several political science models—expected utility theory,

rational choice theory, and the rational actor model— all

incorporate this idea of procedural rationality.”56 The

idea that actors (be they groups or individuals) will

basically perform “cost-benefit analyses” when attempting to

make a choice—for instance, if the costs of committing a

cyber-attack will outweigh the gains of doing so—is of

central importance to deterrence. At the very least, the

decisions that they make must be consistent with the outcome

of a cost-benefit analysis, even if they do not make the

mental calculation in this way.57 The major problem,

however, is that this model has been challenged on several

fronts.58 Though not unique to cyber deterrence, having a 56 Ibid., kindle locations 1596-1597.57 Christopher H. Achen and Duncan Snidal, “Rational Deterrence Theory and Comparative Case Studies,” World Politics, Vol. 41, No. 2 (Jan., 1989), 164.58 Dougherty and Pfaltzgraff, Contending Theories of International Relations, 358.

model whereby one can attain a reasonable degree of

confidence in anticipating the decisions of other actors is

an essential challenge of any deterrence scheme.

Several scholars have found deficiencies with the

rational actor model. Greg Cashman points out that

“information failures due to self-deluding policies stretch

the limits of rationality and point to cognitive biases.”59

Robert Jervis finds that the rationale behind nuclear

deterrence was completely derivative of Western society—

ethnocentric—and therefore deficient.60 Cyber deterrence

must be sure to avoid this charge. Jervis also points out

that group dynamics, such as internal bargaining, “can

undermine the assumption that the state acts as a rational

actor.”61 Christopher Achen "rejects deterrence theory as

logically incoherent.”62 Further, different situations can

also affect people’s ability to behave rationally. Patrick

59 Cashman, What Causes War?, kindle locations 11019-22.60 Dougherty and Pfaltzgraff, Contending Theories of International Relations, 358.61 Robert Jervis, “Rational Deterrence: Theory and Evidence,” World Politics, Vol. 41, No. 2 (Jan., 1989), 204.62 Dougherty and Pfaltzgraff, Contending Theories of International Relations, 358.

M. Morgan cites crisis as a real impediment for traditional

deterrence.63 These are only some scholars who find the

rational actor model wanting.

Further, the assault on the model is not limited to

International Relations theorists. Baba Shiv, of Stanford’s

Graduate School of Business, ran an experiment that showed

attempting to remember a seven-digit number vastly reduced

people’s ability to make a rational choice.64 The

conclusion is that the more mental demands made of a person,

the less rational their subsequent decisions tend to

become.65 When thinking about individuals or groups tasked

with the responsibility to make decisions that affect

millions of peoples’ lives, one would be surprised if this

limitation were not somewhat applicable. Barry Schwartz

claims research has shown that the human brain can only

effectively take account of seven variables at one time.66

63 Ibid.64 Baba Shiv, “Heart and Mind in Conflict: Interplay of Affect and Cognition in Consumer Decision Making,” Journal of Consumer Research, 26 (December), 1999, 278-282.65 Ibid.66 Barry Swchartz, The Paradox of Choice : Why More Is Less (New York:Harper Perennial, 2003).

Just imagine how many variables are involved in many of the

decisions national leaders must make; some important

variables may be ignored due to cognitive limitations.

Again, there are warnings about the rational actor model.

As a response to some of these arguments against the

rational actor model, the work of Keith Payne, among others,

may provide a way forward. Payne admits that the model has

not always been applied well leading to “expectations of

foreign thinking and behavior” that are “grossly

inaccurate.”67 He argues, however, that one of the main

reasons for this failing is not that most actors on the

international stage are irrational, but rather they appear

to be unreasonable to those operating with different goals

and value systems. Payne writes,

Rationality does not imply that the decision-makers’ prioritization of goals and values will be shared or considered “sensible” to any outside observer. The goals and values underlying decision-making do not needto be shared, understood or judged acceptable by any observers for the decision-making to be rational. Nor does “rational” imply that any particular moral

67 Keith Payne, The Fallacies of Cold War Deterrence and a New Direction, (Lexington: The University Press of Kentucky, 2001), 6.

standards guide the route chosen to realize preferred goals and values.68

Rationality, rather, “is a mode of decision-making that

logically links desired goals with decisions about how to

realize those goals.”69 Payne, conjuring up echoes of

Winston Churchill speaking about Adolf Hitler, goes on to

argue that one must truly “know the enemy” to “gain a useful

understanding of an opponent’s beliefs, will, values, and

likely cost-benefit calculations under specific

conditions.”70 This echoes the warnings of Raymond Aron

"that deterrence must be situtation-specifc if it is to have

any real hope of effectiveness" in the post-Cold War

environment.71 If deterrence in the cyber domain is going

to be effective, it must be tailored to specific situations

and actors at the individual and group level.

