Maybe we should treat faking scientific evidence more as a sin than a crime.

E.g. no fines or prison, but "banishment" from scientific circles until the perpetrator repents publicly, explains all the details and asks for absolution.

This would be somewhat shameful, but people fear shame more than death, and there would also be a path towards restoration + more knowledge of how the fraud actually worked and what led to it.

So mob justice instead of trial?

I think it would lead to worse outcomes, and is likely to end up based more on whether people like the person accused, or like their results than on what they actually did.

Social shame is a tricky mechanism.

I agree that it is a tricky mechanism and that there is a potential for abuse. Social network-like pseudonymous mob shaming shouldn't be replicated elsewhere.

That said, what the culprits do is really very shameful and professional organizations used to treat shameful behavior of their members in this way, rather than handing them over to standard public courts.

At the very least, if other scientists are publicly involved, the public gets an impression that the community doesn't "wash its collective hands" over its own bad actors and tries to remedy the problem actively.

Maybe both should be combined. "A trial by your peers" would be trial by a non-judicial panel made of other scientists (who are the real peers of the culprit), and it could only mete our punishments of specific and relevant type: research bans, public apologies etc.

I certainly don't think that people should be jailed for scientific misconduct. Prisons are useful in isolating dangerous people from the rest of the society, but their restorative effect is minuscule even in Europe, much less in the USA. And fraudulent scientists don't have to be isolated from society, unlike rapists or muggers. They just have to lose their credibility.

Perjury is in a sense a more offensive crime than the petty acts that fills prisons everywhere, for it is the one crime that makes criminals of us all. And it is far worse when it is done by a member of a professional class that enjoys the benefit of being assumed credible rather than criminal. Faking data is in a sense much worse than overtly punching someone in the face, because the number of lives I can adversely affect by punching someone in the face is just one, and the number of people I can probably get away with punching in the face before being caught is also just one.

I see your point about wanting to reserve prison for strictly violent offenses, but there is such a thing as serious yet non-violent crime, which is the entire reason why low-security prison is there.

One of the problem with jailing people is that their work ability is no longer available to the rest of the society, or only in a reduced form.

Scientists, even dishonest ones, are very highly educated people with important and hard to replace skills. Locking those skills away is a huge loss, much higher than when you lock up a run-of-the-mill Ponzi schemer.

I would prefer some punishment that would still make use of their talent and skills to benefit humanity. Perhaps they should be required to design better research methods that prevent their own sort of fraud etc.

Pretty incredible to quote yourself in your own post.

I am merely presenting it in an honest manner as having more to it than just rhetoric made up on the spot from the comfort of my keyboard, while also not insinuating the credibility of being able to quote other people who agree.