During the Democratic presidential debates this week, candidates with a particularly thorny national security issue: whether they would forsake the use of nuclear weapons first in a conflict. This policy, known as No First Use, is the policy of just a handful of the declared nuclear powers.

Those arguing for the policy say it would make accidental or impulsive nuclear war less likely. Those against say that, despite overwhelming U.S. conventional military capabilities, certain dire situations might call for the use of nukes and that a stance of ambiguity is the best deterrent. Let’s explore this debate a bit.

For more than 70 years, those in charge of nuclear arsenals have been conflicted over when to use them. The tremendous destruction of Hiroshima and Nagasaki, as well as atmospheric nuclear tests conducted during the Cold War, made it plain that nuclear weapons were on a whole different level than conventional weapons and their use invited the end of civilization as we know it. In a crisis or conflict between two nuclear powers, the use of even one nuclear weapon could lead to the use of even more nukes as both sides seek to terminate the conflict on their own terms. That could lead to an early end of the war, as one side loses its nerve, or gradual escalation to all-out nuclear war.

You rarely—if ever—hear “South Africa” and “nuclear weapons” mentioned in the same breath. That’s,…

The inherently extreme nature of nuclear weapons means that, unlike a machine gun or fighter jet, a country may not necessarily use them right away in a conflict. It also means that, if both sides involved in a war have pledged not to use nuclear weapons first and actually hold to that pledge, a war could remain non-nuclear. This is the concept behind No First Use.

The first country to adopt it was China in 1964. Since then India has adopted NFU, with the stated exemption that the gloves come off if Delhi is attacked with chemical or biological weapons. , however, including the United States, Russia, the UK, and Pakistan, all maintain a level of ambiguity about when they might use nuclear weapons in a conflict.

These countries argue, somewhat reasonably, that “maybe we’ll nuke you or maybe we won’t” is a deterrent to potential adversaries, heading off both conventional and nuclear war.

No First Use is an appealing policy because it takes the pressure off to rapidly respond to nuclear attack. China, unlike the United States and Russia, does not maintain an active nuclear alert force of missiles ready to launch in minutes. China intends to absorb an attack, evaluate the attack, and then launch a devastating nuclear counterblow that would probably include incinerating the attacker’s cities. In the Chinese view this is plenty enough to deter a surprise nuclear attack.

NFU is also seen as beneficial as it would prevent a crazy, impulsive, unpredictable leader (in the view of and others, President Trump himself) from suddenly ordering up a nuclear strike. It would also eliminate possibility of nuclear weapons launched on false warnings, such as the in which Soviet defenses warned that American ICBMs were headed towards the USSR. No First Use would build a useful delay into an American nuclear response while still ensuring the other side gets clobbered.

A pledge not to use nuclear weapons does not, readiness aside, mean the U.S. would let its nuclear guard down. The Pentagon would have just as many nuclear weapons as it had before. It could even have less: China has a reported to the 1,500 deployed weapons in American and Chinese arsenals. China knows that if just a handful survive a U.S. attack to kill millions of Americans, the Americans won’t attack. And they’re right.

The idea of No First Use is a popular one in the United States, the only country to ever use nuclear weapons in war. , 67 percent of the American people supported the adoption of NFU in 2016.

But many nuclear strategists dislike the idea. They argue that some catastrophic loss on the battlefield, such as a war on the Korean peninsula, against Russia, or even China might take place that would require the use of tactical nuclear weapons. They also argue that if the United States had intelligence that a nuclear attack against it was imminent, it should reserve the right to strike first. They argue that nuclear ambiguity creates uncertainty, doubt, and even fear in an adversary that might consider attacking the United States—with nuclear or conventional weapons.

Another problem with NFU: countries lie, and some of them have done terrible things. China’s pledge not to use nukes first may seem mature, principled, and humane, but it’s worth keeping in mind the government in Beijing is the same country that massacred pro-democracy protesters in Tiananmen Square in 1989, it has a million or more Uighurs in concentration camps for their Muslim religion, and that its development of the South China Sea was entirely peaceful. Given the gravity of nuclear war, can the U.S. trust China’s NFU pledge?

Advocates of NFU counter that the overwhelming technological and numerical superiority of the United States in conventional weapons makes it unlikely it would ever be pushed into a corner and forced to use tactical nuclear weapons. As for a surprise nuclear attack, advocates counter the real deterrent is that adversaries know they would be devastated by a retaliatory strike by surviving American forces. Even if Russia could guarantee the destruction of 99 percent of U.S. nukes in a surprise attack, the remaining one percent would make it not worth the effort. This would hold even if the Chinese shouted “surprise” and dropped their No First Use pledge.

Nuclear No First Use is a good idea made more relevant now than any time in the last 27 years. Does it invite instability and nuclear attack? Maybe in the most credibility-stretching of cases, but it also increases stability in the event of a false alarm or a mentally unstable leader with his finger on the nuclear button. It is also apparently what the American people want by a substantial margin.

Whether America adopts the policy or not, it’s past time to have an adult conversation about our nuclear weapons and when we might or might not start a nuclear war.