There may be a few before-and-after studies. I'm not too familiar with them. What's more interesting is having one group with a bracelet and another group without it at the same time, and just tracking their breach records.
The studies seem to be all over the map. Some find a decrease in breaches, because perhaps the offender is afraid and is being a good boy, or sometimes there's an increase in breaches because when there's a false alarm, the police are called, and it's assumed a breach is done. There are many different factors.