I trust the reports, and I trust the data, but the trend has been pretty obvious, I'd wager, and predictable for some time now.
People in the suburbs increasingly migrate to cul-de-sac communities, which isolate them from the rest of the world, to make them feel safer, depriving their children of the much-needed contact with a diverse population that builds trust.
Beyond that, the news is more and more pervasive, and more and more sensationalist than it's ever been. Every tragedy is made public, and the more grave and more egregious, the more media coverage it gets. People grow up believing that child abductions are the norm, and even likely, when the reality is that stranger abductions are less likely than they've ever been -- but the Amber Alert system (at least in the DC area here) broadcasts on highways, making it impossible to believe that a kid isn't abducted every two seconds.
Humans are built to be paranoid. Historically, the person who heard a strange noise outside their shelter and didn't react was mauled by a bear, or lion, or what have you, while the more alert human that took to defensive measures was more likely to survive. Having established our positions as apex predators, in cities this is far less likely to be needed, but our brains still seek danger patterns, cling to them, and those instincts aren't able to be quashed by rationality or statistics. We see something like the Newtown shooting and instinctively feel that all our children are in immediate danger, despite its actual unlikeliness. The media catches another school shooting and reinforces it even more in our brains that there is a very real danger, and breeds distrust and paranoia, triggering an instinct to be even more isolationist, and more paranoid, breeding more and more distrust, which further fuels the cycle for the next news article we see reinforcing our beliefs.
This isn't meant to blame the news, or humans, or any particular person at all, but the idea is that we should all at least attempt to be more data-driven, more pragmatic, and to try and find data that doesn't agree with our assumptions. It's harder than it sounds. In most humans, the brain simply rejects data that doesn't agree with its pre-formed hypotheses, but scientists are better equipped here than the average human, as we revel in data, enjoy being proven wrong, and strive towards better results, not just a reinforcement of the bad data we already believe.