The Limits of Plausible Deniability in Ukraine and Beyond

The Kremlin in Moscow. Photo courtesy of Pavel Kazachkov.

Guest post by Costantino Pischedda and Andrew Cheon

Drone strikes targeted Moscow last week. Though much remains unknown about it, the episode appears to be part of a series of unclaimed coercive attacks that US officials attributed to Ukrainian government personnel, including the killing of the daughter of a Russian nationalist, the sabotage of the North Stream pipelines, and drone attacks on the Kremlin.

With unclaimed coercion, perpetrators impose costs on adversaries to signal their resolve to prevail in disputes while denying involvement or simply not making any claim about responsibility. Unclaimed coercion is not unique to the war in Ukraine. Russia launched cyber attacks in 2007 to extract concessions from Estonia, though Moscow denied responsibility, and in 2010 Seoul claimed North Korea torpedoed a South Korean warship, Pyongyang’s denial notwithstanding.

Unclaimed coercion may have strategic benefits. Without unmistakable evidence about the identity of the perpetrator, the absence of a claim of responsibility creates plausible deniability, which, some argue, allows coercers to send intelligible, credible messages to targets while containing escalation risks. It may also reduce the costs of being seen as a norm violator. For instance, Austin Carson observed that, though both Saudi Arabia and the United States viewed the 2019 attacks on Saudi oil facilities as part of an Iranian coercive campaign, the absence both of conclusive evidence and a claim of responsibility prevented Riyadh and Washington from carrying out a tough military response, which would have looked “illegitimate while jeopardizing allies’ support.”

But there may also be drawbacks. Some research suggests that unclaimed acts muddle communication, leaving targets unsure about what is being demanded, making compliance less likely. Even if the coercive message is clear to targets, it may not be credible. If plausible deniability is seen as a way for perpetrators to shield themselves from the costs of appearing as norm violators by third parties and to contain escalation risks, targets may perceive unclaimed acts as cheap. A cheap action may signal unwillingness to engage in costlier and riskier overt acts to prevail in the dispute—that is, low resolve—which could embolden targets to resist.

Furthermore, the absence of a claim of responsibility could reduce the credibility of perpetrators’ reassurances implied in a coercive act (“if you do as I say, I will stop hurting you”). Targets may interpret perpetrators’ unwillingness to acknowledge responsibility as an indication of their deceitful nature, suggesting that promises to end hostile acts in exchange for concessions would likely be violated, in turn reducing incentives to comply.

Besides offering limited coercive leverage, unclaimed acts may fail to contain the risk of escalation. Targets, angry at being attacked, may confidently attribute attacks to the perpetrators despite a lack of evidence—and decide to retaliate. Incidents can engage targets’ reputation and honor, and thus be provocative, even when culpability is uncertain. For example, after the 1898 sinking of the USS Maine in Cuba, “Remember the Maine, to Hell with Spain!” became the popular rallying cry for war, even though it was unclear whether Spain was behind it.

Despite the growing interest in plausible deniability, there has been little empirical analysis of the effects of unclaimed coercion. Our research helps fill this gap, leveraging a vignette experiment, fielded in May 2022 on a sample of 854 US residents, recruited using sampling quotas to match Census statistics for gender, age, and education. The experiment exposed respondents to a fictional scenario depicting a major explosion at a NATO base in Poland used to funnel weapons to Ukraine during the ongoing war. All respondents were told that President Putin had previously warned of “unpredictable consequences” if NATO continued providing weapons to Ukraine and that both intelligence agencies and independent analysts identified Russia as the likely culprit without, however, ruling out the possibility of an accidental detonation. By randomizing whether Russia claimed or denied responsibility for the explosion, we were able to assess the effects of plausible deniability on targets’ views about complying with Putin’s demands and taking escalatory actions in response.

We found that when the attack is unclaimed, respondents are less likely to favor complying with Russia’s demands for interrupting the flow of weapons to Ukraine. Moreover, the absence of a claim of responsibility does not seem to affect respondents’ preferences for escalation. In our analysis of two possible escalatory responses by the United States to the explosion—an air strike against a Russian military base on Ukrainian soil and going to war against Russian forces in Ukraine—we found no statistically significant difference between the two groups of respondents. Thus, plausible deniability appears to reduce coercive leverage without the benefit of containing escalation.

Unclaimed attacks are among the possible tools at the disposal of governments, whose effects may vary depending on the circumstances. Policymakers considering resorting to unclaimed coercion—or fretting about its use by adversaries—should be aware that the payoffs are likely to be limited. For Ukraine, denying responsibility for attacks on Russian soil might offer the advantage of limiting reputational damage in the West, which would not be captured in our analysis. However, the evidence indicates that unclaimed attacks are unwieldy tools of coercion and are unlikely to reduce the risks of Russia’s escalatory responses.

Costantino Pischedda is Associate Professor of International Relations in the Department of Political Science at the University of Miami. Andrew Cheon is Assistant Professor of International Political Economy at Johns Hopkins SAIS.

Leave a Reply

Your email address will not be published. Required fields are marked *

You May Also Like