Would North Korea consider a first strike?

Vipin Narang says yes:

The strategy turns on Kim’s main calculation that the United States will say it’s not worth losing a major American city to get rid of him.

Of course he could not knock out a major American or allied target, but he could use them somewhere.  And the use would boost his, uh…credibility.  In fact Charles Murray is worried.

I think of the model this way.  If Kim is irrational, we have obvious reason to worry, and of course a first strike could not be ruled out.  Remember Pearl Harbor?  (Or is that “Remember Pearl Harbor!”)

Alternatively, say all involved parties are fully rational in the selfish sense.  Fully rational agents make purely forward-looking calculations.  So if Kim used a nuke to kill a sparrow in North Korea, we would not attack because fear of losing an American city would far outweigh desire to retaliate for the loss of the sparrow.

How about one sparrow in the DMZ?  In Japan?  In the Arctic?  In a Malaysian airport?  Or maybe one sparrow, three sled dogs, and thirty Inuit?

At what point do we give it a go, and risk a poorly aimed North Korean ICBM being shot off into the sky?

What if Kim uses “only” a biological or chemical weapon, designed for minimum but noticeable impact, on a nearby country?  You should think of Kim’s strategy space as a continuous variable, with some noise added of course.

Is the space of “boosts his credibility and domestic stature, but without too much upping the risk of massive American retaliation” really the empty set?

Maybe.  Maybe not.  I give it about one percent, which in expected value terms is still a real worry.


Nuclear explosion in space. EMP attack with nobody killed. Big impact, but we won't go to war over an incident in which nobody is killed. Think Starfish Prime over the United States.


Massive numbers of people would be killed. They just would not be killed immediately or in the same way.

An EMP attack would be worse than one on a major US city and would result in the obliteration of North Korea.

Failure of electronics and electrical equipment would cause some lives to be lost, but what is the basis for saying it would be "worse than one on a major US city"? This seems doubtful.

"Massive numbers of people would be killed"

People get hysterical when the word EMP comes up. They treat it as some sort of magic word. An EMP would probably cause economic and life loss less than what Hurricane Harvey caused.

A North Korean EMP burst, I'm sure the Russians or Chinese could do more damage with a massive multi warhead EMP strike over the continental US.

Sure and people get "hysterical" about the terrorit attacks. Far fewer people were killed on sept 11 than die eachbdaybin the USA. Was the reaction there hysterical or overblown?

We do need to consider context. But it the the entire approach of calculation here wrong. Number of citizens that another country can kill before a response is sensible is not the right calculation.

Starfish Prime didn't kill anybody. Granted the nearest big city was Honolulu and they were much less dependent on electronics, but who exactly would be killed by an EMP over the U.S.? Maybe somebody who is getting open heart surgery when the lights go out? People who need electricity to run their oxygen concentrators? People driving Teslas in autopilot mode?

While an EMP attack from NK would not be even close to as damaging as a NK missile strike on a city, it would still give the US the green light to wipe out NK.

Not if nobody gets killed. We didn't go to war over the massive Sony Pictures hack, and we won't go to war over a zero-casualty EMP attack.

Nonsense. The Sony hack embarrassed some movie executives. An EMP attack is an act of war. If they do it, they're toast. Come on, man.

Are airplanes hardened? Pacemakers and other medical devices? Cars driving 80 mph that suddenly lose all power? Cranes hoisting 20 tons over a downtown street?

"Are airplanes hardened?"

LOL, airplanes are routinely struck by lightening. Yes they are hardened.

"Pacemakers and other medical devices?"

Again, plenty of people with pacemakers have accidentally gotten shocked with 120V and not dropped dead, so they should be fine.

"Cars driving 80 mph that suddenly lose all power?"

Cars are not hardened to the same extent, but it would take a massive EMP pulse to effect a car which does contain a lot of steel.

"Cranes hoisting 20 tons over a downtown street?"

Grounded and hardened against lightning strikes.

Cell phones would probably be the biggest casualty of an EMP pulse. Electronics that are turned off would probably be ok, and devices that are unplugged certainly would be. Even devices plugged in and running would often be protected because the trans

However, cell phones are alway on radios.

You'll note this is the actual damage from StarFish prime:

"he Starfish Prime electromagnetic pulse also made those effects known to the public by causing electrical damage in Hawaii, about 1,445 kilometres (898 mi) away from the detonation point, knocking out about 300 streetlights, setting off numerous burglar alarms and damaging a telephone company microwave link. The EMP damage to the microwave link shut down telephone calls from Kauai to the other Hawaiian islands."

You'll note that the burglar alarms were tripped, not damaged. The phone system went down because it was a microwave link ie a radio system.

"A large number of United States military ships and aircraft were operating in support of Starfish Prime in the Johnston Island area and across the nearby North Pacific region."

You'll note that no airplanes crashed or had their instrumentation damaged.

The biggest result was the eventual failure of 7 satellites. But you'll note that this process took 7 months. A lot of electronic equipment would see damage, but would not immediately stop working.

"These man-made radiation belts eventually crippled a third of all satellites in low Earth orbit. Seven satellites failed over the months following the test, as radiation damaged their solar arrays or electronics, including the first commercial relay communication satellite, Telstar, as well as the United Kingdom's first satellite, Ariel 1.Detectors on Telstar, TRAAC, Injun, and Ariel 1 were used to measure distribution of the radiation produced by the tests."


Lightning strikes usually flow over the skin of what it hits, for the most part. My understanding is that an EMP uses the circuitry as an antenna and generates the current from within, bypassing some of these protections. I would thing the two are not the same. I am willing to say I may be/probably am wrong on this, but that was my understanding.

I would think simple grounding would be insufficient, that something more like a faraday cage would be needed.

"Lightning strikes usually flow over the skin of what it hits, for the most part."

