Bite statistics are very inaccurate. First of all, they only include bites that are actually reported. Second, breed identification is often incorrect. The breed listed is whatever the victim, a witness, or in some cases animal control officers (IF they are called) think the dog is. Many people don't recognize dog breeds or misidentify them-- just look at the thread about what people think your GSD is! This is even worse when you get to the "pit bull" because "pit bull" is not even a breed name, it is a descriptive term which can be applied to several breeds, plus many people think any stocky short-coated dog with a wide head is a "pit bull" so you have a number of breeds that are being reported as "pit bull" and that raises the numbers quite a bit. People have thought my Rat Terrier mix was a pit bull, and my foster Chinese Shar-Pei! You can see a lot of other misidentifications here:
Breed Misidentification
Another bit problem with these statistics is they go by the number of bites without taking into account the popularity of a breed. Of course if there are more dogs around of a certain breed, there will be more bites by that breed. This means that the most popular breeds are likely to be higher on the list. The right way to report this would be using a percentage-- how many dogs of that breed are there in the US for example, and how many of those dogs have been reported for a dog bite? The percentage would give you a much better idea. Unfortunately it is very difficult to know the actual numbers of dogs of any one breed because there is no record of that information available. You could use say information from the AKC, but that would be a problem because a lot of dogs are not registered with AKC (and some popular breeds are not AKC registrable, such as the American Pit Bull Terrier for one.) Plus that would leave out all mixed breeds. So it is very difficult to know the actual numbers, but it is pretty clear that many of the dog breeds often maligned as aggressive are VERY popular, meaning the statistics are skewed. German Shepherds for example are usually in the top 3-4 of the AKC's most popular breeds list, and if you added all the unregistered or non-AKC GSDs I am sure the numbers would be much higher. More individuals=more bites.
Another example- if you count "pit bulls" the way they do on bite statistics-- so the term includes American Pit Bull Terriers, American Staffordshire Terriers, mixes which have either of those breeds in them, plus all the blocky-headed dogs who are mistaken for Pit Bulls those numbers would be quite large. Those breeds/mixes are very common/popular, if you go to any shelter here in Chicago they are full of bully breeds. The statistics don't take that into account though, they just list the number of bites, which gives an inaccurate portrayal.