After New START: The Heritage Plan for U.S. Nuclear Forces

With the formal expiry of the New START Treaty, the question of the adequacy of current U.S. strategic offensive force programs – largely formulated in the early - mid 2010s, when it was assumed that the treaty would actually be observed, Russian force levels would not increase, and China would never rise to even the same league as the U.S. and Russia – will come once more into focus. As of 2023, the unanimous findings of the bipartisan Strategic Posture Commission were as follows: ‘the United States lacks a comprehensive strategy to address the looming two nuclear-peer threat environment and lacks the force structure such a strategy will require.’ ^[1]

Golden Dome or no, this is hardly an academic concern. Although active defense may greatly improve the relative strategic position of the United States, there is no such thing as a leak - proof defense. If the worst-case scenario for an enemy executing a nuclear attack against the U.S. was that their missiles might be shot down, then U.S. adversaries would have no reason not to try their luck. The ability to defend the United States against enemy missile attack will therefore be no deterrent at all unless the United States also possesses a secure and militarily effective retaliatory capability.

Although most observers would probably agree that the Strategic Posture Commission’s 2023 recommendations offer a good starting point, there have been precious few concrete proposals available for public consumption as to what the medium - to - long term plan ought to be. One such proposal was put forward by the Heritage Foundation – an organization which is generally considered to have had some influence on the revival of missile defense under the present Administration – in October of last year. The paper’s author, Mr. Robert Peters, considers it a necessary minimum blueprint for maintaining the nation’s nuclear deterrent, which, as the War Department regularly affirms, is the first and most basic requirement of the national defense. Late last year I sat down with Mr. Peters to talk about the proposal.

On the subject of the New START Treaty, Mr. Peters was quite clear. At various times the lead Strategist at the Defense Threat Reduction Agency and Special Advisor for Countering Weapons of Mass Destruction, he was involved as an employee of the OSD in the process of negotiating the Treaty under the Obama Administration in 2010.

‘We got taken for a ride,’ he said simply. According to the Treaty’s terms of reference, he points out, both sides agreed that New START (which covered only strategic forces) would, upon ratification, be immediately succeeded by a follow - on treaty encompassing all nuclear weapons. This would include non - strategic nuclear weapons; a field in which the Russians possessed at least a ten to one advantage over the U.S., since the U.S. began unilaterally disarmament in non-strategic weapons in 1991. The New START Treaty, says Peters, ‘was never meant to be this kind of overarching arms control structure; it was meant to give President Obama a quick win. And that would pave the way for a more cohesive, all-encompassing arms control treaty immediately after ratification.’

As soon as the Treaty was ratified and the Americans could no longer politically afford to back out, the Russians changed their tune and refused to even consider the previously agreed upon follow - on treaty. Naturally, this meant that Russia would keep its overwhelming advantage in non-strategic weapons, while the U.S. had almost all of its existing nuclear forces regulated by the Treaty.

Although Russia only officially “suspended” New START in August of 2022, the Treaty has been essentially moribund for five years now, as no on - site inspections have taken place since March of 2020. ^[2] In the absence of inspections, there is no way of knowing how many warheads each missile might be carrying, and therefore the Treaty limits are totally unverifiable. Since about 80% of Russia’s total strategic warhead capacity is provided by land - mobile intercontinental ballistic missiles, (ICBMs,) submarine launched ballistic missiles, (SLBMs,) and bombers – all of which can be loaded with stockpiled warheads with no outwardly visible work – there is, says Mr. Peters, no reason to believe that the Russian strategic nuclear force was not uploaded to its optimum capacity long ago, while the United States continued to abide by the Treaty limits in exchange for nothing. Russia, meanwhile, has claimed that it has continued to abide by the Treaty limits, despite being under no obligation to do so.

Mr. Peters concludes, therefore, that President Putin’s offer to renew the New START Treaty is ‘a trap,’ just as the original treaty was. ‘It's not meant to be a good faith agreement in which we'll get back to any kind of bilateral inspections and verification measures,’ he says. And, moreover, ‘The Chinese are very clear that they do not want to engage in arms control. I do not believe that they're going to stop their buildup anytime soon.’ And, he adds, without any U.S. program to match the expansion of their nuclear forces, neither the Russians nor the Chinese have any incentive to come to the table, even if they might otherwise be convinced.

