"Russia and Ukraine Should Keep on Fighting, Not Make Peace": Making the best case for U.S. opposition to a peace treaty (Part I: selfish U.S. interests)

After I wrote, “A Ukrainian Bargaining Strategy for Trump; Alternatively, the Three-Russia Romanov Solution,” (2024), I posted a question on Scott Alexander’s blog, Astral Codex Ten, for his Mantic Monday open thread. I wrote,¹

I am very interested in what vision of the future people have who oppose any peace that doesn’t require Russia to leave 100% of the territory it has conquered. Would you like for NATO to declare war on Russia? Do you have another plan? Or is your position that a Ukrainian victory is impossible, but Ukraine should keep the war going anyway for another fifteen or twenty years, or forever?

That got a discussion of two or three hundred posts going. Three good answers were:

1. We should not make peace in Ukraine, and, in fact, should oppose any effort Ukraine makes to end the war, because we can drain Russia dry, even if Ukraine collapses and Russia wins the war. (Utopian variant: Russia will lose eventually.)

2. We should keep the war going till one side collapses, and hope it’s Russia. (Utopian variant: It will be Russia that collapses.)

3. We should let Ukraine make the decision, even if Ukraine wishes to continue for twenty years and will lose then anyway, because letting Ukraine decide is the moral thing to do.

I’ll address these answers in turn, since though they are good answers they are still wrong.

First, let’s think a bit about how wars begin and end. We game theory scholars have thought about this a lot. Why have a war, if we know who is going to lose anyway and the war will just kill a lot of people? Why not just shift the border appropriately? The scholars’ answer is that the premise is wrong: we don’t know who will win. Or, rather, the two sides don’t agree as to the probability each one will win, and the cost. There is “asymmetric information”, a fancy term for disagreement. The two countries therefore go to war and once they know more about their own and the other side’s capabilities, they make peace.²

Americans are used to winning wars, though. We haven’t had to settle for anything but complete victory since the Korean War in 1952, and even that wasn’t ended by a formal peace, just a ceasefire— formally, we’re still at war with North Korea.³ For a compromise peace, we have to go back to the War of 1812, and we even then we essentially achieved all our pre-war objectives, despite having to give up on conquering Canada. When we think of wars, we think of WW I and WW II and the two gulf wars. “Unconditional Surrender”, not compromise. But not every country gets to be the winning side all the time, and ordinarily even the winning side doesn’t get everything it wants.⁴

Before thinking about Ukraine, please think about Finland. In 1939, Stalin’s Russia, after becoming an ally of Nazi Germany attacked Finland in the Winter War. The Finnish army did marvels, completely humiliating Russia, to the cheers of Britain and France. The situation was hopeless, however, and Finland made peace. Russia could have gone ahead and conquered all of Finland, which 20 years before had been a Russian province, but it settled for just part of Finland, Karelia, which Russia retains to this day. Were the Finns wrong to make peace with Stalin?⁵

Now let’s go on to the arguments for continuing the war in Ukraine.

---

(1) We should not make peace in Ukraine, and, in fact, should oppose any effort Ukraine makes to end the war because we can drain Russia dry, even if Ukraine collapses and Russia wins the war. (Utopian variant: Russia will lose eventually.)

Wrecking the Russian army is indeed a good reason for supporting Ukraine, as I explained towards the beginning of the war in “Good and Bad Reasons Why the United States Should Favor a Ukrainian Victory” (2022). Russia has lost much military equipment and many trained soldiers. The Russian army has been humiliated and many of its generals have been shown up as incompetents, which means that the next time an opportunity arises for a war, the generals will be very reluctant to risk their good names and their jobs. Also, the Russian army has been tied up in Ukraine and is too busy to do mischief in the Caucasus or to invade the Baltics.

I agree with that argument for the first year of the war, which was the most damaging to the Russian military. I think it is the real argument of most thinking people, and also of the European leaders and the State Department, rather than the moral arguments that Ukraine is democratic and virtuous or that it has a more legitimate claim to the territory.

The argument has two big problems. The smaller of the two is that it is obsolete. We are not in the first year of the war any more. By now we have already torn up the Russian army and navy and shown that we would resist if Russia invaded the Baltics. Russia will be deterred. Ex post, the war was a mistake for Russia. Russia has gained territory, but at too high a price. We saw that Russia might have been able to take Estonia by coup de main in 2021; we have now taken precautions against that. Russia has been taught a lesson, and it is time now to make peace, rather than solidify the alliance of Russia, China, North Korea, and Iran.⁶

Also, after the disastrous first year, Russia learned the military lessons of the war and started improving its army, rather like how Ukraine learned its lesson in 2014 when the Crimea was siezed and started improving its army with American aid. Russia also learned how to grow the size of its military supply economy, how to manage civilian discontent, and how to flourish under sanctions. The war is still extremely costly, but it’s improving Russia’s military capability now, not degrading it.

