I'm going to refer back to one of my favorite political scientists ever, because he makes a very clear case that's backed up with some remarkably accurate predictions made with a model based on the theory.
Bruce Bueno de Mesquita calls it Selectorate Theory. In any country, the people competing for power need a certain group of powerful people on their side, called "the winning coalition." In democracies, that coalition is bigger, but it's not the whole electorate by any stretch. In dictatorships, it's much smaller, sometimes only a handful of people in key positions. To get the winning coalition behind you, you have to give them large doses of the things they want: money, opportunities for corruption, important positions when the winning coalition is small. When its larger, you still need to give the coalition what it wants, but it becomes more efficient to do so through policy measures; promising the winning coalition that you'll give them the legislation they favor if they back you. And once they do, you have to keep giving it to them or they'll find someone else who is more likely to get them what they want in return for their support.
For this theory, dictatorships and democracies are only different in the size of the population getting paid off in return for their political support. When the winning coalition is small, the dictator could care less what happens to the rest of the population because he gets literally nothing out of being nice to them. Whatever you do, don't make your coalition suffer because you're being nice to people who don't matter. They'll just get rid of you and replace you with someone more ruthless, who will protect their interests better.
Nonviolence here is transparently useless unless it's a ploy to get a third party to intervene on your behalf - which they will only do if it's in their own interests, but making your struggle look noble will make the decision easier for their public to digest. Being violent has a slim chance of success, but it's much better than non-violence. That's obviously not the point of the Occupy movement, and it's also not the context.
If you're in a democracy, then democratic leaders do have to care more, but again, they don't have to care much about people outside their winning coalition, and being nice to those people at the cost of their coalition may hurt them. However, depending on how close the competition is for that winning coalition, coming across as brutal can do real damage to your ability to hold on to the coalition. The public actually does matter here (not as much as we tend to think, but it still matters), and losing enough of their support by looking ruthless will cost you.
This goes for both players (Occupy and Status Quo, say). The current coalition has to take care not to be obscenely brutal to the Occupy movement lest that start to cost them important backers within the public, and the Occupy movement has to take care not to overplay their hand, lest they start to look over-aggressive themselves.
With that in mind, since many in the Occupy movement believe that big corporations are a key part of the winning coalition, they can be as anti-corporate as they want, but they cannot possibly break enough chain-store windows to force a corporation out of that coalition. And anything more than breaking windows - say, arson, or planting bombs at corporate buildings - is far more likely to lose them critical support than force an undesirable member out of the winning coalition.
I think, based on this logic, that violence and vandalism have no place in the Occupy movement if they'd like to be as successful as they can be. They're already struggling for support, and engaging in violence in a serious, directed way will absolutely, positively cost them more than they could possibly gain from doing it. Until you no longer need the support of the public or are no longer competing for it, violence is a pretty dangerous idea to the survival of your movement. There is a time and a place where it can win, and be a better idea than nonviolence, but Occupy isn't in that position.
