Ohio State's Randolph Roth studies the reasons Americans kill—and it's not what you think.

In January, when I first telephoned Randolph Roth, professor of history and sociology at Ohio State University and author of the book “American Homicide,” it was to talk about the news that the number of murders in Columbus had hit an all-time high. The 2017 body count was 143, breaking the previous record of 139 set in 1991. As a percentage of population, the city's murder rate is still lower than it was in the early 90s, but after years of decline, this new increase in killings had caught the public's attention. I wondered whether Roth, who has spent more than three decades combing through homicide records across the country to learn about what makes some eras more murderous than others, could explain this new phenomenon.

By the time I sat down with him in his campus office in February, two police officers had just been shot to death in Westerville and 17 students and teachers had been gunned down in a Florida high school. Guns and homicide were the topic of the day, and I was that much more curious whether Roth might offer some perspective.

In Roth's 2009 book, 655 pages long and based on close study of tens of thousands of U.S. homicides dating back centuries, the author makes a surprising claim: The best explanation for changes in the murder rate is political. When trust in government and institutions is high and social cohesion is strong, Roth says, the murder rate goes down. When trust declines, homicides increase.

“When we lose faith in our government and political leaders,” Roth wrote in an essay published last fall in theWashington Post, “when we lack a sense of kinship with others, when we feel we just can't get a fair shake, it affects the confidence with which we go about our lives. Small disagreements, indignities and disappointments that we might otherwise brush off enrage us—generating hostile, defensive and predatory emotions—and in some cases give way to violence.”

“As abstract as these sentiments may seem,” he wrote, “they predispose certain people to kill. In fact, they explain homicide rates better than any other factor, including unemployment, guns, drugs or a permissive justice system.”

His essay encourages me to ask Roth the obvious question about Columbus' murder spike: Isn't it a result, mainly, of the opiates and heroin epidemic? He shakes his head. “I just don't think that's a sufficient explanation,” he says. “We've been the center of the heroin trade now for 15 years. Why wasn't that drug trade violent 15 years ago?” He points to a 2016 study of the national increase in homicides, conducted for the National Institute of Justice, which raised the same point. “It is not obvious why the increase in homicide would lag at least five years behind the explosive growth in the demand for heroin, if the expansion of urban drug markets spurred the homicide rise.”

Recent murders attributed to the violent El Salvador-based gang MS-13 have contributed to the body count, Roth acknowledges, but they haven't been numerous enough to explain the increase in homicide locally. In 2016, Columbus saw 106 homicides; the number was 99 in 2015.

So why is murder suddenly on the rise? Roth, whose book covered the period from the Colonial era to the 1920s, is currently studying the modern era. His research, he cautions, is not complete. But he thinks his earlier theory holds up. Across the nation, Roth says, recent murder rates have reflected our fractious political environment, with regional differences appearing to support his argument.

To illustrate, he points to the region that voted more strongly Republican in 2008 than in 2004. On the voting map, it's a stripe of red that stretches from West Virginia across eastern Kentucky and Tennessee through Arkansas and Oklahoma, including portions of Louisiana, Florida and Texas. While the rest of the U.S. was shifting left in a wave that put Barack Obama in the White House, this region was moving to the right. In those counties, Roth says, the murder rate increased 20 percent in 2009—Obama's first year in office. At the same time, murders in urban areas with large minority populations declined by 15 percent.

Now, says Roth, the reverse is happening. White supremacist groups are feeling validated while African-Americans feel increasingly marginalized, and the murder rate in urban areas, especially those with a large African-American population, is increasing. The trend began before Trump's election, in 2015, coinciding with a widely publicized series of police shootings of unarmed black men, Roth points out. The spike in 2017, both here in Columbus (where 111 of the 143 murder victims were black) and in other big cities like Chicago and Cleveland, does not surprise him.

“The homicide rate is going up among African-Americans,” says Roth. “It's going up in the cities. It's about Black Lives Matter and the end of the Obama presidency, or what we should really say is the deliberate, willful destruction of the Obama presidency.”

“Every time you increase one group's trust in the government, you undermine another group's,” he concludes. “It's a hell of a situation.”

