Robert Gates was U.S. secretary of defense and therefore my immediate boss for much of my time as supreme allied commander at NATO. A former director of the CIA and a career intelligence officer, he is brilliant, understated and dry-witted. When asked once about our ability to predict wars, he said, "Our record is consistent and perfect -- we never get it right."
With that as a cautionary comment, I would assess the chances of a full-blown war on the Korean Peninsula in 2018 are in the range of 10% and rising, with an additional 20% chance that some level of ordnance will be exchanged (missiles, torpedoes, artillery) but that both parties will step back from the abyss of a nuclear war. A year ago, I would have placed the chance of war at around 1%. Two things have changed: rapid technical progress by the North Koreans on their missile and nuclear programs; and the unsteady belligerence of the U.S. administration of President Donald Trump -- a combustible and potentially lethal combination.
There are three key indicators to watch in the context of whether or not we see the world sleepwalk into a serious war, which would probably kill between 500,000 and 2 million people. First is the tone of rhetoric from Washington. The Trump administration has had difficulty projecting a consistent message as between the White House, Foggy Bottom and the United Nations mission in New York. If the U.S. government aligns itself around a message of insistence on denuclearization (as opposed to freezing the program), moves away from diplomacy and speaks out via presidential Tweets, the risks of war go up. But there is a high probability that a synchronized approach focused on securing a freeze will emerge in the New Year, assuming Trump's team can prevent him from undoing rational policy via Tweet storms.
The second factor is whether or not Kim Jong Un takes the dramatic and highly dangerous step of detonating a nuclear weapon, presumably a hydrogen bomb, over the Pacific in the open air (as opposed to previous below ground tests). No nation has conducted an above ground test since 1980 (China), and doing so would be the ultimate "thumbed nose" at the international community. Should Kim decide to take that risk, it would probably force the Trump administration to strongly consider a preemptive strike -- controversial under international law and certainly not sanctioned by the U.N. Security Council. The odds of this happening are hard to predict, but many observers say it is a 50-50 probability of an open air nuclear test. If this occurs the odds of a U.S. preventative strike rise measurably.
For Kim Jong Un, the temptation to detonate an above-ground hydrogen device is powerful, especially if it were launched on a ballistic missile. It would firmly cement his presence as a true nuclear power, showing that he was capable of producing a powerful device and launching it over a significant distance. While he knows this would create peak tensions, he probably also calculates it might achieve a negotiation that permitted him to retain his nuclear weapons and freeze his testing in return for guarantees for his regime survival from China and a diminution in U.S.-South Korean military exercises.
China and Russia
A final indicator that risk is spiking would be a ramp-up in great power competition over the issues. The U.S., China and Russia have been thus far loosely in sync in wanting to avoid an all-out war and in seeking negotiations; a change in that roughly synchronized set of views over this issue would increase risk. Thus a decision by China to pull back from its admittedly half-hearted enforcement of U.N. sanctions and fully align itself with the rogue regime in Pyongyang would be a key indicator of tensions rising and the odds of a war increasing.
Russia would probably swim alongside the Chinese policy, seeking leverage with the U.S. in removing sanctions hanging over Russia as a result of their invasion of Ukraine and annexation of Crimea. Turning the Korean Peninsula into a full blow great power confrontation -- much like the Balkans in 1914 -- would signal serious danger, with the potential to extend even beyond a war in Korea. This seems less likely to occur because at root, none of the three nations have interest in a war given the economic and human consequences.
While it seems unlikely North Korea would do anything to precipitate a conflict before the Winter Olympics in Pyeongchang, South Korea, in February, all of the national Olympic teams are assessing the situation and ticket sales are lower than average for this close to a winter event. North Korea at this point does not have a robust team, with only a couple of figure skaters qualifying and mixed signals from Pyongyang about participating. Many Koreans remember the 1988 summer games in Seoul when the North Koreans bombed a commercial airline flight in an attempt to sabotage the games.
Overall, while the odds of an above-ground nuclear detonation and rising tensions are high, the best bet is that there is a chance of around 70% that 2018 will see serious four-party talks (U.S., China, South Korea, North Korea) recommence and we avoid major combat. The most likely overall outcome to such talks is a nuclearized peninsula but with a freeze on North Korean testing, reduced levels of U.S.-South Korean conventional military exercises, and a wary state of mutual deterrence. If there is an above-ground nuclear test, the odds of a negotiated settlement diminish, but do not collapse. Negotiation is still the best bet on how this comes out. Not a perfect outcome, but certainly better than a full-on war, which remains a classic black swan -- low probability but very high and unpleasant impact.
James Stavridis, a retired four-star U.S. Navy admiral, was the 16th supreme allied commander at NATO and spent over half of his career in the Pacific Fleet. He is dean of The Fletcher School of Law and Diplomacy at Tufts University.