Reading our newsfeeds these days, it can feel like society is crumbling around us. But are incidents of violence — particularly gun violence, and shootings by police officers – really on the rise in the United States? Or is it our awareness of these events that's increasing?
The three factors at work here are the actual incidence of violence, the reporting of that violence (to authorities, by authorities and in the media) and the perception of that violence by everyone involved.
Overall Rates of Gun Violence in the U.S. Are Dropping
In 2015, the Pew Research Center published an analysis of the death certificate data collected by the Centers for Disease Control and Prevention (CDC) over the past few decades. This data indicates that, per capita, America's gun homicide rate has decreased by over 50 percent since its peak in 1993. In that year, there were seven gun homicides per 100,000 people. The rate remained more or less flat between 2000 and 2014, at which point there were 3.4 gun homicides per 100,000 people.
The rate of nonfatal violent firearm crime dropped even more precipitously during the same time period, according to Pew data from the U.S. Justice Department: from 725.3 victimizations per 100,000 people ages 12 years and older in 1993, to 174.8 in 2014. That's a decrease of about 76 percent.
However, Gallup polls consistently indicate that Americans believe violent crime is on the rise. When asked "Is there more crime in the U.S. than there was a year ago, or less?" a majority of Americans polled have answered "more" from 1994 through 2014 (with the exception of 2001, in which answers of "less" were more popular by about one percentage point).
Perception: How Our Brains (Don't) Work
To understand the discrepancy between the actual incidence of violence versus our perception of it, we should consider both the way that media reports of violence affect us and the way that our brains process statistics, anecdotes and emotion.
As much as we like to compare our brains to computers, it's a terrible analogy. Most memories aren't cold, hard data: They're data tangled up with whatever emotions we have about that data. Neurobiological research has shown that typically, the more of an emotional response we attach to information, the more vividly and accurately we remember it. And because we rely on prior experiences to process new ones, this creates a bias in how we perceive new information.
Furthermore, when you recall a memory, it's not like a computer opening a file — it's more like trying to reconstruct a tower of building blocks with your hands covered in jam. Every time we call up a memory (i.e., a particular series of electrochemical impulses) we distort that memory slightly based on our current mood and situation.
All of this makes humans — even math-oriented humans — terrible at statistics and probability. Consider the common thought experiment called the birthday paradox: If you get a random group of 23 people together, there's a 50/50 chance that two of them will have the same birthday. It sounds impossible, but the math works out.
The Ways We Get Our Media Are Changing
Layer all of these ways in which our brains mangle data on top of the type of data we have coming in, and it's clear why we're misperceiving the incidence of violence around us. A trend analysis out of the Columbia Journalism School found that since the 1950s, news media have been sweepingly misrepresenting violent crime incidence in their coverage. Just one example: In 1981, 0.4 percent of the crimes committed in New Orleans were homicides, but about half of crime news focused on the subject.
And that was in 1981. The news media has changed a bit since then. Newsroom employment is down over 42 percent from its all-time high in 1990. The Great Recession hit the journalism industry particularly hard because it came during the rise of digital media, which vastly disrupted print media funding. As a result, we have about 33,000 full-time news journalists left in the United States.
Compare that to the fact that by the most recent numbers, about 158 million American adults are smartphone users, all of whom have instant access to publishing tools via social media platforms. And many of whom are using that access to share stories that might have not been reported had they happened in the past, like the deaths of Philando Castile and Eric Garner.
Reported Rates of Killings by Police Are Rising
The decreasing gun homicide rate mentioned above does include fatal shootings by police officers. However, when you break that statistic out on its own, the story is different. According to Pew via the CDC's data, fatal police shootings rose from 333 incidents in 2009 to 464 in 2014.
Investigative reports place the deaths caused by police officers in 2015 at double to triple those numbers. The Washington Post reports 990 deadly shootings by police that year; The Guardian reports 1,134 deaths overall.
Black citizens, and black men ages 15-34 years in particular, are disproportionately the victims of this deadly force: Young black men make up 2 percent of the population but account for 15 percent of these deaths, and are nine times more likely to die in police incidents than any other demographic, according to The Guardian.
It's unclear whether these dramatic increases indicate actual increases in incidence, or whether they're artifacts of how the data were gathered and how sharply these incidents have been underreported or misreported by government agencies. In any given year, the recorded number of killings by police officers varies widely among the CDC, the Federal Bureau of Investigation (FBI) and the Bureau of Justice Statistics.
Again, we must turn to investigative reports. The Wall Street Journal looked into internal records from 105 of the country's largest police departments for the years 2007-2012. They found at least 550 police killings that weren't mentioned in the FBI's official files for that time period. Some local agencies didn't report killings that they considered justifiable homicides. In other cases, confusion between or within agencies prevented reports from being made. For example, Washington, D.C. inexplicably stopped all reporting for a decade beginning in 1998.
Further complicating our perception of the violence happening around us, mass shootings have been on the rise in the U.S. since 2011. And data collected by organizations around the world suggest that global violence has been rising since 2007, driven by conflicts in areas such as Syria, Iraq and Afghanistan. These conflicts align with stress and disenfranchisement resulting from the global financial crisis and climate change, and come in contrast to prior trends of increasing peace since the end of the Cold War.
Can What We Don't Know Help Us?
So gun violence in the U.S. is declining, but is police violence on the rise? We don't have the full answer yet; no one does. The statistics are messy, and the ways that our brains cope with them are even messier. It seems probable, if not inevitable, that all of us — civilians and members of our police force alike — are internalizing these local and global stories, compounding our perception of violence.
It's possible that this could this add to tensions, creating an actual increase in violent incidents. More optimistically, perhaps our awareness will help drive changes to make our most vulnerable populations safer.