Rogue states would be better deterred by our military than the threat of nuclear destruction.

We need nuclear weapons. Let’s just stipulate that at the outset, at least for now. But the United States needs a new deterrent strategy against rogue states that does not require threats of nuclear retaliation.

Since the 1950s, strategic nuclear weapons have served an amorphous psychological condition we call “deterrence.” During the Cold War, when we were struggling with a Soviet opponent for the future of the planet, the U.S. nuclear arsenal was meant to deter Communist aggression both by punishment and denial: if the USSR attempted to seize Europe or to destroy the United States, U.S. strategic nuclear weapons would ensure that Moscow would end up ruling only the ashes of whatever was left – including the burning remains of the Soviet Union itself.

Today, only China and Russia can cause such grievous damage to the United States, and the terrible equation still stands: an attack on the United States might mean the end of America, but it also means the complete destruction of the attacker. But what about small states, the “little guys” with the big weapons? These are nations – like North Korea, and soon, others – who will never be able to destroy the United States, but who could cause immense damage to an American or an allied city. Does the old bargain of a “nuclear eye for a nuclear eye” still hold? And should it?

It’s time to be honest with ourselves. We are not going to inflict nuclear destruction on small states in crowded neighborhoods, killing thousands, maybe millions, of innocent people, and poisoning swaths of territory inhabited by friends and enemies alike. The United States and its allies need to create a new and radically different deterrent against small nuclear powers, one that does not include threats of nuclear retaliation.

This is a bitter pill for American strategists. It upsets all of our traditional and well-worn equations about nuclear use that seemed to serve us so well (or at least, so we think) during the Cold War. If North Korea strikes an American or Japanese or South Korean city, it is simplicity itself to say that the inevitable U.S. nuclear retaliation will vaporize North Korea. Such threats, however, are bluster. They are not only unrealistic, but they violate precepts both of morality and practicality.

There are three reasons we must abandon outdated, Cold War-type deterrent threats against small states.

First, we will not be able to contain the damage from retaliatory strikes, no matter how precise our attacks. Even Richard Nixon admitted as much when he shelved various plans to use nuclear arms against a far stronger North Korea in 1969, and subsequent presidents have all had to deal with the realities of using weapons meant to stop the Soviet Union against much smaller states. As U.S. Air Force General George Butler noted in 1999, “this lesson has been made time and again, in Korea, in Indochina and most recently in the Persian Gulf, [when] successive presidents of both parties have contemplated and then categorically rejected the employment of nuclear weapons even in the face of grave provocation.”

Geography itself is the enemy of small-scale nuclear weapon use. Iran’s most important nuclear reactor, for example, is less than ten miles from the city of Bushehr and its approximately 170,000 inhabitants; likewise, the Iranians have placed a uranium enrichment facility some 20 miles outside Qom, one of Iran’s treasured holy cities. North Korea’s Yongbyon nuclear facility is slightly more than 50 miles from Pyongyang, the capital city of more than three million that itself could be a target for retaliation.

Leaving aside the problem of proportionality within the targeted state, there is also the problem of nearby nations. Nuclear conflict with a vast country like the USSR meant aiming for complete destruction, with little regard for the morning after the war. The goal was to prevent the Soviet regime from emerging from the wreckage, and the damage to other nations would be an unfortunate but unavoidable side effect. That is the nature of strategic nuclear war.

Nuclear war with a small state, however, will affect other countries and could place them in mortal peril, even if the attack is only revenge for a single nuclear strike on the U.S. or its allies. Japan’s shores are less than 700 miles away from North Korea, with Pyongyang less than 800 miles from Tokyo. On the other side of the world, Iran borders the Islamic-majority NATO nation of Turkey as well as nuclear-armed Pakistan. Nor is it far from Iran’s borders to a collection of former Soviet republics, including southern Russia itself. The Fukushima reactor meltdown, an accident in Japan smaller than the 1986 Chernobyl meltdown, still has people across the world, including along the U.S. West Coast, worrying about nuclear contamination nearly three years later. Imagine the pure panic that would ensue from dropping a nuclear weapon – or several – on targets in Asia or the Middle East.

