We at the Stimson Center are celebrating our twenty-fifth anniversary this year. A quarter-century ago, the Cold War was receding at a rapid pace and the Soviet Union was in its last stages of decomposition. Washington and Moscow were on course to reduce their nuclear arsenals by previously unthinkable percentages. It was, in other words, a perfect time to start a think tank. Co-founder Barry Blechman and I were steeped in the practices of strategic arms control. What would we – and the Stimson Center – do now?
After huddling with funders, two new programming initiatives took shape. Barry would convene wise veterans of the Cold War to revive the notion of seeking the complete elimination of nuclear weapons; I would carry the “toolbox” of confidence-building and nuclear risk-reduction measures that helped keep the Cold War from becoming hot to troubled regions. My game plan was to offer countries wishing to avoid dangerous nuclear competitions a menu of choices that could be suitably adapted to fit regional circumstances.
Initially, Stimson convened workshops on CBMs in Argentina, Brazil, Chile, Egypt, India and Pakistan. The need for Stimson programming to promote nuclear-related CBMs in the Southern Cone melted away with the advent of civilian governments. And back in the early 1990s, Stimson was stretched too thin to sustain CBM programming in the Middle East.
In contrast, it was easy to establish comparative programming advantage in South Asia. Very few foreign NGOs were active in the region, and none were involved in programming that addressed the dangers inherent in covert Pakistani and Indian nuclear programs. My early field trips were spent listening, learning, and rebutting arguments that CBMs weren’t needed. Back then, the counter-arguments were that these measures were a Western imposition and that India and Pakistan were too sensible to engage in a nuclear arms competition. The Line of Control dividing Kashmir was a long way away from the Fulda Gap.
Over time, positions softened. One reason was that Stimson hosted over 70 visiting fellows from India and Pakistan — professors, researchers, journalists, policy entrepreneurs at NGO start-ups, and military officers – to delve into the theory and practice of CBMs. A second reason was that Stimson’s analytical products turned out to be prescient. All of the nuclear choices facing India and Pakistan were previously conceived in the West, and some constructs, such as the stability/instability paradox (i.e., possessing nuclear weapon capabilities could actually embolden risk-taking behavior below the nuclear threshold) proved applicable to South Asia.
New Delhi was risk-averse in a pre- and post-nuclear environment, but Rawalpindi wasn’t. With each crisis on the subcontinent, the need for CBMs and nuclear risk-reduction measures became more apparent. Stimson developed a new product line – case studies of crisis dynamics and crisis management. These case studies have been widely read, and have helped the George W. Bush and Barack Obama administrations develop and update diplomatic playbooks. The current focus of our in-house research and commissioned essays is deterrence stability and escalation control on the subcontinent.
Every one of the measures that the governments of India and Pakistan have agreed to – e.g., the use of hotlines between leaders and senior military officers, prior notifications of certain military exercises and flights, as well as notifications of nuclear accidents – were midwifed in Stimson Center workshops, Track II meetings, and publications. We have helped identify and promote many other measures with obvious utility — an Incidents at Sea agreement, the verifiable withdrawal of troops from the Siachen Glacier, a cruise missile flight test pre-notification agreement, and many others — but finalizing these accords has not been a priority Indian and Pakistani leaders.
Nuclear dangers have grown on the subcontinent. The nuclear arms competition continues to pick up speed while Pakistan’s governance and internal cohesion continue to decline. The advent of a muscular government in India could impart a new impetus to improved relations, but it’s unclear at present whether this will be a priority and, if so, whether the usual blocking moves in Pakistan will again result in stasis – or worse.
I hear a sense of worrisome complacency from colleagues in India who confidently predict that the Modi government will strike back after another spectacular act of terrorism that is traced back to Pakistan. Don’t worry, they say, because responsible parties will prevent unwanted escalation. With Rawalpindi’s track record, there will be a presumption of guilt – at least by association – in the event of another crisis-triggering event, and Washington is in no position to offer advice against the targeted killing of extremist groups.
The Obama administration, preoccupied with Iran, Syria, Iraq, Israeli-Palestinian relations, and Ukraine, is very poorly organized to engage in crisis management in South Asia. The State Department’s regional bureau is bifurcated with an Assistant Secretary focused on India and a Special Representative (in the process of retiring) who does Pakistan. Likewise, in the Department of Defense, PACOM handles India while CENTCOM does Pakistan. These structural oddities can be overcome by strong interlocutors, but regional expertise on South Asia at the top rungs is typically thin, Deputy Secretary of State Bill Burns is retiring, and Washington doesn’t have an Ambassador to India. The administration is just finding its footing with the new Indian government, and relations with Pakistan, while improved, are not yet on solid ground.
Whoever gets tapped to be the principal US crisis manager won’t be able to rely on the gambit used to defuse previous crises. Promises extracted from Pakistani leaders to clamp down on those behind the 2001 attack on the Indian Parliament building, and again in 2008 after the carnage in Mumbai, were hollow. They worked to defuse crises because Indian leaders were reluctant to take military action. The next time around, cost/benefit calculations may be different.