Organizational Cultures and the Management of Nuclear Technology

Political and Military Sociology

Russell Kirk, Karthika Sasikumar, Russell Kirk

• Publisher: Routledge
• Year: 2017
• ISBN:
  9781351501156
• Edition: 1
• Topic: Geschichte
• Subtopic: Militär- & Seefahrtsgeschichte

Nuclear technology has been an organizing premise of the international system since 1945. Eight countries have officially acknowledged the possession of nuclear weapons. Many countries have harnessed the atom for electricity generation and other civilian uses. Roughly 440 commercial nuclear reactors operate in thirty countries providing 14 percent of the world's electricity. Volatile oil prices and concerns about climate change have led newly emerging economies in Asia to express keen interest in using nuclear energy to meet growing energy demands. Since the basic technological apparatus for both civilian and military nuclear programs is the same, there are concerns about the potential spread of dual-use technology.The future stability of the international order depends on the responsible management of their nuclear assets by nuclear powers. The relationship between civilian authorities and the military takes on special significance in states with nuclear weapons or near-weapon capability. The constitutional balance of powers, the delegation of authority during wartime and peace, influences from public opinion and bureaucratic structures on the formulation of doctrine, crisis management, and communications with the international media and the general public are influenced by civil-military relations and organizational culture.This volume will be of broad interest to scholars of civil-military relations, political science, and political sociology.

Threats, Institutions, and Nuclear Learning

Behind India’s Veil of Nuclear Ambiguity, 1989–1998

Kampani Gaurav

Cornell University

Political and Military Sociology: An Annual Review, 2011, Vol. 39: 43–59.

This article makes three interrelated arguments. First, nuclear proliferation is not a binary phenomenon, but a process. Second, the proliferation literature’s focus on industrial hardware overlooks the “software” dimension of nuclear weapon acquisition and development—that is, the state’s institutional, organizational, and ideational capacities. Third, understanding how domestic institutions work and how they shape organizational learning is central to grasping the linkages that tie nuclear hardware to software. To illustrate these arguments, this article examines India’s nuclear trajectory from the early 1980s to the present day. It draws on Ernst Haas’ concept of “epistemic communities” and a wider literature to show that Indian institutions were conducive to adaptation but generally hostile to organizational learning.

Most studies of nuclear proliferation tend to treat the phenomenon as a binary phenomenon: that is, a state is, or is not, a nuclear weapon state (NWS) (Singh and Way 2004; Jo and Gartzke 2007; Fuhrmann 2009; Kroenig 2010; Bleek 2010). Framing the question so starkly, however, underplays the more complex reality that proliferation is actually a vast process, which involves the production of fissile material, the development of a nuclear device, the weaponized version(s) of the device itself, the integration of weapons with delivery systems, and the embedding of weapons into organizational and doctrinal webs that enable states to leverage them (Office of Technology Assessment 1993:3–5, 160–61; Carter et al. 1987:1–13). Most studies also singularly focus on the hardware aspects of nuclear proliferation: the fissile material, the weapons, and delivery systems. They generally ignore the software dimension of the proliferation challenge, which includes the organizational and ideational means that states institute to leverage such weapons in crises or wars.

Data logic drives the framing of the proliferation problem as a binary as well as the focus on hardware. Both are easy to grasp. They are also relatively easy to classify, code, and measure. Further, hard material capabilities or tangibles relate to threats more easily. By contrast the soft ideational structures, the intangibles, are more difficult to identify, classify, and grasp. They therefore often elude our attention. However, it would be wrong to ignore them. In fact, hardware has no agency of its own. Agency depends on the successful marriage between hardware and software. States need institutional coordination, organizational planning, doctrinal development, and training of personnel to use military power as an instrument.

When compared to conventional weapons, the significance of such software intangibles assumes even greater significance for nuclear weapons. This is because nuclear weapons constitute revolutionary instruments of power (Jervis 1989). They are revolutionary because the disproportionality of force that inheres in them is so great that it upends the normal logic of using military means to achieve political ends. With nuclear weapons, war in the Clausewitzian sense no longer remains the continuation of policy by other means. Thus, states developing nuclear weapons must also make a parallel revolutionary leap in imagination to leverage them. They must unlearn and then relearn the relationship between military means and political ends. Absent this learning, nuclear hardware remains a relatively mute instrument of military power.

In the current debate a question that political scientists and proliferation specialists often ask is: “when does a state become a nuclear weapon state?” Historically, in the pre-Nuclear Nonproliferation Treaty (NPT) era, a nuclear test marked off a nuclear from a non-nuclear weapon state. Subsequent to the NPT and the institution of a sanctions regime that punished nuclear tests, the criteria gradually shifted to more silent and subtle signals of nuclear proliferation, such as the accumulation of “significant quantities of fissile material” (Hymans 2010:162–79). Similarly, questions concerning measurement of “nuclear latency”—that is, the time that it would take a state to develop nuclear weapons-related infrastructure from scratch or the time that it would take a state to reprogram its civilian nuclear infrastructure toward military ends—have either used proxies for the nuclear fuel cycle or the fuel cycle itself (Sagan 2010:80–101). The essence of proliferation is thus associated with hardware acquisition. The software dimension of proliferation has been given short shrift.

