Most people today feel a lot safer walking around their neighborhood compared to 20 or 25 years ago, when the rates of violent and property crime were at their peak. Likewise, people felt less safe in the early '90s than they did 20 years earlier, when crime was rising but not yet at its peak. People's fear of their surroundings moves along with the crime rate, with a delay of about two years. See this old post, which looks at data on fear from the General Social Survey, alongside homicide statistics.
Something different is going on with teenagers, though, and probably children as well. The Youth Risk Behavior Survey (site) is a national probability survey like the GSS, only given to high school students — a group that is usually passed over in surveys because there are stronger regulations on studying minors. They have asked questions about your experiences with violence at school, and whether you've skipped school because you feared for your safety there. High schoolers, like adults, live in a much safer world today, but unlike adults, they have only grown more afraid.
The question on school violence asks whether you have been in a physical fight on school property at least once in the past year. The one on fearfulness asks whether you've skipped school at least once in the past month because you feared for your safety at school, or going to or coming from school. These make a great comparison since the potentially dangerous location is the same for both, namely the school.
Both questions were first asked in 1993, and most recently in 2013. I've calculated the changes in both violence levels and fear levels between those two years (only, not mapping the micro-changes in between), separated by sex and race (white, black, and Hispanic/Latino).
I've expressed the change in standard deviations rather than how many percentage points they've gone up or down.* This treats proneness to violence and fearfulness as "bell curve"-shaped traits, as most psychological and behavioral traits are. Standard deviations give us a better sense of how far the average teenager has moved in one direction or the other. It's like knowing how many inches taller or shorter they have become, not merely whether they are some percent taller or shorter.
The data are shown in the graph below (click to enlarge).
Change in violence is on the horizontal, and all values are negative, reflecting the overall decline in violence since the peak in the early 1990s. The change of about 0.5 SD for males and about 0.3 for females is huge, given how quickly they have occurred. If safety were a kind of "height," then males today are about an inch and a half "taller," and females nearly one inch taller, than their counterparts 20 years ago.
Change in fear is on the vertical, and five of the six groups are more fearful than they used to be (i.e., lying within the positive upper half). Bear in mind that the question about fear is not simply about feeling afraid, but actually skipping school because you were afraid. Most groups are around 0.1 SD more fearful. If we thought of feeling secure as a kind of "height," it's as though kids today feel about a third of an inch shorter than they used to — despite being taller!
The males (squares) cluster off to the left, where violence levels have plummeted further and kids are only a bit more fearful than before. The females (triangles) cluster to the right, where violence hasn't plummeted by as much (since it was so low to begin with), and kids are more fearful than before — particularly for white females. There's no overall pattern for where the different races lie within the clusters. (Remember these are changes over time, not absolute levels, which do show race differences.)
Is there at least a positive correlation, despite the general confusion that is shown by most points revealing declines in violence and increases in fear? If groups who have seen larger declines in violence showed larger declines in fearfulness, then the points should slope up and to the right.
When you look within race and compare males to females, the lines connecting points of the same color slope up in two cases and is flat in the other. This is the only sign of kids being in touch with reality here: males have seen greater declines in violence than females, and show greater declines in fearfulness, controlling for race.
However, when you look within each sex cluster, the line running through the three races slopes down! Controlling for sex, the racial groups that have seen greater declines in violence show greater rises in fearfulness.
Just about every way you look at it, there is a profound disconnect between school violence and fear of school violence, among the very teenagers who spend their days there. What gives?
It's not the rise in school shootings, since kids don't purposefully skip school because they're afraid that today is going to be the day when that one kid snaps and shoots up the school. The question is phrased to ask about fear of garden variety fights, bullies, and so on (e.g., the part about being afraid going to or coming from school, where spree shootings would not apply).
It's not cocooning and social isolation either, as though they were afraid that no one would have their back in case a fight broke out. That would apply to adults as well — yet adults are cocooning and less fearful of their neighborhoods, as crime has fallen.
My only good hunch is that it's the coddling from helicopter parents and their agents in the school system, especially all this propaganda about the omnipresent menace of bullying. Grown-ups bombard kids with a constant stream of dire appeals to watch out for bullies, report the slightest hint of bullying — IF YOU STAY SILENT, MILLIONS COULD DIE — etc. This makes it sound like school is one great big death trap, and if every grown-up sends you the same message, you'll tend to trust their expert consensus.
Now, if kids were more in touch with their surroundings, perhaps they wouldn't be so naive about what the grown-ups were warning them about. Over-protectiveness prevents the kid from having an awareness of the school's true danger level (or for that matter, how dangerous his own neighborhood is), which would lead him to dismiss the paranoia of his parents.
Adults may be cocooners these days, but they do get out some of the time, albeit with their guard always up. They are keenly aware of how safe every place has become over time, in a way that their sheltered kids are not, for want of exposure to the outside world.
I think this pattern can be found everywhere in young people's lives today — safer environments, but greater anxiety. And for the same reasons — over-protective sheltering has disconnected the kid's inner view of the world, from the actual state of the world.
Your kid is decently less likely to come home from school with a black eye than he was 20-25 years ago, but they're way more stressed-out and fearful than teenagers were back then. "No problem," the parents assure us — just pump them full of mind-altering drugs to calm them down.
Two abominations do not cancel each other out, it just looks even more fucked-up. Some kid who lives in such a safe world, lumbering around like a lobotomy victim because without the drugs he'd feel overwhelmed by the stress of living in such a safe world.
But y'know, it's really no big deal if the kid is warped into a confused, stressed-out, zombie-like freak — at least the parents feel relaxed that their pet child is all under control. Helicopter parents continue their glib, self-righteous boasting about their adherence to "good clean family values," while producing an entire generation of abominations against nature.
Pity the children, and punish the parents.
* This is La Griffe du Lion's "method of thresholds." Convert the fraction who meet some criterion into a Z-score, for each group, and subtract the Z-scores to see how far apart the means are.