A lightning strike generates a pulse of EMP when it hits. And it's far closer. EM radiation follows the square law. So, it's doubtful if any EMP pulse at the edge of the atmosphere is as strong as the lightning hitting the skin of the plane.

"I would think simple grounding would be insufficient, that something more like a faraday cage would be needed."

True to an extent, but again this seems to be an exaggerated threat. You'll note that no major power has ever actually effectively used an EMP pulse in combat. Indeed, how many people have managed to make a targetted EMP pulse generator that could stop a moving car? You'd make hundreds of millions selling such a device to police departments, since they would make high speed pursuits impossible.

The biggest EMP risk is the damage of thousands of transformers, whose lead-time is in terms of months. Less, losing power for 24 hours is not much of a problem, but losing power for a month or two will turn a city into chaos...

"The biggest EMP risk is the damage of thousands of transformers, whose lead-time is in terms of months. "

Agreed. Though again, it's indeterminate if an EMP can be strong enough to seriously damage the transformers. Regardless, that's the weak link in the system.

It would seem prudent for the US to stockpile a few hundred million dollars worth of spare transformers. If the oldest 20% of the stockpile was sold off every year and new equipment purchased to replace it, the stockpile stay up to date and the actual year to year cost would probably be less than 25% of the total capital costs. Say, $50 million per year. And the utility companies get a good deal on some 5 year old equipment that was never used.

Good luck selling that to world public opinion.

Remember, many in Europe considered 9/11 to be like a weather incident. Something unfortunate, but not "war."

But most of the world was completely on board with our response in Afghanistan. That would be the case here too. "Many in Europe" bitch about anything the US does, let them. Some even troll here daily.


Citation please. 30% of the populace in Western Europe is not most of the world. Most of the Muslim world thinks it was a Jewish conspiracy. The overwhelming majority of Afghans think it never happened, let alone due to a whacko anti-Semitic conspiracy theory. As I recall it's because "you can't make buildings that tall." Hmm.

I'd love to see a credible survey. The main issue in the third world is that people who knew what it was were much more likely to be cosmopolitan. Did the Chinese care at all? Can you reliably survey them?

It does appear the EMP attack idea has been oversold as sci-fi.

Another problem with a massive EMP attack, North Korea could not launch one without first appear to be launching an attack against the mainland US. One would assume the moment an incoming NK missile is detected, the US's MAD doctrine would launch multiple retailitory missiles at NK. Before the NK attack is even demonstrated to be a 'less violent' EMP attack NK would be destroyed.

Oh, really? You think launch-under-attack is U.S. policy? Against a single launch vehicle, which as far as you know might be a communications satellite? I don't think you understand how these things work.

You're right, I don't. I'm guessing, though, that long before the warhead reaches its endpoint, the launch will be tracked and the trajectory will clearly be plotted as not one of putting a satellite in orbit. I'm also going to guess that a bit is known about satellite launch vehicles and ICBMs which by design are not strong enough to put something in orbit so early on in the launch I'm going to guess the US will either know or become highly certain that it is an attack and not something innocent.

That makes an interesting question you raise about 'launch on site' versus launch after deliberation. A single nuclear attack would not destroy the US. If you start debating it morally, a nuclear retaliation starts looking more and more problematic. A 'launch on site' type order might have several benefits:

1. MAD: It wouldn't be crazy to think that one way the old communist bloc might try to get around MAD would be to have their secondary states launch decapitation strikes on the US that would be very small and hence not look like a full scale launch of thousands of missiles by the USSR but instead a 'rouge state' act. Perhaps the communist bloc would try to blind the US partially with a small 'rouge' attack and then use the cover to pull off a first strike on the US. Orders might have been and might remain to assume an attack by a secondary communist power on any other nation be treated as a Russian attack on the US. If the USSR knew or assumed this to be the case then that would make even trying to plot such a roundabout first strike pointless.

2. By bypassing the moral debate and assuming even a limited NK first strike on the US requires a launch on site counter-attack, you eliminate the possibility that NK might assume a first strike followed by a plea for mercy might work to give the US pause.

3. The US would probably rightly assume a first strike on the US from NK, esp. one that was an EMP attack or even a strike on a US city, would be the initial volley in more attacks plus a conventional full on war on South Korea. One of the few things Steve Bannon got right was the fact that there's no way for any conventional war in Korea not to kill hundreds of thousands to millions within the first 30 minutes or so. If you assume the first NK missile attack on the US will soon be followed by additional missiles and shelling of South Korea, the only way to bring down NK's military really, really fast would be to hit with lots of nukes.

An EMP attack by North Korea would invite a massive EMP attack on North Korea. We would blind them and leave them weakened to retaliate against America or American allies. The US could then demand the removal of all North Korean leadership. If they do not immediately comply with US demands the US will retaliate with a massive bombing and missile attack that may or may not include nuclear weapons. Within hours North Korea would be a massive grave yard or under new leadership. South Korea will have extensive infrastructure damage but loss of life would be minimized.

BTW the same hold trues for any type of attack by North Korea. EMP attack. Issue demands. A warning that any attempt to launch a counter strike by North Korea will invite the full force of the US military.

North Korea must understand that as long as they remain an annoyance we will tolerate them. If they become an aggressive threat we will destroy them.

This is insane and shows a complete lack of understanding of how nonwestern countries operate.

Also a lack of understanding of electrical needs of artillery (nonexistent), which is the main conventional threat against SK. Also a lack of understanding of military matters to an almost hysterical degree.

Let's break it down.

Leverage. The ability to nuke a US city gives NK enormous leverage. Most liberals in the US are disgusted by our participation in the Korean War and Vietnam, as I recall. And that's ancient history. Imagine what they will say if we threatened actual violence against NK (Trump is not a credible threat). Now imagine if they could wipe out San Francisco. Half the country would surrender before a single shot is fired.