Faced with these realities, Heritage proposes a modest program to upload existing U.S. warheads, (meaning to load already existing warheads currently sitting in storage onto their associated missiles, a task assessed by the CBO in 2024 to cost somewhere in the vicinity of $120 million, on the order of 0.01% of the current annual defense budget;) ^[3] coupled with a relatively small expansion in U.S. strategic launchers, including 50 additional silo -based and 50 road mobile Sentinel ICBMs, 110 - 150 additional B-21 bombers, (depending on whether 100 or 140 B-21s are assumed as the baseline figure,) 4 additional Columbia class submarines and associated missiles, and re-conversion of all remaining B-52s to be dual - capable.

Given a total upload cost of ~$120 million, an estimated Sentinel unit cost of about $113 million, ^{[4] [5]} a marginal cost of around $610 million per additional B-21 and associated LRSO missiles, and a marginal cost of around $10.6 billion for each additional Columbia class submarine and associated missiles, (all costs adjusted to 2025 dollars,) ^[6] the total cost of such an expansion might amount to something like $120 - 150 billion (depending on the number of additional B-21s) over the next 24 years. This increase is equal to approximately 0.6% of total U.S. defense expenditures over that period, if expenditures were held to their present real level of about $1 trillion per year, or 0.4%, if the Administration’s plans for a 50% increase in annual expenditures were successful and maintained into the future. Of course, given the very small size of these expenditures, comparable force levels could be achieved in a much shorter period than 24 years, even with annual military expenditures held to their present level. The B-21s would account for approximately 56 - 63%, the ICBMs 7 - 9%, submarines and SLBMs 28 - 34%, and warhead uploading about 0.1% of additional expenditures.

This would bring the estimated loading of operational strategic warheads from the present force of about 2,200 maximum operational weapons under the Treaty limits (~1,400 ballistic missile warheads, ~536 ALCMs carried by B-52s, and up to 320 gravity bombs carried by B-2s) ^[7] to approximately 3,500 by 2050. Notably, the currently programmed force of at least 100 B-21s and 42 B-52s technically already possesses the capacity, when fully generated, to lift the programmed inventory of 1,000 LRSO missiles, which the Heritage proposal does not increase. However, if the bombers were placed on alert at the Cold War rate of 30%, the alert force of B-52s and B-21s could only deliver about half of these weapons. (Currently, no U.S. bombers are on alert, and therefore all bombers would be destroyed on the ground in an enemy attack.) Furthermore, in any contingency which involves non - nuclear operations escalating to nuclear warfare, it seems highly probable that some portion of the B-21 fleet will be in demand for non - nuclear strike operations and therefore will not be sitting on nuclear alert. A force of 250 B-21s would provide the capacity for the notional alert force to lift the programmed inventory of 1,000 LRSO missiles, while remaining aircraft could be used for other purposes. The cost of increasing the LRSO inventory to the theoretical maximum capacity of all bombers, (about 2,520 additional missiles at a unit cost of $14 million each, for a grand total of 3,520 LRSOs,) would be about $35 billion; ^[8] this possibility is not considered in the Heritage proposal.

It ought to be noted here that, in any case, many of the additional warheads (in both the programmed and Heritage forces, as compared with the present force) will come from simply uploading strategic launchers with already existing reserve warheads, and / or producing additional warheads for those launchers. The former costs essentially nothing, whereas the latter would still be relatively cheap compared with buying new missiles and bombers.

The following tables should be taken as purely evocative rather than definitive. The warhead figures in Table 2 apply the Heritage assumptions for average warheads carried per missile to the programmed force for the sake of comparison, although it is likely that actual average warhead loadings would differ depending on the number of launchers. Heritage, like practically all other open sources, makes no attempt to estimate the number of warheads which might survive an enemy first strike, since any such estimate is highly sensitive to a great many assumptions. Table 3 (Residual Warheads surviving enemy Surprise Attack) is included here, however, along with the relevant assumptions used, because the number of surviving warheads is almost always excluded in public analyses, but it is important to understand that different force structures will have varying effects on this number; and journalists, of course, are under no obligation to be correct.

TABLE 1: TOTAL STRATEGIC LAUNCHERS

[1] Nuclear capable bombers; 30 B-52s have been converted to be incapable of using nuclear weapons.

[2] Operational missiles; 50 silos are currently inactive.

[3] Available missile launch tubes, not operational missiles. This reflects the maximum number of missiles that could be made operational in each force, excluding an average ~15% of subs in overhaul. (2 for present / programmed, 3 for Heritage.) Used for the sake of comparison.

