For the very first lecture, we discussed the Doomsday Clock. The clock acts as a warning of how close we are to destruction. We analyzed the 2024 announcement of the doomsday clock and how it mentioned that the risk of nuclear extinction has been increased as Russia has withdrawn from the New Strategic Arms Reduction Treaty (New START). As a result there is a big risk that the war in Ukraine will escalate to a global conflict and a nuclear war. While I was reading this, I was alarmed, but I was also surprised that the time at the Doomsday Clock showed 90 seconds to midnight closer than ever before, including the period of the Cold War. Two underlying rationales might cause this assessment. First, one might claim that the nuclear threat is actually worse today than it was during the Cold War, albeit seemingly counterintuitive. However, this decision can be challenged given the lack of the formula employed by the Bulletin of the Atomic Scientists to assess existential risk. Consequently, this lack of a standardized process will always allow for different interpretations and opinions regardless of the time on the clock. Alternatively, it is possible that while the nuclear threat may be lower today than during the Cold War, the combined risks posed by nuclear war, climate change, and AI are greater. Thus, I decided on creating an index akin to the Doomsday Clock, which would serve to alert people specifically to the threat of nuclear war while also being standardized and focused solely on the risk of nuclear extinction.

Ideally, this Nuclear War Risk statistic would serve the dual purpose of monitoring the nuclear threat over time and assessing countries’ progress in mitigating this risk. Moreover, we imagine this statistic being utilized in informing policy-making decisions. My goal is to develop a risk index that is straightforward to calculate and provides a rough estimation of the likelihood of a nuclear war. To achieve this, I intend to leverage existing, readily available statistics and manipulate them algebraically to approximate the nuclear war risk. While simplicity is paramount, I also strive for comprehensiveness, aiming to strike a balance between ease of calculation and accuracy. Ultimately, I aim for this index to be a convenient tool for quickly assessing nuclear risk, serving as a practical back of the envelope estimation for policy deliberations and public awareness efforts.

This will be a composite index inspired by the Human Development Index (HDI), and similar to the structure of HDI my goal is to incorporate three key dimensions in assessing the risk of a nuclear conflict: the availability of nuclear weapons by a country, the military readiness for war, and the international tensions between two countries. To model these dimensions as seen in the plot above we can use as indicators the size of the nuclear arsenal, the military expenditure and the perceived level of threat.  Drawing data from these indicators is straightforward, as numerous sources provide calculations for such information. I choose for the size of the nuclear arsenal to use the information provided by the Stockholm International Peace Research Institute (SIPRI) that provided a detailed report towards the end of 2023 describing in detail the nuclear capabilities of every country that is known or suspected to have nuclear weapons. For a country’s military expenditure the latest data are provided by the SIPRI Military Expenditure Database. Measuring the perceived level of threat is harder because it is more subjective and it might not be done in a standardized manner. So we specifically choose to use the yearly data provided from the Munich Security Index that measures the perceived level of threat by using questionnaires but they do it every year something that makes it easy to track any changes.

Using data from these primary sources, we will compute an index to assess the heightened risk of nuclear war escalation due to a country’s nuclear capabilities, military preparedness, and international relations. However, simply relying on the raw numbers from these sources is not sufficient. For instance, when considering nuclear capabilities, possessing more nuclear weapons does not automatically equate to a proportionally higher risk.

We see this In Jerry Brown’s “Nuclear Addiction: A Response,” where it is evident that the Soviet Union and the United States found themselves in a formidable stalemate. Despite continuously expanding their nuclear arsenals to match each other’s power, the risk of nuclear war escalated gradually. The article showcases that the stalemate arose because both countries already had nuclear arsenals capable of mutual annihilation. Thus, under the doctrine of mutually assured destruction, neither country dared to initiate a nuclear war, resulting in a relatively constant or slowly increasing risk of nuclear conflict. This instance highlights that the relationship between the risk of nuclear war and the number of warheads is not simply proportional. Both the total number of nuclear weapons and the relative difference in the nuclear arsenals between two countries hold significance under the doctrine of mutually assured destruction. However, the influence of these factors diminishes as their absolute numbers increase. Therefore, when taking into account the number of nuclear weapons of each country, to calculate the effect of the nuclear capabilities we also need to take into account the diminishing returns. These same two rules under the same logic apply for military expenditure.

I ended up writing the final project in the style of a research paper where I calculate the metric, which I named the Nuclear War Risk Index (NWRI), for potential wars between US and Russia, and US and China. Since this is a new index there is no presupposed intuition of what it means. I compute this value for a conflict in 1962, the height of the Cold War, to illustrate numerically what NWRI value would indicate a very high risk. Similarly, I calculate it for a conflict in 1993, coinciding with the signing of the START I treaty and the conclusion of the Cold War, to provide a value of roughly medium risk. Carrying out these calculations we can see that the risk of a nuclear war with the US and Russia today is greater than that with China. Additionally, we find out that even if the risk is not as high as the peak of the cold war it has crossed the medium risk level which roughly described the end of the cold war. This paper functions as a demonstration of how one can use this index to get a quick estimation of the risk of nuclear war from readily available statistics. Hopefully, this will inspire the creation of further metrics and stimulate discussions about tracking the risk of nuclear warfare. However, it is important to note that this index is based on certain assumptions, such as those related to mutually assured destruction, the law of diminishing returns, and the reliability of the primary data. These limitations are discussed in detail in the full paper. Nevertheless, if someone decides to develop another metric, this paper can serve as a useful blueprint.

For the complete calculations, assumptions and relevant discussion please take a look at the full paper:

Nuclear War Risk Index: A Standardized Doomsday Clock

Works Cited:

Brown Jr, E. G. (1984). Nuclear Addiction: A Response. Thought: Fordham University Quarterly, 59(1), 10-14.

Bulletin of the Atomic Scientists (2024) Current time – 2024 Retrieved from: https://thebulletin.org/doomsday-clock/current-time/ Accessed: 3 March 2024

SIPRI (2023) SIPRI Yearbook Retrieved from: https://www.sipri.org/yearbook/2023 Accessed: 3 March 2024

SIPRI (2024) Military Expenditure Database. Retrieved from: https://www.sipri.org/databases/milex. Accessed: 3 March 2024

Munich Conference (2024) Munich Security Index 2024 Retrieved from: https://policycommons.net/artifacts/11336246/munichsecurityindex2024/12225173/ Accessed:  3 March 2024