Spies have a terrible reputation. As agents of deception and theft, they arouse suspicion and anger wherever they go. States that employ spies risk a hostile response, suggesting that espionage is a precursor to war. Perhaps this is one reason why they have been held in such disdain for so long. Ancient Greek generals were heroic figures; ancient Greek spies were subject to torture and execution. The Arabic tradition viewed foreign intelligence agents as beneath contempt, and the Arabic word for spies, jasous, was originally a slur. The French word espion was a pejorative in Ancien Régime France, associating spies with agents provocateurs who destabilised governments and opened the door to foreign aggression. The Catholic and Anglican churches formerly referred to the day before Passover as ‘Spy Wednesday’, in commemoration of Judas Iscariot’s efforts to see how much money he could receive for betraying Jesus. Historians have linked enthusiasm for espionage to rapacious authoritarian regimes such as the Mauryan and Mogul Empires. Their leaders found inspiration in Kautilya’s Arthashastra, which includes ruthless recommendations on the dark arts.

The presence of spies sparks fear of invasion, and spy scares have a long history. Fear of fifth columnists and saboteurs abounded in France and Britain during the Hundred Years’ War. Fears of German espionage roiled Great Britain before the First World War. Fear of British espionage led to a panic in the Soviet Union in the interwar period, culminating in a brief but intense war scare in 1927. Fear of communist infiltration animated McCarthyism in the United States in the early Cold War, causing some to worry that a hot war was coming. In these and other cases, spies were viewed as both sinister and repugnant. No surprise, then, that they rarely received the protections afforded to diplomats or prisoners of war.

For liberal statesmen at the turn of the twentieth century, the antidote to secret intelligence was transparent diplomacy. Transparency alone would not guarantee international stability, but it would make it easier for leaders to work together. Resisting the urge to hoard secrets helps reduce fear that others are cheating on agreements, shirking their obligations or plotting military aggression. International institutions, not secret intelligence agencies, were the safest places to gather information. National security did not flow from espionage and subterfuge but from open diplomacy. It is no accident that Woodrow Wilson’s post-war vision rested on ‘Open covenants of peace, openly arrived at’. Of his 14 points, this was the first and most vital.

Statesmen, however, proved to be hypocrites. They condemned the practice of espionage while engaging in it with great energy. This was true during Wilson’s time, and long before it. Indeed, this hypocrisy preceded the development of modern intelligence agencies by millennia: the ancients spied on each other all the time despite their expressions of moral outrage. Early modern Europeans treated spying as reprehensible, while simultaneously developing the rudiments of intelligence tradecraft. Indeed, some scholars have suggested that the birth of the bureaucratic state in Europe was closely tied to the need for more reliable methods of gathering and organising secret information. In this respect, spying and state-building were linked from the start. Modern liberal democracies have engaged in covert action repeatedly, perhaps believing that it is a necessary stopgap against illiberal regimes. Leaders have also rationalised their activities by distinguishing military scouting from espionage. Only rarely have they acknowledged the nature of their business. Sir James Harris, the eighteenth-century English ambassador to The Hague, was resigned to the necessity of spying though he was clearly not happy about it. ‘I abhor this dirty work,’ Harris wrote, ‘but when one is employed to sweep chimneys one must black one’s fingers.’

Statesmen thus have a strange relationship with spies and spying. Many view intelligence agents as little more than glorified thieves, and they have long worried that espionage cuts against earnest efforts to make peace. Yet they find espionage irresistible, and they have built increasingly large and sophisticated intelligence bureaucracies in the name of national security.

Scholars are similarly divided. Whether intelligence is a force for peace and security, or a source of crisis and war, remains an open question. Some warn that intelligence is dangerous because it is difficult to distinguish espionage from war preparations. The targets of intelligence cannot be certain that their rivals simply want to gather information. Very often the tradecraft used to learn more about a foreign state is the same as the tradecraft used to lay the groundwork for military action. The target, then, might fear the worst when it uncovers espionage efforts on its own soil. This is an intelligence twist on the security dilemma, the notion that one state’s attempt to improve security makes its rivals feel less secure. Spying might help one state feel better by improving its awareness of rival capabilities and intentions, but the act of spying is unsettling. Observers have recently warned that cyberspace is particularly prone to this problem, because the tools and techniques used to exfiltrate digital information are also prerequisite to cyberspace attack. Attempts to penetrate an adversary’s nuclear command, control and communications network are especially risky.

