Documentation Nuclear Escalation Dynamics

Aim of the Dashboard

The interactive dashboard “Russia’s Nuclear Signaling in the War Against Ukraine” provides users with a comprehensive tool to track, analyse and understand nuclear escalation dynamics in Russia’s war against Ukraine. The dashboard aims to enhance the transparency and accessibility of nuclear signaling patterns between Russia and the West through interactive visualisations and data exploration features. Users can identify peaks and troughs in nuclear rhetoric and gain insights into the positions taken by key actors over time. The data covers nuclear-related statements and actions from autumn 2021 through January 2025.

📄 The dashboard builds on the accompanying research paper.

🎙️ For a quick introduction to the project, check out The Telegraph’s Ukraine: The Latest podcast.

📊 The full dataset is available here.

💻 The source code is available on GitHub and archived on Zenodo.

Timeline Visualisation: The feature enables users to explore the chronological progression of nuclear-related statements and actions. The interactive timeline displays signals from both Russia and the West. Users can zoom into specific time periods, identify peaks of nuclear rhetoric, and track changes in escalation dynamics over time.

Nuclear Escalation Relevant Signals (NERS): Signals are categorised as:

Russia

• R1: Affirmatory (Nuclear) Deterrence Signaling: Statements or communications by Russian or Belarusian officials that affirm, restate, or reinforce established nuclear deterrence principles, doctrines, or red lines, without introducing new elements or signaling change. In practice, this usually involves quoting the content of the official Russian doctrine that states the conditions under which Moscow would consider authorizing the use of nuclear weapons.
• R2: Active Coercive Nuclear Signaling: Communicative acts in which Russian leaders, officials, or affiliated figures set new conditions and redlines for nuclear weapon use that are more specific than the general Russian nuclear doctrine or even go beyond the doctrine. They are formulated as conditional “if-then” or “unless-then” statements, where the “then” usually stands for “we will use nuclear weapons,” “we will consider nuclear weapons,” “we will use all the means available,” or “we will be met with unforeseen consequences.” A good example of the latter is Putin’s statement at the onset of the February 2022 invasion: “Whoever tries to interfere with us, and even more so to create threats for our country, for our people, must know that Russia’s response will be immediate and will lead you to consequences that you have never faced in your history.”
• R3: Tacit Nuclear Signaling: Involves nuclear escalation-related signals that are communicated as actions rather than words. That includes, for example, military exercises with a nuclear component (e.g., Zapad), tests of nuclear or dual-capable launchers, non-routine movements of nuclear or dual-capable launchers, or the demonstrative first use of new dual-capable missiles against military or civilian targets in Ukraine (e.g., we would code the first use of Kinzhal or Oreshnik missiles in Ukraine war as “tacit nuclear signaling,” but we would not code a routine use of Iskander missiles as that is already a routine development that is no longer perceived as “nuclear escalation-related” in the West). In the dataset, we would normally code the description of the action as such (e.g., “Russia tested new ICBM”) or a neutral, official description of the action (e.g., “Russian Ministry of Defense issued a statement that Russia had tested a new ICBM”). Tests and exercises can also be accompanied by additional verbal signals that should be coded separately.
• R4: Anticipatory Policy & Posture Signaling: Teasing/hinting at the possibility of prospective future nuclear policy or posture adjustments, sometimes conditional on Western behavior, external events, domestic dynamics, or other developments. This includes verbal hints that Russia might suspend/withdraw from arms control treaties, revise its nuclear doctrine, resume nuclear testing, engage in nuclear sharing arrangements with allies, and other changes before they were formally adopted.
• R5: Escalatory Policy & Posture Adjustments: Formal modification of nuclear policy framework, strategic posture, or institutional arrangements governing Russian nuclear forces. This includes formally adopted revisions of Russian nuclear doctrine, nuclear sharing arrangements with Belarus, or arms control treaties status changes (e.g., New START suspension).
• R6: Nuclear Escalation Risk Signaling: Communicative acts — usually rhetorical ordeclarative — that highlight, dramatize, or warn of the risk of uncontrolled nuclear escalation, possibly resulting from Western or Ukrainian actions, without expressing an intent to use nuclear weapons or alter policy. These signals possibly aim to amplify the perceived danger of ongoing conflict dynamics, shape Western threat perception, and encourage restraint through fear of uncontrollable consequences. The category includes, for example, many statements by Maria Zakharova, Dmitry Medvedev, and others, suggesting that Western policy could lead to uncontrollable dynamics and unforeseen consequences. Note that these are not explicit coercive signals in the sense of “if you do this, we do that;” rather, these are signals that shape the perceptions of uncontrollable risk, in the logic of Thomas Schelling’s “threat that leaves something to chance.”
• R7: Blame Deflection: Statements that aim to justify the Russian nuclear escalation-relevant actions or policies (e.g., changes in the Russian nuclear doctrine) by framing them as normatively appropriate/responsible steps responding to inappropriate/irresponsible steps by NATO/Western countries.
• R8: De-escalatory (nuclear) signaling: If R1–R7 are generally issued and/or perceived as “raising” or “maintaining” nuclear temperature, R8 signals lead to “lowering” nuclear temperature in the broader escalation dynamics. Under this logic, Russian officials usually issue statements such as “nuclear war should never be fought,” “we are committed to keeping the conflict non-nuclear,” “we want to work with the United States on maintaining strategic stability,” “we are open to new arms control talks,” or “we will maintain New START limits despite suspension.”

