"Though we have acheived progress, our work is not over. That is why I support the mission of the Arms Control Association. It is, quite simply, the most effective and important organization working in the field today." 

– Larry Weiler
Former U.S.-Russian arms control negotiator
August 7, 2018
Good News and Bad News on the NPT

By Michael Krepon and Samuel Black

The past is so much with us that it can sometimes obscure the present. This is certainly true with respect to the nuclear Nonproliferation Treaty (NPT), whose quinquennial review is scheduled to begin on May 3. During past reviews, the health and well-being of the NPT regime has been measured primarily by the actions of the five states recognized by the treaty as possessing nuclear weapons. Under Article 6 of the treaty, all five—China, France, Russia, the United Kingdom, and the United States—are obligated to act in good faith to facilitate the progressive reduction and eventual elimination of their nuclear stockpiles. As permanent members of the UN Security Council, the “P-5” also have special responsibilities to address proliferation dangers that, as reaffirmed by the council on September 24, 2009, constitute “a threat to international peace and security.”[1]

During the Cold War, the P-5 produced, in the aggregate, tens of thousands of nuclear weapons, tested these weapons more than 2,000 times, and threatened to use them during harrowing crises or in warfare. Most of this wretched excess was the doing of the Soviet Union and the United States.

Given this history, past review conferences understandably produced action plans that focused primarily on what the first five proliferators needed to do to strengthen the regime. Continued pressure on the P-5 is certainly warranted because they still possess more than 20,000 nuclear weapons.[2] Nevertheless, well-rehearsed speeches blaming the P-5 for the ills of the global nonproliferation regime obscure recent trends and impressive gains.

Looking back, the NPT has made extraordinary strides. The treaty entered into force on March 5, 1970, with just 43 member states. Among the original absentees were China and France. Initial safeguards at nuclear facilities were rudimentary. The original International Atomic Energy Agency (IAEA) guidance on the structure and contents of safeguards agreements contained three paragraphs on implementation, all of which focused heavily on minimizing the burden placed on safeguarded facilities. For example, this guidance stated that the IAEA should “avoid hampering” the development of nuclear expertise and international cooperation including “international exchange of nuclear material.”[3] By comparison, the IAEA’s 1997 Model Additional Protocol is far more specific, uses stronger language (“shall” instead of “should”), and authorizes far more encompassing inspections, setting a new standard for responsible nuclear stewardship.[4] Over time, the NPT’s membership has grown to almost as many states as belong to the United Nations. In 2008 the IAEA carried out 2,797 inspections, design information verifications, and complementary access visits in 163 countries, which required 14,121.5 calendar-days in the field.[5]

Despite the power and influence nuclear weapons are presumed to possess, the NPT has established nonproliferation as a global norm. States pursuing nuclear weapons programs after the NPT entered into force have been cast as outliers. The IAEA Board of Governors has found two states, Iran and North Korea, to be in noncompliance with their safeguards agreements.[6] Others, including South Korea and Syria, have conducted activities that, when exposed, have been viewed by the IAEA as matters of serious concern, prompting the agency to request clarifications or dispatch inspection teams.

Although a number of non-nuclear-weapon states have strengthened their commitment to the NPT over the past two decades, the most impressive gains during this period have been made by four of the P-5. These gains can be measured by six key indicators of nuclear weapons’ utility: actual battlefield use, threats of battlefield use, overall stockpile size, warheads deployed, nuclear weapons tests, and fissile material production for weapons. Based on these six indicators, compliance with the NPT’s core obligations by four of the P-5 has become stronger, and the value they have placed on nuclear weapons has plummeted since the Cold War ended. Despite these gains, which are described below, the outcome of the 2010 NPT Review Conference is in doubt, as is, more importantly, the health and well-being of the NPT regime. Threats to the well-being of the regime now lie increasingly in the actions of outlier states, the politicization of deliberations by the IAEA board, and the reluctance of key non-nuclear-weapon states to step up to their obligations as guardians of the treaty.

