- Risk of use of chemical or biological weapons in the next 12 months: 12%
- Likelihood of ceasefire by end of year: 21%
- Risk of event involving 25 fatalities inside NATO member territory in the next 18 months: 8%
The most likely outcome is that the war does not escalate either with the use of chemical and biological weapons, or the deaths of significant numbers of people on NATO soil through Russian action. That’s mainly because the risks to all sides of widening the conflict would be significant and the benefits nebulous for all parties. That said, the Russian leadership has shown itself willing to make high-risk decisions in the past, not least launching the invasion in the first place.
On the other side, an imminent ceasefire is also unlikely, because public opinion in both Russia and Ukraine would be against it, and both countries have the manpower and matériel to continue fighting, Russia through its own industrial base, still producing though affected by sanctions, and Ukraine with Western support.
The Russian invasion of Ukraine has been going on for more than four months. After a rapid opening phase, in which most observers thought Russia’s forces would steamroll the much weaker Ukrainian defence in a few weeks, the invaders were pushed back. Now the situation has become a slog with slow Russian gains at enormous military and civilian costs marked by attacks and counter-attacks in the south-east of Ukraine.
We wanted to look at the likely future of the war. Will it escalate – with the use of chemical or biological weapons, or expanding into attacks beyond the border of Ukraine? Or might the two sides reach a ceasefire?
15 forecasters with exceptional track records contributed answers to these three questions:
- Before the 1st of August 2023, will chemical or biological weapons be used in the Russia-Ukraine war?
- Before the 1st January 2023, will there be a full-scale ceasefire declared between Russia and Ukraine?
- Will events involving Russian security forces result in 25 or more fatalities on a NATO member state's territory before the end of 2023?
We’ll look at them in turn.
Before the 1st of August 2023, will chemical or biological weapons be used in the Russia-Ukraine war?
Median forecast: 12%
The use of proscribed weapons of mass destruction – by either side, although for obvious reasons most people would think it more likely from Russian rather than Ukrainian forces – would mark a significant escalation in the conflict and a further breach with international laws and norms.
All forecasters thought this outcome unlikely, although there was a reasonable spread – one put it at a 1% chance, while another went as high as 23%. It’s also worth being clear that 12% is not a trivial chance: if the use of chemical or biological weapons would have really bad outcomes, such as bringing NATO forces into the war (or simply killing large numbers of people with the weapons themselves), then that level of risk would be worth taking very seriously. That said, there are also chemical weapons and chemical weapons: poison gas dropped on a city is very different from phosphorus weapons used against combatants.
One forecaster, who put the outcome at 14% likely, reasoned that Russia clearly has no moral problem with chemical weapons – the country has “not been shy about using chemical weapons for targeted assassinations”, such as in Salisbury – but they don’t seem to have used them in battlefield situations, even the siege of Azov plant in Mariupol, “which would have been an ideal tactical condition to use them in order to kill or force out deeply entrenched opposition forces." “I don't think anyone wants to really escalate in this direction,” the forecaster wrote, “but I can see some borderline cases of phosphorus use.”
Another points out that there are issues with resolving the question. There are rumours that Russia already has used chemical weapons in Mariupol, delivered by a small drone – but “given that the source for these claims was the Azov Battalion and the inability of any external observer to verify them, it’s hard to give them much credence”. They note that much more clear-cut uses of chemical weapons in Syria are still controversial. So the chance of the question being resolved positively are somewhat smaller than the chance of chemical weapons actually being used.
Most of the forecasters felt that the upsides of using chemical and biological weapons were low. “I don't think chemical weapons or typical bioweapons are very tactically useful,” wrote one; “It’s fairly likely they didn’t use them in Mariupol, a long and bloody siege; it raises the question ‘Where else might they use them?’,” wrote another. And they felt the downsides were large and obvious: not simply the risk of blowback from NATO, although that was clear, but the cost elsewhere. “Some level of semi-explicit Chinese political support is required [for Russian war aims], and this gets harder to maintain if WMDs are used,” wrote one. “It would also make it harder for European states to push Ukraine to accept a ceasefire that cedes Crimea and Donbas.”
That said, forecasters whose estimates were on the higher end pointed out that betting on Russia making sensible, rational decisions wouldn’t always make you money. “There are far more negative political ramifications for using chemical weapons than tactical or terror gains from them to make it a good decision,” one wrote. “But I'm not lower than 15% because Russia has already made many bad decisions.”
Escalation of the war via proscribed WMDs would indeed be risky, writes another, but: “Putin’s full-scale invasion of Ukraine was a fairly risky decision, so it seems that he is certainly willing to take measures that Western geopolitical analysts previously thought were extremely unlikely.”
Before the 1st January 2023, will there be a full-scale ceasefire declared between Russia and Ukraine?
Median forecast: 21%
While the use of WMDs would represent an escalation, the war could go the other way, and start winding down. But while the forecasters felt that was more likely than the deployment of chemical and/or biological weapons, they all – even the most optimistic – still felt it was an unlikely outcome. The highest probability given was 30%.
