I recently receive an editorial letter that is unwilling to share referee reports. The letter explicitly mentions the existence of such report but the editor don't want to send it to me.
Is not sending referees reports normal and happen frequently?
I decided to accept the second answer, though all answers are very good. The editor told me that the referee is unwilling to share his report. This is exactly the case suggested by the David's answer.
Contrary to other opinions here, in my experience as an author, reviewer and editor it is entirely normal to receive referee reports that should not be shared with the author. This is an action taken by the referee and not the editor normally. Some reasons why referees don't want the text of their opinion shared with the authors.
They know the author well and their identity would definitely become obvious should the report be shared.
The report contains technical details that impact on the paper, so need to be in the report, but would definitely break anonymity if shared, because it would allow the author to immediately work out the referee.
The referee is unsure about whether they are correct in what they are saying. They might say something like "I think this might have been done before, but I'm not sure where. Try X, they are likely to know for sure."
The report is very short, and the referee simply doesn't think it's particularly useful to give the author. This is particularly true of quick opinions, which many journals in mathematics try to obtain before a full referee's report, which can take a long time.
The referee wrote a report, and it contains personal details about somebody -- the author, referee, or a third party -- that they think would be inappropriate to be shared with the author.
The referee clicked the wrong box in the form.
In my experience, the web sites for sending in referee reports often have spaces for two reports, one for the author (and editor) and one for only the editor. I've often left the second space blank, so that all my comments could (and presumably would) be forwarded to the author. But it's entirely possible for a referee to put almost nothing in the first space and put almost all comments in the second space. (I wrote "almost" because the web site software often requires you to put something in the first space, but it can't detect that your something is worthless.) In such a case, the editor could make a decision but couldn't forward anything useful to the author. (Of course, the editor could also look for a new referee.)
As far as I'm aware of, this is highly not normal. Peer review in journals normally means that you get to see the review as an author. Any other decision may show some lack of trust between the editor and the author. For example,
the editor thinks the author is a "crank";
or does not deserve a serious treatment;
or that the specific paper submitted was clearly below the acceptance threshold of that journal so that a serious review was not even conducted (in this case there is a "review" containing one or two paragraphs, stating the paper is clearly below the bar). This case is called "desk reject".
This is definitely not normal. It implies something went wrong. One possible explanation is that the reviewer wrote something highly confidential and the editor isn't sharing it. Alternatively, the reviewer wrote something really bad (e.g. rude/abusive/sexist/irrelevant) and the editor isn't sharing it.
Still, when that happens (I've seen it as an editor), the more common thing to do is to just not tell the authors of the existence of the confidential review, which avoids the authors asking questions like this one.
It is probably not normal (in my field anyway) not to see any form of the referee's report at all. These days, ( again in my field), a referee usually submits two sets of comments, one for the editor alone, and one which will/may be made available to the author. If the referee is told that part of his/her comments will not be made available to the author without consent, then it is proper that those comments should be withheld from the author if that consent is not given.
