Key points for effective altruists
The idea of Ukrainian sovereignty over Crimea is supported by international law and the desirability of expanding the influence of the EU and NATO. However, it violates the wishes of most people actually living in Crimea, it may prolong the Russo-Ukrainian War and it poses a significant risk of nuclear escalation. Comparison of these factors against each other suggests that the costs and risks of invading Crimea outweigh the benefits.
For the general EA community, the only action I recommend to achieve this goal is to spread messaging similar to what I write in the section below:
Key points for non-effective-altruists
A Ukrainian invasion of Crimea would be militarily costly and would raise grave risks of nuclear escalation, as Crimea may be a "red line" that Putin will not allow Ukraine to cross. If Ukrainian troops do manage to enter Crimea, they won't find themselves greeted as liberators, for reputable polling suggests that the vast majority of Crimeans prefer to be part of Russia. If Ukraine does manage to control Crimea, it will be stuck governing a territory full of anti-EU, pro-Russia malcontents over whom Russia may continue to assert some level of responsibility or exploit as a means for political interference. This will undermine the prospects for Ukraine's postwar integration into the EU and NATO. Without Crimea, Ukraine will have less territory, but also fewer obligations and burdens. A Ukraine that is more at peace with itself and more unified in support of Western liberalism will lead to a happier, healthier Europe. While Ukraine does have legitimate grievances regarding Russia's illegal actions of 2014, militarily invading Crimea would be a costly, risky and inappropriate remedy.
Basic situation
Prior to 2014, Crimea was an autonomous region within Ukraine, but some Crimeans wished to secede from Ukraine and join Russia. In 2014, the Ukrainian president was ousted in a violent uprising. Among other things, the new president favored greater integration with the European Union, which would lead Russia to close many of its ties with Ukraine. This drove many Crimeans, who favored a pro-Russian policy, to fiercely demand secession. While Ukrainian state services attempted to suppress secession efforts, a combination of Crimeans and Russian military personnel seized government buildings, installed a new government and evicted the Ukrainian military. They then held a referendum in which the majority apparently voted in favor of leaving Ukraine to join Russia; outside of Russia, this referendum was rejected as illegitimate. Since 2014 Crimea has been a de facto part of Russia, but Ukraine and nearly every other country continue to assert that Crimea legally belongs to Ukraine.
So far, Russia's 2022 invasion of Ukraine has not altered the basic situation for Crimea as the fighting is still confined to other parts of Ukraine, but things could change. The dominant opinion in Ukraine and in the countries which supply it with military aid (including America) is that Ukraine ought to regain control of Crimea, at least in theory. This is especially true among foreign policy elites who make the relevant decisions. If you've ever heard someone talk about restoring Ukrainian "territorial integrity" or "sovereignty", what they mean in practice is that they want to help Ukraine capture all of their former territory, including Crimea, by force if necessary. And if Ukraine is very successful in the war then it might be able to invade and retake Crimea. In November 2022, Ukrainian President Zelenskiy said there would be no peace until Ukraine got Crimea and the Donbas back.
Note that Crimea as a territory includes both the Autonomous Republic of Crimea and the city of Sevastopol; when I say "Crimea" I refer to both.
Arguments for the Ukrainian side
Russian annexation of Crimea violated the law
As I described here, Russia's 2014 annexation of Crimea violated Ukrainian law, Russian-Ukrainian treaties, and broader international law.
However, this argument is not a very strong justification for Ukraine to retake Crimea. International law is composed merely of the official positions of different countries, sometimes signed explicitly and sometimes maintained implicitly, and it can be made as easily as it is unmade. If countries agreed that Crimea is indeed part of Russia for <reasons> then that would more-or-less become the new international law.
So while the legal argument for Ukrainian control is meaningful (I am no legal scholar and am not at all confident that absolutely no laws would be violated by official recognition of Russian sovereignty over Crimea), it is weaker than it appears at first blush.
Ukraine is a better government than Russia
As I described here in 2021, cross-country indices say that Ukraine is substantially superior to Russia in democracy and civic freedom, though it also has slightly less economic freedom. However, that doesn't necessarily mean Ukraine would be a better government of Crimea. The Ukrainian government might give inferior treatment to Crimea, as many Crimeans seem to believe and as is often the case for minority regions like Crimea. Additionally, it stands to reason that Crimeans (who tend to be very historically and culturally tied to Russia) get more value from ties with Russia and the EEU as opposed to ties with the EU, even as mainland Ukrainians tend to see more value in Western ties.
