MASS shootings are guaranteed to command international attention. After 17 people were killed at a high school in Parkland, Florida on February 14th, American politicians and media outlets have resumed the age-old debate about how to prevent such massacres. Unusually for a Republican president, Donald Trump says he wants to make background checks for buying firearms more stringent, a policy that polls show is overwhelmingly popular. Mr Trump’s other suggestion, however, is far more controversial: arming teachers with concealed weapons.

The bloodshed in the town of Parkland brought the total death toll from mass shootings—defined as those claiming five or more victims, including the perpetrator—to 446 since 2014, according to records kept by the Gun Violence Archive, a nonprofit research organisation. That total inched up further to 451 when a man in Detroit shot four people, including the mother of his child, and then committed suicide on February 26th—a crime that would make headline news in any other Western country, but in the United States registers as little more than background noise. Yet even that horrific body count only constitutes 1% of the total number of gun homicides committed in America during that time period, and about 0.3% of all deaths by firearm (including suicides). The bulk of the country’s shooting epidemic occurs via a steady stream of individual murders, most of which generate little attention.

Recent trends in America’s broad murder rate are mildly encouraging. After an alarming 20% nationwide increase from 2014 to 2016, the tally in 50 big cities fell by 2.5% last year. Although the FBI does not release national data until October, an extrapolation of trends from previous years suggests that the figure for the United States as a whole is likely to have fallen by 6% in 2017.

Sources: FBI; NACJD; cities' police depts; The Economist

Even if the 2016 figure represents a peak, however, America’s murder rate still dwarfs that of other rich countries. If policymakers are serious about preventing gun homicides, they will need to confront the evidence: the best predictor of deaths by firearm is the prevalence of firearms. According to a 2013 paper by Michael Siegel of Boston University and others published in the American Journal of Public Health, after controlling for other factors, a one-percentage-point rise in a state’s level of gun ownership between 1981 and 2010 was associated with the murder rate from guns increasing by 0.9%.