China, Russia, and the United States are engaged in a costly missile race that fuels tensions and increases the risk of nuclear escalation during a conventional conflict. Russia’s and the United States’ strategic forces will continue to be limited, at least until February 2026 when New START expires. However, the demise of the INF Treaty has opened the door to competition in ground-launched missiles with ranges between 500 kilometers (310 miles) and 5,500 kilometers (3,400 miles). In response to the United States’ withdrawal from the treaty, Putin announced that Russia would develop a ground-launched version of the Kalibr SLCM—which is, in fact, what the SSC-8 is believed to be—and a ground-launched hypersonic missile.1 The United States, meanwhile, is developing a number of conventional ground-launched missiles, including a variant of the Tomahawk cruise missile, a surface-attack variant of the SM-6 air-defense missile, a new tactical ballistic missile, and a hypersonic boost-glide weapon.2
Countering China’s large and growing force of regional missiles was one important rationale behind the United States’ decision to leave the INF Treaty.3 China has deployed seven types of ground-launched ballistic, boost-glide, and cruise missiles with ranges between 500 and 5,550 kilometers (some of which are produced in multiple variants).4 Three of these missiles—the DF-21 and DF-26 ground-launched ballistic missiles and the DF-17 ground-launched boost-glide missile—are known or believed to be dual-use. Beijing views the United States’ development of new ground-launched missiles with concern and may increase its own stockpile of such missiles to maintain its large numerical advantage.5 China is also modernizing and expanding its strategic nuclear forces, which are not restricted by any arms control agreement, though are significantly smaller than the corresponding Russian and U.S. arsenals.
Looking forward, this trilateral arms race could intensify, including through the development of new ground-launched missiles capable of delivering nuclear and nonnuclear warheads accurately over ever longer ranges. Intensified quantitative arms racing is another risk—especially if Russia and the United States are unable to negotiate a replacement to New START—with potentially far-reaching consequences. If China perceives that a U.S. buildup increases the threat to its nuclear forces, it may accelerate its program to expand those forces (which, in turn, could have implications for India and hence Pakistan).
Developments in aircraft and air-delivered weapons are less competitive, though still significant. New START limits heavy bombers, defined as aircraft with a range in excess of 8,000 kilometers (5,000 miles) or that are equipped for long-range, nuclear-armed air-launched cruise missiles. However, the development of increasingly capable air-launched missiles is threatening to confer similar capabilities on aircraft that are not accountable under New START. For example, there is concern in the United States that Russia may deploy Kinzhal, a dual-use air-launched ballistic missile with a claimed range of up to 2,000 kilometers (1,200 miles), on its Tu-22M (Backfire) medium bomber—enabling that aircraft to conduct nuclear or precise conventional strikes over distances currently within reach of heavy bombers only.6 China, meanwhile, has developed an air-launched ballistic missile for its H-6N bomber that “may be nuclear capable.”7 (It is unclear whether China’s current bomber meets the definition of a heavy bomber and whether its developmental stealth bomber will do so, but it is also moot given that China is not a party to New START.)
Looking forward, this trilateral arms race could intensify, including through the development of new ground-launched missiles capable of delivering nuclear and nonnuclear warheads accurately over ever longer ranges.
These developments could also increase the danger of escalation. In particular, Beijing and Moscow appear to worry that, in a crisis or conflict, highly accurate U.S. conventional missiles might be used to attack their nuclear forces or military or national leadership with little warning. (Indeed, Moscow worried that the nuclear-armed U.S. Pershing II intermediate-range ballistic missiles deployed to Europe in the 1980s could be used for exactly that purpose.) This concern has been exacerbated by U.S. deployments of missile defense interceptors, which China and Russia fear could be used to intercept any nuclear missiles that survived a U.S. first strike and were fired in retaliation. In a crisis or conflict, this set of fears could prompt China or Russia to launch preemptive attacks on U.S. ground-launched missiles (ideally before their dispersal), missile defense assets, or their associated enabling capabilities. In extremis, it could even add to the pressures on China or Russia to make nuclear threats or to initiate nuclear first use.
In fact, for Russia in particular, concerns about counterforce attacks and missile defense operations are intertwined because of the United States’ deployment of SM-3 interceptors at Aegis Ashore sites (see chapter 3). Russian officials have expressed concerns that these interceptors could be rapidly converted into ground-attack ballistic missiles. They have also highlighted the possibility of Aegis Ashore launchers being used to fire cruise missiles. U.S. statements that those launchers lack that capability do not appear to have assured Russia, which points to the 2019 test launch of a cruise missile from a land-based MK-41 Vertical Launching System, from which the Aegis Ashore system is derived.
Limits on missiles and aircraft could be useful in managing these dangers. They could help to prevent arms racing and mitigate escalation risks by curtailing the threat posed to military and national leadership, and to nuclear forces and their enabling capabilities. Depending on precisely which capabilities were limited, however, agreements could be extremely complex to negotiate and verify—difficulties that would be compounded by the involvement of three parties.
A relatively simple approach, based on an idea by Zhao, would be for China, Russia, and the United States to agree to a single limit on all missile launchers and aircraft that were limited under the INF Treaty or are limited by New START, along with a few other closely related types.8 Such an agreement would build directly on past experience and complement—not replace—the limits on Russian and U.S. strategic forces imposed by New START or any successor treaty.