The profound nature of the positive trends described below has been underappreciated for three primary reasons. First, the residual P-5 capabilities remain large, reflecting the excessive size of their arsenals in past decades. Second, old ways of thinking linger in P-5 nuclear enclaves, where attempts to buck these trends capture headlines, even when they are ineffectual. For example, Bush administration efforts to add new warhead designs to the U.S. nuclear stockpile garnered global attention but died quickly on Capitol Hill. Third, nuclear posture reviews, treaty ratification, and entry-into-force provisions are designed or controlled by those who seek to brake, rather than accelerate, enduring trends. By rightly focusing on how much remains to be done to reduce nuclear weapons stockpiles and to end nuclear testing permanently, supporters of the NPT wrongly fail to appreciate how much the stock price of nuclear weapons has fallen for major powers over the past two decades.

Waning Utility

Consider the particulars, starting with the battlefield use of nuclear weapons. When the nuclear age began, no one was so bold or so foolish as to predict that the “winning weapon” would remain sheathed for more than six decades.[7] Others have analyzed at length the reasons for this unexpected and enormously important tradition of restraint.[8] What matters most, for the purposes of this article, is the inescapable conclusion that a weapon not used on battlefields loses its military utility. Every year that passes without the use of nuclear weapons in crisis or warfare makes it more difficult for a political leader to authorize subsequent use, while making the user more of an international pariah.

A second criterion is the political utility of nuclear weapons. Even if unused on a battlefield, nuclear weapons could still have political utility in that they could influence behavior in crises and in war. During the Cold War, the United States often directed nuclear threats against China during the Korean War and in crises over Formosa/Taiwan. The Soviet Union and the United States also ramped up their readiness to use nuclear weapons during Cold War crises to influence decision-making. U.S. maneuvers were most evident when nuclear-armed aircraft carriers and strategic bombers were forward deployed during a crisis. On some occasions, instructions to increase readiness for use of nuclear weapons were broadcast in ways that could easily be intercepted.

Now, U.S. aircraft carriers do not carry nuclear weapons on board, and Washington typically moves theater missile defense capabilities rather than B-52s into harm’s way when tensions increase in troubled regions. Nevertheless, veiled nuclear threats by the P-5 have not ceased entirely. Russian leaders have not so subtly implied that the CzechRepublic and Poland could become targets for nuclear detonations if they choose to host U.S. theater missile defense components. In addition, the United States has clarified on a few occasions, including the run-up to both wars against Saddam Hussein, that the use of weapons of mass destruction against U.S. forces, friends, and allies would be met with a devastating response. Rare, public threats by British and U.S. leaders in recent decades have been directed only at countries that are not in good standing with the IAEA and are presumed to have weapons of mass destruction in violation of UN Security Council resolutions.[9]

The pattern of nuclear-tinged threats over the past two decades has changed markedly from the Cold War. The most harrowing crises since the Cold War ended have involved outliers to the NPT. India and Pakistan have experienced four crises since 1990 in which veiled or blunt nuclear threats have been exchanged, and North Korea periodically engages in nuclear bluster when it wishes to raise temperatures on the Korean peninsula. Verbal threats in periods of heightened tension in South and East Asia have been complemented by the flight testing or movement of nuclear-capable missiles. The record is clear: states that seek political utility from nuclear weapons during periods of heightened tension now reside primarily outside the NPT.