That's the point. Half the country firmly believes we should only go into a shooting war if we are invaded. Now imagine what the % is if we instantly lose 2-3 major cities upon intervention. Imagine a demand curve.

The question is whether we commit to South Korea and Japan at all in the future. They run afoul of almost every intersectional liberal complaint about society. I cannot for the life of me imagine kids from Berkeley signing up to defend a Christian patriarchical heteronormative nation like South Korea. And they certainly won't trade Zachary's pizza for the freedom of millions of privileged Asians competing for their spots at Harvard and Goldman. And they will reflexively side against the southern rural white trash trump supporting sister raping hillbillies who would actually be fighting the war.

So....solve for the equilibrium. South Korea asks for reapproachment by 2050.

It's peculiar how the idea that democracies would surrender after being dealt a strong military blow keeps going on for centuries despite all evidence to the contrary.

On a different but not unrelated side note, North Korea is unable, for now, to threaten any American city, or to build staged thermonuclear bombs, which could provide the multi-megaton yield necessary for an EMP attack.

Perhaps you are correct. The US is a frail old man only able to wag a stick at those annoying kids without any ability or real desire to do any thing about their antics. Perhaps you and Osama bin Laden are correct

"America appeared so mighty ... but it was actually weak and cowardly. Look at Vietnam, look at Lebanon. Whenever soldiers start coming home in body bags, Americans panic and retreat. Such a country needs only to be confronted with two or three sharp blows, then it will flee in panic, as it always has. ... It cannot stand against warriors of faith who do not fear death."

Your view also explains how leading liberal Democrats like Hillary Clinton, Schumer, Biden, Feinstein, etc kept us out of Iraq.

Perhaps Kim Jong Un shares your view of the US. We are a paper tiger unwilling to defend our allies or even ourself. It is amazing if he doesn't start a ground war next week.

There are many unknowns in a potential conflict. A cyber attack and an EMP attack on North Korea would destroy communications, some transports, and hamper satellite communications. How the North Koreans would respond in the chaos is unclear. The role of remaining Chinese and South Korean assets in covert positions is uncertain given recent purges in North Korea. The redeployment of tactical nuclear weapons in South Korea in the near future is increasingly likely and could target artillery effectively. The deployment of EMP weapons, artillery shells and drones, is likely. How effective they will be in harming guidance systems etc is uncertain but greater then zero.


When is the last time the US voted for military authorization against a country that can strike US soil with a nuclear weapon?

When is the last time the US voted for military authorization against a country that can fire thermonuclear warhead ballistic missiles at every large city in the US?

I'll be waiting. I'm sure you have a long list. And as an aside, the casualty threshold for America surrendering is basically exponentially dropping with every war. And this is military deaths only, which are weighted about at the 0.00001 ratio with real American civilians.

Vietnam was what 60,000 dead? Although we should do it by % of population.

Korea: 37,000 dead / 169 million = 0.02 % we sign for a draw

Vietnam: 60,000 dead / 216 million = 0.02 % we agree to surrender

Iraq: 4,424 dead / 307 million = 0.0014% we "forget to negotiate" SOFA, Iraq says SOFA is in place by executive order, Obama hilariously says exec order doesn't count (hi DACA!!) Long story short we retreat in disgrace. So, surrender.

Afghanistan: 2,351 dead / 326 million = 0.0007%. Still in country but Barack formally surrendered in 2011 before ironically sending us in a surge.

Hilarious by the way. Surrendering while sending more troops into Kandahar. Only a lawyer, amirite? If only Solomon could split the baby with military force calculations with the pacifists. Did I say Solomon , i meant Obama.

In short, you have no fucking idea of what you're talking about. An apt analogy would have been Poland or Hungary. Not willing to risk NYC for Warsaw. And ask people on the street how they feel about trading LA away?

'Remember Pearl Harbor?'

Remember Hiroshima?

Remember Auschwitz?

Not especially, being American, though everyone knows it was a part of the genocidal program of mass murder carried out by the Nazis.

However, if North Korea attempts to use a nuclear weapon in a first strike utilizing surprise, it will discover several things -

1. American anti-missile technology will likely have minimal problems taking out their delivery vehicle, as at least three systems are available - https://en.wikipedia.org/wiki/United_States_national_missile_defense#Current_NMD_program

2. If the device detonates, the North Koreans will likely discover that American anti-bunker munitions, both non-nuclear and nuclear, are much more advanced than they likely planned for. In similar fashion that America's first A-bomb was actually intended to obliterate Berlin, America's bunker busting development has been ongoing for a number of years, with Natanz and associated complexes in Iran being among the intended targets. And Iran is a considerably more capable opponent, much like Germany was a more capable opponent than Japan in WWII. https://en.wikipedia.org/wiki/Massive_Ordnance_Penetrator (Even more speculative would be something along the lines of 'rods from god,' or more prosaically, 'hypervelocity rod bundles.')

3. One framework completely skipped over by Prof. Cowen involves the reaction of the other nuclear powers - all of who are likely to have absolutely zero problem with the U.S. using nuclear weapons after a first strike from North Korea. Though several would be watching such developments closely in terms of their own interests regarding the U.S. and its reaction.

Do the Germans remember Auschwitz? Of course they do. But what does that have to with the North Korean leadership being obliterated in response to a nuclear first strike? As one would expect happening, if only to retain America's credibility in the MAD world we still live in, at least in regards to Russia.

Hiroshima is an example of what happens to an opponent that strikes at the U.S. itself - we can argue whether the U.S. that crushed Imperial Japan is from a previous era, but there is no question to me that any surprise North Korean nuclear attack will show that the American military is as capable of destroying North Korean targets as it was in destroying those in Japan. Up to and including the use of nuclear weapons, though the U.S. currently has non-nuclear options in terms of making sure that the leadership that ordered such a strike is effectively destroyed.