TABLE 2: TOTAL STRATEGIC WARHEADS

[4] For ICBMs, uses the Heritage assumption of an average loading of 2 warheads per missile; the maximum loading of the Sentinel is claimed to be 3 warheads.

[5] For bombers, programmed number of LRSO weapons is ~ 1,000 in both Programmed force and Heritage proposal.

[6] For SLBMs, assumes Heritage average ratio of 7.2 warheads per "available launch tube," for comparison between programmed and Heritage proposals.

TABLE 3: SURVIVING WEAPONS AFTER FIRST STRIKE, (PURELY SPECULATIVE)

[7] Alert bombers loaded to maximum capacity. Assumes a 30% bomber alert rate, and an arbitrary 50% attrition from enemy offensive and defensive weapons. If the enemy possessed sufficient SLBM throw - weight to attack all ICBMs and bomber bases at short ranges, both silo ICBM and bomber attrition rates could easily approach 100%.

[8] Assumes 97.5% probability of kill (equal to missile reliability) for Minuteman silos hardened to 2,000 psi.

[9] Assumes 90% probability of kill (2 warheads per target) for Sentinel silos hardened to 30,000 psi

[10] Assumes an arbitrary 50% survival rate for mobile ICBMs; depending on dispersal rate and area, this number could be higher.

[11] Assumes SSBN at - sea rate of 60%, and none destroyed at sea.

Regardless of any questions as to warhead loading, the difference between the number of strategic warheads in the Heritage and programmed forces will always be relatively small, given that the Heritage proposal does not increase the number of bomber weapons, and only adds 100 ICBMs and 48 - 64 available SLBM launch tubes (depending on the percentage of operational submarines) – a 25% and a 30 - 40% increase over the programmed forces, respectively. However, probably the most consequential difference between the Heritage and programmed strategic forces – and certainly between the Heritage force and the present force – is increased technological diversification. Presently, the U.S. is totally reliant upon its ballistic missile submarines (SSBNs) to survive an enemy surprise attack. Most sources agree that, in such a scenario, the U.S. could not rely on a “launch - under - attack,” for its ICBMs, as there have been previous false alarms during peacetime which might have resulted in a U.S. launch under such a posture; ^[9] and, either way, the present U.S. bomber force is not on alert and would therefore be destroyed at its bases in any attack. For these reasons, any adversary breakthrough in anti submarine warfare would have an outsized impact on current U.S. retaliatory capability. By placing the bombers on alert, adding 110 - 150 bombers to provide an adequate alert force to lift the programmed LRSO inventory, plus a buffer for non - nuclear operations, plus 50 road - mobile missiles, the U.S. would gain a substantially greater hedge against the current rapid progress in ASW technology.

Although the largest absolute increase in warhead totals in both the Heritage and programmed forces is allocated to strategic forces, the biggest divergence between the two, in both absolute and relative terms, is in the realm of non - strategic nuclear weapons, often referred to as tactical or theater nuclear weapons (TNF.) Currently, the U.S. possesses a token force of 200 gravity bombs, clearly intended primarily for political rather than military objectives. Heritage proposes adding to this force an additional 200 gravity bombs, 400 SLCMs, 100 hypersonic weapons, 100 ground launched ballistic missiles, 100 ground launched cruise missiles, and 25 LRASM air launched cruise missiles, for a total of 1,125 weapons. Since the size of the future SLCM - N and B-61 inventories is more or less up in the air, and other weapons are unspecified, no comparison with the “programmed force” can be provided. Regardless, the increase in effectiveness compared with the current force would probably be considerably greater than the numbers suggest, given the ability of fast - flying hypersonic and / or ballistic missiles to strike time - urgent targets, and the value of technological diversification in surviving an enemy first strike and penetrating enemy defenses.

To be sure, reasonable alternatives to the Heritage program exist. It is worth mentioning, for instance, that Congressman Adam Smith, Ranking Member of the House Armed Services Committee, and longtime opponent of the Sentinel ICBM program as presently constituted – has previously floated the rather more radical idea of reducing the overall size of the ICBM force, while exchanging all silo - based missiles for a 100% mobile force. In his view, the programmed (exclusively silo - based) force is too costly for the capability it provides; since few missiles would survive after ‘riding out’ an attack, they must either be launched on warning or lost, necessitating ‘a split - second decision to start a nuclear war.’ If, he says, ‘making them mobile was cost - effective and could be done, then at least you eliminate that threat and make them a more survivable element.’ A fully mobile force might, (depending very much on the particulars of its deployment,) be more survivable in case of surprise attack, more survivable in the case of strategic warning, (i.e. intelligence indicators or crisis prior to actual detection of an incoming attack,) and might also reduce U.S. civilian casualties in a limited nuclear war by largely eliminating the fallout arising from an enemy counterforce strike. Overall, however, while Smith believes that all options should be on the table, he leans towards abandoning land - based ICBMs entirely and focusing on the improvement of U.S. bomber and submarine forces, which he views as more inherently survivable; and, more generally, holding down expenditures on nuclear forces to a necessary minimum in order to free up resources for other forces. ^[10]

