全球智库信息集成服务系统

Since the size of the U.S. nuclear arsenal peaked in the 1980s, American presidents—Democrat and Republican alike—have limited the development of new nuclear weapons. Whether the restriction was written into law, was included in U.S. nuclear policy, or was the result of specific decisions not to pursue new procurement projects, the United States has not built a new nuclear warhead since the late 1980s. This policy decision has generated significant cost savings, restrained strategic competition, and helped to support other stabilizing policies.

With Republicans now in control of Congress and the White House, this policy is at risk. As Russia and China expand their territorial claims as well as their own nuclear arsenals, a growing chorus of U.S. politicians and strategists argue that it is not sufficient to simply replace nuclear systems as they wear out. Instead, they insist that the United States must procure new systems with qualitatively new capabilities. In some cases, appeals for new nuclear weapons are motivated by a sophisticated but mistaken argument about their necessity for deterring potential adversaries from employing nuclear weapons in limited conflicts. Other advocates endorse these programs as a way of winning future arms races or achieving supremacy over other nuclear powers.

Although it has not been specific about its plans, the Trump administration has promised to “greatly strengthen and expand its nuclear capability.”¹ This seems to conflict with the assessment of Gen. John E. Hyten, the commander of U.S. Strategic Command, who has argued that “we don’t need more nuclear weapons, we just need to modernize.”² Are the two statements consistent? What exactly is nuclear modernization? Where should the United States draw the line as it embarks on a program to replace nearly every bomb, missile, submarine, and warhead in its arsenal?

Constructive debate over these plans is often obstructed by imprecise vernacular. The term modernization is variously used to refer to existing programs that refurbish current weapons systems, existing programs that update current systems with improved versions, and to proposals that would create qualitatively new capabilities. Responsible modernization that refurbishes and replaces existing systems with improved variants is necessary in order for such systems to continue to carry out their missions safely and reliably. However, establishing programs that would enable the United States to hold targets at risk in qualitatively new ways would be destabilizing, unnecessary, and irresponsible. Specifically, developing new nuclear capabilities would likely increase global nuclear competition, accelerating a new arms race; create uncertainty for existing modernization programs in the Pentagon budget and also at the national laboratories that maintain the nuclear stockpile; increase the likelihood that new countries could seek to acquire nuclear weapons; and do little to improve the ability of the U.S. armed forces to deter and defend against aggression around the world. As the new administration begins its Nuclear Posture Review, the decades-old bipartisan prohibition against the development of new nuclear capabilities is more important than ever.³

This analysis is informed by a tabletop exercise that was carried out at the Center for American Progress in the autumn of 2016. In it, a bipartisan group of former officials and experts in nuclear weapons policy—including both proponents and opponents of new nuclear options—investigated the role of U.S. nuclear forces in the defense of NATO’s Baltic members. Although the participants were not asked to endorse the findings of this report, their deliberations are instructive in evaluating the case for new nuclear weapons.⁴

In the 1980s, the radical expansion of technological options for delivery vehicles and warhead designs led to concern that fielding these capabilities would be destabilizing to the U.S.-Soviet nuclear balance. Many feared that exuberance in nuclear modernization and the quest for a technological edge were creating an unrestrained arms race that would inevitably lead to these systems being used.⁵ A major international advocacy movement urged the superpowers to freeze the production and testing of nuclear arms.⁶

This proposal was enormously popular: Most U.S. polls pegged public support for a nuclear freeze between 70 percent and 82 percent, and successful votes in state, city, and town governments made it “the largest referendum on a single issue in the nation’s history.”⁷ In 1982, the U.S. House of Representatives passed a resolution in favor of a nuclear freeze by a vote of 273 to 125, but the Senate version failed to win approval.⁸ Although the White House objected vehemently to the proposal, the movement may have helped convert President Ronald Reagan to the cause of arms control, a shift that occurred around the same time.⁹

Arms control initiatives placed limits on certain types of nuclear systems—particularly intermediate-range weapons—but for the most part, the 1980s saw dramatic innovation in each leg of what is popularly known as the nuclear triad: land-based intercontinental ballistic missiles, or ICBMs; strategic bombers; and submarine-launched ballistic missiles, or SLBMs.¹⁰ The Pentagon fielded a new generation of nuclear-capable delivery vehicles, including the MX ICBM, the B-1 and B-2 bombers, the Trident SLBM, and other systems.

