There is no significant risk of lasting negative health consequences after infection

That's simply false. In fact, there is an abundance of evidence of it. 

•  https://www.wsj.com/articles/doctors-begin-to-crack-covids-mysterious-long-term-effects-11604252961
• https://www.who.int/director-general/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19---30-october-2020
• https://www.thelancet.com/journals/laninf/article/PIIS1473-30992030701-5/fulltext

https://chrismasterjohnphd.com/covid-19/finally-confirmed-vitamin-d-nearly-abolishes-icu-risk-in-covid-19

You're... citing someone with a PhD in nutritional science primarily interpreting a study with n=76, and trying to deduce from that a 96% decrease in fatality risk. That's not how those statistics work. You simply can't get that level of information from the studies cited.

What you wrote makes sense, except that you underestimate the risks for young people (others have already provided sources), and I am surprised that you included "not wanting to kill older family members by negligence" almost as a footnote -- is it really so rare for young people to interact e.g. with their parents? You only need to infect them once.

At the end, it is a quantitative decision; being among the first ones to take the vaccine is a risk, getting sick is also a risk; we disagree about the relative sizes of these risks... and also about government incentives (from my perspective, governments are mostly downplaying the risks of COVID-19, because their voters do not want to be inconvenienced).

They value safety of strangers higher than their own safety, and want to take the vaccine for the sake of all the people at risk in the society.

This is the wrong framing. Taking a vaccine does not exchange 5 personal utility against 5 external utility. The amount of damage prevented is way way higher than the amount of personal damage.

Say that, at the time the vaccine is available, $R=0.4$ (tell me if you think this is unrealistic). Then, the expected number of people you will infect is ${\sum }_{k=1}^{\infty }{0.4}^k=\frac{0.4}{1-0.4}=\frac{2}{3}$. So getting vaccinated will save $\frac{2}{3}$ people from getting Covid, in expectation. However, this is a probability distribution that includes outcomes where several people get it, so even if you somehow don't interact with old (or otherwise high-risk) people, you certainly cannot control that such people ultimately get it.

Now, if your net utility of avoiding the vaccine is 5, the average utility for these $\frac{2}{3}$ people might be -300 or something, since they might straight up die from it. Which means you're exchanging a small amount of personal utility for a large amount of negative external utility, in this case -200. It is absolutely possible to value your own safety more than that of other people and still consider it a moral obligation to get vaccinated, as long as you don't value it a hundred+ times more.

You are applying the incentive heuristic inconsistently. On the one hand you infer that if there was string evidence of long term effects, governments would be very vocal about it. But on the other hand, you ignore that these incentives would also apply to the Vitamin D effects that you cite. Governments would also surely have an interest to publicize an intervention that has  a 25 fold reduction in risk.  So the estimate is wrong or your conception of how governments work is wrong.  

I suggest that it is both. Other answers have already mentioned that a 25 fold reduction in risk would be ridiculous, and governments just do not respond to incentives like that.  

This study is a strong reason to fear  prevalent long term consequences for cognitive performance after even mild Covid-19 infections. 

On the other hand, you do not mention the strongest reason for supporting your view: the relatively underexplored long term effects of mRNA vaccines. However, if you worry about those, you should just get the traditional-style Oxford or J&J vaccines. Since they use the same technology as well established vaccines, taking them should be fundamentally as safe as getting your flu shot.

I think cutting the IFR by 25 on the basis of one study is a mistake, the chance of the study being fatally flawed is greater than 1 in 25. On the other hand 0.5% is overall CFR and would be lower for young people.

I think it's hard to cut risk of long term effects by more than a factor of 10 from published estimates. Note there is evidence of long term effects contrary to your claim, i.e. studies that do 6 week follow ups and find people still with some symptom. This isn't 6 months but is still surprisingly long and should shift our belief about 6 months at least somewhat. Also novel disease that attacks many parts of the body is some evidence. I agree the evidence is exaggerated to scare us but it feels like a different situation from reinfection where it actually is almost impossible to find instances except when immunocompromised.

But I think perhaps the most important is that even young people are currently limiting their activities in many undesirable ways in accordance with local government ordinances (which apply equally to old and young). Vaccination allows one to end or partially end these limitations--even if not in a legal sense, probably at least in a moral sense.

There is no strong proof for such effect. Such proof would greatly increase acceptance of governments' policies, so there is a strong incentive to publish any such proof.

Trump spent months speaking about vaccines being soon available. Now we have the evidence that a vaccine works and the FDA will wait three weeks before to think about whether or not to approve the vaccine. 

There would be strong incenties to start vaccinating now and vaccinating as soon as possible is a prime goal of the US president but still not done because the system is dysfunctional to an extend that even after spending years deregulating the FDA and having months to prepare for the moment of the vaccine. 

While we do have slighlty more functioning governments in Europe, our governments are also not spending as much more on science to deal with COVID-19 as would desireable.

There's a study that German researchers did do that says:

Findings  In this cohort study including 100 patients recently recovered from COVID-19 identified from a COVID-19 test center, cardiac magnetic resonance imaging revealed cardiac involvement in 78 patients (78%) and ongoing myocardial inflammation in 60 patients (60%), which was independent of preexisting conditions, severity and overall course of the acute illness, and the time from the original diagnosis.

There's a recent study that suggest 20% of COVID-19 patients develop diagnoseable mental health issues. Here it's worth noting that 20% is not the upper bound as mental illness that gets developed through physical trauma often takes longer to show up (see the literature on depression due to head trauma).

Neither of those are surprising because our priors for this Coronavirus should come from the last problematic Coronavirus which was SARS with produced long-term mental health issues in a even larger number of patients and chronic fatigue syndrome (which might be a result of myocardial inflammation).

