New York state's Rotenberg center uses skin shock to change the behavior of special needs children. It sounds pretty much identical to the use of shock collars to change the behaviors of dogs. The head of the school is a disciple of BF Skinner, of operant conditioning (behaviorism fame).
All animals, including humans, respond to a range of rewards and punishments. Punishments are all a form of "pain", whether than pain is psychic (time out, shame, isolation) or physical (spanking, shocking). Humans are not very different from dogs and other mammals; methods used to train dolphins work quite well on husbands.
Remote electric shock, whether used on dogs or humans is another form of punishment. Some animals might "prefer" it to isolation or time out, others would prefer a time out, still others may respond primarily to positive reinforcement and weakly to negative reinforcement.
The problem with shock is the same as with spanking and torture (the most extreme form of negative reinforcement). It is very hard to do it appropriately -- especially on other humans. I suspect it's easier to be calm and measured about using a shock-collar on a dog, but, on the other hand, I've seen very good and loving dog "masters" unintentionally abuse "training" (choke) chains on dogs who are being obtuse. Another problem is that both animals and humans develop tolerance to negative reinforcement. Voltage levels rise. A dog who responded to a simple snap of a training collar needs a much harder snap. Aversive measures must be metered to avoid tolerance.
This is difficult territory. We need wisdom, science, humility and a good measure of dread to walk these roads with relatively safety. Last resorts indeed.