There are three things that usually contribute to "transparency":
- What is the moderator doing?
- When is the moderator doing that?
- How is the moderator doing it?
Users will want answers to all of these questions. Generally speaking, only one of them can be answered: When.
Telling the user what you're doing gives angles to attack, and lawyer'ing one's way out of it. Telling the user how you're doing something usually has a similar result. At best, you can say "We're doing some stuff behind the scenes, but I assure you we are doing something!". This is, obviously, not all too satisfactory.
Using this primary point (When), you can at least give users a heads-up: "We're looking into this.", or "I'm forwarding this." - This gives the user a small notice that something is going to happen, even if they don't see the actual result. It also doesn't reveal every single action.
Of course there will always be users who aren't satisfied by this either ("Meh, they're just saying that, but they're not actually doing anything!"). That's the point where you can't do much more. Either you very openly document every single action (which is really not a good idea in most cases), or you have to accept that you can't please everyone.
However, this only addresses at-the-moment transparency. Providing a clear set of rules and suggested courses of action (that are well-documented and followed), and a system that clearly informs the user of what exactly they have done wrong will provide an environment that is much more transparent. Users that were on the receiving end of any action will have seen it in action, which spreads to other users by word-of-mouth. New or unaffected users have a reference to rely on.
So, to sum it up:
- Give everyone involved a heads-up.
- Don't try to convince those who don't want to be convinced.
- Provide a clear environment.