I am generally in favor of transparency. In this case I am definitely in favor of more high level transparency about wages. But when it gets to the specific d

I am generally in favor of transparency. In this case I am definitely in favor of more high level transparency about wages. But when it gets to the specific data about one person and their coworkers it gets more difficult due to various cognitive biases (including the Dunning Kruger effect). I would be interested in the results of experiments in complete transparency and how they deal with these issues. I wouldn't currently undertake the experiment because I don't know of a way to structure it that allows it to fail and not be permanent.