You know, I debated writing something along the lines of 'Remember Hiroshima, which followed Pearl Harbor?,' but it seemed unnecessary. Any American, at least, should have no problem remembering what happened the last time a nation directly attacked the U.S., and how the U.S. decisively ended that war.

Hopefully, the North Koreans are a bit more aware of American history and current capabilities than you seemingly are. And while the North Koreans are still fumbling around with primitive post-WWII era technology, the U.S. has spent 7 decades refining its nuclear arsenal, along with developing defenses against nuclear warhead carrying missiles.

Congratulations on more successful no cost trolling on your part, though.

@msgkings, while I have been critical of many of prior's prior comments, I think he raises some very valid points here.

The very real possibility that the North Korean first strike fails would need to be considered in any rational calculation. If the U.S. took the nuke out of the sky, it would be a humiliation that would accomplish the opposite of what was intended. Also, there is quite a decent chance that the rocket or the warhead would not work, again making Kim look bad.

The possibility of a "decapitation" strike to take out Kim also seems excluded from Tyler's analysis. I assume this is what prior is alluding to with bunker-busting bombs. This would be a much smaller response than all-out war, but still a very real concern for Kim. Doing nothing in response to a nuclear strike would be untenable, so attempting to kill Kim seems the likely "small" reaction if the U.S. and allies are not prepared to escalate into a full nuclear war.

"@msgkings, while I have been critical of many of prior’s prior comments, I think he raises some very valid points here."

Agreed. That was a very good post, and he gets the technology and likely political reactions basically correct.

" Doing nothing in response to a nuclear strike would be untenable, so attempting to kill Kim seems the likely “small” reaction if the U.S. "

Yes, the US would launch a massive air campaign against all potential launch sites, all identified NK leadership havens, and all identified major weapons manufacturing sites. Secondly, the US would probably target any military logistic and transportation hubs. Even if the NK's have successfully hidden parts of their nuclear weapons program, there is no way to hide where the missiles are being built.

prior, you are the biggest troll here except you rag on the blog hosts who have the confidence and grace to let you do it, every day. At least show half as much spine when you get yours.

Meanwhile, I LOL'ed at you thinking you are teaching anyone anything. Remember Hiroshima? GMAFB. We are all in agreement, if NK tries anything the US will overwhelm them with nuclear force. Quit posting obvious stuff as if you are so much smarter than everyone else here and you might not get trolled as much. Your shit doesn't cost you anything either.

There can be little doubt that the results of a North Korean attack would be absolutely devasting to North Korea.

But it still would be disasterous for the world.

It is probable that the claim that North Korea can obliterate Seol with Artillary in hours is exagerated.
Regardless, large number of people would die very quickly making what has happened in the Perian Gulf look tiny by comparison.

How can simply being an American make remembering the Holocaust harder than remembering Pearl Harbour or Hiroshima (I'm assuming you weren't actually an eyewitness to any of these events)?

Prior consistently implies that Americans are ahistorical if not downright dumb. But things could improve dramatically if we were to become more like Germans.

Thor nails it, prior is the trolliest troll here


"""But if you graph the number of people killed in all 68 cities bombed in the summer of 1945, you find that Hiroshima was second in terms of civilian deaths. If you chart the number of square miles destroyed, you find that Hiroshima was fourth. If you chart the percentage of the city destroyed, Hiroshima was 17th. Hiroshima was clearly within the parameters of the conventional attacks carried out that summer."""

"""The Soviet 16th Army — 100,000 strong — launched an invasion of the southern half of Sakhalin Island. Their orders were to mop up Japanese resistance there, and then — within 10 to 14 days — be prepared to invade Hokkaido, the northernmost of Japan’s home islands. The Japanese force tasked with defending Hokkaido, the 5th Area Army, was under strength at two divisions and two brigades, and was in fortified positions on the east side of the island. The Soviet plan of attack called for an invasion of Hokkaido from the west."""

"""It didn’t take a military genius to see that, while it might be possible to fight a decisive battle against one great power invading from one direction, it would not be possible to fight off two great powers attacking from two different directions. The Soviet invasion invalidated the military’s decisive battle strategy, just as it invalidated the diplomatic strategy. At a single stroke, all of Japan’s options evaporated. The Soviet invasion was strategically decisive — it foreclosed both of Japan’s options — while the bombing of Hiroshima (which foreclosed neither) was not."""

It was better to surrender to the US than to USSR.

It is always an interesting question about what caused Imperial Japan to bow out of WWII, but there is little doubt that 1 airplane = 1 destroyed city was considerably more impressive than any number of Soviet troops.

An equation that also kept the Soviet Union from 'helpfully' occupying any further Japanese home islands.

"but there is little doubt that 1 airplane = 1 destroyed city was considerably more impressive than any number of Soviet troops."

True, and no one knew if we had used our 2 bombs or if we had a hundred more ready to go.

"It was better to surrender to the US than to USSR."

I think the millions who walked west from the blood lands of Central and Eastern Europe, ahead of the advancing Red Army felt the same way.

Remember Hiroshima?

The bombs in Hiroshima & Nagasaki were a completely unnecessary war-crimes from the point of view of the war with Japan. The bombs we used for two reasons

a) They wanted to know how they worked
b) The wanted to warn the Soviets

The BS arguments about 1 million casualties if the islands were invaded, were made after the bombs were dropped.

"The BS arguments about 1 million casualties if the islands were invaded, were made after the bombs were dropped."

That's completely wrong. My guess is that this was a random thought that crossed your mind, but, on the other hand, if you did read it somewhere, then be aware that information was bogus. It's also ridiculously easy to refute, because the Allies already had a massive organizational plan underway to rival the invasion of Normandy, at the time the Atomic bombs were dropped.