The second and more important argument against the United States and Europe encouraging Ukraine to keep fighting is that it is immoral. We gain from the fighting in Ukraine, but Ukraine pays a heavy price (as does Russia). It was wrong of the US and UK to block the peace negotiations between Russia and Ukraine after the war had stalled later in 2022.

The United States does not always act only in our own interest, and we should not. For example, someone on Twitter noted that the US has more than once informed Germany of an impending Islamist massacre and prevented it. That is against American interests. If we kept silent and the massacres occurred, public opinion in Germany would become more anti-Islamist and more pro-American. Nonetheless, we should continue to warn about impending massacres.

---

(2) We should keep the war going till one side collapses, and hope it’s Russia. (Utopian variant: It will be Russia that collapses, 100% certain.)

John Schilling said:

"If the Russians break first, great. They will eventually rebuild their army, but it will take many years and probably a new President without so much of an appetite for revanchism and conquest. If it is the Ukrainian army that breaks, the consequences for the Ukrainian people will be exceedingly grave, and we don't want that."

This is an interesting argument, one worth discussing.⁷ It is a risk-loving strategy, based on World War I. That war was stalemated. After about 2 years in the war, the Austrian army broke (but Germany propped it up) and the Italians broke after losing a big battle (but France and the UK propped it up). After 3 years, the French army broke, in mutiny (but they covered up the mutiny, did some reforms [Petain], and came back, with the Brits and the Americans to help). After 3 years, Russia broke, and they were knocked out of the war. After 4 years, Germany broke, and they surrendered. The British endured, and the Americans came in too late to wear out.

It is quite plausible that Ukraine will collapse soon. It is also plausible, though much less likely, that Russia will collapse, or at least that its will to fight will collapse. Certainly after 15 years, we might expect that. For Russia, in particular, though, this depends on a certain stupidity. The WW I powers that broke, broke because they kept doing offensives. The Russian offensives were even successful sometimes-- the Brusilov offensive. Germany was doing well until it went on the offensive in 1918. Churchill's WW1 history makes a big deal of this. So if Russia decides to just sit in place, it can probably last there forever.

But suppose it's 50-50 whether Russia breaks first, or Ukraine. We indeed have a 50% chance of total victory then--Ukraine takes back even the Crimea, as Russia collapses in revolution and disorder. This is how the country of Ukraine in 1918 was created (with German help). But we also have a 50% chance of total defeat, which by now means all of Ukraine is annexed to Russia. The potential loss is much greater than the potential gain--- and I haven't even mentioned yet the potential loss from there being a new Russian Civil War in a country with nuclear weapons in places each of the multiple sides could get at.

The utopian version of this argument is that Russia will collapse first. Russia doesn’t think so, and it is improbable. Ukraine is the smaller, poorer, country; the war is being fought on its soil; its regime is less stable.

---

(3) We should let Ukraine make the decision, even if Ukraine wishes to continue for twenty years and will lose then anyway, because letting Ukraine decide is the moral thing to do.

This is especially true because we owe Ukraine a moral debt, since we violated our promise to defend them and to give them buckets of money in exchange for their giving up nuclear weapons.

Argument (3) is much more complex than it looks, and we’ll have to address the complexity in pieces. First, I’ll talk about the moral debt, which has the complexities that (a) we also violated our promises to Russia, and (b) we need to think about whether keeping its nukes would really have made any difference. Then, we need to think about what it means to “let Ukraine make the decision”. I don’t think we mean “let Mr. Zelensky make the decision, even if nobody else in Ukraine agrees with him.” Even if you do think that, there is the question of what Mr. Zelensky really wants, as opposed to what he publicly has been saying— he might just be going along with the US and NATO while really wanting peace. Next, we have to think about whether we want Ukraine to make the decision even if what it would like to do is to extend the war enough (say, by nuking the Kremlin) to make the US send troops and use nuclear weapons, a move which is risky but would be Ukraine’s best chance for an outright win. And, finally, we get to most interesting question philosophically: if the Ukrainian people truly want to continue the war for twenty years even knowing they are going to lose, should we let them do that, or act paternalistically and force them to make peace for their own sake?

I have delayed publishing this too long, so I won’t address those ideas now. I’ll try to address them in a second part on Monday of Christmas Week. Comments welcomed, in the meantime.