Declining Numbers

A third measure of nuclear weapons utility is stockpile size. According to data compiled by Stan Norris and Hans Kristensen, global nuclear stockpiles reached their peak in 1986 at more than 65,000 warheads.[10] Moscow and Washington have reduced their stockpiles by approximately 50,000 warheads below their Cold War peaks.[11] The stockpiles of France and the United Kingdom have been in decline since 1981 and 1992, respectively.[12] The British and French stockpiles are now respectively 43 percent and 32 percent smaller than they were in these peak years.[13] Of the P-5 nuclear arsenals, only China’s is believed to be growing, albeit modestly.[14]

Another important barometer is the number of nuclear weapons deployed. Since the Cold War ended, Moscow and Washington have reportedly removed from operational status more than 19,000 nuclear weapons.[15] The number of nuclear weapons deployed by France has been reduced by at least 125.[16] The United Kingdom has reduced its arsenal by approximately 200 deployed nuclear weapons.[17] Of the P-5, only Beijing is increasing the number of warheads it deploys.[18]

Every test of a nuclear weapon is a declaration of utility. In 1962 alone, there were an astonishing 178 nuclear weapons tests. From 1965 to 1974, there was an average of 62 tests per year. In the following decade, the average was 54 tests per year. From 1985 through 1994, the average number of tests dropped to 18 per year. Since 1996, the P-5 have not conducted a single declaration of utility in the form of a nuclear weapons test.[19]

A final yardstick is fissile material production for weapons. Four of the P-5—again, with the exception of China—have officially declared a moratorium on fissile material production. Beijing may be refraining from new fissile material production for weapons, but it has yet to publicly confirm this.[20]

In summation, by all six key indicators, the utility of nuclear weapons has declined dramatically for four of the P-5 over the past two decades. Stockpiles and numbers of deployed nuclear weapons are declining significantly. The potential for military conflicts between major powers has diminished greatly, along with accompanying threats to use nuclear weapons. Probably none of the P-5 is now producing fissile material for weapons. None of the five has tested nuclear weapons for 15 years, an extraordinary and previously inconceivable stretch of time. Finally and most importantly, no country has resorted to battlefield use of a nuclear weapon since 1945.

Significant Gains

This track record deserves respect rather than denigration. To be sure, Cold War stockpiles and nuclear testing were excessive by any measure. Yet, the extent of reductions in Russian and U.S. stockpile sizes and nuclear force structures reflect significant shifts in perceived utility, not merely the shrinkage of wretched excess. The structure of international relations has changed since the Cold War ended. Berlin, Cuba, and Taiwan no longer evoke nuclear flashpoints. Major powers now lose standing by brandishing nuclear weapons against each other or against non-nuclear-weapon states. Major powers that seek to gain influence now do so by means of economic and energy-related indicators, not by nuclear weapons. When Moscow seeks to pressure its neighbors, it gains more leverage by threatening to cut off natural gas supplies than by threatening nuclear strikes.

The NPT is both a beneficiary and a cause of these positive trends. The extent of causality can admittedly be debated, except in one crucial respect: The treaty’s indefinite extension in 1995 was clearly linked to the subsequent completion of negotiations and entry into force of the Comprehensive Test Ban Treaty (CTBT). This pledge was partially delivered in 1996 but has yet to be honored fully. Nonetheless, the positive linkage between the NPT and nuclear testing is clear: If there were no connection between nuclear testing and the indefinite extension of the NPT, the P-5 would surely have tested old or new weapons repeatedly during the past 15 years.

Nuclear trend lines for four of the P-5 have been decidedly positive over the past four review conferences. Supporters of the NPT, who have pushed long and hard for these trend lines, can be justifiably proud of their work, even though they habitually see the glass as half-empty. Temporary reversals of these enduring trends could occur at any time, which would compound P-5 responsibility for the NPT’s uncertain state. Their most important duty as treaty guardians and as veto-wielding members of the Security Council is to deal purposefully with compliance concerns. In their dealings with Iran and North Korea, however, Beijing and Moscow sometimes have higher priorities than to uphold NPT compliance.

Russian and U.S. nuclear stockpiles remain extremely large, and although further reductions are in store, Moscow has demonstrated limited enthusiasm for widening their scope to include tactical nuclear weapons. The Bush administration did serious harm to NPT norms by championing a civil nuclear deal with India without compensatory steps to shore up the treaty. Meanwhile, Beijing still acts as a free rider to the NPT regime, rather than taking on responsibilities commensurate with its growing power. China’s intentions and force structure remain opaque, while its modernization programs are proceeding at a faster pace than those of other states with nuclear weapons. Beijing, like Washington, has still not ratified the CTBT.