At the same time, the United States began development programs on a new generation of strategic and nonstrategic munitions. The 1980s saw the development of a new low-yield artillery shell, the W82; a warhead for naval aviation to be used for land attack as well as anti-submarine missions, the B90; and two warheads for surface-to-ground missiles, the W89 and the W91.¹¹ While the MX ICBM and both bombers were deployed in the course of the 1980s—albeit in lower numbers than had been requested—the warheads were not. Prototype rounds of each were produced, but none entered full production before the end of the Cold War, and President George H.W. Bush cancelled these programs as part of the Presidential Nuclear Initiatives, or PNI, of 1991–1992.¹² The most recent new warhead is the W88, which first entered the stockpile for use on the Trident II D5 SLBM in 1988.¹³ The most recent warhead model is the B61-11—a variable-yield gravity bomb that adds ground-penetration capabilities to a physics package developed for the earlier B61-7 variant—which was first produced in 1997.¹⁴

In the fall of 1993, Congress attached the Spratt-Furse Amendment to the National Defense Authorization Act for fiscal year 1994, which prohibited “research and development which could lead to the production by the United States of a new low-yield nuclear weapon,” meaning warheads of fewer than 5 kilotons.¹⁵ The Spratt-Furse Amendment did not prohibit all nuclear weapons research and development, but it did inhibit work on the cutting edge of nuclear design and on the systems that critics considered most destabilizing. Coming less than a year after the first President Bush signed a unilateral moratorium on nuclear testing, the amendment also helped to support the case for the Comprehensive Nuclear Test Ban Treaty, or CTBT, which prohibited those systems that critics worried would be most difficult to detect if they were tested in violation of the treaty. For these reasons, the amendment can safely be said to have established a presumption against new nuclear weapons in the United States.

New nuclear weapons proposals in the George W. Bush administration

The election of President George W. Bush posed the first challenge to the presumption against new nuclear weapons.¹⁶ The new Bush administration’s 2001 Nuclear Posture Review, or NPR, took a significant step away from White House policies under President Bill Clinton and the previous President Bush.¹⁷ Citing an increased difficulty of tailoring deterrence to dissimilar adversaries, the NPR called for the capacity to “modify, upgrade, or replace portions of the extant nuclear force or develop concepts for follow-on nuclear weapons systems better suited to the nation’s needs.”¹⁸ Specifically, the George W. Bush administration identified a requirement for earth-penetrating precision nuclear munitions to defeat hard and deeply buried targets, or HDBT, which it would call the Robust Nuclear Earth Penetrator, or RNEP.

The stated rationale was that a low-yield warhead could burn out underground facilities that housed biological or chemical weapons while limiting collateral damage to the surrounding area.¹⁹ In an early attempt to study these requirements, the Bush administration’s fiscal year 2004 request asked Congress to repeal the Spratt-Furse Amendment, to provide small quantities of funding for a feasibility and cost study of the RNEP, and to shorten the time it would take the labs to prepare and conduct a nuclear test—the “rest readiness posture”—from 24–26 months to 18 months.²⁰ After a contentious fight, Congress rescinded the Spratt-Furse Amendment’s limit on research of new nuclear warheads but preserved the prohibition on development.²¹

Although the George W. Bush administration insisted that it was only studying the RNEP, it requested that the prohibition on development be lifted for fiscal year 2005 and requested a large sum for the system—$484.7 million over five years.²² Instead, Congress declined to appropriate funding for fiscal year 2005 or in any year thereafter and the National Nuclear Security Administration, or NNSA, soon abandoned its RNEP teams.²³

Both the NPR and the legislative proposals were subject to considerable criticism. Many observers expressed concern that the document would lower the threshold for nuclear use, making the weapons “more useable.”²⁴ Several experts drew parallels to debates two decades earlier in which European allies and arms control groups successfully prevented deployment of a neutron bomb, a specialized warhead that could kill invading Soviet troops in armored formations without destroying European cities.²⁵ Furthermore, many argued that development of new warheads would raise proliferation pressures around the world, especially if certifying the RNEP meant returning to nuclear testing in violation of the CTBT, which the Clinton administration had signed but the Senate had declined to ratify.²⁶

At the same time, opponents argued that a low-yield earth penetrator would be ineffective against a wide range of HDBT, including those that were too deep, too hard, located in populated areas, or about which intelligence agencies had imperfect information.²⁷ In the case of the RNEP, the country, on balance, found that the cost of holding at risk this specific class of facilities was not worth it. However, the broader debate was a strong demonstration that development of new nuclear weapons, especially those that produced new capabilities to make them more usable in limited circumstances, required special justification and heightened scrutiny. Developing new warheads that were intended not for deterrence of a nuclear attack but to destroy targets on the battlefield was a bright line that many correctly refused to cross.

