New Paradigms Forum Proliferation Issues and Much More …


Guest Blog: Elbridge Colby on “The Substitution Fallacy”


This guest submission arrives from Elbridge Colby, who offered it to New Paradigms Forum in the wake of Dr. Ford's December 19, 2010, posting on conventional weapons as a potential “replacement” for nuclear ones.  Colby and Ford both participated in the November 17, 2010, conference on “Conventional Deterrence in the Second Nuclear Age, at the Carnegie Endowment for International Peace in Washington, D.C.  This is Colby's own take on the issue.

Elbridge Colby has served in several national security positions with the U.S. Government, most recently with the Department of Defense working on the follow-on to the START Treaty and as an expert advisor to the Congressional Strategic Posture Commission.  The views expressed in this essay are his own, and do not necessarily represent the views of any institution with which he is or has been affiliated.

The Substitution Fallacy: Why the United States Cannot Fully Substitute Conventional for Nuclear Weapons


Elbridge Colby

Many serious proponents of moving towards deep reductions in our reliance on nuclear weapons and their eventual abolition point to the promise of conventional forces, and especially high-technology conventional strike capabilities, as “substitutes” for our nuclear weapons.  Hawk grandpere Paul Nitze set forth a vision along these lines in influential op-eds in 1994 and 1999, and the notion has been supported by retired military leaders such as Generals Lee Butler and Eugene Habiger.  More recently, Henry Sokolski of the Non-Proliferation Education Center has carried the banner forward with his “Missiles for Peace” proposal.  In this vision, advances in the precision, rapidity, potency, responsiveness, and other relevant facets of conventional weapons can and should be the basis for progressively substituting them for nuclear weapons in the performance of a broadening set of missions.  Naturally, proponents also point to the importance of other military capabilities, such as ballistic missile defenses, as enablers, but for the core mission of conducting offensive strikes conventional capabilities are posited as the eventual substitute for nuclear weapons.

There is a lot to this line of argument.  Conventional weapons have enabled the United States and other countries to reduce substantially their reliance on nuclear forces. The radical advances in capability ushered in by the Revolution in Military Affairs (RMA) beginning in the 1970s have allowed the U.S. military to transform from a force that relied heavily on nuclear weapons to perform essential military tasks even at the early stages of conflict to one in which nuclear weapons are, happily, reserved for only the most extreme circumstances, as the recent 2010 Nuclear Posture Review stated.  And there is room for more; though the roles of nuclear weapons are now far more limited than in the 1950s or 1960s, there remains ample space for conventional forces to substitute for some missions traditionally reserved for nuclear forces. For instance, today the United States has very limited capabilities to strike promptly at global distances with conventional weapons.  Closing this gap in capability is the primary, laudable purpose for efforts to develop so-called “conventional prompt global strike” (CPGS) assets to hit a range of time-sensitive targets quickly at great distances.

But while there is room for further replacement, there are also reasons to think that there are limits to how far we can go in this direction.  In fact, instead of a linear progression towards substitution, it may be more useful to think of ourselves as traveling on an asymptote in reducing reliance on nuclear weapons.  There are a number of reasons why.  Perhaps the most important is the fact that the tasks left to nuclear weapons today are precisely those that are less susceptible to substitution by conventional weaponry.  In the 1950s and 1960s, when conventional munitions were very imprecise and our C4ISR (command, control, communications, and computers; intelligence, surveillance, and reconnaissance) capabilities were far more modest, nuclear weapons enabled commanders to destroy and disable targets by compensating for these weaknesses through the pure destructiveness of nuclear warheads.  First generation Titan and Atlas ballistic missiles, for instance, had circular error probables measured in the miles, necessitating that they be mounted with nuclear warheads if they were to do damage to a given target.

Nuclear weapons were also relied upon at the tactical level.  Indeed, it is startling to contemplate it today, but in that period both NATO and the Warsaw Pact planned for the large-scale employment of nuclear weapons even at the lower levels of conflict; hence the introduction of atomic landmines, depth charges, and field artillery.  Nuclear weapons gave higher assurance of a kill even if a commander did not know exactly where the enemy was or have confidence in the accuracy or reliability of his weapons systems.

The technological leaps of the RMA starting in the early 1970s changed all this, radically improving the U.S. military’s ability to strike targets accurately.  Because of the RMA, today’s commanders are infinitely better equipped than their forebears to find and correctly ascertain enemy assets; to target them; to hit them reliably and accurately, often within ranges measured in meters or less; and to redirect efforts based upon constantly updated information.  In brief, commanders no longer need nuclear weapons to accomplish a wide range of straight-forward military missions.

A straight line projection from these trends would suggest that such substitution should continue, and there are those who make such prognoses.  The problem, however, is that the targets that have been left for nuclear forces today are those that are likely less susceptible, if susceptible at all, to improvements in precision, responsiveness, or C4ISR.  These include (albeit for differing reasons) hardened or deeply buried facilities, highly valued mobile assets, and other targets that are difficult for conventional weapons to destroy or disable in the required timeframe with sufficient confidence.