The trilateral treaty proposed here would cover three specific categories of weapons:
- Launchers for ground-launched cruise missiles, ground-launched ballistic missiles, and ground-launched boost-glide missiles with ranges over 475 kilometers (295 miles)
- SLBM launchers
- Bombers with ranges greater than 2,000 kilometers
Each accountable launcher and each accountable bomber would count as one unit, irrespective of whether it was armed with nuclear or nonnuclear armaments (or, indeed, was unarmed) and whether or not it was deployed. Warheads and missiles would not be limited. However, because missiles and warheads are not militarily useful without launchers, this approach would impose a meaningful limitation on the parties’ overall conventional and nuclear capabilities.9 It would also enhance the proposed agreement’s feasibility. Limiting launchers, rather than smaller items like warheads or missiles, would reduce verification difficulties significantly. The first U.S.-Soviet arms limitation treaties—the Strategic Arms Limitation Talks (SALT) I Interim Agreement and the never-ratified SALT II Treaty—focused on launchers largely for this reason.
Underlying the proposed agreement would be a key compromise. In a major concession to China and Russia, the United States would agree that, for the purposes of treaty implementation, its Aegis Ashore launchers met the definition of a launcher for ground-launched cruise missiles and were thus accountable. This provision should help to manage Chinese and Russian concerns about the possible future growth of U.S. missile defense capabilities. Meanwhile, in a major reciprocal concession to Washington, Beijing and Moscow would agree that the treaty should not constrain SLCMs or future SLBGMs—an area of U.S. advantage—on the grounds that the relevant launchers typically also accommodate other types of weapons, such as air- and missile-defense interceptors and anti-submarine weapons.
To facilitate this compromise, the proposed treaty would exempt from accountability any SLBM launchers converted into SLCM launchers prior to its entry into force. In practice, this provision would apply only to the launchers on the United States’ four SSGNs (SSBNs converted to carry cruise missiles). Pursuant to New START, Russia currently has limited rights to inspect these launchers. China, which is not a party to New START, would have to rely on verification by Russia to bolster its own NTM. While this approach would not be ideal, especially from Beijing’s perspective, the problem would only be temporary since all four U.S. SSGNs are due to be retired by 2028.10
All other launchers and bombers could be removed from accountability only after they had been permanently eliminated. Thus, the four SLBM launchers on each U.S. SSBN that were “converted” (in other words, rendered inoperative) to meet New START’s central limits would count under the proposed treaty.
Ground-based launchers would be accountable if used to launch missiles with a range in excess of 475 kilometers. This distance—which is slightly shorter than the 500 kilometers threshold adopted in the INF Treaty—would bypass the controversy over the range of the SSC-8. The United States claims (almost certainly correctly) that this missile has a range of “well over” 500 kilometers; Russia has pegged it at 480 kilometers (300 miles).11 The proposed treaty would address U.S. concerns without forcing Russia to acknowledge that it had violated the INF Treaty.
|Table 3: Estimates of Chinese, Russian, and U.S. Accountable Launchers and Accountable Bombers (as of October 2021)|
|Category||Missile or Bomber Type
NATO Reporting Name (national designation)
|Number of Launchers or Bombers|
|GLBMs and GLBGMs||CSS-6 (DF-15) Mods 1, 2, 3
CSS-7 (DF-11) Mods 1, 2
CSS-11 (DF-16) Mods 1, 2
|Notes: The DF-11 Mod 1 has a range less than 475 kilometers, but its launchers would be accountable if they have ever launched DF-11 Mod 2.
Sources (see bottom of table for complete citations): DIBMAC (2020, 21), OSD (2020, 166)
|CSS-5 (DF-21) Mods 2, 4, 5, 6
|Sources (see bottom of table for complete citations): DIBMAC (2020, 25), OSD (2020, 86)|
CSS-4 (DF-5) Mods 2, 3
CSS-10 (DF-31) Mods 1, 2, 3
|Notes: At least 230 new silos, likely for DF-41s, are under construction
Sources (see bottom of table for complete citations): Broad and Sanger (2021), DIBMAC (2020, 29), OSD (2020, 166), Lee (2021), Warrick (2021)
|Sources (see bottom of table for complete citations): DIBMAC (2020, 36), OSD (2020, 56 and 166)|
|Notes: Two additional SSBNs are expected to enter service shortly.
Sources (see bottom of table for complete citations): DIBMAC (2020, 33), OSD (2020, 45)
|Sources (see bottom of table for complete citations): IISS (2021, 255)|
|Category||Missile or Bomber Type
NATO Reporting Name (national designation)
|Number of Launchers or Bombers|
|SS-26 (9K720 Iskander-M||475–1,000||~144|
|Notes: The SS-26 has a range less than 475 kilometers, but its launchers have launched SSC-7.
Sources (see bottom of table for complete citations): DIBMAC (2020, 21), Kristensen and Korda (2021b, 91)
|SS-18 Mod 5 (RS-20V)
SS-19 Mod 4 (Avangard)
SS-25 (RS-12M Topol)
SS-27 Mods 1, 2 (RS-12M1/12M2, Topol-M, RS-24 Yars)
|Notes: Russia also possesses a small but known number of prototype SS-28 (RS-26) launchers.
