Shock absorbers are the pessimists of the automotive world. Their very existence is premised on the belief that roads and highways will inevitably include some bumps and potholes and debris of various kinds, no matter how hard our road designers and construction engineers and maintenance crews work to avoid them. Their purpose is not to correct these defects but to make them less damaging to vehicles and less painful to drivers and passengers. They reflect a recognition that imperfect roads are a chronic condition of transportation, one that we can minimize but not eliminate.
Our interwoven and dynamic world makes shocks—surprise developments that diverge sharply and suddenly from the trends preceding them, sometimes producing disastrous outcomes, sometimes not—all but inevitable. In recent decades, these shocks have included the disintegration of the Soviet Union, the 2008 financial crisis, and the Arab Spring, all discontinuous “black swan” events resulting from complex systems dynamics. The exact form and timing of these shocks are nearly impossible to predict, but their impact can be cushioned, their effects managed. If the first order of dealing with complex systems problems is to avoid treating them as if they were linear problems, susceptible to piecemeal resolution, the second is to build resilience into the system, rendering shocks, when they come, less damaging than they would otherwise be.
A focus on resilience differs in significant ways from efforts to promote stability. Stability is, all things being equal, usually a desirable thing. But in some contexts, the search for stability can be counterproductive. Buildings that are too rigid will crumble under the stress of an earthquake. Similarly, stability strategies that emphasize preventing danger and eliminating risk in the realm of foreign affairs can become too rigid to withstand shocks, too unimaginative to adapt to challenges, too protective of the status quo to adjust to change.
When stability strategies fail, as in the prelude to World War I, their failure can be catastrophic.
In the second half of the nineteenth century, the European system grew increasingly rigid, unable to adjust incrementally to new geopolitical challenges and changing domestic politics. Europe devolved into a system of rival alliances focused on reinforcing the bonds within each camp rather than maintaining broader equilibrium on the continent. Diplomacy lost touch with new technologies and their implications for warfare, and proved unable to cope with the imperatives that were driven by the advent of the railroad and the advantages that would flow from preemptive attack. As a result, the system amplified rather than buffered disturbances and became highly susceptible to shocks generated by relatively minor disputes.1
Resilience strategies, by contrast, acknowledge the inevitability of change and the probability of shocks, and they focus on more general capabilities to respond to hazards when they occur, regardless of what they are.2 California’s Loma Prieta earthquake in 1989, for example, caused the iconic fifty-two-story Transamerica building in San Francisco to sway wildly but produced no lasting damage to the flexible structure, while more rigid buildings not far away collapsed altogether. The early nineteenth-century Concert of Europe featured similar resilience. Prior to the shocks of the Crimean War in 1853 and German unification in 1871, it had been an adaptable and flexible system, able to make incremental adjustments as needed to manage disturbances. It was aimed not at preventing conflicts altogether but at limiting their impact and containing their dangers. How might we build such resilience into the system that is both shaping and shaped by US-Russian relations, improving its ability to bend but not break under stress?
The successful management of any crisis begins with communications. In the wake of the Cuban missile crisis, the United States and Soviet Union established the so-called hotline linking leaders in Washington and Moscow, because it was evident that direct and timely communications were vital to managing crises in ways that minimized prospects for escalation. A similar recognition about the need for American and Russian military commanders in Syria to avoid accidental clashes between their forces led them to establish an official deconfliction channel there in 2016. Extending this narrow channel into broader US-Russian discussions of how we might handle possible security crises in Europe and beyond would be an important preparatory step toward better crisis management. There is much potential danger to discuss in Ukraine, North Korea, Iran, and elsewhere, in our parallel efforts to battle terrorists, and in how we handle the thorny issue of cyberoperations, including false-flag cyberattacks designed to deflect blame or even spark US-Russian conflict.
Communications are not only useful in avoiding accidents and misunderstandings; they are also the best means available for contending with two unavoidable problems that flow from twenty-first-century technology: time and ambiguity. Time played a critical role in the tragedy of World War I, as military leaders pressed to mobilize forces quickly to outflank the enemy rather than be outflanked, and diplomacy relying on the slow transmission of letters and telegrams could not keep pace.3 Time is even more problematic today. News and information flow at light speed across the globe, and the demand for almost instantaneous commentary on cable television and in social media and other digital forums puts enormous pressures on governments to issue statements and adopt policy positions well before the facts of an emerging conflict situation—or their meaning in larger context—are evident. The rapidity with which high-precision weapons systems can strike targets half a world away could force national leaders to make fateful military launch decisions in minutes, rather than hours or days. Moreover, the inherent ambiguity of the cyberworld, coupled with the entanglement of nuclear and conventional weapons systems and the vulnerability of satellite networks, means that they might have to make such decisions with little clarity about the origin or intention of devastating digital attacks. The combination of shrinking time, increasing ambiguity, and burgeoning streams of data is driving increased reliance on artificial intelligence and automation to assess and respond to threats. This can potentially reduce the role of human decision-makers, increase the chances of accidents, and narrow the opportunities for compromise. The time for discussing how to handle such dangers is now, not in the suffocating heat of a crisis.