Divide that by 25 and stockpile large quantities of vitamin D in case you get the disease (https://chrismasterjohnphd.com/covid-19/finally-confirmed-vitamin-d-nearly-abolishes-icu-risk-in-covid-19)

Even if the Vitamin D treatment they did in the study would have such an effect, you don't get that by stockpiling vitamin D and taking it when you get symptoms. 

Orally taken Vitamin D takes a while to be converted into it's active form and in the study they gave that active form intravenously.

Pfizer's vaccine requires extremely low temperatures, so there is a danger that in some locations it will be transported or stored incorrectly, causing greater risk than that suggested by the trials so far

Damaged mRNA doesn't cause additional risks. It just won't produce the desired proteins.

Given the studies that we have that do suggest long-term problems from COVID-19 and no evidence for long-term problems due to the vaccine, you need pretty high double standards to consider the vaccine more risky then getting infected with COVID-19.

It's not often I see someone claim that the US medical regulation system is too lax.

The AstraZeneca vaccine was halted in the US for a month on the basis of a single, potential adverse event. Huge numbers of lives were on the line, and the US regulators were willing to hold up one of the frontrunner vaccine candidates for weeks on the basis of the faintest hint of unsafety.

There might be long-term adverse effects of the vaccine we don't know about, though no-one I've heard speak about vaccines seems to think these are likely to be severe; most vaccines are very safe. But if the FDA gives approval we can confidently assume that, at least over the timescale of the trial, the vaccine is extremely safe. In fact, we can assume we have far too much evidence of safety, that it should have been approved on the basis of substantially less evidence than we have.

As far as efficacy is concerned, as I understand it the Pfizer and Moderna vaccines have very simple designs (which were pre-approved by the – again, extremely over-conservative – FDA) and are overseen by an independent data-monitoring organisation. So while I agree their incentives are perverse, their ability to distort the data should be relatively limited. 

(Here's a piece claiming the same is not true of the AZ/Oxford vaccine; I'm not sure how to evaluate this, but it's worth noting that the author is explicitly contrasting their data with the much more reliable Pfizer and Moderna data.) 

I also think you're excessively sceptical of the evidence of long-term risks from COVID in young people. But in my case, avoiding a significant risk of (a) a really unpleasant and really long (multi-week) illness, and (b) accidentally killing people is sufficient for me to want to take a vaccine as soon as possible, even without a (in my estimation quite small, but nontrivial) risk of long-term sequelae.

I appreciate this question—good to see someone willing to go against the hyperscrupulous LW consensus. I think many people want the vaccine because of a vague idea that it will accelerate the time at which things are back to normal. Most people have suffered more from the indirect effects of the pandemic (job losses, business closures, stress, isolation) than from COVID itself.

Please don't forget that some young healthy people are essential workers who might not really have a choice about this.

they value safety of strangers higher than their own safety, and want to take the vaccine for the sake of all the people at risk in the society.

Quite apart from the actually low personal risk from taking a vaccine, why does this strike you as odd? This is perfectly normal and good human behaviour, and if you don't share it there is probably something quite a bit wrong with you.

A neglected motivation: If I'm vaccinated, and my friends are vaccinated, I can hang out with my friends again

You have a good summary.

Point Against manadatory Vaccination: Decreases trust in Government. There will be rebellion. 

Point for vaccination in young people: Covid does have serious effects other than death. A decrease in brain and cardiovascular function is entirely possible in young healthy people. To just use fatality rates may not be  not an accurate measure of risk. 

Young people may also be more social or work in jobs where they are more  likely to spread it. Taking a vaccine may give you the confidence to go about your socialising and work without fear or guilt of infecting others.

Imagine a bar or restaurant advertising that their entire staff is vaccinated, I imagine that they will be likely to attract more business.

In principle, I want to be vaccinated against every disease I can be, provided the risk of side effects is sufficiently low.

In practice, I will absolutely not be among the first or even second wave of voluntary covid vaccine getters. I got an annual flu shot for 25 years, and stopped after multiple years of my wife pointing out that every single time I did, I got sick within 36 hours, and was unable to work for 10 days. I'd never noticed the correlation before. I've never had unusually strong side effects to any other vaccine, but experiences like this give me pause and make me weigh things differently. I also am sufficiently able to reduce my risk of covid exposure through behavior that the marginal benefit of me getting it, to me and others, is well below average.

There may be a (potentially unconscious) desire for people to signal that they are not an "anti-vaxxer". This term is a very strong cultural pejorative lately. It's similar to being referred to as a conspiracy theorist and almost as negative as being considered a racist. No idea how much this may be contributing to publicly professed enthusiasm for the vaccine but maybe social signaling in this respect is a component.

I think there is totally irrational fear going on in society on vaccines. First of all it is really hard to develope a vaccine that is more dangerous than the infection itself. There have been vaccines that were incredibly dangerous to take in the early 19th century and some vaccines in the early 20th century could kill you too, but they would kill you in one of 10 000 cases. But this vaccine can’t kill you under any circumstances. Phizer vaccine will get ineffective if it’s too warm which means nothing will happen if you let it wait outdoors in warm temperatures and inject it later on. virusu, hence components of the in vaccines loose their potency in warmth they don’t get more dangerous. And you just underestimate the danger too. As long you aren’t 12 years old covid is dangerous. There have been one million infections and 10 men between 20-29 died in Germany. Assuming 100 000 men in this age group where infected and that that 2000 of them ended up in a hospital, you are simply better off taking the vaccine. No one who was vaccinated ended up in a hospital. So how dangerous could a vaccine be