Here are some published journal articles on invisible fences and shock collars.
Polsky, Richard, (2000), Can Aggression in Dogs Be Elicited Through the Use of Electronic Pet Containment Systems?, Journal of Applied Animal Welfare Science, 3(4), 345-357
Schilder, Matthijs B.H. and van der Borg, Joanne A.M., (2004), Training dogs with help of the shock collar: short and long term behavioural effects, Applied Animal Behavior Science 85 (2004) 319-334
Schalke, E., Stichnoth, J. and Jones-Baade, R., Stress Symptoms Caused by the Use of Electric Training Collars on Dogs (Canis familiaris) in Everyday Life Situations, Current Issues and Research in Veterinary Behavioral Medicine,
Interesting articles. The first one appears relevant to the topic, as it deals with invisible fence systems. The author has a PHD, is prominent as a dog bite expert, and is available as a professional "expert witness" for court cases. His subjects for this study are limited and his data is anecdotal. His conclusions are also qualified. ("the implication that shock was associated with, or the cause of, the dog's attack remains tenuous") Reading the actual cases was illuminating. The mistakes made by the owners were obvious, and the training received by the dogs…...all five of them......inadequate. I believe the last incident in particular would have had the same result had the dog been contained behind a physical fence and gotten out. Still, it makes for entertaining reading. www.dogexpert.com/…/Electronic%20fences.pdf
The second, supposedly clinical, study referenced makes use of GSD's undergoing military training. The study is not double blind, and I believe it is not peer reviewed. The conclusions are entirely subjective, as all observed dogs appeared to have some stress from their training, whether they were shocked or not, and the observers (who knew which dogs had had shock training) assessed that the shocked dogs were more affected. Not a very scientific approach. And not directly related to invisible fence.
The third reference is the most interesting, and in fact proves my point. The researchers document that poor timing on the part of the trainer can result in unintended associations (they mention that this is also the case with reward based training, but since reward based training is less resistant to extinction it presents less of a problem).
So I will give you this: invisible fences are not a good idea if the training isn't properly carried out, and since many (most?) people are unlikely to take the necessary time to do the job right, there will be incidents when improperly prepared dogs may make undesired associations, possibly resulting in transferred aggression. I doubt this has resulted very often, but it is possible that it has happened. That said, I believe there are many, many more incidents involving dogs on leashes in public places biting people…...perhaps we should all stop walking our dogs!
(BTW transferred aggression is an interesting thing. I'm sure I am not the only one who has witnessed dogs who normally like each other attack each other because they can't get at the thing they really want. "When you can't bite the one you want, bite the one you're with")