Since the 2022 invasion we have learned much about the conduct of the Ukrainian and Russian states in wartime. In general, both Ukraine and Russia appear to be growing less free and liberal. This is normal in wartime for any country, and doesn't change the fact that Ukraine remains comparatively better than Russia. If anything, the greater degree of Western influence in Ukraine since the war started can be expected to put them on a better long run trajectory, whereas I don't feel hope for Russia to liberalize (domestic criticism of Putin's handling of the war seems to accuse him of being too soft and weak more often than it accuses him of being too cruel).
In addition, there have been massive systematic Russian war crimes, much more than those done by the Ukrainian side. I do believe it is fundamentally illegitimate for a government to rule over people which it has recently subjected to large-scale atrocities. However, the Crimeans do not seem to have been targets of any of this. And to be fair, we haven't seen much of how the Ukrainian military might behave if they were to invade separatist regions. A fairer comparison can be made with treatment of prisoners of war, or discrimination in bombing military rather than civilian targets; in these cases the Ukrainians still seem to be behaving much better. But the relevant question is not whether Ukrainians treat Russians better than Russians treat Ukrainians, it's whether Russians treat Crimeans better than Ukrainians treat Crimeans. While the Russian military is more nasty in general, they may be more benign towards Crimeans, who they may see as their allies and compatriots.
Overall, if I were to be a Crimean citizen, I would surely vote to be a part of Ukraine, as the idea of integrating with the EU (and potentially NATO) sounds vastly better than joining the morass of Russian corruption and international meddling, even if that means paying the price of losing some trade agreements and jobs in the short term. After following politics in other places (Armenia and Syria), I am honestly baffled that so many people in Crimea, who actually had a choice about whether to accept the Russian sphere of influence, would welcome it with open arms. And I feel like the suffering and destruction which Russia's invasion is causing even for pro-Russian separatists in Ukraine serves to vindicate my view. But I accept that my perspective comes from a different culture thousands of miles away, and I haven't found good sources to inform me about governance in Crimea. So I could be wrong here. Ukrainian governance does seem superior for Crimea, but not obviously or greatly superior.
Invading Crimea would weaken Russia
It's not always said out loud (but sometimes it is), people may support Ukrainian military operations not just for the benefit of Ukraine but also because they simply want to bleed the men and materiel of Russia's military. Seizing Crimea would also weaken Russia's strategic position, as it would undermine their ability to project military power in the Black Sea region. It could also lead to the downfall of Putin, which some people hope would lead to liberalization, but more likely would just lead to political turmoil. In the grand strategy of NATO versus the Commonwealth of Independent States, successful Ukrainian counteroffensives certainly advantage the West even though Ukraine is not part of NATO. Finally, Russian short-term problems in Ukraine may take away resources from their strategic forces, including their nuclear arsenal. Even as WW3 between NATO and Russia is almost certainly not going to happen, a weaker Russia will have less political leverage and be less likely to win smaller wars like the current one in Ukraine.
While I do favor the West in this competition, this argument in favor of invading Crimea is tempered by the fact that Russia is already in such a poor position. On top of Russia's basic economic inferiority to the NATO powers, their military is less competent and worse equipped (as shown by recent operations in Ukraine), and things will only get worse for them if Ukraine continues to evict the invaders from the territory they invaded in 2022. Finland and Sweden's movement toward NATO makes Russia's strategic situation all the more dire. Even if America is completely tied down (by domestic issues or by a war with China), the European countries of NATO could handily outpower Russia in conventional forces while also wielding a minimal nuclear deterrent. And with the European situation so bad for Russia, their ability to control events in other regions like the Caucasus, central Asia, the Middle East and Africa has been undermined.
