This past weekend, The Washington Post published a new account of the Obama administration’s struggle to respond to Russian covert activity during the 2016 election — and new details on the intelligence and internal debate about retaliation.

One fascinating revelation is that the White House did not simply punish Russia symbolically by expelling “diplomatic” personnel. It also fought fire with fire via a covert, retaliatory cyber operation.

On its face, a secretive and ambiguous action seems an unlikely choice for deterrence. As when President Trump used the “mother of all bombs” in Afghanistan, standard U.S. procedure is to issue loud and clear warnings to adversaries. While the details of Obama’s cyber operation remain classified, The Post’s reporting suggests that it was designed to be detected by Moscow and to imply Washington’s ability to inflict severe damage should Russia’s meddling increase. Thus, this particular covert response may have allowed the White House to threaten its adversary without creating a public spectacle and the domestic and international consequences.

Russian behavior and the American response is not the first instance of private, secret measures during the coercive contests that often arise in world politics. New research sheds light on why leaders sometimes prefer covert coercion and why such efforts may — or may not — work.

Coercion means bending the enemy’s will

Coercion refers to any attempt to alter another government’s decision-making through the threat of future costs. Such efforts need two essential ingredients to be effective:

1. Intelligibility — A threat of future costs must be expressed in a way that makes sense to the adversary you are trying to influence. This is particularly important if a government coerces through physical action rather than an explicit, verbal “do this or else” statement.

2. Credibility — Anyone can bluff. But, as Thomas Schelling points out, success requires laying out a believable course of action that will be triggered only if the target fails to act as desired.

Leaders have a large coercion toolkit. Usually coercion takes the form of highly visible and symbolic military maneuvers, such as the periodic repositioning of the U.S. Seventh Fleet to East Asia. On land, military mobilizations in the run-up to World War I signaled the resolve of Russia and Germany but also helped propel escalation. As The Post reports, Obama aides generated a “menu” of options that focused on cyber, economic and diplomatic punishments.

Choosing coercive tools with broad visibility can have credibility benefits. As Schelling originally argued, rattling one’s saber in front of a wide audience makes it harder to sheath the saber. Leaders that make threats and then visibly do not follow through can suffer “audience costs,” or a loss of popular support. Knowing this, adversaries may be more likely to believe the threat in the first place. While the significance of audience costs has been challenged by international relations scholars, the fact remains that most coercion efforts are high-visibility verbal and military threats.

The art of coercion in secret

Less well-understood is the art of “backstage” coercion. Privately communicated verbal warnings are one version, which Obama apparently used in a warning to Russian President Vladimir Putin last fall. Easily dismissed as “cheap talk,” these kinds of private warnings, as recent scholarship suggests, can carry diplomatic weight, especially when communicated in meetings among leaders face to face.

Governments can also use covert or otherwise nonvisible military actions to send a targeted message to an adversary. In the Korean War, the Eisenhower administration repositioned bombers used to deliver atomic bombs to quietly nudge the Soviet Union into a negotiated settlement. Eisenhower also used quiet military preparations only detectable by Soviet leaders to demonstrate resolve regarding Berlin in the late 1950s.

During the Vietnam War, the Nixon administration similarly manipulated U.S. air assets to simulate a nuclear alert in the hopes of pressuring Moscow and North Vietnam into a peace deal. The Carter administration paired public protests of the Soviet invasion of Afghanistan with a covert arms supply program intended to show Moscow its resolve to impose costs for aggression.

Do these efforts work? The track record is mixed. While Eisenhower’s atomic diplomacy may have helped with the Korean War armistice, historians have found little evidence that Moscow saw Nixon’s “madman alert” as credible — or even understood it.

Why is coercion harder to do in secret? For one thing, intelligibility is more challenging in the covert sphere. Success requires a careful design that functions like a dog whistle, inaudible and undetected by one set of observers (i.e., the public and third-party leaders) but audible and detected by the target — the adversary.

Making a quietly conveyed message credible is also tricky to pull off. As Keren Yarhi-Milo and I argue, this second ingredient can come from the costs and/or risks generated by a coercive covert action. Although it tends to be more subtle than overt alternatives, covert operations can still provide tangible proof their sponsor is willing to expend precious resources and incur political and other risks, including the risk of inadvertent exposure.

Israel’s covert strike against a suspected Syrian nuclear site in 2007 is a good illustration. The strike showed Israel’s willingness to significantly raise the risks of war and wide exposure to stop suspected nuclear proliferation, even as secrecy helped maintain face-saving opportunities for Damascus to react conservatively.

Assessing the Obama response to Russia

Detecting an ongoing Russian covert operation left the White House balancing competing demands, as The Post’s story describes. Obama needed to deter Russia from further sabotage on Election Day and similar operations in the future. Yet the administration also feared creating a spectacle that would have serious partisan implications at home and risks of a dangerous spiral abroad.

The need to show resolve and cope with constraints helps explain why the White House ultimately opted to pair symbolic public actions with covert cyber-retaliation. By targeting a government system and avoiding public acknowledgement, this cyber operation, like others, had a low public profile.

And how did the Obama operation do in terms of intelligibility and credibility?

Regarding the first ingredient, it seems quite plausible Moscow got the message. By “implanting computer code in sensitive computer systems that Russia was bound to find,” as The Post reported, the designers of the U.S. cyberattack clearly anticipated the need to create a dog whistle effect. The actions Russia was being warned to avoid were likely either clearly intelligible or clarified in private warnings.

Judging the second ingredient of credibility is more difficult. Did Russia find it believable that U.S. leaders would ever exploit cyber-vulnerabilities? A one-off cyber operation seems low cost. The more likely source of credibility is risk. Obama’s reprisal may have been perceived by Russia as representing a new level of cyber-aggressiveness by Washington. Alternatively, it may have simply been seen as a clever but ultimately empty “hello” message by an outgoing president, carrying little risk.

Because of its covert nature, assessing the success of the Obama administration’s response to Russian cyber-meddling will have to wait a few decades for the opening of government archives on both sides. Still, in its messaging structure, the Obama cyber-reprisal closely parallels nuclear alerts and other cases of covert coercion in the Cold War. Yet the 2016 election episode does suggest a larger lesson: Cybertechnology deployed by states may be ushering in a renaissance in the art of covert coercion.

Austin Carson is an assistant professor of political science at the University of Chicago and co-author with Keren Yarhi-Milo ofCovert Communication: The Intelligibility and Credibility of Signaling in Secret.”