However, the issue of new warhead development was not closed. In fiscal year 2005, the year that RNEP funding ceased, Congress appropriated funding to develop a Reliable Replacement Warhead, or RRW. Reflecting concerns that the Stockpile Stewardship Program, or SSP, could not guarantee the surety of the stockpile in perpetuity, NNSA and the armed services proposed a program to construct a new warhead optimized to simplify maintenance and provide confidence in the stockpile without the need for testing.²⁸ NNSA got as far as selecting a design for the warhead from an internal competition.²⁹ Expert studies commissioned to evaluate the concept tended to find that it was technically practicable but carried its own risks.³⁰ However, the project never found sound political footing. Many saw the debate as an extension of the one over the RNEP and applied many of the same arguments: The House Appropriations Committee, then under Democratic control, said “there exists no convincing rationale for maintaining the large number of existing Cold War nuclear weapons, much less producing additional warheads,” while The New York Times editorial board called it a “public-relations disaster in the making overseas” and expressed concern over the cost.³¹

In the end, continued accumulation of experience with the SSP convinced most experts that the labs could sustain the stockpile in perpetuity through surveillance, simulation, and periodic life-extension programs, eliminating the need for RRWs.³² The ability of RRWs to increase confidence in the stockpile did not, in the end, outweigh the diplomatic and fiscal costs of breaking the moratorium on new warhead development. Congressional funding for the program varied between fiscal year 2005 and fiscal year 2008, after which no further funding was appropriated and the RRW program was closed.³³

In short, resistance to the development of new nuclear warheads proved remarkably resilient during President George W. Bush’s first term. Although the 108th Congress included only 49 Democratic Senators and 207 Democratic Representatives, they succeeded in attracting enough Republican support to defeat the RNEP and RRW proposals.³⁴ Neither the problem of HDBTs nor concerns about stockpile surety were sufficient to overcome entrenched resistance. The limited benefits of the new warheads did not justify the fiscal, diplomatic, and stability costs.

Importantly, accumulation of experience has put to rest concerns about the Stockpile Stewardship Program, or SSP, which monitors the stockpile and is required by Congress to certify annually that the arsenal remains safe and effective.³⁵ Directors of the national laboratories routinely report that science-based SSP activities have yielded a greater knowledge of the physics of nuclear explosives than ever before and that the program has proven effective at detecting and correcting faults with the systems.

After President Barack Obama took office, his administration moved to make the moratorium on nuclear warhead development an affirmative policy, stating in its Nuclear Posture Review:

The United States will not develop new nuclear warheads. Life Extension Programs (LEPs) will use only nuclear components based on previously tested designs, and will not support new military missions or provide for new military capabilities.³⁶

In taking this step, the new Obama administration effectively truncated debate in the United States about RRWs and the RNEP, thus allowing the NPR to cover more important questions of arms control and strategic stability. The policy was sometimes explained in public as establishing three limits on new warhead development, or three “no’s”: no new nuclear warheads, no new military missions, and no new military capabilities for existing weapons.

The no-new-warheads policy was an integral piece of a broader approach to U.S. nuclear weapons policy and helped to support and enable other elements of U.S. nuclear policy. Specifically, it helped to limit international concerns among both allies and potential adversaries over U.S. nuclear modernization by announcing that upcoming programs would only replace current capabilities—not expand them. Potential adversaries faced less pressure to modernize their arsenals and were deprived of a public rationalization for the systems they did develop. Allies concerned with the pace and scope of nuclear modernization were reassured that the United States was attempting to avoid a new arms race. In Congress and at the labs, the policy helped to avoid major fights over the cost and structure of the modernization plans before they began. Although the no-new-capabilities policy did not extinguish the concerns of any of these audiences, it did help to enable the armed services to begin research and development work on the administration’s program of record.

Foregoing the development of new capabilities also supported efforts by the Obama administration to define deterrence requirements and set force levels. The general objective was to develop new operational plans that would permit “significantly lower nuclear force levels . . . with reduced reliance on nuclear weapons.”³⁷ To facilitate this, the U.S. Department of Defense was directed to “conduct deliberate planning for non-nuclear strike options.”³⁸ The decision not to seek new missions for nuclear weapons was a natural corollary. At the same time, increasing requests to Congress for nuclear enterprise funding supported SSP and assuaged concerns that the stockpile could deteriorate. In effect, the Obama administration was able to meet the targeting demands for new warheads with adjustments to military plans and the stockpile sustainment concerns by investing directly in these programs.

Interpreting capabilities in the Obama years

Since the 1960s, the term “nuclear modernization” has harbored a deep ambiguity in that it is used to refer both to programs that are needed in order to sustain the existing capabilities of the nuclear arsenal as well as to those that provide new capabilities. This ambiguity can sometimes be used to conceal the extremity of certain positions for nuclear weapons or can inadvertently obfuscate moderate arguments.