West

• W1: Affirmative (NATO) Deterrence Signaling: Communicative acts in which NATO officials and member states’ representatives affirm, restate, or reinforce established NATO deterrence principles, doctrines, or red lines, without introducing new elements or signaling change. The focus of these statements is on the security and territorial integrity of member states rather than Ukraine. As such, the W1 category includes statements such as “we will defend every inch of NATO territory” or “any attack on NATO states will be met with a devastating response.”
• W2: Active Deterrence of Nuclear Use: Communicative acts in which Western officials set Russian nuclear weapon use - against Ukraine or elsewhere - as a red line and actively aim at dissuading the Kremlin from doing that. They are formulated as conditional “if-then” statements, where the “if” usually stands for “any use of nuclear weapons” and “then” usually stands for “will be met with devastating response” or “will completely change our approach to this conflict.”
• W3: Tacit Nuclear Signaling: Involves nuclear escalation-related signals that are communicated as actions rather than words. That includes, for example, military exercises with a nuclear component or tests of nuclear or dual-capable launchers. In the dataset, we would normally code the description of the action as such (e.g., “United States tested new ICBM”) or a neutral, official description of the action (e.g., “U.S. Ministry of Defense issued a statement that Russia had tested a new ICBM”).
• W4: Anticipatory Policy & Posture Signaling: Teasing/hinting at the possibility of prospective future policy adjustments that are or can be plausibly perceived by Russia as an escalation. This includes verbal hints that the West might deploy troops to Ukraine, approve deliveries of long-range weapons or aircraft, relax targeting restrictions for its delivered weapons, or adopt new sanctions against Russia. We also include escalatory rhetoric and signals hinting at the need to “defeat Russia” and similar.
• W5: Escalatory Policy & Posture Adjustments: Formal policy adjustments that have been perceived by Russia as an escalation. This includes deliveries of new types of long-range weapons or aircraft, deliveries of relaxation of targeting restrictions for its delivered weapons, or adoption of new sanctions against Russia.
• W6: Russian Nuclear Policy Assessments: Neutral assessments of the change and continuity in Russian nuclear posture and the probability of nuclear weapon use by Russia. This includes statements such as “we do not see any changes in Russian nuclear posture” or “we assess a greater risk of nuclear weapon use” under certain conditions.
• W7: Stigmatizing and Norm-reinforcing Signaling: Communicative acts that shame Russia for not behaving “responsibly” in the nuclear domain (i.e., portraying Russia as a transgressor of the norms of nuclear order) or verbally reaffirming the validity of the “nuclear taboo” and other norms of nuclear order (e.g., stating that “nuclear war cannot be won and must never be fought.”
• W8: De-Escalatory Signaling: Communicative acts that assure Russia that the West might not be implementing policy adjustments that are or can be plausibly perceived by Russia as an escalation. This includes verbal assurance that the West does not plan to deploy troops to Ukraine, approve deliveries of certain sensitive weapons (e.g., Taurus deliveries), or relax targeting restrictions for its delivered weapons. It might also include signals of willingness to “lower the nuclear temperature,” e.g., to negotiate new arms control agreements or implement steps to reduce nuclear risks.

Data Table: The feature offers an overview of all coded nuclear signals with dates, actors, scale values, and descriptions. Users can search, filter, and download the complete dataset for further analysis.

Methodology and Source

The data used for this dashboard is sourced from the working paper “Russia’s Nuclear Signaling in the War Against Ukraine: A Chronology of the Biden Era” published by the Stiftung Wissenschaft und Politik (SWP), German Institute for International and Security Affairs in cooperation with the Peace Research Center Prague (PRCP).

The chronology identifies, evaluates, and codes over 400 statements and actions related to nuclear issues. The researchers traced nuclear signalling between Russia and the West from autumn 2021 onwards, including both sides’ attempts at managing and controlling escalation dynamics, exercising deterrence, dissuasion, and compellence, and dealing with the domestic and international political implications of these nuclear-related processes.

