General Anil Chauhan's recent observation that India and Pakistan's nuclear arsenals have made it impossible for either side to achieve a decisive political outcome through war sounds at first like good news. No decisive outcome means no all-out war. But read the history of the subcontinent since 1998, and a darker interpretation emerges: the bomb has not pacified South Asia. It has merely changed the rules of the game, and the new rules permit more frequent, more persistent, low-level violence than the old ones did.

This is not a novel paradox. Cold War strategists gave it a name decades ago: the stability-instability paradox, first articulated by political scientist Glenn Snyder. The theory holds that when two nuclear-armed states face each other, the probability of full-scale war between them collapses toward zero, because both sides understand that escalation risks mutual annihilation. But that very stability at the top of the escalation ladder creates a permissive space underneath it. Because neither side dares let a skirmish spiral into nuclear war, both sides can afford to provoke, probe and bleed each other through proxies, border clashes and covert action, confident the other will not risk the ultimate response.

Washington and Moscow Wrote the First Draft

The United States and the Soviet Union never fired a shot at each other directly between 1945 and 1991. Yet those same decades produced Korea, Vietnam, Angola, Nicaragua, Afghanistan and a dozen other bloodlettings fought by proxy across the developing world. Deterrence theory, the scholarly framework built to explain why rational states refrain from attacking nuclear-armed rivals, always assumed that credible threats of catastrophic retaliation would suppress conflict. It did suppress one category of conflict: the direct, existential kind. It did nothing to suppress the other kind, the grinding, deniable, sub-threshold kind that killed millions of Koreans, Vietnamese and Afghans while Washington and Moscow's own cities stayed untouched.

South Asia is now living through its own version of this bargain, compressed into a much smaller and more combustible space.

Before and After the Bomb

India and Pakistan fought conventional wars in 1947-48, 1965 and 1971, the last of which dismembered Pakistan and created Bangladesh. These were wars with fronts, armies, surrenders and territorial consequences. Then, in May 1998, both countries tested nuclear weapons within weeks of each other, and the character of the rivalry changed almost immediately.

The very next year, in the summer of 1999, Pakistani forces infiltrated across the Line of Control into Kargil, and India fought a limited war to evict them, deliberately keeping its air force from crossing into Pakistani airspace and its ground forces from crossing the international border. Both sides, now nuclear powers, fought with one hand tied, because neither dared risk the escalation ladder running all the way up. Since then the subcontinent has seen the 2001-02 military standoff after the attack on India's Parliament, the 2008 Mumbai massacre, the 2016 Uri attack and surgical strikes, the 2019 Pulwama-Balakot exchange, and repeated artillery duels and ceasefire violations along the Line of Control. None of these has been a war in the 1965 or 1971 sense. All of them have been conflicts. The paradox predicted exactly this substitution, and the record bears it out.

There is a bitter irony here that Vajpayee himself seemed to sense even as he tried to defy it. In February 1999, months before Kargil, the Indian prime minister traveled to Lahore and told Nawaz Sharif's banquet guests that "it is unworthy of two nations the size of India and Pakistan to have wasted so much time in mutual ill will," and that "there is nothing in our bilateral relations that can ever be solved through violence." Within weeks, Pakistani soldiers and irregulars were dug into the heights above Kargil, and India was fighting its first war of the nuclear age.

Why Rational Deterrence Does Not Mean Peace

The mistake, understandable but consequential, is to treat "deterrence works" as a synonym for "peace has arrived." Deterrence theory only claims that the threat of overwhelming retaliation makes the deterred party think twice about the specific action being deterred, in this case full-scale invasion or nuclear first use. It says nothing about the hundred other forms of aggression that fall below that threshold: cross-border terrorism, artillery exchanges, cyber operations, assassination, insurgency by proxy. Pakistan's security establishment appears to have internalized this distinction with unsettling clarity since the 1990s, calculating that its nuclear umbrella shields it from Indian conventional retaliation even while it sponsors or tolerates militant groups operating in Kashmir and beyond. India, for its part, has responded not with the massed armor of 1965 but with "surgical strikes" and calibrated air raids designed explicitly to signal resolve without crossing into territory that would force Islamabad's hand toward its own arsenal.

Both governments are behaving rationally. That is precisely the problem. Rational actors operating under a nuclear ceiling do not stop competing; they migrate their competition to the floor, where the tools are terrorism, insurgency and information warfare rather than tank divisions. The 1947 Partition itself, and the communal slaughter that followed it across Punjab, offers a grim reminder that violence in this region has never needed conventional armies to reach horrifying scale. Rajmohan Gandhi's history of Punjab recounts how quickly neighborly coexistence curdled into massacre once political trust collapsed, a reminder that the sub-conventional space the paradox now channels conflict into is not some sanitized alternative to war. It is where ordinary people actually die.

Salman Taseer's assassination in Islamabad in January 2011, gunned down by his own bodyguard for defending a Christian woman accused of blasphemy, had nothing directly to do with the India-Pakistan rivalry, yet it illustrates the same underlying truth about the region: the absence of interstate war has not meant the absence of violence, ideological rigidity or state fragility. Nuclear deterrence stabilizes the border. It does not stabilize the societies on either side of it.

The Domestic Politics of Never-Quite-War

There is a further wrinkle the classic paradox did not fully anticipate: what happens when domestic political incentives reward leaders for provoking crises they know cannot escalate to full war. Every Pulwama, every Uri, every LoC skirmish becomes an opportunity for governments in both New Delhi and Islamabad to rally nationalist sentiment, precisely because both sides trust that the nuclear ceiling will hold. This is a perverse incentive structure: the safer the nuclear umbrella makes limited conflict, the more politically useful limited conflict becomes. Domestic polarization compounds the risk, much as it did in the United States during the Clinton impeachment saga of 1998-99, when bipartisan restraint gave way to a politics of maximal confrontation even over matters far less existential than war and peace. When crisis management becomes hostage to electoral cycles and nationalist theater, the paradox's assumption of purely rational, unitary state actors gets shakier with every round.

What Would Actually Change the Calculus

None of this argues for abandoning deterrence, which remains vastly preferable to the alternative of unrestrained conventional war between two states with populations exceeding 1.6 billion combined. But it does argue against complacency. The stability-instability paradox is not a stable equilibrium; it is a ratchet. Each round of sub-conventional conflict, Pulwama, Balakot, the next crisis waiting to happen, tests the escalation ceiling a little harder, relies on ever more precise signaling to avoid miscalculation, and leaves less margin for the accidents, misread intentions and bureaucratic error that historians of the Cuban Missile Crisis know can turn deterrence into disaster in a matter of hours.

General Chauhan's insight, that nuclear weapons make war unable to achieve decisive outcomes, is correct as far as it goes. But an undecidable war is not the same as no war. It is merely a war fought by other means, in smaller increments, with civilians and soldiers on the Line of Control paying the price for a stability that exists mostly at the level of grand strategy. The lesson of the Cold War, and the lesson now being relearned in South Asia, is that the bomb secures the peace of chancelleries far more reliably than it secures the peace of villages.