The problem is that if Russia becomes too weak then bad things may start to happen. First of all, there are five vulnerable countries outside Ukraine which are threatened by their neighbors but whose security is guaranteed only by Russia: Transnistria, South Ossetia, Abkhazia, Armenia and Artsakh. In all of these cases, Russian weakness may allow Moldova, Georgia, Georgia, Azerbaijan and Azerbaijan respectively to invade these countries (very much against the wishes of the people living there). This has already begun in Armenia and Artsakh; an Azerbaijani blockade threatens to inflict genocide of the entire population of Artsakh. In none of these cases is the international community willing to provide security for the vulnerable unrecognized states. Secondly, in Syria, Mali and perhaps other places, Russia is a negative influence but a rapid Russian withdrawal might lead to even more problems, like how the rapid collapse of the USSR and Yugoslavia created power vacuums which led to many vicious small conflicts. More generally, it's probably not healthy for any country to become too vulnerable. NATO is generally better than Russia, but maybe if NATO had too much power over Russia then they would start to abuse it.
I definitely don't think that Russia is currently too weak but if we assume the hypothetical scenario where the Russian front in Ukraine is collapsing and Ukrainian forces are pouring into Crimea - the scenario which is relevant for this discussion about the future of Crimea - then the idea of Russia becoming too weak is a more serious risk. If Ukraine were to restore its February 2022 borders, that would already be a serious setback for Russia, and if Ukraine were to seize all of the Donbas then that would be a categorical defeat for Russia; I see little need to add Crimea on top of it.
Overall, undermining Russia is a sound argument in favor of invading Crimea, but it's a rather weak one.
Arguments for the Russian side
Most Crimeans want to be part of Russia
As I described here in 2021, a large majority of Crimeans preferred being part of Russia. And I've seen no indication that many have changed their minds since the start of the 2022 war. Wikipedia's articles on wartime violent resistance and protests behind Russian lines show no examples in Crimea. Russia's war crimes in Ukraine don't seem to include anything in Crimea which might turn the population against Russia. Crimeans might feel disappointed in Russia for starting a war that makes life harder for them, but they could also feel grateful for restoring the water supply. Moreover, Crimeans could feel much more aggrieved at Ukraine for fighting back and bombing several targets in Crimea. People don't like going to war, but they especially don't like being ruled by the people against whom they fought a war. So it definitely appears that a substantial majority of Crimeans want to remain in Russia. This is relevant for the principles of democracy and self-determination, which are not only intrinsically important for many nonconsequentialist views, but also serve as guides to good consequences in the long run.
Relinquishing Crimea would shorten the war
By removing one casus belli, the war can end more quickly. If Ukraine is willing to concede Crimea in a peace deal then Russia may be persuaded to end the fighting. Ending the war sooner would save large numbers of soldiers on both sides and civilians in Ukraine, as well as reducing the various economic and social harms of the war, including the disruption of grain supplies to Africa.
To be clear, I don't mean that Ukraine should change its strategy by implementing a ceasefire and negotiating instead; that would likely backfire, and I suspect that more than Crimea must be sacrificed if there is to be a diplomatic end to the war at this current time. I just mean that however this war comes to a close, it can happen sooner if Ukraine agrees to relinquish Crimea.
If Ukraine pursues this war to the bloody completion of its objectives without ever making a deal with Russia, it will happen faster with less loss of life if the Ukrainian columns stop short of invading Crimea.
If Ukraine achieves continued battlefield successes which convince Russia that it must make concessions in a peace deal, then the one Ukrainian concession of Crimea would entice Russia to come to the negotiating table more quickly.
And if Russia turns the situation around and consolidates control over a large swathe of Ukrainian territory, forcing the Ukrainians to eventually accept some kind of defeat, the Ukrainians will accept it more readily if they have already come to terms with giving up Crimea.
Invading Crimea would risk military escalation
Crimea has more historical, cultural, and strategic importance for Russia compared to other territories which Russia has seized from Ukraine. It stands to reason that Russia will go to greater lengths to defend Crimea than it will for other areas. This could mean a generally increased Russian war effort, which will either turn the tide or force the Crimeans to fight harder as well, thus increasing casualties from the war.
Escalation might even include nuclear weapons. While Putin's nuclear saber-rattling over Western aid to Ukraine has been rightly dismissed, that doesn't mean that he will not use nuclear weapons to defend Crimea. It's a different scenario with different stakes.
According to Metaculus's community prediction, the probability of a Russian nuclear detonation in Ukraine rises from 5% to 19% if Ukrainian forces control at least 50 square kilometers of Crimea for at least a week, but both these figures seem a bit high to me, and they don't make sense considering that the overall prediction for nuclear detonation is only 3%. And of course, the invasion of Crimea might be a mere correlate rather than a cause of nuclear war risk. So let's conservatively say that invading Crimea poses a 5% additional risk of some kind of nuclear war (probably not WW3 or even particularly devastating to Ukraine). And if there is at least one nuclear detonation, there is a 20% chance that at least one is used on Kyiv, so let's say there's a 1% additional risk of a detonation in Kyiv. These figures are not contradicted by any of the foreign policy commentary that I have seen.
