This post is part of a series of rough posts on nuclear risk research ideas. I strongly recommend that, before you read this post, you read the series’ summary & introduction post for context, caveats, and to see the list of other ideas. One caveat that’s especially worth flagging here is that I drafted this in late 2021 and haven’t updated it much since. I’m grateful to Will Aldred for help with this series.
One reason I'm publishing this now is to serve as one menu of research project ideas for upcoming summer research fellowships.
Some tentative bottom-line views about this project idea
What is this idea? Why might this research be useful? How could it be tackled?
“A nuclear electromagnetic pulse (commonly abbreviated as nuclear EMP, or NEMP) is a burst of electromagnetic radiation created by a nuclear explosion. The resulting rapidly varying electric and magnetic fields may couple with electrical and electronic systems to produce damaging current and voltage surges” (Wikipedia, 2021). According to the US EMP Commission (2004):
Several potential adversaries have or can acquire the capability to attack the United States with a high-altitude nuclear weapon-generated electromagnetic pulse (EMP). A determined adversary can achieve an EMP attack capability without having a high level of sophistication. EMP is one of a small number of threats that can hold our society at risk of catastrophic consequences. EMP will cover the wide geographic region within line of sight to the nuclear weapon. It has the capability to produce significant damage to critical infrastructures and thus to the very fabric of US society, as well as to the ability of the United States and Western nations to project influence and military power.
Pry (2005) also claims analysts in other states see nuclear EMP attacks as a legitimate use of nuclear weapons or are concerned that their states may be targeted by such attacks.
This project could investigate questions such as:
- How likely is it that one or more nuclear EMP attacks will actually occur?
- What is the probability distribution over various possible numbers of nuclear EMP attacks in a single conflict, if one such attack occurs?
- What are the most likely pathways to such an attack? (E.g., deliberate, accidental, or unauthorised use? Preceded by conventional conflict, by nuclear conflict, or by neither?)
- How likely is it that such an attack is followed by nuclear, conventional, or other military attacks by either side?
- What would the effects of such an attack be?
- In particular, how likely are especially extreme effects (e.g., >50 million fatalities, >1 billion fatalities, existential catastrophe)?
- How, if at all, do (perceived) answers to the above questions affect other aspects of states’ nuclear strategies?
- How does the answers to those questions vary between different potential aggressor and target states (or nonstate actors)?
- What can be done to reduce the chance or harms of such an attack?
- Given all of the above, is this risk worth any further attention from the longtermist community?
I’d guess that nuclear EMP attacks are less likely than other forms of nuclear conflict, that they would be far less harmful than nuclear winter, and that they would have less than a 1 in 1000 chance of causing or strongly contributing to an existential catastrophe. I think this makes this research project idea relatively low-priority. However, I also feel uncertain about those guesses and think that uncertainty might be substantially reducible with even just 0.5-3 weeks of work, which makes me think this project could be worth pursuing for someone with a strong fit for or interest in it.
I expect it would be sufficient to simply:
- Read parts of existing work on the topic, such as Wikipedia articles (e.g.), US EMP Commission (2004), Baum and Barrett (2018), and work cited in or by those works
- Interview several experts on the topic
- Make/commission additional high-quality forecasts on the Metaculus questions I already created on this topic
- Perhaps create and solicit forecasts on some additional relevant forecasts questions
- Summarise your findings and takeaways
I’d intended to briefly research this topic myself, so I made some rough notes on it, which can be found here.
Other people to talk to
- Some staff at ALLFED have researched and are concerned by the possibility of nuclear EMP attacks
Regarding the analogous question for more standard nuclear attacks, see # Nuclear Weapons Used by 2050, if Any?
For example, does this affect states’ desired number of nuclear weapons, missile defence systems, or alert levels for their nuclear weapons?