[content warning: buncha rambly thoughts that might not make much sense]

certainly -- see my bit about how my preferred solution would be to run a volunteer army even if that takes ruinously high taxes on the rest of the population.  (The United States, to its credit, has indeed run an all-volunteer army ever since the end of the Vietnam War in 1973!  But having an immense population makes this relatively easy; smaller countries face sharper trade-offs and tend to orient more towards conscription.  See for instance the fact that Russia's army is less reliant on conscripts than Ukraine's.)

but also, almost every policy in society has unequal benefits, perhaps helping a small group at the expense of more diffuse harm to a larger group, or vice versa.  For example, greater investment in bike lanes and public transit (at the expense of simply building more roads) helps cyclists and public-transit users at the expense of car-drivers.  Using taxes to fund a public-school system is basically ripping off people who don't have children and subsidizing those that do; et cetera.  at some point, instead of trying to make sure that every policy comes out even for everyone involved, you have to just kind of throw up your hands, hope that different policies pointing in different directions even out in the end, and rely on some sense of individual willingness to sacrifice for the common good to smooth over the asymmetries.

One could similarly say it's unfair that residents of Lviv (who are very far from the Ukranian front line, and would almost certainly remain part of a Ukrainian "rump state" even in the case of dramatic eventual Russian victory) are being asked to make large sacrifices for the defense of faraway eastern Ukraine.  (And why are residents of southeastern Poland, so near to Lviv, asked to sacrifice so much less than their neighbors?!)

Perhaps there is some galaxy-brained solution to problems like this, where all of Europe (or all of Ukraine's allies, globally) could optimally tax themselves some fractional percent in accordance with how near or far they are to Ukraine itself?  Or one could be even more idealistic and imagine a unified coalition of allies where everyone decides centrally which wars to support and then contributes resources evenly to that end (such that the armies in eastern Ukraine would have a proportionate number of frenchmen, americans, etc).  But in practice nobody has figured out how a scheme like that would possibly work, or why countries would be motivated to adopt it, how it could be credibly fair and neutral and immune to various abuses, etc.

Another weakness to the idea of democratic feedback is simply that it isn't very powerful -- every couple of years you get essentially a binary choice between the leading two coalitions, so you can do a reasonably good job expressing your opinion on whatever is considered the #1 issue of the day, but it's very hard to express nuanced views on multiple issues through the use of just one vote.  So, in this sense, democracy isn't really a guarantee of representation across many issues, so much as a safety valve that will hopefully fix problems one-by-one as they rise to the position of #1 most-egregiously-wrong-thing in society.

I think that today's "liberal democracy" is pretty far from some kind of ethically ideal world with optimally representative governance (or optimally pursuing-the-welfare-of-the-population governance, which might be a totally different system)!  Whatever is the ideal system of optimal governance, it would probably seem pretty alien to us, perhaps extremely convoluted in parts (like the complicated mechanisms for Venice selecting the Doge) and overly-financialized in certain ways (insofar as it might rely on weird market-like mechanisms to process information).

But conscription doesn't stand out to me as being especially worse than other policy issues that are similarly unfair in this regard (maybe it's higher-stakes than those other issues, but it's similar in kind) -- it's a little unfair and inelegant and kind of a blunt instrument, just like all of our policies are in this busted world where nations are merely operating with "the worst form of government, except for all the others that have been tried".