That is still quite bad. If I lived in Ukraine I don't think I'd accept a 1% risk of my capital getting nuked just so that my country could get a little bigger. And while I'm confident in America and Russia's abilities to probably avoid mutual nuclear escalation, it seems reasonable to suppose that invading Crimea would increase the risk of doomsday by at least 1-in-1000, and that is a risk which should give meaningful pause to those who advocate invading Crimea.
A common retort is that paying heed to these risks is giving in to "nuclear blackmail" and if we start here, there's no telling where it'll end. But in the harsh reality of an anarchic international system, a little bit of leeway for nuclear blackmail may be a necessary lubricant to allow powers to sort out their issues without massive warfare. Moreover, fears of widespread nuclear blackmail imply that it is practical, effective, but restrained by norms, when in reality using nuclear weapons to extort other states is difficult and risky regardless of what the international community says. Situations like this are so special and complicated that they are decided more by calculations and incentives than by precedent. In general I find that when foreign policy commentators make speculative claims about how events on one side of the world will set precedents changing what happens on the other side in another decade, they are usually wrong. America withdrawing from Afghanistan didn't cause lots of Islamic militants to start new campaigns, NATO not sending its armies into Ukraine didn't cause lots of states to start new interstate wars, and Ukraine making concessions to end a war won't cause lots of nuclear powers to start using nuclear blackmail.
Overrated issues which can be somewhat ignored
How would this impact mainland Ukraine and Russia?
Whoever owns Crimea will raise some tax revenue from the peninsula, but also have to spend money on programs for it. If Crimea is part of Ukraine then traveling between mainland Ukraine and Crimea will be easy, but if Crimea is part of Russia then traveling between mainland Russia and Crimea will be easy. In the modern-day real world, having a bigger country is not as beneficial as it was in the 19th century or in a game like Victoria 3, although not everyone seems to have noticed this.
Ukraine's attempt to forcibly retain Crimea contributed to the problems they have faced since 2014, and even if they completely pacify Crimea they will still have a territory full of anti-EU, pro-Russia voters. So I suspect that the Ukrainians who live on the mainland will have better lives if they simply don't bother. I would compare this case to Georgia, where "like the amputation of diseased limbs, the loss of Abkhazia and South Ossetia restored some health to the Georgian body politic".
What precedent would this set?
A common argument is that allowing self-determination in one country will set a precedent in favor of self-determination elsewhere, so there will be a lot more fervor for secession and then the world will become all goofy and chaotic with separatists everywhere. If Crimea gets to change countries, then what about Catalonia, and the Russian-speaking part of Estonia, and the Xinjiang Uygur Autonomous Region - etc. To some people, the prospect of all these places fighting for separatism is a nightmare.
Note that I've been pretty vague about the details beyond 'not invading Crimea'. Ukraine could refrain from invading while still claiming sovereignty over Crimea, Ukraine could fully grant sovereignty to Crimea in a peace deal, or Ukraine could accept an internationally-observed binding referendum on the status of Crimea. Each of these could set a precedent in different ways, but let me make some general points.
First of all, we need to recognize that good actions generally set good precedents. In situations very similar to that of Crimea - a breakaway territory with majority support for secession, currently safely part of its preferred country but threatened by its revanchist former country, and serving as a rationale for an ongoing bloody war - self determination is actually good, so the precedent would actually be a good thing. If you think that Crimea in theory ought to go to Russia then you should recognize that in similar cases, breakaway territories ought to be allowed to secede as well. Of course, if you disagree about Crimea then you probably won't like the precedent. My point is simply that the question of precedent doesn't change much from the immediate question of what's best in this particular situation. I think that in most cases of international conflict, instead of making whatever decisions we think will set the best precedent, we should simply do what makes the most sense in the situations that we immediately understand, and trust that good practices will also lead to good precedents.
