On some level, I do believe blackmail is a kind of coercion, but I fear my structuralist explanations for this view would be deeply upsetting to the average libertarian Joe, so I will keep my dirty little Foucault-inspired secrets to myself.
Here is more from Alina Stefanescu.
And how about the economics of blackmail? If blackmail victims are bad guys, why not allow a horde of potential bounty hunters to profit from uncovering their wrongdoing? We can keep "false blackmail" illegal, while allowing blackmail based on truth, no? We likely underinvest in the gathering of such information, and the profit incentives of blackmail would help correct (and overshoot?) this institutional failure.
Yet I remain convinced there is, somewhere, a sound economic and utilitarian case against blackmail. But what is that case?
My favorite exotic explanation (it is not quite sound) is that legal blackmail would lead to inefficient blackmail. Perhaps the ones who should blackmail you are your family and close friends. That is when transaction costs are low and both parties strike a good deal, often based on an implicit rather than explicit blackmail. ("If you run off with that floozy…") The wrongdoer pays a penalty, and the would-be wrongdoer remains deterred. Nothing gets too messy. But if you open up this business to outsiders, well…trust breaks down. Blackmailers fabricate stories, they send weird threatening letters, and they cause extreme anxiety. Outsiders don’t even know when they should believe the word of a blackmailer, which limits blackmail possibilities from those in the know. Under this hypothesis, we keep blackmail illegal to keep blackmailers we can trust.
Addendum: I have turned on the comments function, in case you have good ideas on this topic.