There have been several posts recently arguing the pros and cons of using prizes in classrooms.
I'm a professional full-time dog trainer and I hear a similar argument from many of my clients. Now before you discredit my viewpoint, let me say that everything I do with dogs is based on universal scientific learning theory. Almost everything that I learned about dog training, I learned from my college psychology classes!
People can be so resistant to using treats/rewards! "My dog should behave because I said so," and "my dog should do what I say because he is a dog!" or "verbal praise should be enough for my dog."
ACTUAL rewards can go a LONG way! I'm willing to bet that if your school stopped paying you money, and instead only offered "good job today!"... it wouldn't take long for you to stop showing up to work. Verbal praise can go a long way, but it's never the same as getting "paid" with a highly motivating "treat." This can vary based on different situations, distractions, and from pupil to pupil. For dogs, a pat on the head and a "good boy" might make him wag his tail, but it's NOTHING compared to a tasty treat!
I'm not advocating giving candy and prizes out for every good deed, however. Although reward-based training/teaching is INCREDIBLY effective, it can easily be misused. Back to the dog training example, one common mistake is to "bait" a dog into behaving. Dangle the treat, and maybe he'll sit. When you make treats THAT obvious with dogs, they will only behave for you when you openly promise a reward. If instead, you are more subtle with your treats, and give them out more randomly, you will likely be able to get your dog to perform a "sit" even if there isn't a treat within a one block radius of where you're standing. It's always about the surprise factor and creating positive expectation. I think that same idea can be applied to classroom rewards with ESL students.
Before I became a positive reinforcement trainer, I was strictly a "punishment" trainer, save for some verbal praise. Although I could still get the dogs to perform basic obedience commands flawlessly, there was a DRAMATIC difference in attitude and even outcome. I am lucky to have that experience because it taught me some valuable lessons, which I plan to apply in my ESL classes when I get to Korea. There is a huge difference between force and inspiration.
Some of you have mentioned taking away homework, etc as the reward. This is called negative reinforcement. "Negative," because you're removing something (in this case, homework), and "reinforcement" because that action makes it more likely that the subject will repeat whatever behavior you are rewarding (in this case, volunteering, or behaving politely, or getting a correct answer, etc). Although negative reinforcement does work, I don't think it's quite as effective as positive reinforcement. "Positive," because you are adding something (games, candy, stickers, etc) , and "reinforcement" because that will make desired behavior more likely to be repeated. From my experience, positive reinforcement is the quickest, most effective way to build solid, long-term learning.
What I've been referring to are some basic fundamentals of learning theory. Humans are a little different because they can reason, which is something that dogs can't do. However, any basic application of positive behavior management will be just as effective with humans as it is with dogs, or dolphins, or fish. In fact, if applied correctly it should be even easier with humans because you actually speak the same language, so to say. The trick is to understand and apply these techniques the right way.
Although I have never taught in a human school, I teach basic obedience and puppy classes to dog owners, and my job is close to 90% about teaching people rather than teaching the dogs. I teach people to teach their dogs.
Anyhow, that's my two cents on prizes and rewards!