I have always defended private unions on the ground that workers have a freedom of association just as much as anyone else. I think the government has tilted the playing field in the union's favor too much, but I will leave that aside for today. I will also leave aside the problem of public unions, where there is no one really representing the taxpayer on the other side of the table in negotiations (many politicians in union states owe their jobs to union support).
Leave all of that aside. The economic problem with unions tends to be that they are such a conservative (little c) force in an economy that needs dynamism to grow and expand wealth. Here is a great example:
The University of California last week tentatively agreed to a deal with UC-AFT that included a new provision barring the system and its campuses from creating online courses or programs that would result in “a change to a term or condition of employment” of any lecturer without first dealing with the union.
Bob Samuels, the president of the union, says this effectively gives the union veto power over any online initiative that might endangers the jobs or work lives of its members. “We feel that we could stop almost any online program through this contract,” Samuels told Inside Higher Ed.
I have said for a long time that negotiations for pay and benefits (in private unions) tend to be the least problematic union activity (different story in public unions, where the relationship to management is not adversarial). Longer term, union imposed work rules and restrictions tend to be much more costly. The reason I think is that corporate executives can easily value the difference between various pay and benefits packages, but have a hard time valuing flexibility and dynamism. If union rules cut off potential future as-yet-unknown growth and cost reduction efforts, the cost of these rules can be huge but equally they can be almost impossible to value (more like options pricing than straight cost-benefit).