Other complaints could no doubt be added to this list, but there is undeniable, impressive evidence that, on balance, the contributions of four of the P-5 to the NPT regime have grown significantly in recent decades. In contrast, outliers to the NPT are relying more heavily on nuclear weapons. Three of the four NPT outliers—India, North Korea, and Pakistan—are the only states to have tested nuclear devices since 1996 and are enlarging their fissile material and weapons stockpiles. Israel, the fourth non-NPT state, has not done nearly enough to support the NPT by means of parallel steps that reinforce the treaty’s objectives and purposes. The NPT regime’s weaknesses are growing for other reasons as well. The IAEA board has not acted consistently and coherently to uphold treaty obligations. Many states resist “second-generation” norms of responsible nuclear stewardship, particularly those associated with strengthened safeguards, materials protection, and accountancy. Key non-nuclear-weapon states that are needed to serve as guardians of the NPT have been missing in action in Vienna and New York. Egypt threatens to hold the NPT hostage to its regional interests. Above all, the nuclear program of Iran, a noncompliant party to the treaty, casts a long shadow over the NPT regime.

These positive and negative trend lines will intersect with uncertain effect at the 2010 NPT Review Conference. Four of the five major powers have contributed most of the significant gains for the NPT regime over the past two decades, but all five still have a long way to go in fulfilling their treaty obligations. Although the P-5’s trend lines are mostly positive, the same cannot be said for the IAEA, for states outside the treaty, or for those undermining it from inside.



Michael Krepon is co-founder of the HenryL.StimsonCenter and author of Better Safe than Sorry: The Ironies of Living with the Bomb (2009). Samuel Black is a research associate at the StimsonCenter.


1. UN Security Council, Resolution 1887, S/RES/1887, September 24, 2009, www.un.org/Docs/sc/unsc_resolutions09.htm.

2. This information was compiled from various sources within the Bulletin of Atomic Scientists’ Nuclear Notebook collection.

3. See International Atomic Energy Agency (IAEA), “The Structure and Content of Agreements Between the Agency and States Required in Connection With the Treaty on the Non-Proliferation of Nuclear Weapons,” INFCIRC/153 (corrected), June 1972, www.iaea.org/Publications/Documents/Infcircs/Others/infcirc153.pdf.

4. See IAEA, “Model Protocol Additional to the Agreement(s) Between State(s) and the International Atomic Energy Agency for the Application of Safeguards,” INFCIRC/540 (corrected), May 1997, http://www.iaea.org/Publications/Documents/Infcircs/1997/infcirc540c.pdf.

5. International Atomic Energy Agency, “Safeguards Statement for 2008,” n.d., http://www.iaea.org/OurWork/SV/Safeguards/es2008.pdf.

6. IAEA Board of Governors, “Report by the Director General on the Implementation of the NPT Safeguards Agreement Between the Agency and the Democratic People’s Republic of Korea,” GOV/2003/3, January 6, 2003, www.iaea.org/NewsCenter/MediaAdvisory/2003/gov2003-3.pdf; IAEA Board of Governors, “Implementation of the NPT Safeguards Agreement in the Islamic Republic of Iran,” GOV/2005/77, September 24, 2005, www.iaea.org/Publications/Documents/Board/2005/gov2005-77.pdf.

7. The phrase “winning weapon” was coined by Bernard Baruch in 1946. See Gregg Herken, The Winning Weapon: The Atomic Bomb in the Cold War 1945-1950 (New York: Vintage Books, 1981), pp. 172-173.

8. See, for example, Nina Tannenwald, The Nuclear Taboo: The United States and the Non-Use of Nuclear Weapons Since 1945 (Cambridge, MA: Cambridge University Press, 2008); T.V. Paul, The Tradition of Non-Use of Nuclear Weapons (Palo Alto, CA: Stanford University Press, 2009).