For the time period until February 2022, the chronology includes:

• Ex post revelations of Washington’s assessments of Moscow’s intentions
• Russia’s public statements on a potential conflict with NATO
• The Kremlin’s nuclear-related statements

From February 24, 2022 onwards, the focus is primarily on tracing nuclear signalling between Russia and the West.

Actor Classification: For the sake of simplicity, the dashboard distinguishes between two actors only: Russia (R) and the West (W). Within the West, the focus is primarily on statements by officials from nuclear-weapon states or from NATO, as these statements were assessed to be more likely to lead to potential nuclear escalation.

Scale Values: The escalation scale uses a diverging format where negative values represent Russian actions and positive values represent Western actions. The absolute value indicates the intensity of the signal (1-8 scale), with higher values indicating more escalatory rhetoric.

Limitations

• The chronology does not claim to be exhaustive. It contains the statements and actions the authors deemed relevant based on their assessment.
• The analysis is based on publicly available information. Classified intelligence assessments and private diplomatic communications are not included unless later revealed by journalists.
• For the period before February 2022, the chronology purposefully excludes numerous contentious issues such as Russia’s demands regarding NATO membership for Ukraine, Western restrictions on weapons deployments, and threats to deploy intermediate-range nuclear missiles.
• The coding of signals involves subjective judgment, though the authors provide clear definitions and criteria.
• The two-actor simplification (Russia vs. West) may not capture the diverse perspectives within each bloc, particularly disagreements among NATO allies or within the Russian leadership.

Further Reading

Horovitz, L. (2025). US Extended Nuclear Deterrence in Europe: Three Scenarios. Crisis of Trust, Breach of Trust and Full Disengagement. SWP Comment 32/2025, 10. Juli 2025. Berlin: Stiftung Wissenschaft und Politik

Horovitz, L., Stolze, M. (2023). Nuclear rhetoric and escalation management in Russia’s war against Ukraine: A chronology. Working Paper Research Division International Security, Nr. 02/2023, August 2023, 257 Seiten. Berlin: Stiftung Wissenschaft und Politik.

Horovitz, L., Wachs, L. (2022). Russia’s Nuclear Threats in the War against Ukraine. Consequences for the International Order, NATO and Germany. SWP Comment 2022/C 29, 20. April 2022. Berlin: Stiftung Wissenschaft und Politik.

Horovitz, L., Arndt, A. C. (2022). Russia’s Catch-all Nuclear Rhetoric in Its War against Ukraine. A balancing act between deterrence, dissuasion, and compellence strategies. SWP Comment 2022/C 60, 19. Oktober 2022. Berlin: Stiftung Wissenschaft und Politik.

About the Authors

Dr Liviu Horovitz is a Research Associate in the International Security Division at the German Institute for International and Security Affairs (SWP) and serves as Head of the project Strategic Threat Analysis and Nuclear (Dis-)Order (STAND).

Dr. Michal Smetana is an Associate Professor at the Institute of International Studies within the Faculty of Social Sciences of the Charles University in Prague. He is also the Director of the Peace Research Center Prague (PRCP) and serves as Head Researcher at the Experimental Lab for International Security Studies (ELISS).

Anna Clara Arndt was a co-author on three previous iterations of this working paper until January 2023. Since then, she has not been involved in the project.

Karina Matvienko worked on this chronology from September 2024 to December 2025 during her tenure at SWP. She is a research assistant in the project STAND in the International Security Division at SWP.

Sara Camussi worked on this chronology during her tenure at PRCP from September 2024 to December 2024. She is a recent graduate of the Double Degree Master’s programme in International Relations at Konstanz University and Charles University.

Ayesha Zafar worked on this chronology during her tenure at PRCP from September 2024 to March 2025. She currently serves as a Project Manager at the Foundation for the Rights of Future Generations (FRFG).

Martha Stolze worked on this chronology during her tenure at SWP from February 2023 to September 2023. She currently works as a research associate at the Weizenbaum Institute and is pursuing PhD research on Russian anti-gender propaganda and gendered disinformation at Freie Universität Berlin.

Pars Tijen Peikert worked on this chronology during his tenure at PRCP from January 2025 to March 2025. He is currently pursuing a Master’s degree in International Security at Charles University in Prague.

Alexandra von Vultejus worked on this chronology during her tenure at SWP from April 2025 to June 2025. She is a Master of International Affairs candidate at the Hertie School in Berlin, specializing in International Security.

Larissa Kirstein worked on this chronology from October 2025 to December 2025. She is currently working as a research assistant in the International Security Division at SWP.

The authors would like to thank Christoph Baron and Pamela Skowron-Mrówka for their support with editing and formatting, as well as Paul Bochtler and Arthur Buliz for preparing the graphic elements of this paper.