Americans are like St. Louis barbeque sauce—sweet, but thick. We’re pleasant enough most days, but we believe in ghosts, aliens, healing crystals, and phony Facebook rumors. Like Kurt Andersen wrote in his book “Fantasyland: How America Went Haywire,” Americans “have an almost promiscuous devotion to the untrue.”
It’s why Trump is president, St. Louis has the 13th highest murder rate on the planet, and a criminal justice “reform” bill moving through Congress may become law: we’re gullible. We believe what we want to believe, evidence be damned. We may recognize a legitimate problem, but we’re likely to come up with precisely the wrong solution just because it sounds like it should work.
The “First Step Act” that’s passed the House has bi-partisan support in the Senate and is supported both by Trump and leading Democrats. It was designed to do something about mass incarceration and the lack of rehabilitative services in prisons. It’s supposed to start to unravel harsh sentences for non-violent drug offenders and do something about the damage caused by so-called “three strikes” laws. It would mean, for example, that around 3,000 federal prisoners in jail for mere possession of crack would become eligible for early release.
It’s opposed by bug-eyed white nationalists like U.S. Sen. Tom Cotton (R-AR), who calls it a “jailbreak” by mostly black inmates sent up under Reagan-era sentencing laws. But it’s also opposed by U.S. Sen. Kamila Harris (D-CA), U.S. Sen. Cory Booker (D-NJ), and East St. Louis native U.S. Sen. Dick Durbin (D-IL), who say the “First Step” bill doesn’t go far enough, and could make things worse.
As the Leadership Conference on Civil Rights noted, First Step does nothing to reduce mandatory minimum sentences. But one of the group’s main reasons for opposing the bill is that First Step plans to use an algorithm to guess which offenders are most likely to offend again, a crime prediction tool straight out of the sci-fi movie “Minority Report”’s Department of Pre-Crime.
First Step would use a mathematical model to predict whether inmates up for release are low- or high-risk for committing crimes once they’re free. But evidence shows criminal risk assessment algorithms don’t work very well. The Pennsylvania Commission on Sentencing has spent over a decade trying to tweak a crime prediction algorithm without much success. They found that the predictions of which inmates were high-risk were wrong half the time. Of the inmates their algorithm flagged as high-risk, 48 percent never offended again.
If First Step uses a similar algorithm, it would mean half the people up for early release could be denied their freedom because a mathematical formula wrongly said they’d commit more crimes. The reason, as U.S. Senators Harris, Booker, and Durbin, along with U.S. Rep. John Lewis (D-GA) noted in a letter to lawmakers, is that the algorithm uses risk factors that discriminate against African-American inmates. Criminal history, work history, and educational achievement are all used to predict future crime, and in all of those factors black inmates would score worse than white inmates on average.
Despite the lack of evidence that First Steps will be a long-term solution, Senate Majority Leader Mitch McConnell (R-KY) has promised a Senate vote on it before Congress leaves DC for the holidays. Supporters, hoping that it might work, are willing to take a leap of faith.
But there’s a substantial difference between trying something because you hope it might work and trying the same thing over and over again, even though you already know it doesn’t work. That’s the situation St. Louis finds itself in with gun crimes and a breathtakingly high murder rate.
St. Louis leads the country in the number of gun crimes prosecuted by the feds. The idea was that since federal sentences are tougher, and inmates have to serve the full sentence, prosecuting gun crimes on the federal instead of local level would reduce crime by scaring the bad guys. The U.S. Attorney’s Office for the Eastern District of Missouri is now prosecuting more gun crimes than any other U.S. Attorney’s office in the country, and the number of those prosecutions has doubled in 2018 from 2017. Over 600 cases were filed through the end of September, as compared with barely 300 the year before.
The evidence, though, shows the tsunami of federal gun prosecutions hasn’t done much—if anything—to cut the Gateway’s City’s violent crime and murder rate. St. Louis still has a per capita murder rate roughly that of Culiacan, Mexico, the city that’s home to the Sinaloa drug cartel, a rate which makes STL the 13th deadliest city in the world.
Federal prosecution of a city’s gun crimes is a tactic that goes back to Reagan and the cocaine cowboy days of Miami. It had some effect then, since murders in ‘80s Miami were a function of profit. Most killings were related to drug shipments, drug turf, and drug deals. Since lengthy federal prison sentences gummed up the cocaine money machine, there was an economic reason to avoid gunplay and federal court. The program became codified under Bill Clinton’s presidency as “Operation Exile,” pushing longer prison sentences as an incentive to reduce gunplay.
A 2003 University of Chicago study produced evidence that Operation Exile, and the similar programs it spawned, had little effect on a city’s crime and murder rate one way or another. St. Louis is proof. And there are plenty of reasons.
Based on anecdotal evidence from cops and prosecutors, most murders in most big cities in the second decade of the 21st century are crimes of passion, not crimes of profit. Shooters may never think about longer prison sentences.
And there are larger issues at work, especially in dying cities hollowed out of jobs and opportunity. Legacies of racism, disinvestment, poverty, and the easy availability of guns combine in a perfect storm of violence.
But whether it’s Operation Exile or the First Steps initiative, doing something has appeal, even when the evidence shows that something may turn out to be nothing at all.
Charles Jaco is a journalist, author, and activist. Follow him on Twitter at @charlesjaco1.