In the wake of the failed recent summit in Vietnam, the United States has reached something of a crossroads in its policy toward North Korea’s nuclear ambitions. Hints of a third summit, or high-level interim negotiations, emerged after the Vietnam meeting, but it is not clear on what basis they would proceed: The two sides seem far apart on the specifics of a deal.  North Korea’s partial halt to its nuclear activities continues but as many reports have made clear, it does not prevent the regime from moving ahead on many fronts. The United States remains committed to complete, verifiable disarmament. The result is a volatile situation that could produce a crisis—or some surprising agreement—at any moment.

But as the world awaits the resolution of the nuclear issue, it is critical to keep in mind that the matter is nested in a larger and more encompassing set of almost equally daunting strategic dilemmas for the United States in Korea—dilemmas we studied in detail in a series of recent analyses at the nonpartisan RAND Corporation. These studies suggest to us that policymakers and the American people need to understand how the security environment in Korea has changed, even as the United States reaffirms its commitment to denuclearization. Simply put, the United States is locked into a series of overlapping missions that could strain the U.S. global defense posture to the breaking point if the operations had to be carried out in a crisis on the Korean Peninsula.

With several colleagues we recently developed an infographic that depicts our major pieces of analysis undertaken at RAND over the last three years. One examined the strategic challenge posed by North Korea’s growing nuclear arsenal. A second assessed the threat of North Korean conventional artillery strikes on South Korea, fired mainly from the dense concentrations of weapons just north of the demilitarized zone in the area known as the Kaesong Heights. A third study looked at the problem of dealing with loose nuclear materials in a fragmenting North Korea, and a fourth at the potential requirement for a noncombatant evacuation operation (NEO) in the event of a conflict. We also reviewed RAND work on a fifth issue connected to all four possible contingencies: the potential for Chinese intervention in these scenarios.

We took one overwhelming lesson from the sum total of this work: Any major conflict scenario in Korea is likely to tax U.S. military capabilities in the extreme at a time when, as strategic competitions with Russia and China are heating up, the United States faces daunting missions in many places. After seventeen years of focus on stability operations, counterinsurgency and nation building—and trillions spent on the efforts—the United States likely cannot afford another decade of distraction from the main strategic events of the present and future security environment. Given the potential stakes and risks of a Korean conflict, that is exactly what it could entail.

For decades, the U.S. national-security community has tended to think of a Korean contingency as a replay of 1950: South Korea and the United States defending against a massive North Korean thrust south. If it was ever a possibility, such a scenario is far-fetched today, given the South Korean and U.S. technological superiority, seventy years of defensive preparations, the crushing veto power of U.S. and South Korean air power, and North Korea’s known deficiencies in supplies, logistics and combined-arms competencies.

That is not the shape a new Korean conflict is likely to take. Our research points to several ways a war in Korea could unfold today and the demanding missions it could impose on U.S. forces that would be extremely costly and challenging.

In the event of a major crisis or U.S. preemptive attack on North Korean nuclear facilities, for example, North Korea could respond with modest—or overwhelming and destructive—artillery attacks across the border, all the way into downtown Seoul. U.S. Defense Department public information suggests that the North has some fourteen thousand artillery systems, with 70 percent forward deployed in over four thousand underground bunkers. All told, they could fire over half a million shells per hour; the long-range systems that can reach Seoul could fire three thousand or more rounds per hour—or a fraction of that amount if it wants to try to keep its guns hidden, firing only occasionally. One DOD estimate suggests that larger-scale artillery attacks could cause over a quarter-million casualties in Seoul.

Such an attack would present South Korea, and by extension the United States, with a terribly difficult dilemma: To continue to sustain damage or go on the offensive to dig out the North Korean guns and stop the firing. (Our analysis suggests that airstrikes alone could not end such a barrage.) Any major offensive action would impose enormous costs on South Korean forces. Very quickly, the relatively modest South Korean assault forces could be hard-pressed to continue offensive actions until the arrival of reinforcing U.S. armored units. That could take months, creating a dangerous interval in which North Korea could be grievously wounded but not yet defeated.

Any such crisis would create a requirement for the evacuation of civilians, following typical U.S. war plans. Our assessment of such a non-combat evacuation operation demonstrated the daunting scale of such a task: there may be as many as 150,000 U.S. citizens in the South, and another four hundred thousand citizens of allied and partner countries the United States might have to help evacuate. Evacuating just the U.S. citizens could require over four thousand helicopter sorties and close to one thousand flights of C-17 transport aircraft—the very same assets the U.S. military would likely be counting on to deliver reinforcements to the peninsula.

Any major conflict could fracture the North Korean state, perhaps turning loose fissile material or even assembled nuclear devices into the swirling chaos of a collapse scenario. A U.S. president facing the risk of such weapons finding their way onto the black market would be highly likely to order the U.S. military to take steps to gather the missing nuclear assets—and the resulting mission would probably impose staggering demands.

A RAND analysis of the possible requirements of a weapons of mass destruction elimination mission and its associated jobs—getting into North Korea, securing large areas for search, guarding nuclear experts as they conduct their forensic investigations—could require over 250,000 U.S. or South Korean ground forces. That’s assuming relatively light resistance. If scattered, fragmented North Korea units fought bitterly across the country, the troop requirements, and time it takes to accomplish the mission, could grow.