• With your FTX thought experiment, the population being defrauded (mostly rich-world investors) is different from the population being helped (people in poor countries), so defrauding the investors might be worthwhile in a utilitarian sense (the poor people are helped more than the investors are harmed), but it certainly isn't in the investors' collective interest to be defrauded!!  (Unless you think the investors would ultimately profit more by being defrauded and seeing higher third-world economic growth, than by not being defrauded.  But this seems very unlikely & also not what you intended.)
  • I might be in favor of this thought experiment if the group of people being stolen from was much larger -- eg, the entire US tax base, having their money taken through taxes and redistributed overseas through USAID to programs like PEPFAR... or ideally the entire rich world including europe, japan, middle-eastern petro-states, etc.  The point being that it seems more ethical to me to justify coercion using a more natural grouping like the entire world population, such that the argument goes "it's in the collective benefit of the average human, for richer people to have some of their money transferred to poorer people".  Verus something about "it's in the collective benefit of all the world's poor people plus a couple of FTX investors, to take everything the FTX investors own and distribute it among the poor people" seems like a messier standard that's much more ripe for abuse (since you could always justify taking away anything from practically anyone, by putting them as the sole relatively-well-off member of a gerrymandered group of mostly extremely needy people).
    • It also seems important that taxes for international aid are taken in a transparent way (according to preexisting laws, passed by a democratic government, that anyone can read) that people at least have some vague ability to give democratic feedback on (ie by voting), rather than being done randomly by FTX's CEO without even being announced publicly (that he was taking their money) until it was a fait accompli.
  • Versus I'm saying that various forms of conscription / nationalization / preventing-people-and-capital-from-fleeing (ideally better forms rather than worse forms) seems morally justified for a relatively-natural group (ie all the people living in a country that is being invaded) to enforce, when it is in the selfish collective interest of the people in that group.
  • huw said "Conscription in particular seems really bad... if it’s a defensive war then defending your country should be self-evidently valuable to enough people that you wouldn’t need it."
  • I'm saying that huw is underrating the coordination problem / bank-run effect.  Rather than just let individuals freely choose whether to support the war effort (which might lead the country to quickly collapse even if most people would prefer that the country stand and fight), I think that in an ideal situation: 
    1. people should have freedom of speech to argue for and against different courses of action -- some people saying we should surrender because the costs of fighting would be too high and occupation won't be so bad, others arguing the opposite. (This often doesn't happen in practice -- places like Ukraine will ban russia-friendly media, governments like the USA in WW2 will run pro-war support-the-trooops propaganda and ban opposing messages, etc.  I think this is where a lot of the badness of even defensive war comes from -- people are too quick to assume that invaders will be infinitely terrible, that surrender is unthinkable, etc.)
    2. then people should basically get to occasionally vote on whether to keep fighting the war or not, what liberties to infringe upon versus not, etc (you don't necessarily need to vote right at the start of the war, since in democracy there's a preexisting social contract including stuff like "if there's a war, some of you guys are getting drafted, here's how it works, by living here as a citizen you accept these terms and conditions")
    IMO, under those conditions (and as long as the burdens / infringements-of-liberty of the war are reasonably equitably shared throughout society, not like people voting "let's send all the ethnic-minority people to fight while we stay home"), it is ethically justifiable to do quite a lot of temporarily curtailing individual liberties in the name of collective defense.
  • Back to finance analogy: sometimes non-fraudulent banks and investment funds do temporarily restrict withdrawals, to prevent bank-runs during a crisis.  Similarly, stock exchanges implement "circuit-breakers" that suspend trading, effectively freezing everyone's money and preventing them from selling their stock, when markets crash very quickly.  These methods are certainly coercive, and they don't always even work well in practice, but I think the reason they're used is because many people recognize that they do a better-than-the-alternative job of looking out for investors' collective interests.
• This isn't part of your thought experiment, but in the real world, even if FTX had spent a much higher % of their ill-gotten treasure on altruistic endeavors, the whole thing probably backfired in the end due to reputational damage (ie, the reputational damage to the growth of the EA movement hurt the world much more than the FTX money donated in 2020 - 2022 helped).
  • And in general this is true of unethical / illegal / coercive actions -- it might seem like a great idea to earn some extra cash on the side is beating up kids for their lunch money, but actually the direct effect of stealing the money will be overriden by the second-order effect of your getting arrested, fined, thrown in jail, etc.
  • But my impression is that most defensive wars don't backfire in this way??  Ukraine or Taiwan might be making an ethical or political mistake if they decide to put up more of a fight by fighting back against an invader, but it's not like conscripting more people to send to the front is going to paradoxically result in LESS of a fight being put up!  Nations siezing resources & conscripting people in order to fight harder, generally DOES translate straightforwardly into fighting harder.  (Except on the very rare occasion when people get sufficiently fed up that they revolt in favor of a more surrender-minded government, like Russia in 1918 or Paris in 1871.)
• To be clear, I am not saying that conscription is always justified or that "it's solving a coordination problem" is a knockdown argument in all cases.  (If I believed this, then I would probably be in favor of some kind of extreme communist-style expropriation and redistribution of economic resources, like declaring that the entire nation is switching to 100% Georgist land value taxes right now, with no phase-in period and no compensating people for their fallen property values.  IRL I think this would be wrong, even though I'm a huge fan of more moderate forms of Georgism.)  But I think it's an important argument that might tip the balance in many cases.

Finally, to be clear, I totally agree with you that conscription is a very intense infringement on individual human liberty!  I'm just saying that sometimes, if a society is stuck between a rock and a hard place, infringements on liberty can be ethically justifiable IMO.  (Ideally IMO I'd like it if countries, even under dire circumstances, should try to pay their soldiers at least something reasonably close to the "free-market wage", ie the salary that would get them to willingly volunteer.  If this requires extremely high taxes on the rest of the populace, so be it!  If the citizens hate taxes so much, then they can go fight in the war and get paid instead of paying the taxes!  And thereby a fair equilibrium can be determined, whereby the burden of warfighting is being shared equally between citizens & soldiers.  But my guess is that most ordinary people in a real-world scenario would probably vote for traditional conscription, rather than embracing my libertarian burden-sharing scheme, and I think their democratic choice is also worthy of respect even if it's not morally optimal in my view.)

Also agreed that societies in general seem a little too rearing-to-go to get into fights, likely make irrational decisions on this basis, etc.  It would be great if everyone in the world could chill out on their hawkishness by like 50% or more... unfortunately there are probably weird adversarial dynamics where you have to act freakishly tough & hawkish in order to create credible deterrence, so it's not obvious that individual countries should "unilaterally disarm" by doving-out (although over the long arc of history, democracies have generally sort of done this, seemingly to their great benefit).  But to the extent anybody can come up with some way to make the whole world marginally less belligerent, that would obviously be a huge win IMO.