Conclusion

68 Ibid., 8.69 Ibid., 7.70 Ibid., 111.71 Dougherty and Pfaltzgraff, Contending Theories of International Relations, 385.

After analyzing the challenges of deterrence in the

cyber domain at the international system, didactic, unit,

group, and individual levels, it is clear that these

challenges are nothing short of substantial. Some of the

problems, like attribution, may get much easier with

technological advancement. Will Goodman offers at least a

partial way forward when it comes to problems of the

didactic level, like super-empowered individuals, and some

of the anarchic elements of the international system. Many

of the issues at the state level—like the civil liberties

versus security debate—are not theoretically intractable,

but do demand serious compromise from different domestic

groups; if the problem of cyber aggression is serious

enough, however, middle ground will likely become more

attractive. At the group and individual level, more

detailed, situation-specific research could provide a way

past many of the problems levied at the rational actor model

and allow effective deterrence policies to be developed,

tailored to specific actors, as Keith Payne argues. Cyber

deterrence, though by no means easy, may not be impossible,

but at every level of analysis there are issues to be

addressed. As Keith Payne argues,72 pointing out the

challenges of deterrence can be met with resistance by those

who would prefer to more or less continue on the same path

with only minor alterations, but if progress is to be made—

and security to be increased—these challenges must be

addressed.

72 Keith B. Payne, Deterrence in the Second Nuclear Age (Lexington: The University Press of Kentucky, 1996), 160.

Works Cited

Achen, Christopher H. and Duncan Snidal. “Rational

Deterrence Theory and Comparative Case Studies.” World

Politics, 41, 2 (1989), pp. 143-69.

Baker, Stuart. “The Attribution Revolution: A five-point

plan to cripple foreign cyberattacks on the United

States.” Foreign Policy, June 17, 2013.

Cashman, Greg. What Causes War?: An Introduction to Theories of

International Conflict. Kindle Edition: Rowman & Littlefield

Publishers, 2013.

Center for Strategic and International Studies Commission on

Cybersecurity. “Rebuilding Public Private

Partnerships.”

Dougherty, James E. and Robert L. Pfaltzgraff, Jr.

Contending Theories of International Relations, 5th Edition. New

York: Longman, 2001.

Etzioni, Amitai. “Cybersecurity in the Private Sector.”

Issues in Science and Technology, Fall 2011, pp. 58-62.

Goodman, Will. “Cyber Deterrence: Tougher in Theory than in

Practice?.” Strategic Quarterly, Fall 2010, pp. 102-35.

Hathaway, Oona, et. al. “The Law of Cyber Attack.”

California Law Review, 100, 4 (2012), pp. 819-85.

Hayes, Carol M. and Jay P. Kesan. “At War Over CISPA:

Towards a Reasonable Balance between Privacy and

Security.” Illinois Public Law Research Paper No. 13-03;

Illinois Program in Law, Behavior and Social Science

Paper No. LBSS13-04, August 1, 2012. Available at SSRN:

http://ssrn.com/abstract=2135618 or

http://dx.doi.org/10.2139/ssrn.2135618.

Ilves, Toomas Hendrik. “What Keeps Me Awake at Night:

Worries and Challenges for a Small European Ally.”

Speech at Tufts University, October 9, 2013.

Jervis, Robert. “Rational Deterrence: Theory and Evidence.”

World Politics, 41, 2 (1989), pp. 183-207.

Kshetri, Nir. “Cyber-Victimization and Cybersecurity in

China.” Communications of the ACM, Vol. 56 No. 4, 35-37.

Libicki, Martin. Cyberdeterrence and Cyberwar. Santa Monica,

CA: RAND, 2009.

Langner, Ralph. “Stuxnet's Secret Twin.” Foreign Policy,

November 19, 2013.

Payne, Keith. Deterrence in the Second Nuclear Age. Lexington: The

University Press of Kentucky, 1996.

________. The Fallacies of Cold War Deterrence and a New Direction.

Lexington: The University Press of Kentucky, 2001.

Pfaltzgraff, Robert L., Jr. “Conflict, War, and Deterrence

Theories: Part II.” Lecture delivered at Tufts

University, December 12, 2013.

________. “Paradigms, Theories, and Levels of Analysis.”

Lecture delivered at Tufts University, September 12,

2013.

Rosenzweig, Paul. “The Internet ‘Kill Switch’ Debate.”

Lawfare, February 2, 2012.

Shiv, Baba. “Heart and Mind in Conflict: Interplay of

Affect and Cognition in Consumer Decision Making.”

Journal of Consumer Research, 26, December (1999), pp. 278-

292.

Swchartz, Barry. The Paradox of Choice: Why More Is Less. New York:

Harper Perennial, 2003.

Weber, Max. “Politics as a Vocation.” Essays in Sociology.

New York: Oxford University Press, 1946.