For a ceasefire to become a realistic possibility, one pointed out, at least one side in the conflict would have to see diminishing political support for the war, and/or would have to see some sort of tactical advantage in ceasing hostilities for a period. But Ukrainian public opinion is unlikely to shift, at least until the winter, if limited gas supplies and damage to electricity generation lead to widespread energy shortages. (As a second forecaster writes: “The will of the public in Ukraine is definitely not in favour of accepting Russian aggression at this time.”)
And despite Ukrainian losses being heavy, they have manpower reserves, a constant flow of Western weapons, and training support to use those weapons.
On the Russian side, “casualties seem to be running at an unknown but politically manageable level, and the Russian state has the financial resources to shore up at least temporary population support with higher levels of social spending,” the same forecaster writes. The regime is insulated from the worst of the political consequences of casualties since most of the urban warfare has been fought by troops from the Donetsk and Luhansk “People's Republics” rather than those recruited in Russia itself. And Russia’s industrial base should be able to support the sort of war that we’re currently seeing – “slow concentrated troop movements preceded by the application of massive artillery bombardments, the kind of thing the Russian military is good at” – for “many years to come”, even with Western sanctions.
There is, writes one forecaster, “some fairly low chance the Ukrainians could come under pressure from the central Europeans to move towards peace” as the economic consequences of the war continue to bite, and as concerns grow about a widening of the war. “But I'm reasonably pessimistic, as I think Ukraine will have to be in a different position or hold quite a different mentality to be able to give up any of its land for a ceasefire to hold,” they continue. “I think the Ukrainian leadership will pretty much fight until they are not supported by the West.” The West’s stockpiles of weapons and munitions are depleting, but there’s still a lot left, especially if you include lower-tech weapons than those currently being sent.
In the credit column, the political leaders might not be implacably opposed to a ceasefire. Volodymyr Zelenskyy, the Ukrainian president, has said that he wants the war to end, notes one forecaster, “and seems somewhat willing to compromise for that to occur,” while Putin “could be open to a deal taking place if he still has some way to save face and convince Russians that it was still a victory, such as the capture of a small but strategic piece of territory.” And “six months is a long time,” one notes, and the political and military situations could change significantly.
But nonetheless, it’s probably unlikely. “I could see a small-scale ceasefire for a couple of cities for a short humanitarian window – early March had one for two cities,” says a forecaster. “But a full-country ceasefire would be a much less likely scenario.”
Will events involving Russian security forces result in 25 or more fatalities on a NATO member state's territory before the end of 2023?
Median forecast: 8%
Not long ago, it looked as though there was a realistic chance of a widening conflict. The Russian port of Kaliningrad on the Baltic Sea is cut off from the rest of Russia by a 40-mile stretch of Lithuania known as the Suwalki Corridor, and in June Lithuania said that it would restrict exports of goods into Kaliningrad through its territory. Russia responded with sabre-rattling threats to seize the Corridor and join its territories together.
That seems to have eased off. “The chance of something going wrong in Lithuania seems much lower than it did a couple of weeks ago,” one forecaster says, “but it was a good example of how things can escalate due to nested games and mistakes. There is plenty of scope for conflict spillovers and accidents as well as intentional escalation.”
Nonetheless, the forecasters think most of the Russian rhetoric is “chest-thumping without action'' and that “actually Russia is very careful about going to war with NATO.” If they were to do it, Lithuania is the likely starting point, but “it would be a massive escalation and I think both sides really want to avoid it.”
That said, there are other ways for the question to be resolved positively other than a deliberate action in a Baltic state. One would be a stray missile hitting a NATO country – unlikely, but not impossible. Another would be the accidental shooting-down of a passenger plane.
Another would be conflict elsewhere. “Russia and Turkey have a ceasefire in Syria,” notes one forecaster, and Turkey is a NATO member. But that is also unlikely: “There have been no deaths caused by Russian soldiers on Turkish soil. A Turkish F4 was shot down by Syrian forces in 2012, but that incident was over international airspace and did not involve Russia.”
Russian attacks on military bases near the Ukrainian city of Lvov could also lead to the question being resolved positively. Lvov is about 70 kilometres from the Polish border and there are military bases within 10 kilometres of that border, at least one of which has been attacked already. A missile launched from the Black Sea that misses its target could conceivably hit Poland.
The forecasters agree that intentional escalation seems far too risky. “If this were to happen, I think it would either be the result of an accident or because a small amount of Russian forces or a NATO member's forces attacked the other side without permission of higher-ups,” writes one. Even that is unlikely, but if it were to happen, “I think both leaders would try to de-escalate just because the ramifications would be too terrifying (World War III, a nuclear exchange) if they didn't defuse the situation.”
That said, one forecaster pointed out that “One and a half years are an eternity here, and the situation could change dramatically,” so they kept an 8% chance to represent the uncertainty.