Footnotes

1. Astral Codex Ten and Mantic Monday sound kooky, but this is in fact a well-regarded “rationalist” “rational altruism” blog by one of the smartest people around, Scott Alexander. It is one of only two blogs I know— Steve Sailer’s being the other— which has a big enough fraction of good comments for the comment section to be worth reading. (Volokh Conspiracy, in contrast, has some excellent comments, but they are so diluted by mediocre, stupid, or rude comments that the comment section is not worth reading.)

2. There’s a lot more to it, of course. One other consideration, a huge one, is that there are people, not “countries”. Everybody in the German Army knew in August 1944 that Germany was going to lose, but Hitler, recalling Frederick the Great’s amazing survival in 1762, kept his hopes up till his Battle of the Bulge offensive failed in January 1945. After that he knew Germany would lose, but he preferred a Twilight of the Gods Gotterdammerung ending to being shot by a firing squad. And he still knew how to control his subordinates.

3. What about Vietnam? We won that war. We entered it to defeat the Viet Cong, and after the 1968 Tet Offensive, the Viet Cong were crushed. The North Vietnamese army remained, but its 1972 offensive was soundly defeated by the South Vietnamese army and the US Air Force (the US troops had pretty much all left by then). Not a single provincial capital was in Communist hands. What happened then was that in 1974 the North Vietnamese tried again, South Vietnam started to crumble, the U.S. Congress refused to let the US Air Force help this time, and the crumbling quickly accelerated into collapse.

4. We actually cheated, quietly, with “unconditional surrender” in the case of Japan. We let Hirohito remain Emperor.

5. In 1941, Germany backstabbed Russia and invaded it. Finland joined in and retook its lost territory, but prudently stopped there. When the tide turned, Finland made peace with Russia again, in 1944 I think, and gave the lost territory back to Russia again.

6. I see people arguing the Domino Theory that supported the Vietnam War. It’s a bit humorous, because that used to be a conservative’s argument. In the case of the Vietnam War, the argument was that if North Vietnam took over South Vietnam, that would not be very bad in itself, but the next domino to fall would be Laos, and then Cambodia, and Siam, and Malaya, and so forth. And, of course, Laos and Cambodia did fall, but Siam held firm. Partly that was because China and North Vietnam got into a little war that prevented expansion to the east, partly it was because the United States did seem as if it might defend Siam.

The argument doesn’t apply to Ukraine. Russia has had enough trouble conquering a small fraction of Ukraine that it won’t be tempted after a peace agreement to try to conquer the rest, much less to go on to Moldava, Estonia, and Poland. Note how this differentiates a Ukrainian peace from the Munich Agreement too. At Munich, Hitler agreed not to invade Czechoslovakia in return for being given its frontier regions, which were German populated and which also crucially, contained all the fortifications and mountain passes. Thus Hitler got something for nothing, and when he moved to take over the entire country the next year, it was unable to resist. After a 2024 peace agreement, Ukraine would be more able to resist than it was in 2022, not less.

7. The 50-50 breakdown argument is similar to the War of Attrition from game theory. See Chapter 3 of my Games and Information. In a War of Attrition, each player has the strategy of Fight or Exit in each of many periods. If both Fight, each gets a negative payoff that period. If both Exit, both get a zero payoff. If one Fights and the other Exits, the one who Fights wins and gets a positive payoff, while the one who Exits gets zero. In the unique equilibrium, they each randomize whether to Fight or Leave, and the expected payoff is zero because on average the game will continue so long that on average the sum of the payoffs is zero for each player. For example, suppose (Stay, Stay) by the two players yields (-1,-1) and (Stay, Leave) yields (10, 0), the winner getting a prize of 10. The expected payoff from each player playing a probability X of Exit is computed from

Payoff (Exit) = X (0) + (1-X)(0) = 0 = Payoff (Stay) = X(10) + (1-X)(-1) = 10X -1 + X = 11X -10, so 1 = 11X and X = 1/11 = .090909. . .

Thus, each player choose Exit with about 9% probability— which means they each choose Stay with 91% probability. The war will last N years with probability .91^N. With the aid of ChatGPT, we can calculate that for different values of N. We would guess the war would last 7 years from the table below. If it did, and Russia was the country that chose Exit in year 7, Russia’s payoff would be -6 (= 6[-1] + 0) and Ukraine’s would be 4 (= 6[-1]+ 10). The sum of their payoffs is

Modelling an asymmetric War of Attrition is tricky, because if, say, the per-year payoffs from (Stay, Stay) were (-1, -2) for Russia and Ukraine. And suppose that (Exit, Exit) does not have a (0,0) payoff, but splits the prize, but (5,5). I don’t have time to do it here, but I will in a future Substack, and perhaps make a paper out of it, since it applies to the enforcement of much of litigation in courts.