9. For evidence of noncompliance prior to 1991 threats, see UN Security Council, Resolution 707, S/RES/707, August 15, 1991, www.fas.org/news/un/iraq/sres/sres0707.htm. For evidence of noncompliance prior to 2003 threats, see UN Security Council, Resolution 1441, S/RES/1441, November 8, 2002, http://www.un.org/Docs/journal/asp/ws.asp?m=S/RES/1441(2002).

10. Natural Resources Defense Council (NRDC), “Table of Global Nuclear Stockpiles, 1945-2002,” November 25, 2002, www.nrdc.org/nuclear/nudb/datab19.asp (hereinafter NRDC table).

11. Ibid. The information in this table has been supplemented by data from the Bulletin of the Atomic Scientists’ Nuclear Notebook collection.

12. Ibid.

13. Ibid.

14. Robert S. Norris and Hans M. Kristensen, “Nuclear Notebook: Chinese Nuclear Forces, 2008,” Bulletin of the Atomic Scientists, July/August 2008, p. 42-45.

15. For current figures for Russian and U.S. deployed strategic nuclear weapons, see the Bulletin of the Atomic Scientists’ Nuclear Notebook collection. For peak figures for Soviet and U.S. deployed strategic nuclear weapons, see NRDC, “Table of U.S. Nuclear Warheads,” November 25, 2002, www.nrdc.org/nuclear/nudb/datab9.asp; Pavel Podvig, ed., Russian Strategic Nuclear Forces (Cambridge, MA: MIT Press, 2001), pp. 139, 251, and 351.

16. For French deployed warheads, we take “deployed” to mean the number available for immediate use: carried on patrol by ballistic missile submarines (SSBNs), stationed at airfields, or mounted on land-based ballistic missiles. We assume, conservatively, that only one French SSBN is on patrol at any given time. See Robert S. Norris and Hans M. Kristensen, “French Nuclear Forces, 2008,” Bulletin of the Atomic Scientists, Vol. 64, No. 4 (September/October 2008), http://thebulletin.metapress.com/content/k01h5q0wg50353k5/fulltext.pdf. At its historical peak in 1992, we again assume a lone SSBN on patrol as well as the other deployments, including of the Hades missile. Robert S. Norris and William M. Arkin, “French Nuclear Forces 1993,” Bulletin of the Atomic Scientists, Vol. 49, No. 8 (October 1993), p. 56.

17. For British deployed warheads, we take “deployed” to mean the number available for immediate use: carried on patrol by SSBNs or stationed at airfields. Current British SSBNs reportedly carry 48 weapons, and one is on patrol at any given time. See Robert S. Norris and Hans M. Kristensen, “British Nuclear Forces, 2005,” Bulletin of the Atomic Scientists, Vol. 61, No. 6 (November/December 2005), http://thebulletin.metapress.com/content/k52425n3320m8644/fulltext.pdf. At its historical peak in 1975-1982, there were 48 weapons on patrol on SSBNs and roughly an additional 200 gravity bombs at airfields. See Richard Moore, “The Real Meaning of the Words: A Pedantic Glossary of British Nuclear Weapons,” UK Nuclear History Working Paper, No. 1, March 2004, www.mcis.soton.ac.uk/Site_Files/pdf/nuclear_history/Working_Paper_No_1.pdf.

18. U.S. Department of Defense, “Military Power of the People’s Republic of China 2009,” March 25, 2009, pp. 24-25, www.defense.gov/pubs/pdfs/China_Military_Power_Report_2009.pdf.

19. For raw data, see NRDC, “Table of Known Nuclear Tests Worldwide,” November 25, 2002, www.nrdc.org/nuclear/nudb/datab15.asp.

20. Sharon Squassoni, Andrew Demkee, and Jill Marie Parillo, “Banning Fissile Material Production for Nuclear Weapons: Prospects for a Treaty (FMCT),” July 14, 2006, www.fas.org/sgp/crs/nuke/RS22474.pdf.