Interview
The most common misconceptions about nuclear war | Jeffrey Lewis (2022)
- Core Critique of Existing Discourse: Jeffrey Lewis argues that the effective altruism (EA) community's focus on "counterforce vs. countervalue" targeting is analytically obsolete, treating Cold War-era terminology as if it applies to current US policy, which is more complex and less distinct.
- Myth of Mutual Assured Destruction (MAD): Lewis asserts that "Mutual Assured Destruction" is a rhetorical construct used to mock Kennedy-era constraints, not current US policy; official US strategy (e.g., O Plan) explicitly aims to "dominate" and prevail in a nuclear war, even if presidents privately doubt its viability.
- Decision-Making Realities: US nuclear decision-making is driven not by pure game theory but by bureaucratic pathologies, intra-service rivalry (Air Force vs. Navy), and the "Lin Biao pathology" where decisions are framed by the relative absurdity of alternatives on the table rather than objective strategic merit.
- The "Good Idea Fairy" and Gatekeeping: The nuclear field suffers from significant gatekeeping and linguistic gamesmanship (e.g., strict adherence to terms like "employ" vs. "use") designed to maintain orthodoxies and exclude dissent, often stalling progress on new ideas or risk-reduction strategies.
- Intra-Service Budget Rivalries: Distortions in the US arsenal, such as the persistence of vulnerable land-based ICBMs, are driven largely by the Air Force's desire to maintain budget and autonomy rather than strategic necessity, with the Navy often opposing these systems to protect its own deterrent capabilities.
- China and the US Arms Race: The US and China have no nuclear arms control treaties due to US unwillingness to limit missile defenses or accept "mutual vulnerability," a strategic choice that has inadvertently accelerated China's nuclear build-up; Lewis advocates for a dialogue on shared stability rather than deterrence based on numerical superiority.
- Nuclear Winter Credibility: Lewis confirms that nuclear winter is a plausible and credible risk, dismissing ideological objections as often rooted in denialism or a lack of technical understanding, and argues that even a limited exchange could trigger catastrophic climate effects.
- Deterrence Efficacy: Traditional deterrence is less effective than assumed because adversaries can exploit "salami tactics" and information asymmetries, and the system relies on an "irreducible risk" of human error or misperception rather than perfect rationality.
- Three Proposed Policy Changes:
- Usage Restriction: The US president should publicly declare nuclear weapons will only be used if no conventional weapon can achieve the same result, effectively reducing the target set to a handful of hardened facilities.
- Comprehensive Test Ban: The US must ratify the Comprehensive Nuclear Test Ban Treaty (CTBT) to prevent a renewed testing race with China and Russia, which are already modernizing test sites.
- Missile Defense Limits: The US must accept the strategic reality that offensive and defensive systems are linked; limiting offensive arms requires limiting defensive systems (ABMs) to prevent destabilizing arms races.
- Recommendations for EA and Researchers: The EA community can contribute most by bringing analytic rigor, quantitative risk assessment, and open-mindedness to the field, while avoiding the "disarmament" framing that has historically become a partisan wedge in US politics.
- Funding Priorities: Lewis recommends directing resources toward open-source intelligence (OSINT) to track nuclear programs (e.g., Chinese silo construction) and archival research to prevent the repetition of historical policy mistakes driven by forgotten context.
- Future Outlook: While the immediate risk of nuclear war remains low on a daily basis, the compounding risk over decades makes the current reliance on deterrence unsustainable, necessitating a long-term transition away from nuclear weapons dependence.