Government sources

Nuclear Operations, Air Force Doctrine Document 2-12, U.S. Air Force, 7 May 2009.

During the Cold War the US provided for the security of its allies by threatening a nuclear response in the event of an attack on them by the Soviet Union. This policy, based on the threat of retaliation, served as the foundation for what is now called extended deterrence. Extended deterrence remains an important pillar of US policy; however, its application in the context of the 21st century is very different from the Cold War. Today, extended deterrence is less about retaliation and more about posturing to convince an enemy that they are unlikely to achieve the political and military objectives behind any attack on the US or one of our allies.

The US employs extended deterrence on a daily basis to project deterrent effects in key regions across the globe. These forward-deployed assets combined with the global reach of continental United States (CONUS)-based nuclear forces provide theater-level assurance to allies abroad and deterrence to adversaries.

Foreign Perspectives on U.S Nuclear Policy and Posture: Insights, Issues and Implications, Lewis Dunn, Gregory Giles, Jeffrey Larsen, and Thomas Skypek, Defense Threat Reduction Agency, 12 December 2006.

Analysis

Extended deterrence on the way to a nuclear-free world, George Perkovich, Research Paper, International Commission on Nuclear. Non-proliferation and Disarmament, May 2009

Complex Deterrence, Patrick M. Morgan, P.M., T.V. Paul, and James Wirtz, (eds.), University of Chicago Press, 2009.

The Case for No First Use, Scott D. Sagan, Survival: Global Politics and Strategy, 1468-2699, Volume 51, Issue 3, 2009, pp. 163 – 182

Concerns about extended deterrence have thus often been cited as a reason to maintain current US nuclear declaratory policy concerning first use options and ambiguity about when nuclear weapons might be used. But extended nuclear deterrence can be made compatible with a no-first-use doctrine if changes in US nuclear security guarantees were made to fit current conditions of US conventional military superiority. Indeed, the term ‘nuclear umbrella’ is highly misleading and should be dropped from the strategist’s lexicon. The umbrella metaphor implies the existence of a defensive protection strategy (rather than a retaliation deterrence commitment) and the term also fails to differentiate between a commitment to use nuclear weapons first, if necessary, to defend an ally attacked by overwhelming conventional force or nuclear weapons (the Cold War policy in NATO, Japan and South Korea), and the more tailored guarantee to use nuclear weapons in retaliation against a nuclear attack, but only a nuclear attack, on US allies. This second guarantee is consistent with a no-first-use doctrine, and would maintain extended deterrence for nuclear threats to allies and would not therefore encourage them to develop their own nuclear weapons. This doctrine would also be consistent with a broader diplomatic strategy to encourage non-nuclear states to devalue the role of nuclear weapons and to see the first use of nuclear weapons as illegitimate. Finally, since such a doctrine would emphasise conventional responses to conventional acts of aggression, the credibility of the deterrent threat would be maximised because lingering fears about whether the United States might renege on its commitments, if they include the use of nuclear weapons, would be reduced. In short, a no-first-use doctrine would not undermine the part US nuclear forces play in deterring nuclear strikes and threats to allies in NATO, Japan and South Korea, and could provide broader diplomatic and non-proliferation benefits.

U.S. Nuclear Deterrence in the 21st Century: Getting it Right, The New Deterrent Working Group, Center for Security Policy Press, July 2009

The number and character of the forces the United States can deploy under a START follow-on treaty must also be sufficient to continue effectively to meet the Nation’s security commitments to allies through extended nuclear deterrence. Warhead limits that are too low: encourage near-peers to seek to become nuclear peers; encourage rogues to push ahead with nuclear programs; worry allies that the U.S. will not honor its nuclear guarantees; and encourage the latter to think about developing their own nuclear weapons.

Good and Bad Nuclear Weapons: Berlin’s Part in Shaping Nuclear Reality, Michael Rühle, Körber Foundation for International Affairs, Körber Policy Paper No. 3, April 2009

NATO and Extended Deterrence in a Multinuclear World, Michael Rühle, Comparative Strategy, 28:10-16 (2009)

Panel on Are the Requirements for Extended Deterrence Changing? Panel discussion at the 2009 Carnegie International Nonproliferation Conference: The Nuclear Order—Build or Break, 6 April 2009. [Audio] [Transcript]

• Scott Sagan, Stanford University (chair)
• Lukasz Kulesa, Polish Institute of International Affairs
• George Perkovich, Carnegie Endowment for International Peace
• Ambassador Yukio Satoh, Japan Institute of International Affairs

Good and Bad Nuclear Weapons: Berlin’s Part in Shaping Nuclear Reality, Michael Rühle, Körber Foundation for International Affairs, Körber Policy Paper No. 3, April 2009

Reduce US Nukes in Europe to Zero, and Keep NATO Strong (and Nuclear), A View from Poland, Lukasz Kulesa, Polish Institute of International Affairs, PISM Files 7, March 2009.

These changes are timely because a retaliation-based extended deterrence policy is prone to breakdowns in today’s complex and multi-polar world. This is why the agreements include steps for bilateral reinforcement of NATO commitments. The relative clarity of the bipolar world permitted carefully designed signals about which actions by a potential aggressor would result in retaliatory and escalatory steps by the U.S. The multi-polar world makes sending these signals much more difficult, because the signals must apply to multiple actors operating in different contexts and with different perceptions of the U.S and its allies.

Washington’s Apparent Readiness to Start Nuclear War, Andy Butfoy, Survival, 50:5, pp.115 — 140 (2008)

It could also be argued that the supposed extended-deterrence and related non-proliferation benefits of maintaining a US first-use option are mostly based on extrapolation from the Cold War and are less relevant given contemporary circumstances. There is little evidence of official American analytical work on the likely impact of a no-first-use declaration on contemporary extended deterrence or proliferation. The common conflation of first-use policy with generalised notions of the nuclear umbrella look more rhetorical than analytical. The implied choice offered by defenders of the status quo is that the United States can either keep its first-use policy, or abandon extended deterrence and face the acquisition of nuclear weapons by Japan and others, as if the threat of first use is the pillar on which all else rests. Implying that the current world order would come tumbling down if the first-use option was abandoned, or merely modified in accordance with tighter NSAs [negative security assurances], might be a good debating tactic, but oversimplifies and overdramatises the matter.

A state’s preference to live under the US nuclear umbrella, and its contentment with non-nuclear-weapons status within the NPT, should not – nearly two decades after the end of the Cold War – be confused with contemporary attachment to the first-use option. Today, a logical and sensible case can be made for extended deterrence which need not encompass threats of first use. Frontline countries such as Taiwan or South Korea may well wish to remain under the US umbrella, but it does not necessarily follow that they also wish the Pentagon to base its deterrent strategy (not to mention plans for their defence) on a readiness to escalate conflict in their neighbourhood to nuclear war. Given the weight placed on the putative association between first use and extended deterrence by conservative analysts, and the frequency with which it is repeated, it is remarkable how little contemporary evidence there is for the link.