In its 2005 report on the Robust Nuclear Earth Penetrator, for instance, the National Academy of Sciences found that “[m]any of the more important strategic hard and deeply buried targets are beyond the reach of conventional explosive penetrating weapons and can be held at risk of destruction only with nuclear weapons,” even as such facilities swell above 10,000 worldwide.  Meanwhile, U.S. capabilities to target mobile assets effectively remain, to say the least, imperfect, a fact vividly illustrated by the astoundingly unsuccessful performance of Coalition forces against mobile Scuds in the first Gulf War.  And while U.S. ISR capabilities are likely to improve, so too is the ability of our adversaries to elude detection and acquisition. Fundamentally, if conventional weapons cannot physically deliver the destructive power required to destroy or disable a hardened or insulated target or if a target cannot be located within range of conventional weapons’ destructive radius then conventional weapons will be of only secondary utility against them.  Of course, there are likewise targets that today’s nuclear weapons cannot destroy or disable, but the limits on today’s nuclear and earth-penetrating capabilities are less often imposed by technical constraints than by policy decisions.

That said, some of these missions may be amenable to technological breakthroughs.  But we know that nuclear weapons are different.  Conventional weapons will never be capable of equaling the awesome destructiveness of thermonuclear weapons.  Yet the threat to wreak this awesome power upon things that an adversary values is the central, defining character of nuclear weapons and of the nuclear deterrence that rests upon them.  There is no doubting that the United States has built intricately elaborate war plans envisioning the use of U.S. nuclear forces against military and war-related targets and Washington and Omaha [where the U.S. Strategic Command has its headquarters] have made repeatedly clear that we do not target civilian populations per se – but no one serious is so foolish as to think that the execution of even a substantial portion of this “counterforce” strategy would involve anything but unimaginable devastation.  Some may bemoan or wish to rid us of the need for this basic type of deterrence, but there can be no pretence that conventional weapons can substitute for this unique role.

A second basic issue with the notion of substituting conventional for nuclear weapons is the effect they will have on strategic stability.  Classical understandings of strategic stability, at least among the great powers, have posited that states would behave more cautiously and sensibly if each knew the other had the assured capability to strike back with punishing force even in the event of a first strike.  Survivable, second-strike systems were therefore favored over those weapons best-suited for disabling an opponent’s retaliatory capabilities (including both weapons and command and control systems).  This perception was rooted in the assessment that, if countries feared they could be subject to the destruction or dismemberment of their strategic capabilities, they would be pressured to posture and behave in riskier ways in order to ensure they could strike before losing the ability to do so.  The implication of this was that “strategic” weapons that had substantial military effectiveness – that is, that could quickly destroy or disable an opponent’s silos, bombers, submarines, mobile ICBMs, missiles in flight, etc. – were disfavored.  Yet these are the very missions that analysts such as Sokolski argue conventional can substitute for nuclear weapons.

This analysis is not merely theoretical.  Russian and other foreign analysts have been very quick to note the potential of conventional strike technologies, and that they regard them as “strategic” in the sense that they pose a threat to their strategic capabilities.  Russian analysts have, for instance, claimed repeatedly that modern U.S. cruise missiles launched from submarines or standoff distances by aircraft and which can evade Russian early warning and air defense radars can pose a serious threat to Moscow ’s retaliatory capabilities.  Other possible future conventional systems could barrage areas where suspected mobile ICBM launchers are located.  Chinese analysts have expressed similar anxieties.  Together, these analyses strongly suggest that merely substituting conventional for nuclear munitions in the performance of strategic missions would not convince the Russians, the Chinese, or others that they are less threatening to their core strategic capabilities.

Further, because conventional weapons lack the destructiveness of nuclear weapons, substituting the former for the latter would make a truly devastating prompt second strike extremely difficult, if not impossible.  Thus conventional weapons could substitute for the offensive, counterforce aspects of nuclear targeting, but not for the second strike, retaliatory aspects – characteristics that are precisely the inverse of what has traditionally been considered stabilizing.

Proponents of substitution might retort that such fears reflect a mindset locked in old thinking.  But that is precisely to concede the point.  Assuming that proponents do not wish to return to the pre-1945 world of conventional military competition, safely substituting conventional for nuclear weapons would require a new understanding of stability, one fundamentally different from the one that prevailed not only in the Cold War, but has formed the basis for arms control treaties since then, such as START and New START.  This understanding would need to describe how stability could rest on something other than the assured ability to strike back.  Surely a world of conventional competition and advantage, such as the pre-nuclear world, would not satisfy this criterion, but perhaps a fundamental transformation of the world political environment might.