Mr. Peters, when asked about this, said that he’s open to changing the composition of the U.S. ICBM force, and hopes that the Heritage proposal for 50 mobile missiles can serve as a starting point for further discussion; but he does not favor reducing the total size of the ICBM force, much less eliminating these weapons altogether.

How, then, do the present, programmed, and Heritage forces measure up against the current and projected opposition? Unfortunately, that question is difficult to answer. Although U.S. targeting requirements are not strictly pegged to the number of enemy warheads, such comparisons are useful because, if the U.S. wishes to simultaneously deter both the PRC and Russia, one useful rule of thumb is that the U.S. should be able to inflict an amount of damage on PRC and Russian target sets that is at least comparable to the amount of damage which these adversaries can inflict upon U.S. target sets. Otherwise, the PRC and Russia (especially if acting in concert) will enjoy escalation dominance and, depending on the degree of their relative advantage, will have an incentive to escalate to the use of greater and greater numbers of nuclear weapons if, in a crisis or limited war, they are ever at risk of being defeated at lower rungs on the escalation ladder. During the Cold War, the United States generally sought to prevent this eventuality by maintaining a total nuclear force that was at least somewhat comparable to the total Soviet nuclear force, (“essential equivalence,”) and to this end provided regular “Soviet Military Power” reports which included relatively extensive information on Soviet nuclear forces in order to justify U.S. programs. Contrary to popular belief, however, even classified U.S. Cold War estimates serially underestimated actual Soviet stockpiles of both warheads and fissile material, as confirmed by post Cold War sources. ^{[11] [12]}

In the post - Cold War era, most U.S. political leaders were strongly averse to nuclear weapons; and furthermore, they found themselves (for the first time since 1945) with both a political opportunity and a compelling military incentive to reduce their reliance on these weapons as much as possible, (since it was at that time the United States, and not its adversaries, which enjoyed increasing advantages in non - nuclear forces.) Consequently, the objective of “essential equivalence” was allowed to lapse; and, beginning with President George H.W. Bush’s unilateral reductions of 1991 (which began the elimination of U.S. nonstrategic weapons without any requirement for Russian reciprocity,) U.S. disarmament began to run ahead of bilateral U.S. - Russian agreements. In recent years, most sources (including no less an advocate for strategic forces than Senator Deb Fischer, Chair of the Senate Strategic Forces Subcommittee,) have assessed that the objective of parity with the combined nuclear forces of the PRC and Russia is unrealistic, ^[13] (meaning that the old standard of “essential equivalence” is more or less a non - starter, even in the pro - nuclear circles of Congress,) although, of course, all previous assessments of what constitutes an optimal or achievable balance of spending will be outdated if and when the Administration gets its wish for fifty percent more money. At any rate, perhaps as a result of the fact that “essential equivalence” has remained out of vogue for one reason or another, (and therefore the previous desire to secure funding for U.S. nuclear programs was generally replaced by a desire to hold down or reduce funding for these programs,) the U.S. government more or less ceased detailed public reporting on adversary nuclear capabilities after the end of the Cold War, and has not since resumed.

TABLE 4: CURRENT & FUTURE U.S. FORCE LOADING vs CURRENT & FUTURE ADVERSARY FORCES

[1] Includes 1,400 missile weapons, 536 ALCM, 304 B-61s for B-2s, and 200 B-61s for tactical fighters

[2] 3 warheads per ICBM, 9.3 warheads per operational SLBM, 20 per B-52, 16 per B-2. Theoretical maximum capacity of SLBMs is 12 light warheads. See CBO, The Potential Costs of Expanding U.S. Strategic Nuclear Forces If the New START Treaty Expires, https://www.cbo.gov/publication/56524 (This figure is included for the sake of making an apples - to - apples comparison with open - source estimates of PRC and Russian arsenals, which are largely based on maximum force loading.)