This article makes three interrelated arguments. First, nuclear proliferation is not a binary; it is a process. Second, the proliferation literature’s focus on industrial hardware overlooks the “software” dimension of the nuclear weapon acquisition and development process—that is, the state’s institutional, organizational, and ideational capacities. Third, and finally, domestic institutions critically shape the software dimension of a state’s nuclear equation. They govern the process of organizational learning. Understanding the process of how institutions work and how they shape organizational learning is central to grasping the linkages that tie nuclear hardware to software. Absent successful institutional and organizational learning, the nuclear agency of proliferating states in the international system is likely to be limited.

The article proceeds in two parts to illustrate these arguments. In the first, it examines India’s nuclear trajectory from the late 1980s, when weaponization began, until 1998 when India conducted nuclear tests and formally claimed NWS status. The period from 1989 to 1998 was India’s decade of nuclear “opacity.” The consensus among scholars at the time and today remains that India during this period was a de facto NWS—that it possessed a basement arsenal of disassembled weapons (Perkovich 1993; Hagerty 1995). Through the use of extensive interviews with some of the leading participants in India’s weaponization program in the 1990s, this paper presents new historical evidence to show that such claims are grossly overstated. In the second part, the article makes a distinction between nuclear adaptation and nuclear learning. The term adaptation here means policy changes that are piecemeal responses to environmental pressures without analytic oversight to tie means to ends. Learning, in contrast, refers to fundamental changes in the state’s method of conducting business—changes that instrumentally tie policy means to ends (Haas 1990:23–28). Using the Indian case study, it draws on Ernst Haas’ concept of “epistemic communities” (1990:40–46) and a wider literature on decision-making structures and processes to show that the Indian state’s domestic institutions in this period were conducive to adaptation and at the same time generally hostile to organizational learning. The net effect of this, on both the hardware and software counts, was that the Indian state’s capacity for instrumental nuclear agency during the decade of the 1990s was severely limited.

Unpacking a Nuclear Weapon State: India’s Technical and Institution Disarray during the Decade of “Opacity” (1989–1999)

India rejected the NPT in 1969 and signaled its strategic autonomy by testing a fission device in 1974. However, it did not follow up the 1974 test with a wider nuclear test program. Nor did governments in New Delhi build weapons on the basis of the 1974 test design. However, the Indian weapons program was revived in 1980 following Pakistan’s attempts to acquire a nuclear capability. India had planned nuclear tests to validate advanced nuclear designs in 1982–1983, but the tests were canceled under pressure from the United States (Perkovich 1999:226–28, 242–44). India finally embarked on a weaponization program in 1989 once it became clear that Pakistan had acquired nuclear devices (Chengappa 2000:332–33). However, India’s nuclear status remained murky until May 1998 when the government conducted two sets of nuclear tests.

During the late 1980s, proliferation scholars and analysts used hardware templates—both nuclear testing and the accumulation of significant quantities of fissile material—as measures of India’s nuclear latency. India’s 1974 nuclear test, its large and unsafeguarded nuclear estate, rumors of planned nuclear tests, and veiled boasts by Indian prime ministers of the country’s nuclear capabilities left few doubts that New Delhi had the technical means to make a relatively quick transition to the status of an NWS. Observers also interpreted India’s acquisition of nuclear-capable combat aircraft and the inception of a ballistic missile program in the early 1980s as evidence that it would have little trouble developing an operational nuclear capability (Perkovich 1999, 226–403). By the late 1980s and early 1990s, scholars considered India to be a de facto NWS (Perkovich 1993).

The term typically used to categorize India is “opacity.” The definition and grammar of opacity were developed by Avner Cohen and Benjamin Frankel to distinguish first-tier (NPT-5) from the second-tier (emerging) nuclear weapon powers (Cohen and Frankel 1991). The key difference between the first and the second-tier nuclear powers, they argued, was that of visibility. Emerging nuclear weapon powers deferred to the NPT’s norm of nonproliferation and pursued their private security goals of weapons acquisition invisibly, in hived-off secret enclaves within the bowels of the state. Using Israel’s example, Cohen and Frankel proposed opacity as an ideal Weberian construct from which new proliferators could show some deviations. But the model overall had general applicability across the system. In this model, states did not conduct nuclear tests, denied possessing nuclear weapons, did not issue direct nuclear threats, did not deploy weapons with the military, did not promulgate nuclear doctrines, and insulated the nuclear enclaves within the state apparatus (Cohen and Frankel 1991, 16–23). But although opacity’s defining feature was invisibility, its premise was the existence of nuclear capability and some institutional capacity to wield it.