But this leads to questions: How, then, are the conditions in which conventional weapons could be truly substituted for nuclear weapons different from those in which all serious offensive weapons could be eliminated?  If the conditions required for conventional weapons to substitute fully for nuclear weapons also would obviate nations’ need to retaliate and to conduct strikes against especially hard targets, then presumably serious offensive conventional weapons would not be required either.  In such a world, could not nations simply field forces like today’s Japan ’s Self-Defense Forces, postured for basic defense and the maintenance of public order but certainly not optimized for serious warfare?

This illuminates the reality that the basic issues that the concept of substitution is addressing are not in essence technical; they are political.  The central challenge facing advocates of nuclear abolition is not to replace the purely military functions performed by nuclear weapons with conventional substitutes – though this too would be exceptionally difficult and costly, and perhaps impossible – but to find a model of stability and security that can provide the stability and security that nuclear deterrence has for more than a half-century.

Ultimately, the analytical approach of considering conventional strike capabilities as substitutes for nuclear weapons has real but decidedly limited and probably declining utility in thinking about our strategic forces.  In the context of strictly military missions such as the destruction of hard targets, the concept is useful; the United States should, presuming the satisfaction of considerations of capability, cost, strategic stability, seek to substitute conventional weapons for nuclear weapons.  But when considering the fundamental purpose of nuclear weapons, that of delivering a devastating second strike upon an aggressor, the concept is of little help.  As Americans deliberate on the future of our strategic forces, we should therefore be careful not to neglect investing in our nuclear capabilities because of an overreliance on a decidedly limited concept.

-- Elbridge Colby

About Dr. Ford

Dr. Christopher Ford presently served until December 2016 as Chief Legislative counsel for the U.S. Senate Foreign Relations Committee, Chief Investigative Counsel for the Senate Banking Committee, Republican Chief Counsel for the Senate Appropriations Committee, Senior Fellow at Hudson Institute, U.S. Special Representative for Nuclear Nonproliferation, Principal Deputy Assistant Secretary of State, Minority Counsel and then General Counsel to the Senate Select Committee on Intelligence, and Staff Director of the Senate's Permanent Subcommittee on Investigations. A graduate of Harvard (summa cum laude), Oxford (as a Rhodes Scholar), and the Yale Law School, Dr. Ford was also ordained by Roshi Joan Halifax of the Upaya Zen Center as a lay chaplain in a lineage of Soto Zen Buddhism. He was a jujutsu student of the late Grandmaster Dong Jin Kim of the Jigo Tensin Ryu lineage, and is a member of Dai Nippon Butoku Kai with Sandan (3rd degree black belt) rank. Dr. Ford served from 1994 until 2011 as an intelligence officer in the U.S. Navy Reserve, and is a member of the International Institute for Strategic Studies, Chatham House, and the Council on Foreign Relations. Dr. Ford is the author of the books "China Looks at the West: Identity, Global Ambitions, and the Future of Sino-American Relations" (2015), "The Mind of Empire: China's History and Modern Foreign Relations" (2010), and "The Admirals' Advantage: U.S. Navy Operational Intelligence in World War II and the Cold War" (2005). He also co-edited "Rethinking the Law of Armed Conflict in an Age of Terrorism" (2012). For a list of his publications, see The views he expresses here are entirely his own, and do not necessarily reflect those of anyone else in the U.S. Government.
Comments (2) Trackbacks (0)
  1. Contrary to the suggestion above, I should point out that I do not believe that total substitution of nuclear weapons is possible without a loss of destructiveness against key targets that, in some key cases, means these targets cannot be held at risk. My article “Missiles for Peace,” cited in Mr. Colby’s essay, does not claim otherwise. For a deeper discussion of these issues in work commissioned by my center, the Nonproliferation Policy Education Center (NPEC) [link:, click here [link:

    – Henry Sokolski
    Executive Director, NPEC

  2. It sounds like Henry Sokolski, Elbridge Colby, and I are actually in violent agreement that conventional weapons are incapable of fully “substituting” for nuclear weapons. (See also, e.g., my own remarks at the Carnegie Endowment for International Peace in November 2010 [link:

    Logically, I suppose, this incapability is not a “showstopper” argument against complete nuclear disarmament, but it helps frame the starkness of the choice offered by disarmament advocates. To give up all nuclear weapons is perforce to give up being able to hold certain types of target at risk — a particularly noteworthy point given the contemporary proliferation of hardened and deeply-buried targets (HDBTs) around the world. Alas, though I think it still offers a way to help possessors such as the United States rely less — and perhaps much less — upon nuclear weapons, the improvement and increased deployment of strategic-range precision conventional tools is no “silver bullet” answer to the modern targeteer’s disarmament dilemma.

    Many thanks to Henry for writing to NPF!

    – Chris

Trackbacks are disabled.

NPF Pages

Recent Additions to NPF

NPF Discussion Pages

Calendar of NPF Postings

February 2017
« Nov    

NPF Archives (by month)