Sources (see bottom of table for complete citations): DIBMAC (2020, 29), Kristensen and Korda (2021b, 91)
3M-14 (and others from the Club-K system)
|Notes: The number of launchers is very uncertain and could be much larger because of an unknown number of Club-K systems from which cruise missiles with a range in excess of 475 kilometers, including the 3M-14, have been launched. Whether the range of the SSC-5 (~60 launchers) exceeds 475 kilometers is unclear. SSC-7 launchers are counted under SS-26.
Sources (see bottom of table for complete citations): DIBMAC (2020, 36), Kristensen and Korda (2021b, 91), Rosoboronexport (2021)
|SLBMs||SS-N-18 M1 (RSM-50)
SS-N-23 M1(RSM-54 Sineva)
SS-N-32 (RSM-56 Bulava)
|Sources (see bottom of table for complete citations): Kristensen and Korda (2021b, 91)|
|Sources (see bottom of table for complete citations): Kristensen and Korda (2021b, 91) IISS (2021, 198)|
|TOTAL||>800 (possibly significantly larger)|
|Category||Missile or Bomber Type
NATO Reporting Name (national designation)
|Number of Launchers or Bombers|
|Notes: HIMARS and perhaps M270A1 launchers are expected to launch missiles with ranges in excess of 475 kilometers in the future.|
|LGM-30G Minuteman III||>5,500||458|
|Sources (see bottom of table for complete citations): DOS (2020)|
|Notes: Launchers are considered to be accountable for the purposes of treaty implementation because the sea-based MK-41 Vertical Launching System has launched cruise missiles with ranges in excess of 475 kilometers. Additional launchers in Poland are under construction.
Sources (see bottom of table for complete citations): LaGrone (2013)
|SLBMs||UGM-133A Trident II D5||>600||336|
|Notes: Converted SLBM launchers on SSBNs, but not on SSGNs, are included.
Sources (see bottom of table for complete citations): DOS (2020) Kristensen and Korda (2021a, 51)
|Notes: Test bombers are excluded.
Sources (see bottom of table for complete citations): DOS (2020), IISS (2021, 57)
GLBM: ground-launched ballistic missile
GLBGM: ground-launched boost-glide missile
GLCM: ground-launched cruise missile
SLBM: sea-launched ballistic missile
|a There are significant uncertainties associated with the figures for China and particularly Russia.
Sources: William J. Broad and David E. Sanger, “A 2nd New Nuclear Missile Base for China, and Many Questions About Strategy,” New York Times, July 26, 2021, https://www.nytimes.com/2021/07/26/us/politics/china-nuclear-weapons.html; Department of State (DOS), Bureau of Arms Control Verification and Compliance, “New START Treaty Aggregate Numbers of Strategic Offensive Arms,” December 1, 2020, https://www.state.gov/new-start-treaty-aggregate-numbers-of-strategic-offensive-arms-15/; Hans M. Kristensen and Matt Korda, “United States Nuclear Weapons, 2021,” Bulletin of the Atomic Scientists 77, no. 1 (2021): 51; Hans M. Kristensen and Matt Korda, “Russian Nuclear Weapons, 2021,” Bulletin of the Atomic Scientists 77, no. 2 (2021): 91; International Institute for Strategic Studies (IISS), The Military Balance 2021 (London: IISS, 2021); Defense Intelligence Ballistic Missile Analysis Committee (DIBMAC), “Ballistic and Cruise Missile Threat 2020,” July 2020, https://media.defense.gov/2021/Jan/11/2002563190/-1/-1/1/2020%20BALLISTIC%20AND%20CRUISE%20MISSILE%20THREAT_FINAL_2OCT_REDUCEDFILE.PDF; Rod Lee, “PLA Likely Begins Construction of an Intercontinental Ballistic Missile Silo Site Near Hanggin Banner,” Air University, August 12, 2021, https://www.airuniversity.af.edu/CASI/Display/Article/2729781/pla-likely-begins-construction-of-an-intercontinental-ballistic-missile-silo-si/; Office of the Secretary of Defense (OSD), “Military and Security Developments Involving the People’s Republic of China 2020,” Annual Report to Congress, 2020, https://media.defense.gov/2020/Sep/01/2002488689/-1/-1/1/2020-DOD-CHINA-MILITARY-POWER-REPORT-FINAL.PDF; Rosoboronexport, “Club-K,” 2021, http://roe.ru/esp/catalog/marina-de-guerra/armas-de-la-nave/klab-k/; Sam LaGrone, “Inside Aegis Ashore,” USNI News, August 8, 2013, http://roe.ru/esp/catalog/marina-de-guerra/armas-de-la-nave/klab-k/; and Joby Warrick, “China Is Building More Than 100 New Missile Silos in Its Western Desert, Analysts Say,” Washington Post, June 30, 2021, https://www.washingtonpost.com/national-security/china-nuclear-missile-silos/2021/06/30/0fa8debc-d9c2-11eb-bb9e-70fda8c37057_story.html<.