For maximum resilience, formal lines of communication are not enough. In high-stakes crises, emotions can run hot and the mooting of compromises can be a most delicate matter. Lower-level representatives engaging in off-the-record discussions have greater freedom to explore possible off-ramps and convey the limits of flexibility than do heads of state. In the Cuban missile crisis, the “confidential channel” between US Attorney General Robert Kennedy and Soviet ambassador Anatoly Dobrynin played a crucial role in averting disaster. In fact, the ultimate resolution to the crisis—a secret understanding that Kennedy would remove American nuclear missiles stationed in Turkey after the Soviets dismantled missile sites in Cuba, an understanding that would not be put in writing or acknowledged publicly—was brokered between Kennedy and Dobrynin.4
The fact of a confidential channel is less important, however, than the quality of the communication within it. That quality, in turn, depends both on the skills and wisdom of the specific personnel involved and on the recognition by leaders in both Washington and Moscow that they must invest in building and safeguarding the relationships that the channel requires for effectiveness. Dobrynin observed in his memoirs that a channel that ensured “possibilities for a candid if not always pleasant dialogue between the leaders of both countries … appeared to be the only way of preventing the Cold War from turning into a hot one.” But to work, he added,
[the confidential channel] has to be permanently available, and its immediate participants must possess a certain level of diplomatic and political experience and knowledge. Above all, the channel should never be used by any government for the purpose of misinformation. Of course, a diplomatic game is always being played, but deliberate misinformation is inadmissible, for sooner or later it is going to be disclosed and the channel will lose all its value.5
Dobrynin’s caution against transmitting misinformation in the channel reflects the critical but fragile role of trust in its effectiveness. The belief that one’s interlocutor is lying can quickly prove fatal to a relationship, particularly one in which the stakes are as great as those in the US-Russia relationship. A certain degree of personal trust—including that conversations in the channel will remain confidential—is a prerequisite for using the channel to clarify intentions, to communicate key redlines and vital interests, and to moot potential formal and informal understandings that might be too risky to explore directly at the presidential level. Sound relationships take time to establish, however; they cannot be created on the spur of the moment during a crisis. Incorporating that feature as an enduring element of the system is a critical part of enhancing resilience.
One of the factors contributing to systemic fragility is ambiguity about the sides’ strategic intentions. As discussed in this book’s first two chapters, part of this ambiguity derives from the inherent uncertainties involved in cyberoperations, and part is a result of each side’s growing conviction that the other is intent on its destruction. These perceptions are fueling escalatory spirals of suspicion and distrust and increasing the likelihood that each side will overreact in a crisis. There is no way out of this perceptual spiral other than strategic dialogue that directly addresses each side’s intentions and its perceptions of its most important national interests. Such a dialogue cannot eliminate genuine conflicts of national interest, but it could conceivably narrow ambiguity and help to minimize the role of misperception in driving each side’s actions. Given the present domestic political obstacles, pursuing a direct dialogue between the White House and Kremlin is not a near-term possibility, but so-called Track II discussions of prominent nongovernment experts could nonetheless help.
