Chinese intrusions in Eastern Ladakh earlier this year have provoked a review of India’s intelligence apparatus. After all, how could the PLA mobilise and deploy a significant number of troops from its interior for an ingress across a heavily militarised border without a timely Indian counter-response? All signs point to an intelligence failure.
Understandably, a number of analysts have reiterated the need for reforming India’s intelligence apparatus that can plug loopholes and avoid a similar ill-fate in the future. But, all these policy suggestions — while valuable in some sense — rest on an erroneous assumption: Intelligence can be made full-proof if certain operational and organisational infirmities are overcome. I argue, it cannot.
M K Narayanan, former National Security Adviser, argues that it is “axiomatic that leaders make better decisions when they have better information”. The Chinese made no effort to conceal PLA troop mobilisations in Pangong Tso and Hotsprings-Gogra and so there was no question of them having evaded India’s high-quality imagery intelligence (IMINT) and signals intelligence (SIGINT) capabilities. He concludes, therefore, that the intelligence failure occurred at the “interpretation” or “analysis” stage. Indian intelligence analysts failed to “decipher China’s intentions in time”. On this front, he criticises the decision of the National Security Council Secretariat (NSCS) to dismantle the Joint Intelligence Committee (JIC), and laments the lack of China experts in India’s premier foreign intelligence agency, R&AW.
Richard K Betts in his oft-cited paper from 1978 agrees with Narayanan’s argument that intelligence failures are seldom the result of a shortcomings in tasking or data collection. Rather, they are mainly produced due to mistaken analysis on behalf of intelligence analysts or political elites. However, how will more “expert” information or a JIC resolve this issue?
Firstly, more often than not, the problem is too much information rather than too little. While regional expertise would improve the quality of intelligence processing, it is not a panacea for good political judgement. In fact, China experts are still struggling to explain Beijing’s intentions. For example, it was hardly predictable that China, in the midst of facing global censure for mishandling the outbreak of a pandemic, a slowing domestic economy, and after two (reasonably successful) informal summits with their Indian counterparts in Wuhan and Chennai, would risk escalation on the border and yield India’s goodwill for a few inches of un-inhabited land in Ladakh. If Beijing planned to use strategic surprise to catch New Delhi off-guard when the latter would least expect an ingress, more China expertise would not have proven remedial. Historically, as Taylor Fravel has argued, China has been more “compromising” in moments of domestic duress and international pressure. The wolf-warriors are an unpredictable and erratic turn in Chinese foreign policy whose proximate causes are still unclear.
Secondly, as Betts has argued, “altering the analytic system” (such as setting up bodies for intelligence processing such as the JIC) does not guarantee better judgement. In response to intelligence failures, the reflexive impulse is often to push “recommendations for reorganization and changes in operating norms” (see Kargil Review Committee Report). As Sumit Ganguly and Frank O’Donnell claim, increasing accountability and parliamentary oversight could have potentially provided “meaningful external scrutiny” of Indian intelligence findings and prevented Chinese aggression. These measures can indeed be useful in enabling faster intelligence sharing and processing, both forward and backward in the chain of command. But, they are only marginally effective since we do not have a full-proof framework for “accurate” intelligence processing.
Betts says that there is one primary reason behind this. As was shown above, intelligence findings can contradict strategic estimates or assumptions leading to their dismissal. This is why “wishful thinking, cavalier disregard of professional analysts, and, above all, the premises and preconceptions of policy makers” are most often at fault for intelligence failures. This is not because analysts or leaders are lazy or incompetent, rather it is because they are cognitive misers and often subjected to the pressures of time-sensitive decision-making. For example, when interpreting an excess of mutually-contradictory data inputs, analysts’ risk assessments can either oversimplify reality or be too vague to be actionable. In response to ambiguous information, leaders are in turn likely to rely on intuition, beliefs, images etc. since there is “usually some evidence to support any prediction”. Regardless of how many layers of scrutiny are institutionalised and how accountable and robust the intelligence apparatus is, such limits of human cognition and the complexities of the intelligence process cannot be eliminated. Even if leaders became self-aware of their prejudices and biases they would struggle to consistently internalise this.
In light of these circumstances, it is tempting to rely on “worst-case assumptions”. It could be argued that if there is any cause for apprehension, analysts should simply assume the worst and proceed with extreme caution. However, this too is unsustainable. Firstly, constantly preparing and mobilising for worst-case outcomes would be prohibitively expensive for a recession-hit Indian state. Secondly, with an increase in false alarms (since at least some of the worst-case intelligence will be dubious) there is a risk of routinisation eroding sensitivity to actual crisis situations (cry-wolf syndrome). Thirdly, precautionary mobilisation can lead to counter-mobilisations and a security dilemma spiral. For example, in response to doubtful reports that Pakistan was planning an offensive in Jammu and Kashmir to isolate it from the rest of India, Jawaharlal Nehru decided to mobilise Indian troops along the Punjab and Kashmir borders in 1951. This lead to counter-mobilisation by Pakistan. At the time, the CIA reported that “almost 90 per cent of India’s and 70 per cent of Pakistan’s ground forces were deployed against each other”. This crisis only passed with the assassination of Pakistan’s Prime Minister, Liaquat Ali Khan, in October 1951. Worst-case thinking could cause avoidable escalation, more frequent crisis situations, and even accidental war.
Assessments of India’s intelligence apparatus, therefore, require reviewing how many external threats were foiled in comparison to instances of intelligence failure. Instead of ritualistically blaming its intelligence agencies after every let-down, Indian leaders need to aspire towards a suitable ratio of success. Since only cases of failure become public knowledge, such calculations are a difficult task for non-governmental scholars or commentators. This does not mean external recommendations for improving the analytic system should be ignored. I only suggest — in light of what I have discussed here — that there should be greater recognition, both amongst policymakers and domestic audiences, of the implausibility of “accurate” intelligence processing.
Does this mean India should, as Betts concluded in his article, “live with fatalism”? Not entirely. Firstly, in both, the Kargil crisis and the current stand-off on the Sino-Indian border, India’s adversaries purposefully chose moments when expectations of aggression would be minimal. While India’s forces cannot constantly remain in a heightened state of military preparedness, a base level of vigilance in low-threat situations moving forward is advisable. Learning from each intelligence failure is essential. Secondly, for Indian strategists, the limits of intelligence should mean a greater focus on deterrence strategy. As I have argued elsewhere, if India cannot rely on its intelligence to counter Chinese moves on the border in every situation (necessary for deterrence by denial), it must have effective back-up options for deterrence by punishment (such as a larger and better-equipped counter-offensive Mountain Strike Force).