"Operation Downfall was the proposed Allied plan for the invasion of Japan near the end of World War II. The planned operation was abandoned when Japan surrendered following the atomic bombings of Hiroshima and Nagasaki, the Soviet declaration of war and the Soviet invasion of Manchuria. The operation had two parts: Operations Olympic and Coronet. Set to begin in November 1945, Operation Olympic was intended to capture the southern third of the southernmost main Japanese island, Kyūshū, with the recently captured island of Okinawa to be used as a staging area. Later, in the spring of 1946, Operation Coronet was the planned invasion of the Kantō Plain, near Tokyo, on the Japanese island of Honshu. Airbases on Kyūshū captured in Operation Olympic would allow land-based air support for Operation Coronet. If Downfall had taken place, it would have been the largest amphibious operation in history.

Japan's geography made this invasion plan quite obvious to the Japanese as well; they were able to predict the Allied invasion plans accurately and thus adjust their defensive plan, Operation Ketsugō, accordingly. The Japanese planned an all-out defense of Kyūshū, with little left in reserve for any subsequent defense operations. Casualty predictions varied widely, but were extremely high. Depending on the degree to which Japanese civilians would have resisted the invasion, estimates ran up into the millions for Allied casualties.

At the time, the development of the atomic bomb was a very closely guarded secret (not even then-Vice President Harry Truman knew of its existence until he became President), known only to a few top officials outside the Manhattan Project, and the initial planning for the invasion of Japan did not take its existence into consideration.

While the geography of Japan was known, the US military planners had to estimate the defending forces that they would face. Based on intelligence available early in 1945, "

Estimated strength of forces: US: 5 million, UK: 1 million, Japan: 4.3 million + conscripts


Utter rubbish. The War Department had over 500,000 Purple Hearts made in anticipation of the casualty counts expected for invasion of the Home Islands. This was enough purple hearts to cover all the applicable medal awards for the next 60 years of combat operations by the US military. Military planners thought they would have 500,000 American casualties so they made up that many medals. That excludes allied casualties and those of the Japanese.

And let's be honest here. Okinawa saw 50,000 American casualties and Japan saw 70,000 (up to 100,000). A single additional battle on that scale would be more deadly than an atomic bombing, let alone the mass starvation that the blockade was causing.

Or look at it from another angle, it took two nuclear bombs to convince the Japanese to call it quits. In the mean time they captured Americans and tortured false confessions out of American airmen. The Japanese leadership wrongly concluded that we would keep nuking them. Anyone who did not surrender after the firebombings was delusional. Anyone who did not surrender after the first nuclear weapon was insane. Yet Japan continued merrily along after the first bombing even though we had pamphlet bombed the nuclear targets and doing so in preparation for the second bomb.

In reality we already knew how the bombs worked. Trinity was detonated back in July and we had far better data from that test than we ever could from a hostile explosion over enemy territory.

So you have a crap theory which is actively undermine by physical evidence that does not square with any period history. Sounds about typical for ignorant internet commentary.

The tragedy is that they could have surrendered after Midway. Or Guadalcanal. They were a 20 years too early, or they would have won.

Obviously I may be wrong but Kim knows any attack on the US or any close ally of the US (S Korea, Japan, etc) and North Korea is turned to glass. We'll have our excuse to do what we want to do now but can't unprovoked. I don't think he's that crazy. I think he really is just playing the game after seeing what happened to Hussein and Qaddafi: if you have nukes, you are basically safe and don't have to use them.

Trinitite-style fused glass was produced because Trinity was on a tower only 30m tall. Hiroshima and Nagasaki were airbursts at 500m. No glass at those sites as far as I know. Air bursts would likely be used in a nuclear attack on North Korea to avoid fallout blowing to S Korea, China, and Japan.

I know, I was being all metaphorical like

Wait, you didn't mean that literally an entire country would be turned to glass? I'm so confused now.

Hussein and Qaddafi may be bad examples. Hussein was toppled and hanged partly because we thought he may have an active nuclear program, and Qaddafi gave up his program because he thought what happened to Hussein might happen to him.

"Hussein was toppled and hanged partly because we thought he may have an active nuclear program": who's 'we'?

All the major intelligence agencies in the world. Iraq was found to have been trying to buy yellowcake, than having a large stockpile of it after the war. Hussein's military leadership also thought he had a program going on.

"if you have nukes, you are basically safe and don’t have to use them."

That would be a rational thought process. But Kim continually launching Ballistic missiles into the Pacific over portions of Japan is not rational. Alarming the Japanese is not helping him.

On the contrary, Kim is being pretty rational. As Dr. Strangelove pointed out the point of having a Doomsday device is letting the world know about it. Kim needs to signal in no uncertain terms what they are building so no one gets confused like we did with Hussein. Kim knows launching missiles into the ocean shows he has the ability to attack without doing so, so we would not be justified in wiping them out.

If he actually hits an ally's city, he's done. He knows this. I believe this current escalation eventually peters out with a face saving announcement of some kind. NK's regime will never be invaded Iraq/Libya style, which was always his goal.

Yes, I agree with the basic premise. And the North Korean launch clearly tried to minimize the path over Japanese occupied islands.


""Kim needs to signal in no uncertain terms what they are building so no one gets confused like we did with Hussein. "

True, but Hussein's mistake was to pretend to have a higher capability than he had. No one doubts that Kim has nuclear warheads.

This is exactly the issue in question, and it's an old issue w.r.t. nuclear deterrence. Suppose NK nukes a Japanese city. Presumably the next thing that happens is that the skies over NK darken with incoming nuclear missiles from the US. If the Japanese government *doesn't* believe that will happen, or really if they don't believe *Kim* believes that will happen, they've got a really strong incentive to go openly nuclear themselves. Because it's *easy* to believe that Japan will nuke you back for nuking one of their cities.