While the murder rate in some cities has tilted upward the past couple of years, it's important to note that it's still at a historic low for the U.S—although much higher than in other affluent nations. The aggregate rate in 2016 was 5.9 killings per 100,000 people; during the Colonial period it was sometimes over 100. After the Revolution, the murder rate dropped dramatically. In some parts of the country, it dipped as low as 3 per 100,000. Then, in the tumultuous years leading up to the Civil War, murders increased again. They remained high throughout the bitterness of Reconstruction, although medical advances tempered the increase, as some survived who might earlier have died. Throughout the 20th century, the rate has fluctuated between about 4 and 10 per 100,000 population, with a protracted low during the 1950s and peaks in the early '80s and '90s. Currently, the national rate is less than half what it was then.

Today, Roth is working, along with colleagues Wendy Regoeczi, a criminologist at Cleveland State, and Rania Issa, an analyst for the city of Cleveland, on a study of murder in 34 Ohio counties from 1959 to 2010. It's painstaking work. To know what causes changes in the murder rate, they have to figure out what that rate is—and the data reported to the FBI, says Roth, is often wrong, undercounting murders for bureaucratic reasons such as coding errors, cold cases, and high coroner caseloads. So they gather the facts themselves, in much the same way as he gathered the data when researching “American Homicide.”

They start at the county courthouse. “You look at every scrap of paper,” says Roth. “You go through the docket books, you go through the jail records. You go through the case files. You go through the inquests. You go through everything. And then you read every issue of the local and county newspaper. And then you go to the vital records and look at all the death records.” Roth, a former Stanford math major, uses a formula to compare the official and unofficial records and come up with a more accurate total.

He posts the raw data, containing accounts of thousands of murders, on the web. It's called the Historical Violence Database, hosted online by Ohio State's Criminal Justice Research Center. “I'm trying to give future historians a chance to re-think things without having to do the massive amount of work I've done,” he says.

Once he establishes whether murder is rising or declining, Roth looks at the political and social climate. In the period since World War II, he has polling data to rely on. But in the 17th, 18th and 19th centuries, he's had to look at other evidence. First, there's the known history: war vs. peace, prosperity vs. strife. But Roth looks to other clues as well. A rise in executions for treason or sedition indicates challenges to government legitimacy, while executions for witchcraft or moral offenses suggest feelings of division and hostility. Deadly riots are a symptom of political instability. And when people name lots of counties and other places for national heroes, patriotism and trust in government is likely higher.

It is not an exact science, Roth concedes. But he says that after testing all kinds of hypotheses—immigration, changes in race or gender relations, weapons laws, policing efforts—the only consistent connection over time and geography was the one tying murder to trust.

Academics look at problems in order to understand them, not necessarily to solve them. If Roth is correct in his assessment that today's increase in the murder rate arises, just as it did in earlier eras, from political instability and social division, it's hard to know what to do with that information. Do we need to accept that increasing violence is here to stay until we reach some kind of political detente?

Again, the historian wants us to look at history for a clue. The recent rise in mass killings, for instance, is not as new as we may think. “Mass murders have happened all through American history,” says Roth. The difference? “It used to be a group activity.”

Rapid-fire guns and high-capacity ammunition changed the equation. Roth thinks the most effective policy response to the current scourge of mass shootings would be to limit the number of rounds of ammunition in a magazine. “Every time we've had a mass murder that's based on technology, we've acted. 9/11, we hardened the cockpit door, required two crew members at all times in the cockpit,” he says. After the Oklahoma City bombing, the government placed restrictions on fertilizer purchases. Earlier in the 20th century, the government responded to anarchist bombings by regulating dynamite, and gangland murders resulted in taxing silencers so heavily that they disappeared from the marketplace. “There are things you can do,” he says.

Guns don't make us violent, Roth says, but they increase the carnage. He supports the Second Amendment, but sees a need for some restrictions. In the end, Roth sets aside his professorial stance and invokes his other persona: former Eagle Scout and assistant scoutmaster. In the Boy Scout handbook, he says, rights and responsibilities are laid out on pages directly opposite one another. “We have rights, but we have responsibilities,” he says. “If people who want to exercise their rights would realize it's not just a one-sided thing, we could do better.”

Want to get Columbus Monthly in your inbox? Sign up for our e-newsletter.