[3] Assumes ~1,475 warheads loaded on ICBMs / SLBMs; ~1,000 total bomber weapons, (equal to the capacity of bombers,) plus the minimum consensus estimate of about 2,000 nonstrategic weapons.

[4] Based on figures in Schneider, How Many Nuclear Warheads does Russia Have? https://nipp.org/wp-content/uploads/2023/09/Vol.-3-No.-8.pdf Note that Schneider himself does NOT provide an estimate for Russian warhead totals; only a breakdown of component data, from which the reader may draw his own conclusions.

[5] Fisher, 23rd NUCLEAR TRIAD SYMPOSIUM https://www.youtube.com/watch?v=3-77jWb8mH8&t=20396s

The above table contains a comparison of U.S. government (CMPR) and open source estimates of present and near future Russian and Chinese nuclear forces. Obviously, the range of uncertainty is very great. For context, the DIA assessed last year that the PRC would possess 10,000 long range missiles of various kinds by the year 2035, [14] (though the range of each of these systems is unclear;) warheads are cheap to produce compared with missiles, and the PLA maintains a general policy of keeping its long – ranged missiles dual – capable, (with swappable conventional and nuclear warheads,) in contrast to the U.S., which has decided to segregate its nuclear and non – nuclear missiles. The CMPR figure of 1,000 warheads by 2030 therefore seems to imply that fissile material production is and will remain a bottleneck, although by that time the PRC’s nuclear buildup will be at least 10 years old; both the U.S. and USSR were already capable of producing thousands of warheads per year by the 1960s, and about 50% of the known nuclear reactors currently under construction worldwide are in China. [15]

What can be said with a relatively high degree of confidence is that, if the Russians have chosen for the last five years to abide by the New START Treaty limits despite the lack of inspections, the United States today may still be inferior to Russia alone in operational nuclear weapons by a margin of around 2 : 1 or more. If Russia has not held itself to the New START limits, then Russia’s superiority today is greater still. Practically nothing can be said about China with a high degree of confidence, precisely because the PRC has pursued a strategy of “hiding and biding,” and to this end has carefully concealed its actual strength in nuclear weapons.

Regardless, it can at least be surmised that the United States is inferior to its combined adversaries in nuclear forces today to a degree that was never considered acceptable on either side of the U.S. political spectrum at any time during the Cold War, and that that balance is deteriorating monthly with the current rapid expansion of PRC nuclear forces. With that said, any assessment of the overall nuclear balance beyond the next five years or so would require concrete projections of the size and effectiveness of U.S. (and adversary) strategic defenses, and such assessments do not presently exist. If the entire Heritage program were completed next year – with no corresponding change in Russian and PRC forces – it might easily still fall far short of the Cold War standard of “essential equivalence” with Russian forces alone – never mind the PRC and Russia together – and may therefore leave Russia and / or the PRC with strong incentives to escalate to the use of nuclear weapons against U.S. forces in a crisis or war. Importantly, the stronger U.S. general purpose (that is, non - nuclear) forces become – all else being equal – the greater the adversary’s relative incentive will be to capitalize on nuclear advantages, either by launching a nuclear surprise attack, or by escalating to nuclear use if non - nuclear operations should fail to achieve their objectives. As such, any overall program of defense which does not adequately address the nuclear balance may only increase the probability of a nuclear attack against the United States, by affording the PRC and Russia both the opportunity and the incentive to escalate to win.

Consequently, Heritage chose to title its recommendations for the U.S. nuclear arsenal of 2050 “a proposal for American Survival.” It represents, in the estimation of Mr. Peters, a necessary minimum effort, and a reasonable starting point for further discussion and iteration. In the final analysis, the only way to know for sure whether or not he is right is to suffer a nuclear attack, in which case we will know with one hundred percent certainty that the deterrent margin was not sufficient.

---

Ben Ollerenshaw is a journalist specializing in defense matters whose work has appeared in the Hill, the National Interest, RealClearDefense, and other publications. He may be contacted at benollerenshaw2@gmail.com

Read the full text of the Heritage Proposal here: https://www.heritage.org/defense/report/the-us-nuclear-arsenal-2050-proposal-american-survival

Note: The cost and residual warhead estimates appearing in this article are the author’s, and do not reflect the views of Mr. Peters or the Heritage Foundation. This article also cites the author’s interviews with Congressman Adam Smith, Ranking Member of the HASC, and Senator Deb Fischer, Chair of the SASC Subcommittee on Strategic Forces, which have not previously appeared elsewhere.

Notes: