When journalists on the fashion beat accept conventional wisdom as fact and don’t look too closely at whether the trends they describe are real, the worst that happens is unstylishness. When reporters covering addiction, homelessness, and mental illness do the same, it can lead to policies that do enormous harm, especially when mainstream media defaults to the idea that cops and coercion are always the most effective way to deal with these issues and refuses to reckon with the abundant research that shows otherwise.

To promote policy that actually works, reporters and editors need to act more like science journalists and less like stenographers who—whether implicitly or explicitly, accidentally or deliberately—bolster political campaigns that use ignorance to drive fear.

It would be hard to find a better example of this problem than Nellie Bowles’ recent essay in The Atlantic, which argues that San Francisco is a “failed city,” in large part because liberal policies have worsened addiction and mental illness. These policies persist, she suggests, because local politicians refuse to confront the empty-headed but well-intentioned delusions of the hippies and their descendants who just want to let it be. She also claims that the recall of progressive district attorney Chesa Boudin in a June 7 election demonstrates that the city is finally awakening from this daze.

Bowles’ work is far from alone in its failure to look at evidence of effectiveness of various policies when discussing the politics around them. In one 24-hour period in June, a columnist for The Washington Postargued that “Boudin’s recall proves that Democrats have lost the public’s trust on crime”—without any mention of data on which policies work best. A similar news analysis from The New York Times also mentioned no actual data. And a New York magazine essay on “Chesa Boudin and the Debacle of Urban Left-Wing Politics” similarly ignored the question of whose preferred approaches are supported by evidence—and whose aren’t.

Bowles writes that her hometown “became so dogmatically progressive that maintaining the purity of the politics required accepting—or at least ignoring—devastating results.” She describes the city’s de facto supervised injection site in the Tenderloin as a place that looks like “young people being eased into death on the sidewalk, surrounded by half-eaten boxed lunches.”

Her argument falls apart in the face of scientific data. Hundreds of studies support the “harm reduction” approach used in clean needle programs and supervised injection sites—and none of them show that it makes drug use or civic life worse.

Indeed, harm reduction was deliberately adopted based on research evidence, not platitudes from the 1960s. Further undermining her analysis, studies overwhelmingly illustrate the counterproductive nature of using cops and coercion first. For one, red states with old-school tough prosecutors actually have worse crime rates than liberal ones like California.

However, since Bowles apparently assumes that harm reduction tactics were adopted because they seemed groovy, she ignores this research base. (Which, ironically, is the type of mindless approach she critiques San Francisco policymakers for supposedly having used.) What she and many other journalists frame as the failure of harm reduction is actually the failure of criminalization.

A brief tour of the data: According to the Centers for Disease Control and Prevention and dozens of other obscure organizations like the World Health Organization, the National Academy of Medicine, and the American Medical Association, clean needle programs dramatically reduce HIV transmission without increasing drug use rates. One study published in the Journal of Substance Abuse Treatment showed that, compared to people on the street who do not, those who participate in syringe service programs are five times more likely to seek more traditional forms of recovery and three times more likely to quit injecting.

What about supervised drug consumption? Here are threereviews of the literature, which show that it reduces HIV risk, injecting, harm associated with injection, and overdose death rates—while not increasing and sometimes reducing local crime and needle litter. (A 2018 review widely touted by critics for suggesting that supervised consumption did not have a significantly positive outcome had to be retracted by the International Journal of Drug Policy due to poor methodology.)

How about the “problem” that Bowles identifies with reduced penalties for drug possession and the increased use of incarceration and coerced treatment she apparently prefers?

Incarceration for drug crimes increases the spread of HIV, Covid, and other infectious diseases. It decreases the odds that people will get or return to effective drug treatment and can double the risk of overdose death and raise suicide risk. Incarceration also impedes people’s ability to obtain jobs and housing, which makes recovery far more difficult.

And what about drug use itself? There’s no association between rates of drug arrests and rates of drug use, as there should be if this served as a deterrent or a prod toward recovery. (Incidentally, leading experts on addiction and the United Nations support drug decriminalization because criminalization harms health without any apparent benefits.)

As for coerced treatment following arrest, this is neither an effective way to diagnose nor to treat addiction and the other mental illnesses that frequently drive it. Less than 5 percent of those sent to treatment by the criminal legal system receive methadone or buprenorphine, which are the only treatments proven to cut the death rate from opioid addiction by 50 percent or more. Mental health care for incarcerated people is also abysmal.

At the same time, research shows that over three-quarters of the most troubled homeless people can be housed successfully if they are treated respectfully and not subjected to demeaning curfews and guest restrictions or required to maintain perfect abstinence. This approach, known as “housing first,” has been responsible for a 50 percent decrease in homelessness among veterans between 2009 and 2019—though you wouldn’t know it from the mainstream media.

Coercion itself undermines treatment success. Effective treatment relies on the ability to be open about relapses, but coercive systems punish this and therefore deter honesty. Further, good therapy typically requires sharing difficult life experiences, which, if forced, can be abusive, traumatic, and counterproductive. High-quality treatment is also compassionate and welcoming, which is exactly what carceral settings are not.

If we want better outcomes for people with addiction, people who are unhoused, and people with mental illness, the answer isn’t to force them into our jails or prisons or our frequently fraudulent and poor-quality addiction treatment system. Instead, spending less on coercion and more on improving quality is both more compassionate and more effective—including at reducing the street disorder, chaotic behavior, and low-level crime that makes affected communities less safe and livable.

To get better policy, we need better journalism. Reporters on all beats need to be trained in scientific and critical thinking. Editors and writers must be genuinely skeptical of conventional wisdom and seek out studies that explore policy solutions, even if they already think they know what works or where an idea originated. They also need to understand how to vet this research—and rather than picking studies to support their anecdotes, choose anecdotes that reflect the best data. (Generalizing from a small sample—i.e., a few district attorney races—is one genre of articles that a more scientific approach would at least temper.)

We need journalism that covers policies on crime, homelessness, and addiction as empirical questions that can be understood through research rather than as purely political battles between loosey-goosey lefties and law and order. In this case, however, the hippies have the data, and the cops don’t.