Last week, militants killed 26 civilians, mostly tourists, in Kashmir, the disputed region between India and Pakistan.
The killing of 26 civilians would be bad news anywhere in the world, but in Kashmir, it could be far worse. Tensions between the two nations have always been high, ever since they were split up in the partitioning of British India in 1947, but in recent years they have increased still further.
Kashmir is often at the centre of those tensions. Although formally part of India, it has been granted special, largely autonomous status since partition, but in 2019 India effectively ended that, leading Pakistan to downgrade diplomatic relations and suspend trade.
Last week’s attacks have seemingly brought the two countries to the brink of conflict: India’s government accused Pakistan of supporting the attack, while Pakistan said it expected a retaliatory assault by Indian forces.
A war could have profound consequences. India is increasingly a manufacturing powerhouse; notably, it is the world’s largest supplier of generic pharmaceutical drugs. India also sits upstream of its neighbor in the huge Indus river system, which Pakistan relies on for its agriculture and energy generation: after the attacks, New Delhi suspended the 65-year-old Indus Water Treaty pledging not to limit water flows, raising fears that it could use water as a weapon. And, of course, both countries have nuclear weapons: an escalating conflict between the two could lead to history’s first nuclear exchange.
All that makes it vital to understand the likely outcomes of the attack. The Swift Centre, a not-for-profit which helps organisations improve their ability to predict and respond to the future, commissioned its global network of professional, award-winning forecasters to make rapid predictions of those outcomes. Specifically, the organisation’s forecasters looked at the chance of conflict itself; and then, if that conflict does happen, how likely a major escalation would be, the likely impacts on pharmaceutical exports, the Indus Water Treaty, and the risks of nuclear war.
Q1: Will the national military forces, militia, and/or law enforcement personnel of India and Pakistan engage in a lethal confrontation resulting in at least 100 uniformed casualties before May 15?
Average forecast: 8.5%, range 2% to 28%
The forecasters were first asked how likely they thought a significant conventional conflict was: specifically, a conflict causing at least 100 uniformed casualties between the two countries in the next two weeks.
They thought it was relatively unlikely, although not out of the question, and some thought the risk was significant.
Usually, forecasters start by trying to assess the base rate of such events: how often they happen in ordinary times, an initial position which can then be updated with new information.
One of them noted that while tensions are high and occasional border disputes happen, large armed conflicts between official state forces are rare: “the 1999 Kargil Conflict was the last time that fatalities reached three digits. Before that, it was in 1990.”
There have been smaller skirmishes since then, but deaths have never gone beyond double figures. As a starting point, then, some forecasters thought that the base rate should be low: In the 35 years since 1990 there have been only two conflicts that would have counted, giving an annual base rate of less than 6%. Since the forecast only looks ahead two weeks, the base rate should be correspondingly lower.
Others thought that since some cross-border skirmishing has already begun, a better way to find a base rate would be looking at how many such skirmishes turn into larger conflicts. Since the last official war between the two countries in 1971, one forecaster counted six smaller conflicts, of which, they argued, only one – 1999 – unambiguously reached the threshold for today’s forecast, giving a per-incident base rate of about 15%.
But India has unequivocally said it would respond, and Pakistan has begun moving troops into Kashmir. “Clearly this is the most significant crisis facing the two countries since 2019,” one said. Another argued that “India-Pakistan conflicts have generally been nothingburgers over the last few decades” but there are reasons to think that this time is different: levels of nationalism are extremely high in both countries, and animosity has been growing; Pakistan is in the grip of a long-running economic and social crisis, and war with India could boost its regime’s faltering legitimacy; and Israel’s attack on Hamas has provided precedent for major cross-border attacks in response to terrorism incidents. Also, the US, usually a broker between the two, appears not to be engaging.
On the other side of the ledger, two weeks is a very short time, and “nuclear deterrence, economic fragility, and third-party diplomatic pressure” all act as brakes on escalation, a forecaster argued.
Q2. Will the total combined number of fatalities among the national military forces, militia, and/or law enforcement personnel of India and Pakistan exceed 1,000 before June 30, 2025?
If there are 100 or more uniformed casualties by May 15: 19% (range 3.6% to 35%)
If there are fewer than 100 uniformed casualties by May 15: 2.3% (range 0.5% to 8.5%)
The forecasters were also asked about the chance of a larger war breaking out this summer. Specifically they were asked the probability of it if the smaller 100-casualty conflict happened, and if it didn’t.
In short, they thought that if there weren’t 100 casualties by May 15, it was very unlikely (2.3% chance). Some went as high as 8.5%, but on the whole they thought that it is “really unlikely that there will be nothing for 15 days then something this huge.”
But if there are 100 or more casualties, then the probability of a serious war starts to get very worrying. One of the more optimistic forecasters said that even if the number is reached, “the goal of both sides will be to tamp down hostilities before they get out of control,” but others were far less sanguine: 100 casualties would imply “fierce and fraught” fighting, and emotions would be running high in both countries, increasing the chance of escalation. “If there are 100 or more fatalities in the next two weeks, then I see a one-in-three probability of reaching four digits by June 30.”