This approach to building systemic resilience through formal and informal channels of communication stands in contrast to the approach the United States has employed in recent years, which has treated official contacts with Moscow as something to be turned on and off in response to Russian behavior, either as a reward or punishment. Following the Russian-Georgian war in 2008, for example, the Bush administration took a step that was unprecedented in the post–Cold War period, cutting off all official US-Russian contacts above the level of deputy assistant secretary, in order to show the Russians “that they could not get away with something like invading Georgia.”6 Only a handful of years later, after full diplomatic contacts had been restored under Obama’s “reset” policies, Washington again attempted to send Moscow to the diplomatic sidelines following its annexation of Crimea and covert war in eastern Ukraine, joining other G8 states in kicking Russia out of that great power club, suspending cooperation with Russia in the NATO-Russia Council, and imposing travel bans and economic sanctions on named individuals in and out of the Russian government. These policies have largely continued under the Trump administration in the aftermath of the 2016 election-meddling controversy. The newly appointed Russian ambassador to the United States complained of difficulty getting meetings in Washington,7 and media reports that presidential son-in-law Jared Kushner attempted to create a “secret channel” for discussions with Moscow produced an avalanche of criticism.8
However justified the United States has been in imposing a price for Russian aggression, attempting to isolate Russia in our globalized world has failed in practice. Judging from Moscow’s continued belligerence against NATO in general and Ukraine in particular, banning Russia’s presence at the European “adults’ table” has done little to prompt reconsideration in Moscow, but it has almost certainly reinforced Russian perceptions of Western hostility and added to its incentives to subvert the Western alliance. And as doors have shut to Russian diplomats in Europe and the United States, Russia has responded with such steps as deepening ties to China, intervening militarily in Syria, and using undeclared agents of influence to establish relationships with prominent Americans, circumventing its diplomatic isolation—none of which has served US interests.9
Part of designing resilience into the system involves identifying critical “single point of failure” features and building backups that can compensate for their damage or loss. As processes that were once performed by humans have increasingly grown more automated, we have realized stunning efficiencies. Automated inventory and delivery systems allow “just-in-time” provision of food and consumer goods and massively reduce waste. Modern ports employ automated cargo-handling to improve capacity and reduce the costs of manual labor. These systems are not only vulnerable to cyberinterference from Russia and a range of other state-sponsored and individual cyberactors but also to extreme weather events flowing from climate change.
The American Global Positioning System (GPS) is but one example of a system on which numerous critical military and commercial functions depend, but which is highly vulnerable to disruption and at present is not adequately backed up by emergency alternative systems. Plans have recently been put in place to provide such a backup, but they have been moving slowly. Most experts have long agreed that the “eLORAN” system (Enhanced Long-Range Aids to Navigation), developed by the US Coast Guard, can provide an effective non-GPS-dependent timing and navigation system that can sustain critical national communications and other GPS-related functions in the event of a sustained outage, and provision for its further development and use was signed into law by President Trump in 2018.10 But discussions about the need for a backup system have been under way for nearly twenty years,11 and even now the new legislation makes implementation “subject to the availability of appropriations.” The pace of progress has lagged vastly behind the urgency of the problem.
Electoral systems are another area where we have been slow to grasp the magnitude of our vulnerability and slow to put backup systems in place. Research by cybersecurity professionals has long revealed numerous electoral system vulnerabilities, dating back more than a decade. “Ethical hackers” investigating system problems have been able quickly to penetrate voting machines even when they are air-gapped from the internet, suggesting that hostile “black hat” hackers could, if they chose, alter the casting of votes inside polling places.12 Despite belated efforts to improve election system security since the revelations of Russian interference in the 2016 election, voter databases remain susceptible to intrusions and data corruption, and the tabulation and reporting of vote tallies remain vulnerable to interference. There is no single fix for these problems, but one low-cost step that would significantly improve the resilience of our electoral process is the use of paper ballots to back up electronic voting and the employment of low-cost scanners to record and tally those ballots. Many—though far from all—state electoral commissions have begun to adopt this approach.13
Other parts of America’s critical national infrastructure are in similar need of greater resilience. Many of our power plants, water systems, ports, and energy supply networks are dependent on automated industrial control systems (ICS) and supervisory control and data acquisition (SCADA) systems that are highly vulnerable to cybersabotage. Too few of them have independent backup systems in place that could function in the event of an emergency. One of the factors that allowed Ukraine to limit the impact of Russian cyberattacks on its power grid in 2015 and 2016 was that its relatively unsophisticated network could function manually when automated systems failed.14 Building such fail-safes into America’s power grid and other critical infrastructure would require a high-cost, long-term effort.15 Without one, however, our vulnerability to potentially paralyzing cyberdisruptions will continue.