During the decade of the 1980s, proliferation analysts and scholars assumed correctly that India had the technical infrastructure to develop a nuclear arsenal rapidly. What escaped their notice, however, was the Indian state’s chronic institutional fragmentation and the absence of any institutional and organizational coordination mechanism within it that could quickly transform latent capabilities into usable capacities. What they also underplayed were the asymmetric power institutions that allowed the prime minister’s office to tower over and impose its will, not just on the military, but also often on the nuclear scientific institutions in determining the program’s timing and direction. The state’s prevailing institutions therefore created conditions in which prime ministers’ indecision imposed organizational paralysis. These features of the institutional make-up of the Indian state created conditions conducive to adapting but not learning.

This became manifest in several ways. Successive prime ministers approved research on nuclear weapons in the 1980s without consulting the armed services (Chengappa 2000:246–61). The Air Force imported combat aircraft in this period without seriously thinking through nuclear mission requirements (Chengappa 2000:305, 327, 357–58). Similarly, the ballistic missile program launched in 1983 did not seriously incorporate operability benchmarks. Until 1989, aside from the prime minister and occasionally the cabinet secretary, there was no individual inside government to knock heads and bring the agencies into a serious coordination effort (Deshmukh 2006). It is unclear if the principals—in this case, successive prime ministers—even understood the time lags between a research and development program, prototype development and the finished weapon product. Prime Minister Rajiv Gandhi, for example, boasted in 1985: “[I]f India decided to become a nuclear power, it would take a few weeks or a few months” (Spector 1985:95). Yet from the time the Indian government launched a crash weaponization effort (in the spring of 1989) until 1994–1995, when aircraft modified for nuclear operations came online (Chengappa 2000:382–84), the process took six years to accomplish.

The state’s institutional fragmentation produced extraordinary negative downstream consequences in the 1990s. Nearly a year after the launch of the weaponization program, Arun Singh, a former defense minister who briefly participated in nuclear planning in 1990, disclosed: “I thought it was crazy that BARC¹ didn’t know where DRDO² stood or vice versa. Nothing had been worked out as to who was to control the weapons and under what circumstances and time frame we would strike back” (Chengappa 2000:356). Ideally, prime ministers should have overseen the project and resolved personality and institutional clashes. However, spasmodic political attention marred the project. The prime ministers of the 1990s placed the program on autopilot and delegated responsibility to senior bureaucrats (Chengappa 2000:332–33). The weapon that emerged from this process therefore met with minimal end-user requirements. It turned out to be too large for the designated Mirage aircraft (Chengappa 2000:327). Other interfacing challenges concerned the Mirage’s center of gravity, electronic and electrical systems, sighting system, electromagnetic interference, electronic jamming system, and secure communications. All these took the greater part of the decade to resolve.³

The Air Force for its part resented the scientists’ “hijacking” of the weaponization program. They were unhappy with not being raised to the status of full partner.⁴ Denied sensitive information, the service was partially in the dark about the trials and methods the scientists were using to certify sensitive subsystems on both the weapons and the aircraft. Some of these—notably the bomb’s delayed detonation fuse and accurate height burst fuse—were critical to mission safety.⁵ One senior defense official with critical knowledge of the efforts expressed concern about the “hand-hammered quality” of the modifications on the Mirage.⁶ So persistent were some of the safety and reliability concerns in the 1990s that senior Air Force officials demurred from calculating the statistical odds of a successful nuclear mission.⁷ It would surprise many to learn that India had only a single nuclear-capable aircraft and test pilot during the period 1989–1994 (Chengappa 2000:382–84). New Delhi’s so-called basement nuclear arsenal was, in the words of a senior defense official, “a paper tiger.”⁸

The belief in the 1990s that India was a de facto NWS was premised on its possession of a small arsenal of nuclear bombs whose fissile cores and non-nuclear trigger assemblies were disassembled and dispersed across several secret locales in the country. However, the belief that Delhi could also deliver them successfully was an assumption. Rarely asked was the question whether a state possessing weapons also had the institutional and organizational capacity to wield them. These capacities refer to the civil–military chain of command, standard operating procedures, and practice drills and ground rehearsals to coordinate action across the various agencies tasked with responding to a nuclear emergency. They also refer to operational planning and use doctrine in the military’s tactical meaning of the term. External assumptions notwithstanding, new evidence suggests that successive Indian governments in the 1990s did not think through the institutional and organizational logic of their nuclear arsenal.

For example, there was no institutionalization of the information regarding the program in government. A small network consisting of about a half-dozen scientists and bureaucrats controlled the information flow. Outside the group, information was shared partially with prime ministers, sometimes with their principal secretaries or the cabinet secretary, and, after 1996, with the Air Force Chief of Staff on an informal basis (Perkovich 1999:377). The method ...