It's a trade-off. There are some scenarios where you can expect that armed teachers may save lives and some where they'll cost lives. For example, they may save lives in a situation where a student goes on a shooting rampage and an armed teacher manages to take him out before he kills as many as he would have if nobody had been there to take that shot. On the flip side, you have scenarios where a teacher, say, leaves a gun in her purse when she goes to the bathroom and a students gets his hands on it and shoots someone, or the gun goes off accidentally during class and kills someone, or the teacher has a breakdown and shoots someone, or the teacher gets overpowered by someone who then uses the gun, or the teacher gets spooked and shoots someone who wasn't actually a threat (e.g., working late, grading papers, when surprised by a kid who the janitor let in to get something out of his locker, and in a darkly lit hallway the teacher freaks and a child winds up dead).
With this kind of thing, it's inevitable that there will be events on either side of that divide, over a long enough period. The question is which side comes up more often. Judging by population-wide statistics, cultures with lower rates of gun ownership tend to have lower overall homicide rates (in addition to fewer accidental shootings), so my bet would be that armed teachers are likely to create more risk than they prevent.
Some of this depends on licensing standards. If it's relatively hard to get a concealed carry permit, and there are requirements for people to take courses in safety and non-violent conflict resolution, and they have to pass tests (of marksmanship and gun safety), and there are background checks, etc. then the kinds of people you'll have carrying guns are going to be lower risk, and potentially more able to do something positive with guns. If, on the other hand, pretty much anyone can pack heat, then it's likely to be more dangerous.