The problem is what happens after Kim has nuked Japan: we know we should retaliate and destroy NK, but also that if we do, we're liable to get several of our cities nuked by NK, as well. The question is, how do we make a credible commitment that we will destroy NK at the cost of millions of American lives, after Japan has been nuked? If we can't make that commitment, then Japan has a big incentive to get their own nukes so they don't have to worry about us protecting them. And so do lots of other countries, all over the world, who are living under the US or NATO nuclear umbrellas.

If Kim tries to nuke Alaska, the US/Trump will nuke every square CM of North Korea within 30min, and the US will test its systems by trying to shoot down anything they launch in the interim. Politically it is no different than New York.

This is almost certainly true.

What makes anyone think Trump would be weak in the face of a **foreign nuclear attack**?

The entire argument here is ludicrous.

"Alternatively, say all involved parties are fully rational in the selfish sense. Fully rational agents make purely forward-looking calculations."

The hypothesis is false. Human beings are known not to be fully rational in this sense. Dan Lavatan-Jeltz's comment is basically right: any nuclear attack anywhere by North Korea will be met in the same way as a nuclear attack on an American city.

In this situation, it might be a benefit to be perceived as a possibly irrational. NK first strike is worse if Kim thinks Trump might blindly retaliate. If we think Kim is a little bit irrational, then there are some actions we won't take that we otherwise would.

A relavant econ paper is Kreps, Wilson "Reputation and imperfect information" Journal of Economic Theory, 1982.

No way. There is only one country which considers first strikes, preventive wars, which drops unnecessary nuclear bombs (in the words of their own officials, not mine!), and considers financing and training radical islamists to be "awesome" (otherwise why repeat the experience of Afganistan in Syria?) ... nuclear is an insurance policy for his regime, not an attacking weapon...

"Is the space of 'boosts his credibility and domestic stature, but without too much upping the risk of massive American retaliation' really the empty set?"

Yes, I think so. I think the attempt here to construct a rational argument for a North Korean first strike fails pretty badly.

Geopolitically, Kim already has "credibility". He is known to have nuclear weapons and everyone believes he is willing to use them, at least in self defense. Also there is already a fear that he is crazy so that everyone needs to tiptoe around him. How can his "credibility" get any higher? If he actually uses a weapon for attack, even with minimal damage, the effect on his credibility is minimal, but it greatly increases the likelihood he will get destroyed. Also, I don't think the "it's not worth retaliating" logic being suggested here applies. If he makes one attack, everyone will think it much more likely that he will make another attack, possibly with greater loss of life. An attack that harms few people might also be seem as a failed attempt to do greater damage. A first strike greatly increases the chance that Kim is toast, and as such, is irrational on the international stage.

What about his domestic grip on power? The nuclear weapons tests, coupled with threatening rhetoric, already give him domestic stature. A coup from high ranking officials seems the most likely threat to his rule, and surely risking a war that would utterly destroy the country greatly increases the chance that people are willing to oppose him.

If Kim launches a first strike, it will be because he is irrational.

Good stuff. Except, what we don't know enough about -- presumably -- is the inner workings of the Kim court.

Is he all powerful or are there competing factions, within the military, for example, that he must contend with but cannot fully dominate?

That is true, but I'm having trouble imagining a hypothetical scenario where launching a strike would help him solidify control among competing factions. One can't escape the basic calculus that a nuclear attack weakens Kim's position on the world stage. It seems like that could only harm him at home.

Yeah, getting your entire country nuked back to the Stone Age (which is exactly what would happen if Kim strikes anywhere inhabited with his nukes) definitely would "harm him at home" LOL. As in he'd be dead.

Why did he kill his own brother?

Because the biggest threat to kings are others with royal blood who could become rulers.

This suggests he's weak enough to need to off his brother.

Great analysis.

That's not the point. The point is what does he ask for now that he can murder millions. You insult this blog by not adhering to marginal thinking.

Does he ask for the removal of US troops? When he threatens SF and does a test flight how does that change the political situation in the US?

I'd guess US troops will be gone in 15 years. Why is it worth it to kids ? They don't know Korea. They're not attached to Korea. It's racist. It's not feminist. And now that we've achieved our dream of immigration, I'm probably the only one on this entire thread who had family fight that war. Why fight for Korea?

Why would Kim try to attack the US when juicy targets like Seoul and Tokyo are so much easier to hit with his weapons?

The US could have attacked DPRK at any time in the 80s, or 90s, or 00s; but we didn't because the cost of doing so wasn't worth the benefit. At those times, the cost would have "merely" been maybe a division of dead Americans, several divisions of dead ROKs, and significant destruction of Seoul. Now, a US attack would likely cost millions of dead, massive destruction of Seoul and likely a city or two in Japan.

I can't see that the allies, for which the US is supposedly protecting, would want to pay that kind of prices to destroy Kim and the DPRK.

Has Kim yet realised that if he destroyed Washington DC many around the world, and many around the US, would probably cheer? True, North Korea would then be destroyed; win-win. Except for the unfortunate South Koreans of course.

Has Kim yet realised that if he destroyed Washington DC many around the world, and many around the US, would probably cheer?

I take it neither you nor prior_approval ever get tired of being insufferable twits.

"Hundreds of thousands of innocent civilians who happen to live in Washington DC deserve to die because Trump says mean things on Twitter." -dearieme

Can you hear yourself?

Kim lost his nerve on this option two weeks ago when he decided not to bomb Guam. He had his chance and he blinked.