To simplify negotiations and reduce the need for intrusive inspections, if one ground-based launcher of a given type has been used to launch a cruise, ballistic, or boost-glide missile over a distance of greater than 475 kilometers, then all launchers of that type would be accountable—even if loaded with missiles of a shorter range or with no missile at all. Russia’s 9P78 (Iskander) launchers, for example, can be deployed with the SSC-7 cruise missile (which has a range of 490 kilometers or 305 miles), the SS-26 ballistic missile (which has a range of 350 kilometers or 220 miles), and perhaps also the SSC-8 cruise missile. All such launchers would be accountable.
The proposed treaty would limit all bombers with a range greater than 2,000 kilometers—a somewhat broader array of types than are limited under New START. Specifically, medium bombers, such as Russia’s Tu-22M bomber, would be accountable because improved long-range munitions could confer them with capabilities similar to heavy bombers.12 Similarly, U.S. bombers that have been converted to deliver only nonnuclear munitions and are not accountable under New START would be accountable under the proposed treaty.
Given these provisions, it appears that China, Russia, and the United States currently possess roughly equal numbers of accountable launchers and accountable bombers—though there are large uncertainties in the estimates for China and particularly Russia.13 (See table 3 for rough estimates of the number and types of accountable systems each state currently possesses.)
Looking forward, each state’s force of accountable launchers and accountable bombers is likely to grow. The United States possesses a few hundred ground-based M142 High Mobility Artillery Rocket System (HIMARS) launchers and Multiple Launch Rocket System M270A1 launchers that would not be accountable today. However, it plans to test the former type, and perhaps the latter type too, with a missile that has a range longer than 475 kilometers—at which point all launchers of each type used in any tests would become accountable. China, meanwhile, is expanding its missile and bomber forces, including through the construction of at least 230 new ICBM silos, and may keep pace with the United States. The opacity around Russia’s nonstrategic missile force makes its current composition and future trajectory difficult to assess—though it is likely that Russia currently possesses more than the 800 accountable launchers and accountable bombers listed in table 3 and is constructing more. As a result, it is possible, albeit far from certain, that the rough equality in accountable launchers and accountable bombers will persist, which would enhance the treaty’s political feasibility.
In any case, the central limit chosen should be slightly higher than any state’s arsenal of accountable launchers and accountable bombers at the time of entry into force. This approach, which mirrors the idea behind the SALT I Interim Agreement, would forestall the nascent missile arms race. Moreover, by allowing states to partially implement their modernization plans, perhaps without having to dismantle existing weapons, it would be more politically palatable than a treaty that required reductions. The trade-offs imposed by the limit would enhance security. If China, for example, sought to expand its strategic forces significantly, it would have to reduce its regional forces, mitigating U.S. concerns about the military balance in the Western Pacific. Meanwhile, if the United States decided to build up its regional forces significantly, it would have to reduce its strategic forces, mitigating Chinese concerns about the survivability of its nuclear arsenal.
Central Limits for Accountable Launchers and Accountable Bombers
The treaty should limit China, Russia, and the United States to an equal number of accountable launchers and accountable bombers. The limit chosen should allow each state to make modest increases in such launchers and bombers after the treaty’s entry into force.
The treaty should exempt from accountability any launchers for SLBMs that were converted into launchers for SLCMs prior to entry into force.
Each accountable launcher or accountable bomber should first become accountable at the following stage of development:
- For fixed launchers, when silo doors are attached to a missile silo or when a missile is deployed within a silo, whichever happens earlier
- For mobile launchers, when a launcher leaves the production facility for the first time
- For SLBM launchers, when an SSBN puts to sea for the first time
- For accountable bombers, when a bomber takes flight for the first time
In implementing the treaty provisions, the following definitions should apply:
- “Accountable bomber” means a bomber of a type, any one of which has an unrefueled range—when carrying an ordnance load of at least 5,000 kilograms (11,000 pounds)—in excess of 2,000 kilometers.14
- “Accountable launcher” means an accountable ground-based launcher or an SLBM launcher.
- “Accountable ground-based launcher” means a ground-based device of a type, any one of which has been used to launch a cruise, ballistic, or boost-glide missile with a range in excess of 475 kilometers against a ground-based target.
- “SLBM launcher” means a device on a ship or submarine of a type, any one of which has been used to launch a ballistic missile with a range in excess of 600 kilometers (370 miles) against a ground-based target.
- “Ballistic missile” means a weapon-delivery vehicle that has a ballistic trajectory over most of its flight path.
- “Boost-glide missile” means a weapon-delivery vehicle that sustains unpowered flight through the use of aerodynamic lift over most of its flight path. A reaction control system designed to change a vehicle’s attitude is not considered capable of powering flight.
- “Cruise missile” means an unmanned, self-propelled weapon-delivery vehicle that sustains flight through the use of aerodynamic lift over most of its flight path.
The parties should agree that, for the purposes of treaty implementation, the definition of an accountable launcher includes any ground-based launcher of a type derived from a ship-based vertical launch system from which a cruise missile with a range in excess of 475 kilometers has been launched against a ground-based target.
The definitions for an accountable launcher or accountable bomber are not intended to include systems used to launch target missiles for missile defense tests, provided such systems are of a different type and distinguishable from any accountable launcher or accountable bomber.
Definitions of missile ranges are discussed in appendix A.
The verification and monitoring regime should consist of two main parts: information exchanges and confirmation through NTM and on-site inspections.