In a recent example of the latter, a letter signed by several retired four-star U.S. Air Force and Navy officers who had commanded strategic forces that called for “modernization” of the nuclear arsenal was published in The Wall Street Journal under the headline “The U.S. Nuclear Triad Needs an Upgrade.”³⁹ In fact, the letter did not recommend qualitatively new capabilities or new missions for the arsenal, but it was difficult to tell as much due to an imprecise vernacular. In another example, Linton Brooks, former U.S. ambassador and former administrator of the NNSA, argued for continuing “the policy of not developing new nuclear weapons with new military capabilities,” but suggested “interpret[ing] this policy in a way that permits sensible modifications to current weapons during the life extension process (for example, by fielding some primary-only ballistic missile warheads).”⁴⁰ It is not clear whether this proposal would run afoul of the policy on new capabilities that was set by the Obama administration, which underscores the need to define these terms precisely.

In general, the Obama administration understood nuclear modernization as a means of sustaining the existing capabilities of the nuclear arsenal by directly replacing aging systems with new versions that are capable of conducting the same missions. President Obama explained his modernization policy as representing a balance between “making sure that the triad and our systems work properly” and guarding against “new and more deadly and more effective systems that end up leading to a whole new escalation of the arms race.”⁴¹

In fact, the Obama administration’s modernization plan did require that the new generation of delivery vehicles meet higher performance specifications with respect to survivability, precision, and surety. The Air Force has declared that the B-21 Raider bomber will gain new stealth capabilities that enable it to “penetrate the increasingly dense anti-access/area denial environments developing around the world.”⁴² Similarly, the new Columbia-class ballistic missile submarine, or SSBN, will be equipped with a quieter electric drive and a variety of other improvements to enhance its survivability.⁴³ Each of these innovations—as well as similar enhancements to the Long-Range Standoff Weapon, or LRSO, (an air-launched cruise missile) and the Ground-Based Strategic Deterrent, or GBSD, (the program to replace the Minuteman III ICBM)—certainly improve the effectiveness of these nuclear delivery vehicles by increasing their ability to accurately deliver warheads to their targets.

Although Obama administration officials sometimes gave the impression in public remarks that nuclear modernization would not increase the capability of the nuclear triad, this was not the actual policy. The NPR only prohibited new capabilities in the warheads themselves, which could be life-extended but not modified in a way that would improve their effectiveness or that provided for new military missions.

Matters become even more complicated when one considers a series of upgrades to existing weapons systems authorized by the Obama administration. For example, the B61 life-extension program, which is sometimes explained as a program to extend the life of the warhead, is more properly understood as an effort to consolidate four of the five existing variants of the B61 gravity bomb into a single variant, the B61-12.⁴⁴ The NNSA hoped that the program would allow for significant reductions in the number of B61s retained in the stockpile, as well as the retirement of the B83 gravity bomb.

Although this program did not modify the explosive package of the B61 gravity bomb, it did provide for the development of a new guided tail kit assembly that will improve the free fall accuracy of the weapon: According to the Government Accountability Office, “The guided capability will enable the weapon to meet military requirements with a lower nuclear yield.”⁴⁵ Hans Kristensen and other civil society critics have argued that the B61-12 represents “an upgrade that will also increase military capabilities to hold targets at risk with less collateral damage.”⁴⁶ As the first guided nuclear gravity bomb, the lower yield may also reduce collateral damage estimates and may make it more usable in certain contingencies.⁴⁷ Moreover, if the weapon is deployed to Europe as part of the NATO extended deterrent as planned, it would be able to hold targets at risk that were previously beyond the capability of the Mod-3 and Mod-4 nonstrategic B61 variants deployed there.⁴⁸

Although the life-extension program does not modify the explosive package of the B61, it clearly provides the weapon with a new ability to hold targets at risk and offers new options to planners tasked with nuclear targeting and escalation control. In this instance, the Obama administration interpreted its policy against new warhead capabilities to pertain only to the explosive effects, not to upgrades to the guidance system.

Modifications to the fuzing and precision of the Trident II D5 SLBM have also dramatically increased the lethality of U.S. strategic forces. For example, in an article in Bulletin of the Atomic Scientists, Hans Kristensen, Matthew McKinzine, and Theodore Postol report that “deployment of the new MC4700 arming, fuzing, and firing system on the W76-1/Mk4A [SLBM warhead] significantly increases the number of hard target kill-capable warheads on US ballistic missile submarines.”⁴⁹ The new fuze allows the W76 to detonate at a variable height within a lethal column of space over a target, where previous iterations of the warhead may have fallen short or long of the zone necessary to destroy a hard target. Kristensen, McKinzine, and Postol argue that this improves the U.S. capability to execute a disarming first strike of an adversary and increases the risk that “Russian nuclear forces will be used in response to early warning of an attack—even when an attack has not occurred.”⁵⁰

The cumulative effect of these upgrades is that President Obama left his success