U.S. INTELLIGENCE AND INDIA'S NUCLEAR TESTS: LESSONS LEARNED
CRS Report for Congress
Received through the CRS Web
U.S. Intelligence and India’s Nuclear Tests:
Richard A. Best, Jr.
Specialist in National Defense
Foreign Affairs and National Defense Division
The U.S. Intelligence Community did not have advance knowledge that India
intended to conduct nuclear tests beginning on May 11, 1998. Although intelligence
agencies cannot have foreknowledge of every significant development in world affairs,
many observers (and senior intelligence officials) believe that, in view of the election of
an Indian government committed to “inducting” nuclear weapons, much greater attention
should have been given to indications of impending nuclear tests. A commission headed
by retired Admiral David Jeremiah reviewed the Intelligence Community’s performance
and recommended a number of measures to improve future performance. Some of the
most important of these were addressed in the FY1997 Intelligence Authorization Act
(P.L. 104-293) but implementing the legislation has been delayed as filling key positions
has been delayed. A persisting problem is the tendency of analysts to discount seemingly
irrational initiatives by other countries.
The series of nuclear tests conducted by India beginning on May 11, 1998, was not
predicted by the U.S. Intelligence Community. Although foreknowledge would not
necessarily have enabled U.S. policymakers to head off the explosions, the absence of
information has raised fundamental questions about the nation’s intelligence effort, which
will cost some $26.7 billion in FY 1998. Defenders of the Intelligence Community point
out that, even though many in the public have come to expect constant coverage of all
regions of the world, resources are finite, information of adversary value is routinely
hidden, other concerns are pressing, and South Asian developments do not directly
threaten U.S. security. Most observers—including Members of Congress and senior
officials in intelligence agencies—agree, nonetheless, that the failure to have provided
warning of the Indian tests before they occurred was a major lapse with significant
implications for predicting nuclear developments in other countries and that actions must
be taken to improve intelligence performance.
Congressional Research Service ˜ The Library of Congress
Indian Nuclear Policy1
India, in a previous surprise to the U.S. Intelligence Community, exploded its first
nuclear device in 1974. From that time on, proliferation experts realized that India has had
the capability to explode nuclear weapons at any time. In late 1995, U.S. intelligence
agencies detected preparations for a nuclear test at Pokharan, a site some 350 miles
southwest of Delhi, leading to diplomatic representations that many observers credit with
persuading the Indian Government to call off the explosions. The protest also may have
provided the Indians insight into the capabilities of U.S. intelligence collection efforts.
Parliamentary elections, conducted in February-March 1998, resulted in a minority
coalition government headed by Prime Minister Atal Behari Vajpayee of the Hindu
Nationalist Bharatiya Janata Party (BJP) whose election manifesto had included a
commitment to “exercise the option to induct nuclear weapons.” It should be noted that
all major Indian parties are committed to maintaining a nuclear option and that Vajpayee,
who once served as foreign minister, is considered a moderate among BJP leaders.
The BJP government that took office in late March was soon made aware of
international concerns about the possibility of nuclear testing and proliferation. The
American U.N. Ambassador Bill Richardson in an April 14th meeting in New Delhi and
National Security Advisor Samuel Berger in a subsequent Washington meeting with the
Indian Foreign Minister underscored U.S. opposition to nuclear testing and proliferation.
In response, Indian leaders indicated that no precipitous actions were likely to be taken and
that a full-scale review of the country’s national security policies was underway. The
Indians also made great efforts to avoid test preparations that would be readily detectable
by overhead satellites. The planning for tests was apparently kept within a small circle of2
senior officials and ministers.
Although the Intelligence Community monitored Indian developments and scrutinized
activities at Pokharan on a regular, but not constant, basis, the official confirmation of
Indian tests at Pokharan on May 11 caught U.S. intelligence agencies by surprise.