Data exchanges. China, Russia, and the United States should exchange baseline information, comprehensive semiannual updates, and regular notifications about accountable launchers and accountable bombers. This exchange would enable each party to maintain an identical database that it would attempt to verify. A party would be required to declare and provide notifications about all its accountable launchers and accountable bombers, regardless of whether they were based on its own territory.
Verification: Data Exchanges and Notifications
As part of the baseline data exchange thirty days after the treaty’s entry into force, each party should:
- List all its types of accountable launchers and accountable bombers
- Declare how many accountable launchers and accountable bombers it possesses, in aggregate and disaggregated by type
- Identify externally observable distinguishing features (should any exist) to differentiate accountable launchers and accountable bombers from similar nonaccountable systems (should any exist)
- Identify all bases for accountable launchers and accountable bombers
- Declare, for every base for mobile accountable ground-based launchers, each fixed structure and unenclosed storage area (whether or not it typically contains accountable launchers)
- Provide a unique identifier (UID) for every mobile accountable ground-based launcher and assign each such launcher to the base on which, and the specific fixed structure or unenclosed storage area in which, it is typically stored (where possible, the parties may utilize preexisting UIDs assigned pursuant to New START or any future arms control treaty)
- Provide a site diagram for each base for accountable ground-based launchers, showing—with geographic coordinates—the location of each fixed launcher, each fixed structure (whether or not it typically contains accountable launchers), and the boundaries of each unenclosed storage area (whether or not it typically contains accountable launchers)
The parties should plan an implementation meeting to discuss any questions arising from the baseline data exchanges.
Following the baseline data exchange, the parties should exchange the same categories of data every six months.
The parties should notify each other of the production, first deployment, or dismantlement of accountable launchers and accountable bombers. Deployment notifications for fixed accountable ground-based launchers should specify the coordinates of the deployment location. Deployment notifications for mobile accountable ground-based launchers should specify the base on which, and the fixed structure or unenclosed storage area in which, the launcher is typically stored.
The parties should also notify one another of any change to the fixed structure or unenclosed storage area in which a mobile accountable ground-based launcher is typically stored. Such a notification should specify the new fixed structure or unenclosed storage area in which (and, if necessary, the new base on which) the missile is typically stored.
In implementing the provisions of this treaty, the following definitions should apply:
- “Base” means a facility at which accountable launchers or accountable bombers are typically stored and maintained. A facility designed, intended, or used to accommodate accountable launchers or accountable bombers temporarily is not considered to be a base, provided that no accountable launchers or accountable bombers are typically stored or maintained there.
- “Fixed structure” means a unique structure within a base designed, intended, or used to store mobile accountable ground-based launchers.
- “Unenclosed storage area” means any area, other than a fixed structure, designed, intended, or used to store mobile accountable ground-based launchers.
National technical means. The parties should be able to use NTM, including satellite imagery, to verify numbers of accountable bombers, fixed accountable ground-based launchers (silos), and SLBM launchers. Because of developments in satellite technology, the NTM capabilities of China, Russia, and the United States have improved dramatically since the 1970s. Even back then, the SALT I Interim Agreement and SALT II Treaty, which limited similar categories of arms, did not include on-site inspections because those arms were large enough to be visible with overhead imagery. The parties should agree, however, not to interfere with one another’s NTM.
Verification: Non-Interference With NTM
Each party should agree not to interfere with the NTM of other parties by, for example, using concealment measures to impede the verification of accountable launchers or accountable bombers located on bases. This treaty provision is not intended to require any party to modify any existing facility.
On-site inspections. Although on-site inspections of fixed launchers would be unnecessary, on-site inspections to verify mobile accountable ground-based launchers would be needed. Although they can be stored in unenclosed areas, such launchers are usually based and maintained inside structures (garages) when they are not deployed in the field, limiting the utility of overhead imagery.15
The United States already inspects Russia’s mobile ICBM launchers under New START. Inspections under the proposed treaty here would be similar but less intrusive in one important respect. Inspectors would aim to confirm the assignment of mobile accountable ground-based launchers to a specific fixed structure or unenclosed storage area as well as their absence from fixed structures and unenclosed storage areas declared to not contain them. However, there would be no requirement, as in New START, to display missile front sections, revealing the attached reentry vehicles and other sensitive objects.
Inspectors would have the right to verify all the mobile accountable ground-based launchers in a given fixed structure or unenclosed storage area—typically at least two and sometimes more. By contrast, under New START, inspectors can verify just one launcher on each inspection. For this reason, the proposed treaty provides for fewer inspections than New START. Specifically, each party would be entitled to conduct ten inspections annually against each of the other parties, whereas New START permits eighteen inspections annually against the other party. Inspections could take place on the territory of a third party if mobile accountable ground-based launchers are based there. For example, HIMARS launchers, which would likely have become accountable by the time the proposed treaty entered into force, are currently based with U.S. Army and Marine units on allied territory.16
Verification: Inspections of Mobile Accountable Ground-Based Launchers
To verify declared information, each party should be permitted to conduct up to ten inspections per year at bases for mobile accountable ground-based launchers of each other party.
The inspecting party should inform the host state of a pending inspection at least twenty-four hours in advance of the inspection team’s arrival in country.
No later than one hour after an inspection team arrives in country and designates for inspection a base for mobile accountable ground-based launchers, the inspected party should cease any movements of such launchers into or out of the base.