In situations different from the current one, the precedent simply might not apply. It's all a question of legal (mis)interpretation. No doubt some people will try to apply the precedent to different situations - for instance, the independence of Kosovo was quite incorrectly used by Russia as a legal precedent to justify their annexation of Crimea. But just because someone makes such an argument doesn't mean it actually persuades anybody or motivates their behavior. I don't think that if Kosovo had not become independent then Russia would not have annexed Crimea. Whenever I look into conflicts like this one, I find more and more evidence that they are ultimately caused by local domestic factors, not so much by great power politics and certainly not by people who truly care much about international law.
Anyway, I do agree that some kind of weak-but-nontrivial precedent broadly in favor of self-determination would be set by relinquishing Crimea, but that isn't a bad thing. In my opinion, self determination is broadly good so that would be a good thing - but for the sake of argument to the skeptical, let's just say that it is unclear.
Dismissing two bad arguments
"Crimea belonged to Russia back in the old days, but in 1954..." I don't care, be quiet. Who used to control a territory generations in the past has no direct bearing on who ought to control it now.
"If you don't live in Ukraine then you can't judge this topic - the Ukrainian people should decide this for themselves." Look at the sick irony of pretending to support the democratic voice when you think the status of Crimea should be decided by Ukrainians. I don't live in Crimea, but nor does anyone else who lives in the current de facto borders of Ukraine, so their opinions are not really more legitimate than mine or the Russians'. If you actually believe in the principle of self determination, that the people of <place> should be allowed to decide things for themselves, then you should agree with me that the Crimean people should be allowed to decide things for themselves via referendum. If not then don't pretend to echo the voice of the people. In any case, Ukraine will act how it chooses, but it is absolutely morally legitimate for people in America, Europe and South Korea to question the conditions under which our countries send military aid to Ukraine.
Weighing up the arguments
There are three arguments in favor of Ukraine not invading Crimea: first, that a large majority of Crimeans want to be part of Russia; second, that conceding Russian demands in Crimea is likely to shorten the war; and third, that attempts to regain Crimea may escalate the war to the point of nuclear attacks.
Effective altruists should agree that these are all pretty important. I suspect that most learned people in our community would rank nuclear risk relatively high and self-determination somewhat low. Meanwhile, the scholarly and political mainstream is perhaps not quite so worried about nuclear risk and hardly seems to even care about self-determination. Personally, I would rank self-determination as most important.
Meanwhile, the three arguments in favor of a Ukrainian invasion are that Russia's annexation of Crimea was illegal, that invading Crimea would weaken Russia, and that Ukraine is better at governing than Russia is.
To weigh pros against cons, it's easiest to start with the question of governance versus popular will. Most Crimeans want to be in Russia, but us enlightened Westerners think it would be better to be part of Ukraine - how to resolve this dilemma? I would say that if you are the kind of person who prefers actual democracy to attempting a benevolent dictatorship, who believes in respecting the outcomes of elections even when we don't like the winners - as a moral principle, as an antidote to hubris, or as a mechanism to better governance in the long run - then you should prioritize the will of Crimeans over the apparent superiority of Ukraine's government. At minimum we should agree that respecting popular will is either equally or more important compared to our perception of which side is better.
Next, let's compare the benefits of having a shorter war against the loss of the opportunity to weaken Russia. Here we should obviously favor having a shorter war - if you disagree then you want war purely for the sake of weakening Russia, which is not something that normal people believe. The only reason to want to weaken Russia in the first place is so that they can more easily be defeated in some hypothetical future war, and of course a real current war is a bigger problem than a hypothetical future one. So we should agree that in this matchup, weakening Russia is less important than shortening the war.
Finally, let's compare the benefits of reinforcing international law against the risk of nuclear war. So, how many civilians would you allow to be vaporized in order to satisfy lawyers? I guessed a 5% probability of some kind of nuclear detonation and 1% of detonation over Kyiv, which in turn can be taken to suggest perhaps a 1-in-250 risk of widespread nuking of Ukrainian cities and a 1-in-1000 risk of international doomsday. Widespread use of nuclear weapons against Ukrainian military targets away from cities is also a possibility, a more likely one than a detonation over Kyiv, and still very worrisome even if it wouldn't cause as many casualties.
