As I have already said a few times, the corner cases where a union might make things better involve correcting other market failure like monopsony (which is usu

As I have already said a few times, the corner cases where a union might make things better involve correcting other market failure like monopsony (which is usually caused by the barriers to entry item you listed, which in turn causes the participant size problem). None of these apply to the specific case in the article we are talking about (i.e. this specific union is bad for society) because everyone is already highly paid relative to similar work elsewhere in the economy. Of course there are assumptions in the theory, those drive the corner cases (of which I mentioned the most relevant one in my first post) and is also present in the theory you are relying on to say what the corporations are doing is bad. You can't like the stuff that says monopoly is bad when it applies to corporations and then turn around and think it's good when it is clearly happening in labor via unions, that's utterly illogical. It's either clearly bad in both cases, or the theory is worth nothing and we have no justification for any market intervention, which is obviously not true. We need market interventions (arguably a lot more than we are currently doing, but likely less than we used to do pre Rhenquist supreme court) to correct market power abuse and there are no sacred cows. It is largely bad both when capital and when labor does it.