Prior to departing for the designated base, the host state escort team should inform the inspection team if 50 percent or more of the mobile accountable ground-based launchers declared to be typically stored at the base are temporarily absent.17 If so, the inspection team should have the right to (1) proceed with the inspection as planned, (2) designate for inspection a different base for mobile accountable ground-based launchers, or (3) decline to conduct an inspection with no reduction in the number of inspections the inspecting party is entitled to conduct.
After arrival at the designated base, the escort team should inform the inspection team of (1) the total number of mobile accountable ground-based launchers located at the base; (2) for each fixed structure or unenclosed storage area in which mobile accountable ground-based launchers are located, the number of such launchers present and their UIDs; and (3) a list of any fixed structures or unenclosed storage areas in which no mobile accountable ground-based missiles are located.
During an inspection, the inspection team should have the right to designate and inspect one fixed structure or unenclosed storage area declared to contain mobile accountable ground-based launchers for the purposes of confirming the total number of such launchers in that fixed structure or unenclosed storage area and their UIDs. Additionally, if applicable, the inspection team should have the right to designate and inspect one fixed structure or unenclosed storage area declared not to contain mobile accountable ground-based launchers to confirm the absence of such launchers.
Within thirty days of the first deployment of a new type of mobile accountable ground-based launcher, one launcher of this type must be exhibited to inspectors to enable them to establish the launcher’s physical dimensions and to take photographs to facilitate future verification. If this exhibition is not possible during a routine inspection, then it should not count against the inspecting party’s inspection quota.
Eliminations. Parties would be permitted to eliminate accountable launchers and accountable bombers as necessary to meet the treaty’s numerical limits by following specified procedures—the results of which could be confirmed through NTM.
Verification: Elimination Procedures
Any party that plans to eliminate accountable launchers or accountable bombers must notify the other parties thirty days in advance of the elimination process, detailing the type and number of accountable launchers or accountable bombers to be eliminated and the planned location for the elimination procedures. For fixed accountable ground-based launchers, the notification should include the launchers’ deployment locations. For mobile accountable ground-based launchers, the notification should include the launchers’ UIDs.
The elimination procedures should depend on the type of system being eliminated and closely follow those specified by New START, with the requirement that no SLBM launcher on an SSBN should be considered eliminated until all the SLBM launchers on that SSBN have been eliminated.18
Eliminated accountable launchers and accountable bombers should remain visible to NTM for a period of sixty days after elimination procedures have been completed. After this sixty-day period has expired, eliminated accountable launchers and accountable bombers should no longer be considered accountable.
Technical feasibility. The proposed agreement is built on Russia’s and the United States’ extensive experience in implementing limits on heavy bombers and various types of missile launchers pursuant to New START, START I, the INF Treaty, and the SALT I Interim Agreement. Currently, under New START, Moscow and Washington regularly exchange detailed information on ICBM launchers, SLBM launchers, and heavy bombers, and inspect all accountable systems, including the launchers for mobile ICBMs.
Trilateral arms control would represent a significant increase in complexity, but most of the additional challenges would be political. While China might operate its mobile missile forces in a crisis or conflict in a quite different way from Russia (including the extensive use of underground facilities), its day-to-day operations appear to be fairly similar.19 Thus, from a purely technical perspective, the inclusion of China would represent only a modest challenge. While a trilateral treaty would likely be somewhat more difficult to negotiate, implementation would only be marginally more complex. After all, data exchanges and inspections would take place bilaterally between each pair of parties; they would not be orchestrated by an international inspectorate.
The biggest technical challenge would be verifying the absence of undeclared accountable launchers—mobile ground-based launchers, in particular. Some of these launchers are small and highly maneuverable and hence tricky to detect with NTM. A party could therefore attempt to cheat and retain more weapons than permitted by the treaty by failing to declare some of them. Nonetheless, the United States has previously assessed this problem to be manageable. In 1988, during congressional hearings about the INF Treaty, the U.S. Air Force chief of staff, General Larry Welch, testified that
to maintain a militarily useful force of covert missiles, [the Soviets] would need an elaborate hidden infrastructure to hide the missiles, periodically test them and provide an effective means of employing them in a crisis. They would run a high risk of detection over time. . . . Apart from these technical considerations, the military utility of such a cheater force would be low compared to the risk.20
The NTM capabilities of all three parties are almost certainly more sophisticated today than U.S. capabilities were at the time of this hearing. Moreover, the verification challenges facing the INF Treaty were likely greater than the ones facing the proposed treaty here. The INF Treaty sought to eliminate an entire class of weapons; the proposed agreement would limit each party to somewhere in excess of 1,000 accountable launchers and accountable bombers. The size of a militarily significant violation—the kind of violation that a verification regime should be able to detect if a treaty is to be viable—was smaller, therefore, under the INF Treaty than it would be under the proposed agreement. (If states are seriously concerned about undeclared launchers, they could agree to use perimeter portal monitoring to track all launchers leaving production facilities. This technique, which was used in START I and the INF Treaty, involves continuous monitoring of all a production facility’s exits that are large enough for a missile to pass through to ensure that no undeclared missiles are leaving the facility. In practice, however, all parties would likely agree that this approach would be unacceptably intrusive.)