Now, if he had the ability to do so, it might have been clever to bomb *near* Guam, just to prove that he could, and exactly along the same lines of this argument, test the limits of whether bombing near Guam counts as hitting it. Probably not.

Trump has the same latitude though. He could also bomb something nearby, daring Kim to really escalate in a way that matters (other than just cheap talk). (Although, Trump called his bluff without the need for shooting some deserted NK island with Guam, so, no need.) This is the other aspect of 'everyone has nukes', you can stretch the definition of what counts as a retaliate-able act, because so long as you don't form the view that the attack will end his reign, then he may not choose to use them (because that really will guarantee the end of his reign).

"Trump has the same latitude though. He could also bomb something nearby, daring Kim to really escalate in a way that matters (other than just cheap talk)."

Yes exactly. What about a few pinprick hits on the Nork railway system, to aggravate the fuel shortage by a tiny bit but mainly to show we can?

Not enough to cause Seoul to be attacked though.

Why would we want to dare Kim to escalate? Why poke the bear. Right now Kim is all signaling, he's letting the world know he's got nukes and can use them. But he won't ever use them. It's a shame that the most awful regime in the world gets to just keep on truckin', but that's how it is. He and his descendants probably rule NK for decades longer at least, Castro-style, but no one is getting nuked. South Koreans understand this.

+1, this is right.

Also, I don't think the assumption that Kim "can" do something like hit Guam with a nuke right now matches the consensus about their capabilities. They have launched missiles. They have also tested nuclear weapons. But that doesn't mean they yet have the technology to deliver a nuclear weapon thousands of miles away.

All things being equal, Trump would probably prefer a quieter NK than today, even if they have nukes. Daring him to escalate can call his bluff - then you can point out just how cheap his talk is.

There's maybe two ways from here. Option 1 would be maturing the conversation: everyone (both NK and RoW) might prefer a way to stop the cheap talk (which is still irritating and hogs newspaper column inches that people might prefer pointed on other issues). If everyone can maturely say, 'yes, we get it, you have nukes', then hopefully Kim doesn't feel the need to constantly signal, and RoW can move on and deal with other things. Calling a bluff may be a way to help this along.

Option 2 may be for Trump to use this constant NK signalling as evidence of a continued and real threat, as a galvanising tool to gain support from allies to strengthen sanctions, shut down offshore NK/Kim family businesses which generate cash for the country, and otherwise pile pressure on them - while still recognising that they have nukes but they're not going to use them to defend these minor matters. Based on the UN resolution passed overnight, this appears to be the preferred strategy right now. (So long as Trump is keen on this option, calling a bluff may not be helpful, until these sanction type options are exhausted.)

The US is unlikely to do any serious damage to North Korea that might also do harm to the USS Pueblo that's still docked there.

There are near 40 million foreign born citizens in the US today.

When the NK's leader threatens the US, he's threatening the whole world. Bombs don't distinguish between US citizens, Latinos, Asians or Europeans. It would be good if other world leaders came up to clarify this point.

Um....what about China? Why hasn't anybody mentioned them yet?

Specifically, even if Kim went ahead with this scheme and it "succeeded," he then would have to deal with a China who just witnessed him court a nuclear exchange in their backyard and would presumably not be too pleased about that. The Kim regime relies on China's tacit cooperation and has to be careful not to do anything so drastic that they decide to just cut their losses and pull the plug entirely.

Trump should inform China (indirectly of course) that if North Korea does this, the US would have no problem with China taking matters into their own hands and dealing with Kim in their own way.

+1, China does seem like another significant problem with this analysis.

Undiscussed risk - Kim sells nukes to terrorists or other bad actors.

The Norks were involved with both the Iraqi program and the Syrian program. Pakistan too? I can't recall.

The ability of a puny backward economy like North Korea to build a couple of nukes and missiles with a 500 mile range shows us that any nation on earth can do the same thing. The technical barriers to entering the nuclear weapon club is low in our era. If lots of countries can do it, then some certainly will. US strategic thinking must adjust to a world where Cuba, Venezuela, and Argentina have nukes. Saudi Arabia, Egypt, and Turkey will have nuclear weapons which will certainly change our ability to influence their region of the world. None of us like North Korea's nuclear program; think about a world where 25 other nations have those abilities. That world is on our doorstep.

See Leslie Gelb and Richard Betts, "Paranoids, Pygmys, Pariahs, and Nonproliferation", published 40 years ago. They make a satisfactory case that few countries have any incentive at all to develop nuclear weapons because nuclear weapons draw heat. The only places which have aspired to do so in the last 40-odd years have faced existential threats (Israel), were facing a regional rival with such technology (Pakistan), or are run by kooks (Baathist Iraq, Islamic Iran, and North Korea).

I don't follow the logic. What does striking first accomplish from the NK side? Doesn't the US risk losing a city even without a NK first strike?

Does it just automatically escalate the stakes to nuclear and preclude a non-nuclear attack by the US? But wouldn't we already expect a nuclear response from NK for a conventional attack.

I've read the article 3 times but sure wish the logic were spelled out explicitly.

In a recent interview with Sean Illing published by Vox, Tyler estimated a 5% chance of a limited nuclear exchange in any given decade. Sounds like he's becoming more optimistic by rating this at 1% (five times more optimistic, no?).

But, in the same interview, he rated the odds that God exists at the same 5% (I'm one of the "many" who consider this absurdly high odds), so what does he know.

How about one sparrow in the DMZ? In Japan? In the Arctic? In a Malaysian airport? Or maybe one sparrow, three sled dogs, and thirty Inuit?

Always enjoyed that episode of Yes Prime Minister.
However, remember that "even though they [the DPRK, here] probably certainly know that [Trump] probably wouldn't, they don't certainly know that, although [he] probably wouldn't, there is no probability that [he] certainly would."

