Respectfully, I think your critique is entirely misguided. What you wrote above sounds more like a explanation in search of a phenomenon rather than the reverse

Respectfully, I think your critique is entirely misguided. What you wrote above sounds more like a explanation in search of a phenomenon rather than the reverse. Couching a social phenomenon in a socioeconomic context seems like an especially poor fit, IMHO. For instance, do you really think the police, SNAP, and federal assistance for hurricane victims incline people to flame others on the Internet? Folks who need food stamps are probably among the least likely to spend "all their time on the Internet," and more likely to be working two jobs. I'm not sure the presence of cops, SNAP, or FEMA matter to folks who don't need any of those things in a concrete fashion. More importantly, though, it's a fallacy to assume that this sentiment is new merely because it's newly-ish observed. I do think there's an economic component, given how shitty the economy is compared to previous generations' experience; people are unhappy, but I suspect some of that stems from the idea that they're on their own, whereas "everyone else" (or perhaps the rich) is getting a handout. Actually, I'm not even sure folks like millennials believe that in aggregate, but I don't know for sure. That said, anonymity combined with increased connectivity is novel, and I think it's an enabling factor. I think our social mores and our tools haven't caught up. People can say whatever comes to mind with little or no consequences, and the onus is on the recipient to filter that. There's a many-to-one relationship there and it's hard to manage. A technological, "dumb" answer would be applying sentiment analysis to emails, tweets, et al. Incorporate as signals stuff like: whether it's a first-time sender; whether the Twitter account has followers; and so on. It's possible these would be band-aids for a real social problem. Or perhaps they would at least lower the percentage in harassing people online. Socially, it might help if escalated harassment (i.e. persistent, involving death threats, etc) were treated as legal matters. As it is now I suspect it's too hard for victims to do much, and that it can be difficult to get the police to pay attention to it. I don't really have a concrete proposal; I just have a intuition that reporting harassment or stalking to the police is unlikely to be productive as it stands now, and I think the law is unfortunately a really blunt instrument, so this would be hard to get right. (It wouldn't have to be perfect, since no solution which will scale to hundreds of millions of people could possibly be.) Internet mob justice is probably the answer that will happen from time to time, when people decide to deliberately out trolls. I am not a big fan of Internet mob justice, but sometimes it seems like it's that or no justice at all.