Political feasibility. The emerging arms race in ground-launched missiles and somewhat less competitive but still significant developments in aircraft technology appear to disquiet Beijing, Moscow, and Washington. Both parties in the U.S.-Russian and U.S.-Chinese dyads express concerns about the other’s programs. Below the surface, there may also be some concern, particularly in Moscow, about the possibility of renewed competition between China and Russia over the long term. Moreover, and perhaps most importantly, technological developments are driving up the risk of escalation should growing tensions spark a conventional conflict.
The treaty proposed here could help restrain an adversary’s threatening capabilities, mitigate the costs and tensions associated with arms racing, and manage the escalation risks created by growing nonnuclear threats to nuclear forces. While the first two of these advantages, in particular, may be understood by Beijing, Moscow, and Washington, crafting a treaty that each participant viewed as beneficial to its specific interests would likely be more difficult when there are three parties rather than two. These difficulties would be magnified by the poor state of U.S.-Russian and U.S.-Chinese relations.
One threshold question is whether each party could accept the basic compromise underlying the proposed treaty: the United States’ Aegis Ashore launchers would be accountable; sea-based missiles, other than SLBMs, would not. (As argued in chapter 1, greater transparency about SLCMs and SLBGMs would be both possible and desirable and could enhance the viability of the proposed agreement.)
In evaluating this trade-off, the United States should recall that it has always justified its Aegis Ashore deployments on the need to combat Iranian missiles and that this agreement would not stop it from retaining, or even modestly expanding, such capabilities. Meanwhile, China and Russia should note that, although they lag the United States in sea-based missiles, they have made considerable efforts to enhance their capabilities in this area. Indeed, limits on SLCMs (and air-launched cruise missiles, for that matter) are noticeably absent from Putin’s proposal for a verified moratorium on Russian and U.S. deployments in Europe (including European Russia) of any missiles that were formerly prohibited under the INF Treaty.21 This proposal hints that Russia may entertain limits and transparency on land-based missiles by themselves (though the treaty proposed here would, by limiting bombers, also limit air-launched missiles indirectly).
Another challenge in securing a treaty is Beijing’s relative lack of experience with arms control, particularly negotiating and implementing limits on missiles or aircraft. Specifically, inspections of mobile ground-based launchers would represent a radical increase in transparency for China and hence could be a potential deal breaker. Two characteristics of the proposed agreement, however, should make it somewhat more palatable to Beijing than, say, limits on missiles or warheads. First, the parties would use NTM to verify silo launchers, SLBM launchers, and accountable bombers. Second, the intrusiveness of inspections, which would be limited to mobile ground-based launchers, would be mitigated since inspectors would not need access to missile front sections, as in New START. Indeed, China would not be required to provide any information at all about its warheads.
If China concluded that this basic treaty concept would advance its interests but that it could not accept on-site inspections, the parties should, at least initially, consider a non-legally binding agreement and rely exclusively on NTM for verification for some period of time. While this approach would present significant challenges—in trying to verify mobile launchers, in particular—the corresponding benefit of engaging China in a mutually beneficial arms control agreement would be considerable. If this arrangement proved successful, the parties could then enshrine the limits in a treaty with full verification provisions.
An additional consideration for China is that it appears to target states besides the United States and perhaps Russia with its missile forces. Various DF-21 battalions in central and southern China, for example, appear to hold Indian targets at risk.22 China may therefore view the proposal’s failure to constrain Indian forces as a significant flaw—though China’s ongoing missile competition with the United States is likely a more immediate concern than its competition with India.
Finally, within the United States, opposition to the proposed treaty may come from officials and analysts who are concerned about a new missile gap—in particular, in regional missile forces vis-à-vis China. Some would probably worry that the proposed treaty would enshrine U.S. inferiority in regional missiles and would likely argue that the United States should build up its force before seeking limits through arms control.
One problem with this approach is that deploying mobile ground-launched missiles to allied territory, where they would be militarily useful, would likely prove politically fraught. Local populations would almost certainly oppose such deployments, even for missiles that were not nuclear-armed. As a result, the deployments would likely be quite small—if, that is, they could be agreed on at all. The treaty proposed here would therefore foreclose an option—a major buildup of forward-deployed mobile ground-launched missiles—that, in practice, the United States would likely not to be able to execute anyway. (Instead, the United States would probably rely on SLCMs and perhaps SLBGMs, which would not be limited by the proposed treaty, to meet its deterrence requirements.)
A second problem is the costs and risks of a buildup. Even more important than the direct financial costs is the risk of escalation created by apparently unlimited deployments that lead China and Russia to fear that, in a conflict, the United States might attack their leadership, nuclear forces, and nuclear C3I systems. There are psychological, bureaucratic, and political reasons why this danger tends to get ignored but no good reason why it should be.23
The costs of building up forces before pursuing arms control would increase further if it stimulated arms racing. Further arms buildups may be unavoidable in any case since an arms control agreement would take time to negotiate, even if all three parties were willing to negotiate in good faith. And this near inevitable buildup might actually facilitate an arms control process by allowing states to feel that they have met their deterrence requirements. However, given China’s rapidly growing economy and Russia’s likely ability to further redirect domestic spending to the military, there is a clear risk of serious arms racing here.