Director of Central Intelligence (DCI) George Tenet stated:
While the Intelligence Community has for years closely followed the Indian
nuclear program, there is no getting around the fact that we did not predict
these particular Indian nuclear tests. We did not get it right. Period. We have
a professional responsibility to stand up, acknowledge that, and learn from it.3
1 For detailed background on India’s nuclear tests, see Richard P. Cronin, Barbara Leitch LePoer,
Jonathan Medalia, and Dianne Rennack, India-Pakistan Nuclear Tests and U.S. Response, CRS
Report 98-570F, June 18, 1998; for background on the Indian elections, see Barbara Leitch
LePoer, India’s 1998 Parliamentary Election Results, CRS Report 98-324F, April 2, 1998.
2 See James Risen and Tim Weiner, “U.S. May Have Helped India Hide Its Nuclear Activity,” New
York Times, May 25, 1998. P. A3.
3 Press Statement by the Director of Central Intelligence George J. Tenet on the Release of the
Jeremiah Report, June 2, 1998, reproduced at http://www.odci.gov.
Intelligence analysts as well as policy-level officials throughout the government shared the
blame for failing to anticipate the tests. Phyllis Oakley, the Assistant Secretary of State
for Intelligence and Research, was quoted as telling a reporter after a Senate hearing,4
“Look, we were wrong. We were all wrong.” Members of Congress were harshly critical
of the lack of warning; Senator Shelby, Chairman of the Senate Intelligence Committee,
termed it “a colossal failure.”5 It was small comfort that the Intelligence Community was
able to point to the likelihood of the Pakistani tests that did occur in late May.
The Inevitability of Intelligence Failures
Scholars who have studied intelligence history recognize that surprises can never be
eliminated. The modern intelligence community,, including the Central Intelligence
Agency (CIA), arose in large measure out of the failure to anticipate the Japanese attack
on Pearl Harbor in December 1941. There have been nonetheless a number of significant
events since the end of World War II for which there was little advance knowledge, e.g.,
the North Korean attack on South Korea in 1950, the introduction of Soviet nuclear
missiles into Cuba in 1962, and the Iraqi invasion of Kuwait in 1990. The quantity of
information available has varied in each case, but, like the attack on Pearl Harbor, these
intelligence “failures” resulted in part from analysts’ inability to realize that hostile
countries pursue policies that appear illogical and counterproductive. With a mindset
disposed to believe that a potential action is irrational, even almost unthinkable, an analyst
may ignore indicators that, in retrospect, make the action appear inevitable.
Sherman Kent, who was for many years in charge of the preparation of National
Intelligence Estimates, wrote in the aftermath of his failure to detect the installation of
Soviet missiles in Cuba in 1962: “It is a melancholy fact of life that . . . man will often6
blind himself to truth by going for the comforting hypothesis, by eschewing the painful.”
In the end, surprises may not achieve their ultimate aims; there is usually no escape from
the hard realities of capabilities and resources—the attack on Pearl Harbor, after all, was
a colossal strategic error that ultimately led to the total defeat of Imperial Japan.
Nevertheless, the Intelligence Community was created to prevent future Pearl Harbors and
that is still its main mission.
Analysts studying India, which has never been at the center of U.S. security concerns,
may also have to overcome some unconscious biases. India, as the “world’s largest
democracy” and a non-participant in most of the Cold War, garnered a certain amount of
uncritical sympathy from Western academicians who have trained intelligence analysts.
Some observers suspect among intelligence officers an institutional bias in favor of India’s
4 Quoted in Risen and Weiner, “U.S. May Have Helped Indian Hide Its Nuclear Activity.” The lack
of foresight was not limited to government officials; for example, writing after the 1998 elections,
one respected scholar claimed that “the BJP must act cautiously if it is to both preserve its coalition
and deliver on its promise. It is unlikely to antagonize Pakistan by curtailing Kashmir’s autonomy
or the United States by going openly nuclear.” Marshall M. Bouton, “India’s Problem is Not
Politics,” Foreign Affairs, May-June 1998, p. 85.
5 Quoted by R. Jeffrey Smith, “CIA Missed Signs of India’s Tests, U.S. Officials Say,”
Washington Post, May 13, 1998, p. A1.