Finally, the scenario with the most destructive potential would be one in which North Korea decided to use even a part of its nuclear arsenal. Media reports and published analysis by U.S. defense experts suggest the North may already have as many as sixty weapons, and within two to three years could be approaching one hundred devices. The North could employ these in many ways, as an individual demonstration attack for coercive purposes, a limited but still devastating attack on ports and bases to slow U.S. reinforcements, or—in the event regime survival were at stake—an all-out employment of every weapon in its arsenal. Its public statements have signaled a possible reliance on such comprehensive strikes given its vulnerability to U.S. attack.

No matter how they occurred, any nuclear attacks in Korea or the surrounding region would have catastrophic results. If a device were employed against Seoul, for example, then even a small weapon (ten kiloton yield) could produce ninety thousand fatalities and over three hundred thousand casualties. A weapon ten times that size—which North Korea has already tested—could cause over four hundred thousand deaths. U.S. and South Korean military forces and facilities would be severely damaged and face a colossal burden of post–attack humanitarian relief.

As unthinkable as each of these scenarios can be individually, in the event of a major war, many or all of them could unfold at the same time. They are, to a certain degree, interrelated; any of them could easily metastasize to bring about the others. An artillery attack would require a noncombat evacuation operation, for example, and if South Korea and the United States struck north, then that could spark North Korean nuclear use. Even initiating a noncombat evacuation operation during a crisis could trigger escalation, as the North could take it as a sign of impending military action. Any large-scale conflict could produce the collapse of North Korea and the resulting weapons of mass destruction elimination mission.

In the event of real instability in Korea, the United States and its ally, South Korea, may face an interlocking set of missions whose combined demands on U.S. military power are likely to be overwhelming.

These risks point to a related issue, an awful truth about the current composition of the U.S. defense establishment: It is simply not built to fight long, extremely costly conventional wars. The U.S. defense industrial base is not able to surge production if hundreds of tanks or aircraft are lost. Massive numbers of dead and wounded troops could overwhelm a system for dealing with casualties that has atrophied over time, and the military intake and training mechanisms are not designed to quickly make up severe losses.

On top of this, major crises or conflicts in Korea come with a potent additional danger: China is likely to become drawn into any major crisis or conflict on the peninsula. Many people forget the country has a formal military alliance with North Korea dating from 1961—one that pledges mutual assistance in the event of “unprovoked attack” in more absolute terms than the wording of the U.S.-South Korea alliance. An estimated one million Chinese citizens live and work in South Korea, creating the potential need for a massive Chinese noncombat evacuation operation. Loose nuclear weapons on its doorstep would be an even greater threat to China than the United States. And leaders in Beijing could be looking for opportunities to step in as peacemaker or mediator, seizing the geostrategic leadership from the United States.

China’s role could be helpful—or put China on a collision course with the United States as happened in the last Korean War. The risk of escalation would be ever-present and possibly exacerbated by China’s current foreign-policy mindset, which is to increasingly press its demand to be the ultimate arbiter of events in Asia.

Given other U.S. commitments and responsibilities, the lesson is unavoidable: a major crisis or war in Korea is something that would be a massive strategic handicap for the United States. (If a war resulted from an unprovoked U.S. first strike on North Korea, and the United States was blamed for the resulting devastation, then that effect could be fatal to U.S. global legitimacy.) Dramatic steps may need to be considered to transform the security context in Korea, to sidestep these risks before they emerge to undermine U.S. security posture.

For U.S. interests on the peninsula, the United States may want to aim to maintain the credibility of the U.S.-South Korea alliance while endorsing significant steps to reduce the risk of conventional war. North Korea may balk, but North Korean leader Kim Jong-un’s focus on economic development is clear—and if such steps could allow him to reduce his military burden and escape some sanctions through confidence-building, he might be willing to make real concessions. The momentum for major agreements seems to have stalled, but U.S. interests at stake suggest that it should make real efforts to join general tension reduction and conventional arms control to the goal of denuclearization.

The program of tension reduction and confidence-building could have multiple components: Overall reductions in conventional forces; specific, targeted cuts in forces deployed near the DMZ; a large, true demilitarized zone on either side of the border in which military forces are not allowed to be stationed; installation of sensors and transparency measures; and even the deployment of official UN peacekeepers in the border area. The agreement could be formalized with a multilateral endorsement—including China—of non-aggression. In the process, the United States could convey a willingness to consider significant reductions in forces deployed in Korea as well as take other confidence-building steps to build trust with North Korea, China, and other regional allies.

This proposed U.S. goal—along with that of its ally, South Korea—aims to make conventional conflict all but impossible in Korea, and in the process vastly reduce the potential for the multiple intersecting scenarios described. Combined with some form of nuclear deal that reduces the risk of conflict from that avenue, the United States could likely reduce its planned requirements for a crisis contingency on the Korean Peninsula.

In this era of global competition in which U.S. resources are stretched thin, the United States should consider looking for opportunities to scale back potential overcommitment. The current inter-Korean dialogue presents an opportunity. If it is not taken, then the alternative may be more costly than U.S. global posture could sustain.

Michael J. Mazarr is a senior political scientist and Gian Gentile is a senior historian at the nonprofit, nonpartisan RAND Corporation.

Image: Reuters