Against this background, arms control could complement deterrence by reducing the risks—escalation and arms racing—associated with arms buildups. For arms control to have any chance of success, treaties must ultimately be fair; any proposal that enshrined superiority for the United States in regional and strategic forces would be summarily rejected by both Beijing and Moscow. The proposal outlined here recognizes the rough equality in accountable launchers and accountable bombers that already exists between China, Russia, and the United States—and that may continue for at least some time into the future. The question that decisionmakers in Washington—and Beijing and Moscow, for that matter—must ask themselves is how the risks of trying to break away from this rough equality compare to the risks of living with it.
1 Presidential Executive Office, “Meeting With Sergei Lavrov and Sergei Shoigu,” transcript, Moscow, February 2, 2019, http://en.kremlin.ru/events/president/news/59763.
2 Andrew Feickert, “U.S. Army Long-Range Precision Fires: Background and Issues for Congress,” R46721, Congressional Research Service, March 16, 2021, https://crsreports.congress.gov/product/pdf/R/R46721.
3 David E. Sanger and Edward Wong, “U.S. Ends Cold War Missile Treaty, With Aim of Countering China,” New York Times, August 1, 2019, https://www.nytimes.com/2019/08/01/world/asia/inf-missile-treaty.html.
4 Defense Intelligence Ballistic Missile Analysis Committee, “Ballistic and Cruise Missile Threat 2020,” 21 and 25.
5 David Lague, “Special Report: U.S. Rearms to Nullify China’s Missile Supremacy,” Reuters, May 6, 2020, https://www.reuters.com/article/us-usa-china-missiles-specialreport/special-report-u-s-rearms-to-nullify-chinas-missile-supremacy-idUSKBN22I16W.
6 Kristensen and Korda, “Russian Nuclear Weapons, 2021,” 102.
7 Office of the Secretary of Defense, “Military and Security Developments Involving the People’s Republic of China,” 51.
8 Zhao, “Opportunities for Nuclear Arms Control Engagement With China.”
9 This limitation would be “meaningful” rather than “total” because, in a conflict, an accountable launcher could be reloaded after launching a missile.
10 Department of Defense Inspector General, “Evaluation of Nuclear Ballistic Submarine (SSBN) Sustainment,” DODIG-2018-127, June 15, 2018, 2, https://media.defense.gov/2018/Jun/28/2001937172/-1/-1/1/DODIG-2018-127.PDF.
11 Daniel Coats, “Russia’s Intermediate-Range Nuclear Forces (INF) Treaty Violation,” November 30, 2018, https://www.dni.gov/index.php/newsroom/speeches-interviews/item/1923-director-of-national-intelligence-daniel-coats-on-russia-s-inf-treaty-violation; and Joseph Trevithick, “Russia Shows Off Parts of Its Controversial Cruise Missile System, but Not the Missile Itself,” The Drive, January 23, 2019, https://www.thedrive.com/the-war-zone/26138/russia-shows-off-parts-of-its-controversial-cruise-missile-system-but-not-the-missile-itself.
12 For a bilateral management approach, see Vaddi and Acton, “A ReSTART for U.S.-Russian Nuclear Arms Control,” 12–13.
13 Zhao, “Opportunities for Nuclear Arms Control Engagement With China.”
14 A payload of at least 5,000 kilograms was chosen as it would be very modest for any modern bomber.
15 Decker Eveleth, “Mapping the People’s Liberation Army Rocket Force,” A Boy and His Blog: OSINT and National Security (blog), March 29, 2020 (updated July 2, 2020), https://www.aboyandhis.blog/post/mapping-the-people-s-liberation-army-rocket-force; and Matt Korda, “A Rare Look Inside a Russian ICBM Base,” Federation of American Scientists, November 19, 2019, https://fas.org/blogs/security/2019/11/a-rare-look-inside-a-russian-icbm-base/.
16 John Todd, “History in the Making: Reactivating the 1st Battalion, 77th Field Artillery Regiment,” Defense Visual Information Distribution Service, October 28, 2020, https://www.dvidshub.net/news/382094/history-making-reactivating-1st-battalion-77th-field-artillery-regiment.
17 This provision is based on Protocol to New START, part 5, section IV. Mobile launchers may be absent during exercises, for example.
18 Protocol to New START, part 3.
19 Eveleth, “Mapping the People’s Liberation Army Rocket Force.”
21 The Russian proposal also included an offer not to deploy the SSC-8 in this region, though Russia continued to deny that this missile had violated the INF Treaty. Kingston Reif and Shannon Bugos, “Russia Expands Proposal for Moratorium on INF-Range Missiles,” Arms Control Today, November 2020, https://www.armscontrol.org/act/2020-11/news-briefs/russia-expands-proposal-moratorium-inf-range-missiles.
22 Frank O’Donnell and Alex Bollfrass, “The Strategic Postures of China and India: A Visual Guide,” Belfer Center for Science and International Affairs, Harvard Kennedy School, March 2020, 22–23, https://www.belfercenter.org/sites/default/files/2020-03/india-china-postures/China%20India%20Postures.pdf.
23 James M. Acton, “Is It a Nuke? Pre-Launch Warhead Ambiguity and Inadvertent Escalation,” Carnegie Endowment for International Peace, April 9, 2020, 48–49, https://carnegieendowment.org/files/Acton_NukeorNot_final.pdf.