In addition, espionage sometimes reveals opportunities for attack. Such knowledge is tempting, especially if the state believes that its information advantage is a wasting asset. Good intelligence encourages hope of a quick decisive victory over an unsuspecting enemy, and fear of missing out. The goal of intelligence agencies is asymmetric knowledge — one side gains the other’s secrets while protecting its own — but this asymmetry encourages opportunistic violence. Israel launched the Six-Day War, for example, in part because it had excellent intelligence on Arab armies.

But in other circumstances, espionage may be a force for peace. Suppose, for instance, that lousy information is a cause of instability. Conflict is costly so states should rationally seek to avoid it, yet they do not know enough about the balance of power to arrive at reasonable compromises. Unequal access to information also makes it difficult for states to commit to peace, because they have a sensible fear of being the sucker. Intelligence agencies could resolve this problem by making critical information available to policy makers and diplomats on all sides. Private information asymmetries that inhibit peaceful bargaining will disappear if peacetime adversaries maintain a competent and productive intelligence bureaucracy. Armed with reasonable knowledge, they will not have to take up arms.

Intelligence might also help resolve the security dilemma, which has more to do with uncertain intentions than with the objective balance of power. States might have good information about adversary forces, but little insight into how adversaries plan to use them. Espionage that reveals this knowledge might help alleviate the problem, reducing the risk of war by misperception or miscalculation. War is more likely when information is hidden and obscure; intelligence agencies are in the business of finding it so that decision makers are not operating in the dark. States are less likely to lash out if they have a reasonable understanding of their rivals.

None of this is to suggest that intelligence collection leads to friendly relations. What it does suggest is that espionage can channel political rivalries away from violence. War is less likely if states are locked in an intelligence contest among dedicated secret services. Spy-versus-spy battles preclude actual battles. States might tolerate foreign espionage because they understand that the alternative is much worse. During the Cold War, the United States and the Soviet Union grudgingly accepted that spying was inevitable, and intelligence professionals on both sides developed ‘rules of the game’ about acceptable and unacceptable methods. In some cases, they went further, allowing their rivals to act covertly and giving them the benefit of plausible deniability. Allowing the other side to operate in the shadows served as a good way of controlling escalation. Covert action was a useful release valve.

States also advertise their intelligence capabilities to disabuse rivals of the idea that they can remain hidden. This should give would-be aggressors a moment of pause. Sometimes states surveil their rivals — and quietly let them know about it — to reduce their enthusiasm for war and military escalation. In other cases they reveal intelligence in public, as in the ongoing Ukraine crisis, where US leaders have sought to remove Russia’s ability to manufacture a phoney pretext for war. Military plans depending on deception and surprise should become less appealing if planners cannot operate in the shadows.

Finally, secret services can serve as conduits for subterranean diplomacy among adversaries who find it difficult to negotiate in public. They can search for signs that long-time rivals are interested in better relations, and set up back channels that might lead to meaningful talks. Domestic hawks usually bash policy makers for reaching out to states they view as implacably hostile. In order to prevent these critics from becoming diplomatic spoilers, secret intelligence services facilitate quiet talks that let the air out of crises. Because they spy on each other, they know each other. The information channels they create to steal information are also useful for sharing it.

Intelligence sometimes lends stability to international relations. Sometimes it makes things worse. How do we know when spying leads to war or peace?

One possible answer has to do with innovation and technology, both of which are topics of enduring interest in the intelligence world. Espionage has always relied on clever technologies for stealing secrets and for avoiding revelation of its activities. Intelligence agencies try to gain access to valuable information without being discovered. But the more valuable the information, the higher the risk of discovery, because states pay closest attention to their dearest secrets. The consequences are especially dire for individuals who spy on behalf of foreign intelligence agencies. Other costs of discovery include the loss of control of sources and methods, which can leave secret services in the dark, or vulnerable to deception. Most relevant here is the danger that revelation of spying activities can lead to unnerving diplomatic crises.

