Not at all! I appreciate your engagement. Addressing selection effect is hard, and your comments make me think about what I could have done differently and what future studies can do. I see my paper as a first step toward this important question, but definitely not the final word on ED.
Thank you, Dan!!
Thank you!
I expect that with more data on other aspects of conflict, security environment, etc. we will be able to test our theories from novel angles. It is possible that future studies would find that nuclear superiority matters in different contexts, therefore I need to qualify my conclusion!
Regarding your broad point, yes, confict (or MID) initiation may be not the only thing extended "deterrence" wants to deter; but it's an important one (and arguably the most basic one), and my approach is that if we did not find evidence in that context, that says something about the broad claim.
Multiple times, so I think that it's still useful as a measure of the security environment. And combined with other analyses (e.g., selection model estimation, Soviet alliance formation apttern), I think that the data suggests that the evidence for the selection argument is not strong.
Thank you for the point. Yes, MID is not a perfect way of measuring the broad security environment; in that sense, that evidence alone may be not strong enough to challenge the selection effect claim. That said, MID is a widely used indicator and has been externally "audited" ...
Thank you, Brian! This means a lot to me!!
Special thank you to Giles David Arceneaux, Do Young Lee, and Iain Henry for constructive feedback!
Of course, they do not rule out the possibility that nuclear superiority may generate other benefits; but US policymakers would need a serious net assessment regarding the costs and benefits of a potential nuclear buildup across several different contexts.
Taken together, these findings demonstrate the need to reconsider the idea that nuclear superiority would strengthen peacetime deterrence against nuclear-armed adversaries.
My previous JPR paper, which tests the deterrent effect of nuclear superiority in a direct deterrence context, also found no evidence that nuclear superiority reduces the risk of a crisis initiated by another nuclear-armed state.
doi.org/10.1177/0022...
Using five different indicators of nuclear balance, my analysis finds that across different model specifications, there is no strong evidence that a favorable nuclear balance reduces the risk that an ally of a superpower becomes a target of an MID initiated by another superpower.
This is a favorable empirical setting for the thesis. Why? Because the risk of nuclear escalation was perceived as serious between the US-led bloc and the USSR-led bloc. Those concerns were especially strong during the Cold War. This makes nuclear balance especially relevant.
I tested this “superiority-credibility-thesis—the idea that nuclear superiority strengthens extended deterrence because it makes threats of nuclear retaliation credible—using data on superpower-led alliances during the Cold War.
The idea that nuclear superiority strengthens deterrence is not new; scholars have debated about it since 1949. And experts have claimed that nuclear superiority is particularly useful for extended deterrence, where concerns over credibility are especially strong.
New START has expired; China is building up its nuclear arsenal. Calls to expand U.S. nuclear forces are growing. Proposals vary, but the underlying message is the same: the United States needs to maintain a nuclear advantage over adversaries to strengthen deterrence.
Does a favorable nuclear balance of power strengthen extended deterrence? My new JPR paper tests this claim and finds no strong evidence for it.
doi.org/10.1093/jopr...
A 🧵..
The New START treaty doesn't expire until February 5. Stationing Oreshnik in Belarus prior to it's expiration is one last violation of the treaty. 🧵
1. Crises, acquisitions, and policy all suggest that conventional weapons are taking on a larger role deterring nuclear use. My new article in @intsecurity.bsky.social examines why US officials would consider conventional options when they have nuclear options available and how they might use them.
🚨 New article out in @jeppjournal.bsky.social with @svenhegewald.bsky.social
“The changing geography of support for European integration in the shadow of the Ukraine war."
How did Russia’s invasion reshape public support for EU policies?
www.tandfonline.com/doi/full/10....
How well could the US track Soviet mobile ICBMs in the cold war through SIGINT/ELINT satellites? A good debate here. "...it is more clear now that NSA tested the idea and had resources to pursue it." strategicsimplicity.substack.com/p/sigint-and...
Congratulations!
Fantastic resource for historians: a spreadsheet of NARA's digitized microfilm, with links to the files in the Archives catalog. May it help you as much as it has helped me.
www.archives.gov/files/colleg...
Thank you Shana!
Special thanks to @gdavidarceneaux.bsky.social, @sgadarian.bsky.social, @lanoszka.bsky.social, Ryan D. Griffiths, and Amy King.
In sum: during crises, allies want to see credible, costly signals, and the United States has a range of military and diplomatic toolkits. Insights from costly signaling theories apply well to alliance reassurance.
Interestingly, all three effective tools had substantively similar effects. Military signals (conventional or nuclear) and public remarks seem substitutable—in contrast to claims that actions always outweigh words. But diplomatic overtures towards adversary fail to reassure.
I examined how each U.S. policy response influences the perceived safety of South Korea from DPRK aggression—a direct measure of reassurance.
The results:
Conventional and nuclear signals work;
Public statements of support also work; but
Diplomatic overtures do NOT work.
To address this gap, I fielded a survey experiment with South Korean citizens during a hypothetical North Korea–ROK–U.S. crisis. I tested four tools: conventional signals, nuclear signals, statements of support, and diplomatic overtures to the adversary.