There are many instances where statistical data is misused, from vaccines to GMO foods, and breed specific legislation. Speaking on the recent trends against pit bulls, though in the past it was dobermans and even chihuahuas.
Cities across the country are banning breeds based on one reported statistic, a statistic that changes every few years. If they were smart they'd look at the trends, a dog with the highest reported incidents of bites gets banned one year, a few years later it's a different dog. The previous dog bans are lifted and behold, that dog does not go back to the top spot again.
If breed had any real impact on it the bans would only have an effect when enacted, but we see that, in reality, they do not. Someone once said that when you observe something, the act of observing it alters it. This is the most probable reason the statistics keep changing.
Let's use an example we can all understand, the occurrence of autism. Autism was thought to be very rare in the past, due to a lack of attention on the subject only a small portion of the population knew much about it and the symptoms are usually difficult to spot.
So everyone just thought it was as rare as a human born with a tail. Medical scientists raised the awareness of autism as an attempt to garner more funds for research, suddenly almost everyone knew the disease existed, and the research got funding.
With the extra funding methods for diagnosing it improved drastically, combined with the now heightened awareness, more people with the illness were discovered. It was reported, and thus recorded, far more than before. A bunch of jerks then used this increased reporting as a scare tactic to gain money and fame.
It worked because people are panicky when they don't know the facts, and when in a panic they are easily swayed by pretty faces. Dog breed bans have also had the same kind of leeches, feeding on the panic they create by using a very simple phenomenon to create the appearance of a problem which does not really exist.
Statistics, for the most part, should only be read by scientists until the general population learns what confirmation bias means, and the media is banned from reporting false causes. Any animal, even the human animal, can be dangerous if they feel threatened. Any animal, even a rodent, can seriously harm the perceived threat.
Every dog is a pit bull, and every human is a pit bull as well. You just need to slant the statistics by altering the perception of the general public in a way to change what they notice.