States often turn to technology to reduce these risks. Reliance on technological collection platforms, as opposed to human sources, reduces the physical risk to intelligence personnel. Innovative concealment techniques — from camouflage to codes — protect sources and methods from exposure. And some technologies permit collection from very long distances. Because they are less intrusive, they might be less provocative. Ideally, distant platforms allow intelligence agencies to collect important information about the balance of power without inadvertently causing a crisis. In this way, intelligence technology can ease the problem of misperception and reduce the danger of war.

This argument is appealing. It is logically coherent and consistent with day-to-day international relations. States are hostile to foreign agents trying to recruit human sources, but they usually seem less upset by the notion that foreign satellites are watching them from long distances. Perhaps there is an unspoken acknowledgement that states can keep an eye on one another, as long as they do it from afar. Less intrusive collection methods are less irritating. And sometimes states even welcome this kind of collection, as it gives them an opportunity to show off new capabilities without having to parade them publicly. Doing so bolsters deterrence by quietly revealing their military strength.

Yet there are limits to this logic. One is the historical record. States try very hard to conceal their activities, contriving elaborate camouflage schemes and creative efforts to misdirect foreign observers. These efforts benefit from secrets elicited from foreign spies. In the late 1970s, for example, the Soviet Union deliberately targeted US defence contractors for this purpose. Such activities are inconsistent with the idea that states understand the value of foreign surveillance, much less that they welcome it.

In more extreme cases, there have been attempts to destroy adversaries’ distant platforms. This has led to a peculiar sort of arms race, as states invest in new technologies that allow them to collect from ever greater distances, while their targets invest in technologies to stop them. In the late eighteenth century, for example, states started to use hot-air balloons for battlefield observation. Military commanders liked the idea of being able to see behind enemy lines, but there were practical limits to what early balloons could accomplish. Nineteenth-century innovations increased the altitude and manoeuvrability of balloons, yet they were still large and slow and inherently vulnerable. The First World War saw the use of powered flight, which allowed for more aggressive collection, but defenders soon developed tactics that made overhead reconnaissance a very dangerous trade. A similar story played out in the Second World War, when increasingly higher altitude aircraft met increasingly long-range defences.

The technology race continued into the Cold War. Seeking information about Soviet military and economic activities, the United States developed very high-altitude aircraft such as the U-2, an aeroplane that some believed was invulnerable to interception. It was not. In 1960, Soviet forces shot down an American U-2 and captured the pilot, causing a diplomatic crisis during a period of high drama. Increasingly lethal air defences encouraged both countries to escape the atmosphere altogether and deploy space-based imagery satellites. Today, orbital imagery is a fixture of intelligence, yet there are real concerns about the safety of satellites. Recent anti-satellite tests by China and Russia reinforced these fears. Stand-off intelligence platforms might offer a safe way of eliciting information that increases stability, but what counts as a safe distance remains unclear.

What do we make from this curious history? Intelligence contests might reduce the risk of war by providing information that reduces insecurity, and by providing non-military options for states in competition. Sometimes states seem to act according to this logic, letting intelligence contests play out according to tacit rules of the game, and quietly accepting covert action as an alternative to military force. At other times, however, states are often infuriated by espionage and covert action, and spy scares lead states closer to conflict. The evolution of imagery suggests that technology alone cannot account for this shifting attitude towards intelligence.

This is not to say that technology is irrelevant; states pay close attention to new innovations and spend a lot of time and money devising counter-measures. The nature of their response, however, probably depends on deeper political factors. High-tech spying is tolerable when underlying relations are dull and predictable, and when states are content to let intelligence agencies monitor the peace. But when relations sour, states are likely to fear the worst. In these cases, they are more likely to view foreign espionage as a prelude to hostile action. Even distant technical collection activities look sinister in a crisis. Intelligence activities do not cause political conflicts, but they might exacerbate existing tensions.