You might say that it would not be a wise policy to break international law every time it seemed like a way to reduce nuclear risk. But as stated previously, this is mostly not a question of breaking international law in order to risk nuclear war risk - it's more of a sacrifice of the opportunity to shore up the international law that was already broken. Furthermore, if nuclear war does happen, is that not itself a violation of international laws or at least norms which are far more important than stuff like the Budapest Memorandum? Admittedly, a nuclear war probably won't happen even if Ukraine invades Crimea, but even from a purely legal standpoint it seems like a severe event. And then consider the severe humanitarian and other consequences.
It should be noted that the EA community has largely reached a consensus that nuclear risk is an important cause area, but generically upholding international law is not a cause area that we consider much at all.
Of course some international laws are very important. But when reinforcing international law leads to problems such as the current one in Crimea, we ought to consider whether these particular international laws in favor of territorial integrity are actually good in the first place. I am almost in favor of actively undermining these international laws, but you don't need to agree with me on that to recognize that these international laws are at least not so wonderful that they justify risking nuclear war.
To summarize. Governance versus popular will: either no points, or points against invading Crimea. War duration versus weakening Russia: points against invading Crimea. International law versus nuclear risk: points against invading Crimea. So not invading Crimea is the dominant choice (i.e. better no matter how you prioritize these various factors).
To revise this with cardinal scoring and a little more margin for disagreement, I would say that governance versus popular will counts as somewhere between -3 and 1 points for invading Crimea, war duration versus weakening Russia counts as somewhere between -4 and 0 points for invading Crimea, and international law versus nuclear risk counts as somewhere between -5 and 1 point for invading Crimea. Best case for invading Crimea is 2 points in favor; worst case is 12 points against. Our uncertainties about these scores don't follow strong skews in favor of invading Ukraine, so we can easily conclude that the expected value is negative.
In conclusion, no, Ukraine should not invade Crimea.
Implications
Crimea deserves to be part of Russia, and Ukraine should not launch a major military operation to seize control of the peninsula.
However, there are many other issues at stake in the current Russo-Ukrainian War, and on these other issues Ukraine is mostly the correct side. Even when this war effort involves striking targets within Crimea, that is completely legitimate, just as it is legitimate for Ukraine to strike targets in mainland Russia. And even if Ukraine does overrun Crimea, the good of liberating other Ukrainian territory would outweigh the bad of taking over Crimea. Therefore, support for the Ukrainian war effort should continue, with lethal military aid.
Given that Ukrainians mostly don't want to relinquish their claim to Crimea, countries such as the United States can influence Ukraine in various ways. First, through diplomatic backchannels, they can advise Ukrainian officials against invading Crimea. Second, they can place export controls on arms sent to Ukraine, with the idea that they are not allowed to be used in support of an invasion of Crimea. Third, they can say that military aid will be curtailed or cut off if Ukraine invades Crimea. Fourth, they can immediately give diplomatic recognition to Crimea being a part of Russia, which a handful of small countries have already done; this may make sense for uninvolved countries, but for those which are more involved in the Ukrainian war effort I think it is more sensible to act pragmatically and use Crimea as a bargaining chip to help make Russia retreat on the mainland.
What exactly should Crimea's status be?
I've been vague about exactly what should happen beyond "not invading Crimea." To be clear, there are numerous options.
First, most generously to Russia, Ukraine could recognize Russian sovereignty over Crimea in a peace treaty. This would most effectively hasten the end of the war.
Second, Ukraine could accept an internationally observed, binding referendum for the status of Crimea - one with more credibility than the one Russia held in 2014. In theory this will lead to the same outcome, but Russia's pride may be wounded and so it may be less effective as a diplomatic bargaining chip. At the same time, it would arguably set a better legal precedent for Crimea's status to be decided by a trustworthy referendum rather than by a diplomatic deal.
Third, Ukraine could continue claiming sovereignty over Crimea while not actually invading it for the foreseeable future. This would pose the least challenge to existing international norms, while also being less attractive to Russia. Ukraine could try to wring some diplomatic concessions by making secret or implicit guarantees that they will never invade Crimea, but Russia might not believe those promises. Additionally, the broken legal status could pose challenges for the people of Crimea. And in the long run, it leaves open a legal space for yet another war. After seeing the results of so many so-called "frozen" conflicts with unrecognized de facto borders, which are so often not actually frozen and lead to more warfare, this does not seem like an option that anyone would truly prefer even as a compromise. If Crimea is going to indefinitely remain in Russia, we should accept the facts and officially recognize it. Insisting on a legal claim that is very unlikely to ever be enforced is not an effective way to uphold the law, and in this case it has meaningful negative consequences for the people who have to deal with the ramifications of de facto borders being unrecognized.