One of the domains most at risk—yet designed as if it were immune to threats—is not on the ground but in space. The United States is increasingly dependent on space-based assets for both civilian and military communications, early warning of missile launches and other military threats, and delivery of precision-targeted weapons to their targets. Nearly every aspect of America’s “smart warfare” capabilities on which our military superiority depends is in turn dependent on space. But our potential adversaries, including the Russians, have begun to regard this dependence as America’s Achilles’ heel, and they have been rapidly increasing their ability to target our space-based assets, turning one of our biggest assets into a problematic vulnerability.16 Ground-based kinetic anti-satellite weapons can increasingly threaten satellites in a variety of orbits, laser weapons and other high-technology devices can disable or disrupt them, and cyberweapons potentially can corrupt or interfere with their operations. Moreover, much of this satellite network is dual-purpose, supporting both conventional and nuclear operations. This commingling means that Russian actions meant to undermine our conventional military effectiveness in a regional conflict by targeting vulnerable satellites could, by accident more than design, threaten our strategic systems and spin into a broader nuclear confrontation. According to General William Shelton, former commander of US space forces, “Our satellites were not built with such threats in mind.”
New threats to our space-based assets are a reality that we cannot eliminate. Building greater resilience into our systems could help, however. Since the end of the Cold War, our satellite technology has grown more capable and better able to endure the stresses of a nuclear war, but to save scarce budgetary dollars, this has come at the cost of redundancy.17 According to Elbridge Colby, a former deputy assistant secretary of defense in the Trump administration, building separate space-based systems to handle nuclear and conventional applications could not only add valuable redundancy but also reduce the chances of inadvertent escalation from conventional to nuclear war, and provide tactical warfighting advantages to the United States in a potential confrontation with Russia.18 Such an effort would necessarily require an expensive, long-term program, but the benefits could be well worth the costs.
This overall approach to building technological resilience would represent a significant shift in the way the United States has approached its critical infrastructure. During the Cold War, our infrastructure was largely invulnerable to attack outside the context of an unlikely nuclear war. We were able to treat our space-based assets as secure, and we viewed our ground-based infrastructure as an essentially economic asset beyond the reach of our adversaries. But the advent of the cyberage requires a change of perspective, so that we begin to see our critical infrastructure as a key component of national security even when the connection to military operations is indirect, in much the same way that the federal government decided in the 1950s that the country required an interstate highway system as a matter of national security to help move troops and weapons, not just for more efficient personal and commercial transportation.
A system without rules is inherently fragile. The US-Soviet relationship was most vulnerable to instability during the early Cold War period when rules to govern it had not yet been clearly established. The crises over Berlin and Cuba had their roots in Soviet probing of the limits of America’s tolerance and of its willingness to run risks. As those limits became evident, a set of formal and informal rules began to take shape that reduced the chances that third world proxy wars and other forms of superpower competition in the 1970s and 1980s might escalate into broader conflict. Today, both the cyberworld and the space domain are in an analogous state to the early Cold War period, with no consensus among the great powers on rules that might contain dangers.
But a system with rules that are too rigid can fail to adapt to change and manage shocks. The arms control regimes and Helsinki Rules that characterized the mid-to-late–Cold War world have crumbled in recent years because they failed to adjust to the stresses of emerging multipolarity, new technologies, and political change. Europe’s security architecture and the governance of international financial institutions are current examples of areas in which America’s preferred norms are under challenge by Russia, China, or other emerging powers. Finding a balance between too much rigidity and too little structure is critical to enhancing systemic resilience in the twenty-first century.
Formal arms control arrangements can continue to have a place in an evolving set of resilient rules, but they are unlikely to play as large a role as they did during the Cold War. A weapons-based arms-control approach is ill-suited to the cyberworld, where malware is constantly evolving and where states have enormous disincentives to put their capabilities at risk by unveiling any specific features. Even in the nuclear domain, traditional approaches to arms control are growing more impractical as the lines between nuclear, cyber, and conventional warfare blur. Limits on weapons systems were effective during the Cold War, when each party to a bilateral treaty could balance its arsenal against the other, and anti-ballistic missile systems that might otherwise fuel an arms race were rudimentary and sure to be overwhelmed. But in a multipolar world, in which the United States and Russia must each contend not only with the other’s nuclear weapons but also with those of China and other nuclear powers, such a balance starts to break down. Modern ballistic missile defense systems can play a meaningful role in contending with the small arsenals of emerging nuclear powers, but in so doing, they inevitably affect the perception of balanced bilateral capabilities and mutually assured destruction that underpinned Cold War arms control. If one side can effectively shoot down at least some portion of the other’s missiles, the rival side is incentivized to improve its arsenal both quantitatively and qualitatively to compensate, complicating prospects for treaty-based limitations. Moreover, when the United States, Russia, and China all see each other to varying degrees as rivals, attempting to limit their nuclear arsenals to common numerical ceilings becomes unviable absent a mutual willingness to adopt a “minimal sufficiency” approach to deterrence rather than a “warfighting” approach, based on the belief that deterrence depends on the ability to fight and win a potential nuclear war against any combination of opponents. As defense expert Andrew Krepinevich put it, “In a world with three nuclear great powers, none can maintain parity with the combined forces of the other two.”19
Where formal treaty limitations might fail, however, informal understandings and agreed codes of conduct might help. Following the Cuban missile crisis, both sides implicitly recognized the dangers of stationing strategic nuclear missiles near the other’s borders, in large part because short distances to targets would greatly abbreviate the warning times within which to make counter-launch decisions and leave them dangerously prone to war by accident. The exception to that norm—the American deployment of Pershing II missiles in Europe in the early 1980s—proved the rule and ultimately produced the INF Treaty’s formal ban on intermediate-range missiles. A geographic restriction on the deployment of strategic weapons today could be helpful in reducing tensions.
