Earlier this week I finished drafting a paper on the Korean peninsula crisis for a talk in Japan next week. I concluded, somewhat glumly, that the prospect of a deal between the US and the other parties involved was slim, and that military action was a very real possibility. Within hours of typing my last full stop, President Trump was telling the Koreas ^[1] that North Korea could ‘come to the table’ and ‘make a deal’.

Sometimes it’s good to be wrong, and I’d really like that to be the case here. But I’m not confident that there’s really a deal in the making. We have no detail of what might be involved, but an analysis of the positions of all the parties suggests that someone is going to be left unhappy.

In any situation ‘do nothing’ is an option—sometimes the best one. Here that would mean accepting North Korea’s weapon programs and maintaining the status quo posture for American and allied forces and diplomacy. The upside of ‘do nothing’ is that there’s no risk of preventive action spiralling into disaster. The downside, obviously, is that North Korea would develop weapon systems that risk a future catastrophe. The calculus is whether the total future risk from doing nothing is greater or less than that of taking action. An added complication is that both the assessment and reality of short- and long-term risks vary between different parties.

For those countries that are within striking distance of short- and medium-range missiles, we’re already at the point where active steps can bring a catastrophic response. South Korea, China and Japan have to factor the possibility of a nuclear counterattack into any estimate of costs and benefits. That necessarily makes doing nothing more attractive than it would have been before North Korea had nuclear capabilities. But we have to deal with current realities.

Countries whose mainlands are beyond the range of North Korean missiles, but which will come within range of long-range missiles under development, have an incentive to act now. The US doesn’t yet have to worry about a nuclear missile hitting Los Angeles, so its calculus is different from that of Seoul or Tokyo. The US is yet to pass the risk threshold already crossed by more proximate nations.

A crucial question is how the US will weigh its own safety against the safety of its north Asian allies. A multilateralist US that put a high premium on the safety of Japan and South Korea would be loath to take steps with high escalatory risk. A more self-centred US might weigh its own physical safety higher than that of its partners. The Obama administration had the former characteristic. It was prepared to back sanction regimes—even when unlikely to have a decisive effect—and wasn’t especially aggressive in the use of force. The approach was often ‘do nothing’ or, more accurately, ‘do nothing likely to be either effective or directly harmful’. They even had a name for it: ‘strategic patience ^[2]’.

The ‘America First’ Trump administration has been talking up a more aggressive approach, while criticising Obama’s ^[3]. It’s possible that President Trump’s instincts are more belligerent than those of the people advising him. But his rhetoric—at least until this week’s visit to Seoul—certainly suggests a US that won’t accept vulnerability to North Korean nukes.

On the other side, Kim Jong-un has seen other strongmen toppled by the US, and he’ll be noting Trump’s threats to abandon its deal with Iran ^[4] despite no evidence of Iranian non-compliance. And the US failed to deliver its part of a previous nuclear agreement ^[5] with North Korea (though both sides were at fault). All that will feed Kim’s desire for an effective deterrent to an American-led regime change.

If the US and North Korea could agree to the status quo, there might be a stable equilibrium. For it to work, Kim would have to believe that he had sufficient deterrent power to forestall American efforts to remove him, without an ability to strike the continental US. And Washington would have to believe that North Korea wasn’t covertly working on further systems, and wasn’t likely to precipitously attack US allies or territories.

Only one of those—North Korea not working on long-range missile systems—is objectively verifiable, but it would require Kim’s compliance with an inspection regime (backed by intelligence work for added assurance). Even then, it wouldn’t be unreasonable for the North Koreans to recall the ‘Iraqi WMD’ case that led to Saddam Hussein’s downfall.

Everything else is even less certain. The North Koreans would have to trust the reliability of American guarantees—and have their own experience, as well as Iraq’s, Iran’s and Libya’s, to learn from. And the Americans would have to be convinced that North Korea would be content with mutual vulnerability and not be inclined to use its weapons against other targets—the most likely of which are American allies.

Maybe Trump had something like that in mind when he spoke of a deal, but I’m not confident that the trust required for a status quo solution can be generated. Kim likely wants more deterrence than he currently has, and the US doesn’t want him to have it. And it’s not unreasonable for Japan and South Korea to regard North Korean short- and intermediate-range nuclear missiles as unacceptable. So, even if the US can strike a deal that stops North Korea short of an ICBM, it will be of little comfort for the rest of north Asia. And leaving Tokyo and Seoul vulnerable in exchange for safeguarding American cities would probably weaken the US alliance framework. That would be an outcome to none of our benefits—though it’s preferable to a war with uncertain escalation prospects.