Subscribe for 33¢ / day
The Latest: Admiral wants more US missile defense in Hawaii

This image made from video of a still image broadcast in a news bulletin by North Korea's KRT on Wednesday, April 26, 2017, shows leader Kim Jong Un at what was said to be a "Combined Fire Demonstration" held to celebrate the 85th anniversary of the North Korean army, in Wonsan, North Korea. Independent journalists were not given access to cover the event depicted in this photo. (KRT via AP Video)

The Associated Press

The Trump administration has been floating the idea of a "bloody nose" attack on North Korea that advocates believe could punish its aggressive threats without starting a war. This is dangerous: North Korea will respond, Americans will die, and nobody knows how bad a war could become. Rather than try a fanciful and extremely risky attack, the United States and its allies should continue deterrence. North Korea is already deterred from striking first; we can deter North Korea indefinitely, while working on other options to change the regime.

The bloody-nose option imagines that it is possible to launch a limited, surgical strike on North Korean nuclear or missile facilities that would force Kim Jong Un to chart a new course. But this is wishful thinking, based on two deeply flawed assumptions: that North Korea won't respond and that the United States can control the escalation ladder if it does.

Proponents of the bloody-nose option also hope to attain two highly unlikely goals: convince North Korea to come to the negotiation table to denuclearize and to stem proliferation.

But North Korea is the most consistent country in the world: It always meets pressure with pressure of its own. There is overwhelming, decades-long evidence that North Korea will fight back. From the sinking of the South Korean corvette Cheonan in 2010, to skirmishes along the border at Yeonpyeong in 2010, 2002 and 1999, to shooting incidents over the decades, North Korea has not only shown a willingness to use force but also has done so repeatedly. There are exactly zero examples of a time North Korea caved in to pressure. To think that this time it would not fight back is wishful thinking based on no evidence and weak logic.

There is a logical reason for North Korea to respond to an attack, as well. If it does not, all it will have done is show that its leadership will accept a minimum level of punishment. There is no way for the United States to credibly promise to attack only once - so North Korea has to assume that if there is one bloody nose, another will follow, perhaps a little bloodier.

The second faulty assumption is that the United States can control the escalation ladder if North Korea does respond. What are the next steps if North Korea attacks a U.S. base in Japan or South Korea? What if 500 American civilians and military personnel are killed in a North Korean retaliatory strike? The pressure for an even greater military reaction will be overwhelming, which will then lead North Korea to conclude that all along the United States was preparing for a major war, and it will respond accordingly by escalating - and so on.

What's at stake? The lives of 300,000 Americans - roughly the population of Pittsburgh.

On any given day, there are between 250,000 and 300,000 U.S. citizens in South Korea. This includes 23,000 military personnel, their families and thousands and thousands of Americans who work, study or live in South Korea. The largest evacuation the United States has ever done was to remove 60,000 people from Vietnam in 1975. It is unrealistic to believe that the United States can evacuate 250,000 people in a week or two, under conditions of total war.

Some argue: "If there is a war, at least it will be over there." Putting aside the astonishing callousness of that attitude, if there is a war on the Korean Peninsula, thousands of Americans will die, many of them civilians.

Military strikes are only likely to reinforce the desire in North Korea for a military deterrent. Strikes are thus likely to spur and increase North Korea's missile development, not slow it. After all, if someone is attacking you, there is a powerful logic to get more weapons, not fewer. The Olympics have provided a pause in the discussions, but now some are even talking about a massive first strike - guaranteed to lead to even more American and Korean casualties than a bloody nose.

The latest sanctions the United States has placed on North Korea, along with deterrence, are an appropriate way to pressure the regime. Both these tools can limit North Korea's influence without inciting a war.

North Korea won't attack first, because to do so would be regime suicide. But it will fight back if attacked. Deterrence has worked for 65 years, and it can continue to do so indefinitely. In the meantime, the United States, South Korea and other powers need to work to find other ways to affect the regime and its strategic perceptions, and to help the 25 million people in North Korea who suffer human rights abuses on a daily basis.

David C. Kang, author of "American Grand Strategy and East Asian Security in the Twenty-First Century," is a professor of international relations and business at the University of Southern California. He wrote this for The Washington Post.