6 Quoted in Donald P. Steury, ed., Sherman Kent and the Board of National Estimates
(Washington: Center for the Study of Intelligence, Central Intelligence Agency, 1994), p. 178.
policies that derives from the work of the CIA’s predecessor organization, the Office of
Strategic Services, many of whose officers gave tacit and overt support to the Indian
independence movement during World War II.
Balanced against such difficulties, however, are the facts that nuclear proliferation
is one of the highest concerns of the U.S. Government, that the basic parameters of the
Indian nuclear program were well known,7 and that an Indian government had just come
into office with no hesitancy about “inducting” nuclear weapons into its strategic arsenal.
It is clear—admittedly with 20/20 hindsight—that more collection efforts should have
been directed to the nuclear test site and that analysts (and policymakers) should have had
a better sense of the dynamics of Indian politics and Delhi’s international ambitions.
The Jeremiah Report
In the wake of widespread concern about the Indian tests, DCI Tenet appointed a
committee headed by retired Vice Admiral David E. Jeremiah, who had previously served
as vice chairman of the Joint Chiefs of Staff, to assess the Intelligence Community’s
performance and report back within 10 days. Although the Jeremiah Report has not been8
made public, its major conclusions were discussed in a June 2, 1998, press conference.
It made recommendations in four principal areas:
•analytical practices•manpower and training
To ameliorate the “mindset problem,” Jeremiah recommended that the Intelligence
Community use outside regional experts in a more systematic fashion to ensure that
government analysts are kept abreast of the latest work by academic and other
nongovernmental specialists. It was further suggested that experts in the process of
analysis be brought in to encourage consideration of contrarian hypotheses. The report
also recommended the establishment of mechanisms to guarantee stronger integration of
analysis from different agencies and disciplines.
In regard to collection, Jeremiah acknowledged an imbalance between the
Intelligence Community’s ability to collect imagery, and the ability to read and analyze it.
“In everyday language, that means there is an awful lot of stuff on the cutting room floor
at the end of the day that we have not seen.” Observers consider this imbalance to be a
fundamental problem of Intelligence Community management that has yet to be effectively
addressed. It is perversely easier to justify expensive collection programs that produce
vast quantities of data than more analysts to review and report the information obtained.
In addition, it is difficult to have an adequate number of collectors and analysts
skilled in the foreign languages that become important at specific times. Technical
specialists with advanced academic training related to sophisticated collection and
7 See Jonathan Medalia, Indian and Pakistani Nuclear Tests? Potential Test Ban Risks and
Technical Benefits, CRS Report 96-631F, July 17, 1996.
8 The comments by Admiral Jeremiah quoted herein are taken from the transcript of the Jeremiah
News Conference of June 2, 1998, reproduced at http://www.odci.gov.
communications systems may be able to command salaries that cannot be matched by
government agencies. Observers have suggested greater use of outside experts and even
reserve military personnel.
Jeremiah also addressed the need to coordinate collection systems more effectively.
Imagery, signals intelligence, human intelligence (including diplomatic reporting), and
other disciplines along with open sources (press, television, etc.) yield the evidence on
which analysts base their work. Separate U.S. agencies manage different types of
collection systems, and it is administratively challenging to ensure that information on
important topics is collected by the most appropriate collection means.
The Jeremiah Committee review concluded that the organization of the Intelligence
Community needs to be examined thoroughly to assess the process by which
responsibilities are assigned and resources allocated, and whether resources are adequate.
DCI Tenet indicated his approval of the review, stating that “I accept all of Admiral
Jeremiah’s recommendations. I am making it my highest priority to implement them as
quickly as possible.”9
Coordinating Intelligence Efforts
Many of the conclusions of the Jeremiah Committee are consistent with
recommendations made in 1996 by two wide-ranging reviews, viz. the House Intelligence
Committee’s Staff Study, IC21: Intelligence Community in the 21st Century and the
report of the Commission on Roles and Missions of the Intelligence Community, usually
known as the Aspin-Brown Commission (after its two chairmen, the late Les Aspin and
Harold Brown, both former Secretaries of Defense). The fundamental conclusion of these
reviews was that there should be better coordination by the DCI of the various agencies
that comprise the Intelligence Community, but without full-scale unification.