Similarly, although legally binding limits on cybercapabilities are not feasible, a proscription against cyberintrusions or other forms of cyberattack on critical infrastructure—including electoral infrastructure, nuclear command and control systems, and nonmilitary satellite systems—should be possible, particularly if it were backed up by credible threats of retaliatory punishment should it be violated. Both in the cyberworld and in space, the United States and Russia share an interest in protecting their critical assets from damage or destruction, and each is largely dependent on the self-restraint of the other in refraining from attacks. Negotiating where the lines are drawn can be a basis for establishing and enforcing a resilient set of rules.
Agreeing on rules to govern cyberinfluence efforts, in addition to those involving cyberintrusions and attacks on critical infrastructure, should also be possible. Joseph Nye, a Harvard scholar who once chaired the National Intelligence Council, argues that the United States and Russia could draw a line between the “permitted soft power of open persuasion and the hard power of covert information warfare.”20 Openly acknowledged and properly labelled content would be declared in-bounds; deceptive, illegally procured, or covertly produced content would be out of bounds.
Skeptics in both the United States and Russia scoff at the notion that rules might somehow limit the actions of either side, which they see as having little respect for rules or treaties, while harboring deadly intent toward its adversary. And it is indeed true that when states believe their vital national interests at stake, neither formal nor informal rules will stand in the way of actions that they regard are necessary. Russia bemoans NATO’s bombing of Serbia during the Kosovo war, accusing the United States and its allies of end-running UN charter requirements for Security Council approval for such action, and ultimately changing established international borders by force.21 The West in turn charges Russia with flouting both the UN charter and OSCE principles in annexing Crimea and launching covert warfare in Ukraine. For the skeptics on each side, these perceived violations demonstrate the futility of pursuing a rules-based order.
But history shows that even deadly enemies can reach understandings that encourage mutual restraint and reduce the likelihood of a crisis arising from misinterpretation or miscommunication. Moscow and Washington certainly viewed each other as untrustworthy adversaries during the period of détente, yet both saw their interests served by understandings that limited and contained the dangers of their rivalry. Even if rules ultimately carry little weight in a conflict, they can reduce the chances of winding up in that conflict, particularly if they are bolstered by strategies to incentivize their observance.
When engineers design a building, bridge, or dam, they typically plan against past catastrophes. Constructions are sometimes built to withstand a “one-hundred-year storm” or worse, depending on the consequences of failure and the parameters of their budgets. Today, we live in a world of increasing ferment in which extreme events, the equivalent of five-hundred-year storms, are becoming increasingly common. And as these threats have multiplied, technological advances have made the consequences of disaster existential. Designing resilience into the complex system surrounding US-Russian relations is rapidly becoming more difficult, even as it grows in importance.
The common temptation amid periods of rapid change and growing peril is to attempt to “stand athwart history, yelling, ‘Stop!’” to borrow William F. Buckley’s memorable turn of phrase. But history has repeatedly been unkind to uncompromising defenders of the status quo. Looking back on modern history, one can see a steady pattern of one-hundred-year extreme geopolitical events among the world’s great powers, including the bloody Thirty Years’ War, the Napoleonic conquests, and World Wars I and II. And while these events have come with impressive regularity, they have also increased in their destructiveness over time. That the United States and Soviet Union managed—through a combination of good policies and good luck—to avoid such destruction during the Cold War is an anomaly historically. Those attempting to engineer our complex post–Cold War system would be wise not to base their designs on the assumption that our relatively recent string of good fortune will long continue.