Since the 1980s, forensic investigators have found examples of mass killers emulating their most famous predecessors. Now, there is growing evidence that the copycat problem is far more serious than is generally understood. Ever since the 1999 massacre at Colorado’s Columbine High School, the Federal Bureau of Investigation has been studying what motivates people to carry out these crimes. Earlier this year, I met with supervisory special agent Andre Simons, who until recently led a team of agents and psychology experts who assist local authorities in heading off violent attacks around the country, using a strategy known as threat assessment. Since 2012, according to Simons, the FBI’s unit has taken on more than 400 cases—and has found evidence of the copycat effect rippling through many of them.

Evidence amassed by the FBI and other threat assessment experts shows that perpetrators and plotters look to past attacks both for inspiration and operational details, in hopes of causing even greater carnage. Would-be attackers frequently emulate the Columbine massacre; one high-level law enforcement agent told me that he’s encountered dozens of students around the country who say they admire the Columbine killers. “Some of these kids now weren’t even born when that happened,” he said. The 2007 massacre at Virginia Tech and other attacks that generated major publicity have also spawned many copycats, according to several law enforcement officials I spoke with.

As part of our investigation into threat assessment, Mother Jones documented the chilling scope of the “Columbine effect”: We found at least 74 plots and attacks across 30 states in which suspects and perpetrators claimed to have been inspired by the nation’s worst high school massacre. Their goals ranged from attacking on the anniversary of Columbine to outdoing the original body count. Law enforcement stopped 53 of these plots before anyone was harmed. Twenty-one of them evolved into attacks, with a total of 89 victims killed, 126 injured, and 9 perpetrators committing suicide. (See more about this data here.)