My biggest fear for NK would be:

1) Accidental war. They believe they are being attacked and launch.
2) Proliferation.
3) Regime death spasm attack. "If the people don't want me, I'll nuke America/Japan as a final F-U!"

The second on your list -- proliferation -- is most likely occurring already and it is also easily the most troubling.

Agreed. This is the most understated issue I think. The president being set is that of global nuclear weapons and a certainty they will be commonly used at a tactical weapon level.

But what is the solution to that kind of proliferation? The genie is sort of out of the bottle.

The worry I have is not irrationality, it is brinksmanship.

Suppose the world truly believes that Kim would never use his nukes for any reason because he then gets erased from the map in a counter strike. Okay, makes sense.

So why then would the world ever pay into any sort of blackmail? After all, we know Kim would not actually use nuclear weapons so why pay him off (e.g. with aid or diplomatic concessions)?

But Kim is also a rational actor here, so he knows we know and hence would change strategy. Classically this means that he must get rid of the 0% chance we give him of using nuclear weapons. Say by making accidental launches more possible with deadman switch firing. We then enter a bit of deterrance theory centered on risk tolerance. In theory North Korea should be more risk tolerant than us (less to lose, fewer checks on the system preventing intentional recklessness).

Playing around with risk tolerance is a much more difficult calculation, having intentionally fragile C3I or an aggressive posture leaves many places where launch decisions will be made with incomplete information by peripheral actors (like lower ranked military officers like Stanislav Petrov and Vasili Arkhipov). This is, after all, the whole point of adopting a high risk tolerant strategy.

If there is no risk of launch, then it would not be rational for Kim to pursue nuclear weapons in the first place. He and his father both had more than enough conventional hostages to make them simply superfluous. Defensively, he gains nothing, if anything his nuclear program is easier to decapitate than a chemical or biological program.

Ultimately Kim is either irrational, or it is rational for him to behave with high risk tolerance as though he were. Both have non-zero chances of a launch and given that the USSR had a few close calls with launch I do not particularly feel like waiting and hoping with the DPRK.

So, invade or nuke them preemptively now?

How high is your risk tolerance?

The Soviet Union had delegated command decisions for nuclear missile launch to individual submarine captains in the Cuban Missile Crisis with the concurrence of their political officers. One such submarine, B-59, had both its captain and its political officer concur that a lack of contact with Moscow, American depth charges, and a loss of US civilian radio signal meant that should fire nuclear torpedos.

By luck, B-59's second officer was also the flotilla commander on this sub, and this sub only, it required a three man consensus to launch nuclear weapons. Vasili Arkhipov, the flotilla commander voted nyet, and we did not suffer a nuclear launch.

American response doctrine called for massive retaliation. It is highly likely that death tolls would be in the millions.

This is what happens with brinksmanship - rapid escalation with minimal chances for the central controlling agent to reign escalation in. Kim is quite likely to adopt this type of posture intentionally. If he is not perceived as having a non-zero chance of launch by either accident or intention, he gains no leverage.

This sort of accidental launch scenario happened several times to both superpowers, and they settled into a much less fragile C3I framework with tighter control of missiles ... but neither of them were using nuclear weapons to blackmail resources from the others. Given the hostages the DPRK already has (its own population, Seoul) they will want something more than a simple survival assurance, which they have had for 50 years and has not been questioned by any of the last 4 American regimes.

If you can tolerate the risk then you can keep ignoring Kim's nukes. Trading thousands of assured dead now for potentially millions later is a dicy proposition. More worrisome is if we keep paying the blackmail and establish a precedent of paying off nuclear blackmail and induce states like Uzbekistan, Turkmenistan, Central African Republic, Armenia, Azerbaijan, Georgia, Venezeula, etc. to go down the nuclear road.

Doing nothing may be the right response, but let's be honest by "containing" the DPRK we are selling out the millions of DPRK civillians who live brutal lives and risking a North Korean nuclear launch at some odds.

It is not at all clear Kim is rational in the usual usage. Sociopaths and psychopaths are also internally rational. That doesn't mean assessing them as sane and rational as that word is often used - as I think you use it here - is smart of produces the desired results. At least result most here would think good

Is China's threat not to protect DPRK if it strikes first credible?

Is Kim's goal merely protection of his position at the top and of the regime?

How constrained is the USA response or military option by the ROM?

Of course Kim's option set with a zero meaningful response by the USA is not an empty set. Has not been as evidenced by history and current events. Has that changed now? Probably smaller but not empty.

What everyone seems to be missing it that this has been the 21st centuries Chambelaun event and very soon it will be past the point of political solutions if they really ever were a real solution.

Either Kim is rational or he's not. If he's rational, find something he wants more than nukes. If not, terrify him.

It seems to me that the rationale is quite clear.

We are not attacking North Korea, because the retaliation might be bad (Seoul). We also know the it would have been less bad, had we intervened before (nukes). We are also quite certain that it will be more bad if we wait further (icbms).

We further know that North Korea is willing to use this bargaining position to get resources. So it seems fair to extrapolate this behavior into a stronger negotiating position.

It seems to me, given these options, that this is a problem that at some point become unbearable, and that the only decision to make it when to act, where sooner means less cost.

Then people say, but even if a first strike were successful, it would be a mess of a country. This makes little sense: the only thing that would change would be the visibility and responsibility of the problem.

So, seem to me the logic is crystal clear, the logical position to take is first strike.

To Tyler's question: if NK is rational, will they retaliate. Well, actually that one is also easy. North Korea has maximum incentive to give the impression of being crazy (if rational). But if a first strike actually happen, they at that point have zero incentive to attack Seoul or anywhere else, as the value of the threat would in that scenario already be obsolete.

If they are irrational or motivated by malice, well in that scenario the argument for first strike pre-icbm is just that much larger.

Comments for this post are closed