In practice, it has proven very difficult for DCIs to exercise coordinative
responsibilities. In part, this reflects the diverse priorities of individual DCIs, some of
whom devoted considerable attention to covert operations or to participating in the policy
deliberations of the National Security Council. It is, however, also reflective of an absence
of Community-wide mechanisms based on statutes that give the DCI the ability to
mandate, effectively pressure, or at least strongly encourage, cooperation among the
different agencies. For many years, the only Community-wide officials were the DCI and
the Deputy DCI both of whom allocated most of their time to the CIA itself.
The fundamental reality of the Intelligence Community has been the fact that the DCI
does not have “line authority” over agencies other than the CIA; in particular, he does not
have operational control of the large Defense Department intelligence agencies such as the
Defense Intelligence Agency, the National Reconnaissance Office, the National Imagery
and Mapping Agency, and the National Security Agency. Amendments over the years to
the National Security Act of 1947 gave the DCI some authority to coordinate the work
of all agencies and to establish standards and priorities. In reality, however, budgets of
DOD agencies are controlled by the Secretary of Defense and their operations are heavily
influenced by the pressing needs of military commanders. A number of legislative
9 Tenet, Press Statement, June 2, 1998.
initiatives have been introduced to give DCIs the authority to execute budgets and
exercise line authority, but, according to many observers, Congress has chosen not to do
so as a result of the close operational nexus between DOD agencies and the military
services. Some Members of Congress, however, continue to believe that the DCI should
have authority to execute the intelligence budget even for agencies in the Defense
Enacted in the aftermath of the IC21 Study and the Aspin-Brown Report, Title VIII
of the Intelligence Authorization Act for FY1997 (P.L. 104-293) included provisions
designed to strengthen the authority of the DCI.10 The DCI was given greater authority
to develop annual budgets for intelligence and intelligence-related activities; he was also
given statutory authority to approve collection requirements and priorities, and to resolve
conflicts in collection priorities. Additional positions requiring Senate confirmation were
created: a deputy DCI for Community Management, and three assistant DCIs—for
collection, for analysis and production, and for administration.
The changes envisioned by the FY1997 Act were not implemented quickly. In late
1996, DCI John Deutch resigned; subsequently, there was a protracted confirmation
process that resulted in George Tenet not being sworn in until July 1997. In addition, the
Administration (and some Members of Congress) strongly opposed the requirement that
the three new assistant DCIs be confirmed by the Senate. As a result of this disagreement,
the first Deputy DCI for Community Management—Joan Dempsey, a career intelligence
professional—was not confirmed until May 1998 and assistant DCIs have only recently
been appointed (it being agreed that they will serve for a year without Senate
Many observers believe that sufficient statutory authority now exists to enable the
DCI to create a more “corporate” and better coordinated Intelligence Community, even
if it is not a consolidated single entity under his direct control. According to this view, the
problems reflected in the failure to anticipate the Indian nuclear tests give DCI Tenet the
opportunity to align collection and analytical resources more effectively and to
demonstrate whether or not additional statutory authority might be required.
The failure to predict the Indian nuclear tests of May 1998 provides an opportunity
to review the priorities and capabilities of the Intelligence Community and to implement
legislation signed in October 1996 that gives the DCI more specific authorities and
additional staff to manage the national intelligence effort. Out of this review and the
implementation process may emerge a consensus for further legislative initiatives.
A more intractable problem are the mental biases of analysts who often downplay
unpredictable and (in their view) irrational and counterproductive undertakings by foreign
countries. Efforts to introduce outside expertise into the intelligence process and to
encourage contrarian analyzes may help, but the challenge to overcome the constraints
of analysts’ own mindsets or the consensus of an agency will likely remain.
10 See Richard A. Best, Jr., Intelligence Reorganization in the 104th Congress: Prospects for a
More Corporate Community, CRS Report 96-681F, September 13, 1996.