This logic might tell us something about cyberspace competition. Some have warned that the domain’s peculiar technological characteristics make digital spying a dangerous business. The reason is that cyber-space espionage and sabotage use the same methods. From the defender’s perspective, it is impossible to tell whether an intrusion is simply designed to steal information, or whether it is a prelude to a damaging attack. The result is a ‘cybersecurity dilemma’ that increases the chances of conflict. Escalation by miscalculation is more likely, according to this logic, when it is impossible to tell the difference between normal intelligence gathering and preparations for offensive action.

Yet cyberspying has been a fact of great power politics for decades, and there is no evidence that it has led to great power crises. The United States and China routinely accuse one another of large-scale espionage, for example, but this has not led to crises or military escalation. Instead, their cyberspace activities appear to be part of a long-running intelligence contest rather than a test of raw power. Intelligence contests are competitions in which great powers fight over information rather than military dominance. Indeed, they can be a release valve when great powers are locked in competition but prefer to avoid conflict. Rather than foreshadowing violence, aggressive cyberspace espionage might suggest a durable cold war.

Aggressive collection could be dangerous in a crisis in which adversaries are mobilising for possible military action. In such a case, leaders might view revelations of cyberspace espionage as evidence that their enemies are preparing for war. They might reasonably fear losing important tactical information, or, in the worst case, losing control of their own communications. This is certainly plausible, and leaders seeking peaceful off-ramps from dangerous crises would be wise to take the danger into account. But again, the factors leading to instability in war are more political than technical. Misperception can occur in a number of ways in a crisis because leaders are deeply mistrustful of one another, by definition. They might misunderstand the purpose of military movements or mis- construe diplomatic rhetoric. Fractured and unstable political relations, not technology, are the underlying cause of trouble.

The relationship between diplomacy and intelligence is akin to the relationship between diplomacy and force. Scholars have long debated whether certain military technologies are inherently destabilising. According to this logic, the deployment of offensive weapons makes war likely because it makes conquest easy. Insecurity abounds when the offence is dominant, increasing the risk of arms racing and military conflict. Similarly, some have warned about highly complex militaries relying on elaborate coordination schemes. Technology and bureaucracy conspire to encourage risky behaviour because leaders need to move quickly, or their military forces will be disorganised and vulnerable to attack.

Others have challenged these claims. The same weapon system might appear to different observers as offensive or defensive. Much depends on their view of the adversary’s intentions. If observers believe that their adversaries are bloody-minded and risk-acceptant, they are more likely to view weapons as offensive. If they believe their adversaries are insecure and cautious, however, they will view the same weapons as less threatening. Technology matters on the battlefield, but the political implications depend on perceptions of international politics. Stable relations, even if unfriendly, will predispose leaders to view enemy technologies as relatively unthreatening. Unstable relations will cause them to assume the opposite.

The same might be true for intelligence. I say ‘might’ because this is very much an open question among researchers, and there are plenty of other candidate hypotheses about intelligence and war. But if this intuition is correct, then the implication is straightforward: intelligence is neither a force for peace nor a cause of conflict. Sometimes intelligence agencies can reinforce stability by revealing information that reduces misunderstandings. Rival states may coexist peacefully, letting their secret services battle for information rather than letting their militaries prepare for battle. In these cases, intelligence not only reduces the risk of war by clarifying the balance of power, it also provides a way for states to compete without risking too much. In other cases, however, espionage can make conflict more likely by aggravating mutual hostility and mis- trust. When relations are particularly nasty, states may be convinced that their rivals are bent on political agitation and regime change. Under these conditions the discovery of espionage efforts will look ominous and threatening, especially if intelligence chiefs are amateurs and their agencies are bureaucratically immature. Ham-fisted collection may be indistinguishable from attempted subversion, reinforcing everyone’s worst fears and creating powerful incentives to prepare for war.

All of this rests on the nature of state rivalries at any given time. Secret intelligence is a peculiar business, to be sure, but it is ultimately a tool of statecraft. How leaders use this tool, and how others respond, depends on their policy goals and strategic beliefs. In this sense, intelligence contests reflect and amplify international relations. Understanding the effects of spying will depend on grasping the politics behind the spies.