Q3. Will the Indus Waters Treaty (IWT) between India and Pakistan, which governs the distribution of water from the Indus River basin, be fully reinstated by December 31st, 2025?
If there are fewer than 100 uniformed casualties by May 15: 67% (range 40% to 98%)
If there are 100 or more uniformed casualties by May 15: 22% (range 2.3% to 66%)
The IWT is a point of significant tension between India and Pakistan, as noted above. India’s cabinet announced on April 23, after the Kashmir attacks, that the treaty was “held in abeyance with immediate effect, until Pakistan credibly and irrevocably abjures its support for cross-border terrorism.”
It’s the first time the treaty has ever been suspended since its 1960 introduction, including during the 1965 and 1971 Indo-Pakistan wars and many smaller conflicts.
The forecasters therefore felt that assuming the skirmishing dies down before 100 deaths are reached, it is likely to be reinstated. However, if there is a significant conflict – as in the 100 deaths scenario – “it will be politically more difficult for both sides to reinstate the treaty before the end of the year,” one forecaster argued.
One forecaster thought that India might see the conflict as a pretext to end a treaty which it no longer wants to abide by, “to claim more of the water for consumption and power generation.” A larger-scale attack makes that pretext stronger.
On the other hand, there does not seem to be provision in the treaty for either side to unilaterally exit, making the pretext weaker, and China – which controls the headwaters of the Indus and thus has leverage over India – would likely put pressure on New Delhi.
Some forecasters noted that the outcome is more likely to be symbolic than practical, since India probably does not have the capability to significantly alter the course of the Indus, but one argued that recent infrastructure improvements “provide the capacity to store the water on a massive scale, which allows India to more meaningfully leverage water as a weapon of war in the longer term.”
Q4. What will be the total value, in USD, of pharmaceutical exports from India during the calendar year 2025?
If there are 100 or more uniformed casualties by May 15: $29 billion (80% confidence interval: $22 billion to $34 billion)
If there are fewer than 100 uniformed casualties by May 15: $31 billion (80% confidence interval: $24 billion to $35 billion)
India’s generic pharmaceuticals industry accounts for around one-25th of its total exports, and the sector has been growing at spectacular rates for the last several years. If a war halted or significantly reduced them, it would hurt both India’s economy and several countries’ medical supply stocks.
The forecasters felt that even in the event of a conflict causing 100 or more casualties, the impact on Indian pharma was likely to be low. “Only a major, massively destructive full-blown war will significantly affect Indian pharmas and their exports,” one forecaster.
Partly that is because Pakistan is only a small export market, while the US accounts for around $7 billion of the roughly $30 billion exported a year, and Washington would be unlikely to reduce those purchases as a result of a war.
One forecaster noted that two major Indian ports are in Gujarat, a state on the border with Pakistan, which might be a problem for exports should a war break out. But they also said that the bulk of pharmaceutical exports go through the port in Mumbai, further south and likely outside any naval conflict zones. They also said that the war would be unlikely to involve naval battles: “Air strikes and special forces/ground forces have been the preferred responses in previous conflicts.”
Several forecasters thought that a much more imminent threat to the Indian pharma sector was US tariffs.
Q5. Will either India or Pakistan use a nuclear weapon, including tactical or battlefield nuclear weapons, in the context of a military conflict between India and Pakistan?
If there are 100 or more uniformed casualties by May 15: 3.3% (range: 0% to 5%)
If there are fewer than 100 uniformed casualties by May 15: 1.6% (range: 0% to 1.4%)
Surely the question that weighs most heavily on people’s minds. Reassuringly, the forecasters felt that a nuclear war was unlikely even in the event of a significant conflict, although a one-in-30 chance is still worryingly high. “I just don’t think India and Pakistan want the current conflict to escalate,” one said, “and I can’t imagine either side wanting or planning to use a nuclear weapon first.” So unless the two sides misread each other’s intentions, or there’s some sort of accident – not impossible in a tense and volatile environment with two countries with poorly defined chains of command – then nuclear war is unlikely.
“The forecasters felt that a nuclear war was unlikely even in the event of a significant conflict.”
One raised the possibility of a “mad major” scenario, in which a local commander has launch control over tactical nuclear weapons and uses one to slow an enemy advance, but “while there is uncertainty over Pakistan’s safety mechanisms on nuclear weapons,” it does seem that the warheads are kept separate from launch vehicles, making such an outcome less likely.
Another forecaster argued that the “mad major” might not be necessary: Pakistani military doctrine explicitly envisages using tactical nuclear weapons if they are losing a conventional war. And its threshold of when to use nuclear weapons is “unclear and relatively uncommunicated”, meaning India’s efforts to fight a limited war against it could accidentally trigger it.
All that said, the forecasters all felt that in either scenario, nuclear escalation was an unlikely outcome. The two sides have had nuclear weapons for decades now, and even previous major conflicts, like 1999, did not lead to nuclear exchange. “The Kargil War has shown that it is possible to prevent nuclear escalation even at higher casualty levels in conventional conflict,” one said, although “a massive conventional escalation threatening the territorial status quo in Kashmir could indeed trigger a Pakistani nuclear reaction, even if only tactical.”