This policy of continuing to make a revanchist legal claim but never trying to enforce it can be the most politically pragmatic option for leaders who have to walk the line between their realistic judgment and their hawkish, nationalist constituents. That seems to be what happened in Georgia. However, in the long run I don't think it is healthy for the body politic. The lingering issue of territory waiting to be reconquered can itself be the thing that stirs up hawkish sentiment, and new politicians can gain followings by promising to reconquer the lost territories, and leaders then find themselves pressured to deliver on the demands of nationalists.
There is also potential for legal compromises like shared sovereignty, such as the status granted to the Aland Islands in 1920. Technically Crimea already had a compromise, that of being an autonomous region, but it could go even further in the direction of integration with Russia. Maybe something like this would be ideal, but I would caution that a compromise can have the downsides of two extremes as well as the upsides so it isn't necessarily the best option. And it may be difficult to implement or unpopular. I'm not against it in principle, but it's certainly hard to communicate, so it seems more effective in most contexts to focus on the simpler idea that Crimea should be left alone with Russia. When Ukrainian and Russian diplomats discuss these issues they may consider the complicated compromise options, and when doing so they may be influenced by the balance of simple sentiments for or against Ukrainian control of Crimea.
Fundamentally, the issue of adherence not to international law in general, but the norm of not allowing territorial conquest by war, seems critical. Regardless of Ukrainian claims, if Russia is allowed to keep territory it gained militarily, it weakens the single most important reason for states not to engage in military conflict, returning to the pre-WWII status quo. The indirect implications for China and Taiwan, and for leaders around the world in the coming decades, contribute to - or detract from - international stability in ways that are far more important than the direct implications.
That said, I think that considering a binary question is mistaken, and in fact the international economic and other pressure against Russia, is enough to reinforce the norm, while hopefully allowing a stalemate or indefinite ceasefire in the region, addressing the other concerns.
The norm of not allowing territorial conquest by war does not really exist as stated. The norm is to not allow conquest into the recognized borders of other sovereign states. Conquering unrecognized states and breakaway regions is fully normalized, as shown for instance by the nonexistent international reactions to recent wars in Tigray and Karabakh.
China invading Taiwan would not be classical territorial conquest by war because Taiwan is not recognized as a sovereign state. According to international law, China already has the legal right to invade Taiwan. If anything, Ukraine invading Crimea, not Russia invading Crimea, is analogous to China invading Taiwan.
Are there other cases where something similar to the 2014 invasion of Crimea might happen - sure. But the idea that invading Crimea now will help make leaders decide against such stunts is speculative and unsubstantiated. Different parts of the world, different scenarios with different stakes. And in cases like Crimea, where the domestic majority welcomes it and there is little to no bloodshed, it's hardly something to be upset about.
It is a highly dubious view of norms: that they are so powerful that they govern state behavior, but so fragile that when one country violates them many others will as well.
Yes, well, I would think that annihilating their military and sweeping them out of mainland Ukraine would be enough to reinforce the norm.
Crimea was a part of Ukraine when it was conquered by Russian troops. Unambiguously. The claims by Russia that it really would have wanted to break away may even be correct - but they don't get to invade first, then run an election to say they declared independence. It certainly needs to be the other way around for it to be at all similar to those other cases.
If you only enforce the rules when there is already a military conclusion, you're not enforcing international law, you're saying that might makes right. There needs to be continued pressure on Russia about the fact that they are in continuous violation of international law. And at this point, if they want Crimea legally, they would need hand it back to Ukraine and let them vote on succession. (Or they need Ukraine to recognize their claims.)
That's assuming norms are binary, is reductive, and makes no sense as a response. Yes, norms are degraded by violations, and yes, they are important guides to state behavior in a wide variety of cases. If you don't think either one of those claims is true, I'd be happy to defend it.
That's beside the point, I wasn't claiming otherwise. The point is that Taiwan is more like those other cases.
I wasn't arguing against the use of sanctions to punish countries for violating international law (or some laws, at least).
And enforcing law requires might, and sanctions are might of a different form, so this doesn't make sense anyway.
I don't think we fundamentally disagree there but I'm saying this stuff is very tenuous as a rationale for foreign policy - one norm violation doesn't make a great deal of difference.