Another reason we cannot credibly threaten to use nuclear arms against small states is that our retaliation would seem disproportionate, even to our own allies. This matters both strategically and morally. If the United States uses a nuclear weapon, whatever started the war will quickly become irrelevant, as attention will immediately turn to the casualties from the U.S. retaliation. The burn victims alone will present a ghastly moral dilemma we have never encountered before; like it or not, the United States will end up responsible for the care of its defeated enemy, and troops advancing through areas destroyed by an American nuclear response (who themselves will be at risk in a nuclear environment) will quickly realize that they will have no humane choice but to euthanize many of those civilian casualties on the spot. That footage, and not the initial attack on the United States, will be the images that will run in perpetuity on the world’s television screens, and perhaps might even achieve the propaganda victory the enemy wished for in the first place.

Finally, there is no way to guarantee, short of burning the targeted country to the ground and committing nuclear genocide, that we will eliminate the enemy regime with nuclear weapons. What if a retaliatory strike on the North Korean regime fails, and some member of the Kim family emerges from the rubble only to declare that the Dear Leader, or the Young Leader, or the Great Leader, or Dennis Rodman’s Best Friend, or whoever is left, is still in charge of the country? Do we then strike a second time? One strike will be difficult to contemplate. Two, especially if purely for revenge, will be almost impossible to conceive.

So what is to be done? Shall we simply allow rogue states to gain nuclear arms, impose a Cold War deterrent template on them, and hope they believe in it, even if we do not? Should we try to tell countries like North Korea and Iran that we really are cold-blooded enough to destroy their cities and execute their innocent populations as revenge for American deaths, even if the existence of the United States itself is not in danger? If we make these threats, we’d better hope they work, because if we cannot persuade enemy leaders that we are as barbaric as they are, a horrible miscalculation could ensue.

None of the scenarios for nuclear use against small states make much sense because they do not respond to what the leaders of rogue regimes fear most: their own deaths. That is why an effective and credible deterrent – that is, one we can plausibly threaten and then actually carry out – rests not on nuclear weapons but on a far more realistic threat of conventional war. With regard to Russia and China, I have already made the case elsewhere for a minimum deterrent. But where smaller rogue powers are concerned, America needs a new deterrent, one in which we warn that WMD attacks, including the use of even a single nuclear weapon against the U.S. or its allies, will precipitate major conventional war with the explicit goal of regime change and the apprehension or death of the enemy leadership.

The key to deterring rogue states is to remember that their leadership cares little about their own people. (This was the assumption of the “countervailing strategy” against the Soviet Union developed by Jimmy Carter and later embraced by Ronald Reagan in the late-Cold War.) They care, rather, about their own survival and their control over their territory. Threatening them with nuclear arms is difficult not only because such threats require them to believe that Americans share their stomach for mass murder, but because these regimes, in effect, hold all their neighbors hostage to the effects of a U.S. attack. Nuclear threats sound hollow to such men, and they are.

A deterrent based on engaging in war on Western terms, however, is another matter. This is a threat that U.S. and allied forces have already carried out, repeatedly. For confirmation, one might ask Saddam Hussein, Slobodan Milosevic, or Muammar Gaddafi, were they not all deposed and dead. A promise of unspecified nuclear horror is not credible, but a vow that says “no matter how this ends, you will be dead and your regime gone” is one that enemy leaders can grasp because they’ve seen it happen in front of their own eyes.

The costs of this change will be enormous. To alter the U.S. deterrent from vaporous nuclear threats to concrete promises of conventional war will require reinvestment in our military forces – and real investment, in guns, boots, and personnel, not exotic weapons systems that will never see battle. It also means a recommitment of purpose on the part of the American people, who of late have withdrawn from the world rather than poke the various nests of snakes that sooner or later will threaten U.S. and allied security. “Kicking the can down the road” is now the meme for U.S. foreign policy, but we’ve already seen the results of that short-sighted approach: the Clinton administration declared an empty victory over North Korea in 1994 and just over a decade later Clinton’s successors found themselves facing a nuclear-armed regime run by sociopaths.

It is time to put away the nuclear crutch we’ve relied on since the end of the Cold War. Real threats require real solutions, and rogue leaders considering the development or use of nuclear arms against the United States need to be spoken to in the only terms they understand: if they use a nuclear weapon, they will precipitate a war with the most powerful alliance in the history of mankind, one that will end with their own deaths from a soldier’s bullet, a bomb from the sky, or perhaps even at the end of a hangman’s noose raised by their own people. There is no nuclear alternative, and we must not shirk either the moral or strategic necessity of finding a deterrent and a strategy that actually will work in the increasingly brutal jungle of the twenty-first century.

Tom Nichols is Professor of National Security Affairs at the Naval War College and a professor at the Harvard Extension School, and the author of No Use: Nuclear Weapons and U.S. National Security (Penn, 2014), from which this article is adapted. The views expressed are his own.