For years, the homelessness industry and its advocates insisted that the science behind Housing First—the policy stipulating that homeless people need housing without strings attached or behavioral conditions—was settled: arresting people for public drug use, or requiring treatment, would only push these people deeper into homelessness. The better answer, they confidently maintained, was to give out apartment keys unconditionally and trust that housing alone will put them on a path to stability.
But confidence is not the same as evidence, and the Department of Health and Human Services is now questioning their central claim. Congress, too, is putting decades of dubious federal homelessness spending under the microscope. Meantime, California is providing something the debate has too rarely had: a real-world experiment showing what happens when enforcement is pulled back.
Finally, a reason to check your email.
Sign up for our free newsletter today.
The Golden State’s Proposition 47, passed by voters in 2014, was not intended to affect homelessness. Prop. 47 was a criminal-justice reform that reduced certain drug and property felonies to misdemeanors. But since 37 percent of California’s homeless population uses drugs at least three times a week, and nearly one-third use methamphetamine regularly, the measure’s impact extended far beyond the courthouse.
The unintended consequence, new research suggests, was more homelessness and drug use.
In August, the Public Policy Institute of California (PPIC)—hardly a bastion of law-and-order conservatism—published the most rigorous causal study yet of what happened after the nation’s most populous state largely stopped enforcing its drug laws.
The results weren’t pretty. PPIC found that since Prop. 47’s passage, the state has seen a roughly 10 percent increase in the unsheltered homelessness rate, or around 7,000 additional people sleeping on the streets. This accounts for about one-fifth of the state’s total unsheltered population growth between 2015 and 2019.
Serious drug use rose by about 8 percent, and drug treatment admissions fell by about 20 percent, driven almost entirely by the collapse of one specific pipeline: referrals from courts, jails, probation, and parole. In the two years alone after Prop. 47 passed, criminal-justice referrals to treatment dropped 14 percent in California while falling only 2 percent nationally. Between 2008 and 2017, the state’s incarcerated population fell by about 50,000, while annual treatment admissions from criminal-justice referrals fell by about 45,000—roughly 90 percent of the state’s entire decline in treatment admissions.
Voters did not anticipate that Prop. 47 would change the calculus of law enforcement, making police intervention harder and less effective. Police departments, however, adjusted their priorities accordingly, and felony drug arrests fell from roughly 200,000 in 2013 to under 100,000 by 2015.
The message can’t be missed: when law enforcement waned, it severed the single most reliable connection between addicted Californians and a treatment bed—and drug use and homelessness grew because of it.
Access to treatment for homeless Californians remains strikingly low: fewer than 5 percent of those with opioid histories receive medication-assisted treatment, while about one in five people who wanted treatment in the past six months couldn’t get it. The data in PPIC’s analysis now backs up what common sense suggests: for struggling addicts, reducing enforcement is not a lifeline. It becomes a trap door, dropping them deeper into social isolation and the cycle of addiction.
The PPIC economists also tested (and ruled out) housing costs as an explanation for the spike in the number of homeless Californians after Prop. 47 passed. They found no statistically significant relationship between changes in homelessness and changes in rents, vacancy rates, poverty, or unemployment across California counties. And when researchers added fair-market rent as a control, the estimated impact of Prop. 47 on homelessness remained largely unchanged. Housing costs matter for people on the margins, but they don’t explain why California’s unsheltered population diverged from the rest of the country beginning precisely in 2015.
The PPIC study’s authors are careful not to overstate their findings. They flag that homeless-count methods improved around 2015, California’s rents rose faster than in comparison states, and fentanyl entered the various markets on different timelines. These caveats warrant humility about the precise magnitudes. But they do not redeem the Housing First claim that enforcement makes homelessness worse. No serious study design has ever found this to be the case.
Rather, the PPIC’s analysis provides the strongest case yet for keeping the justice system engaged on drug use and property crimes so that the pathway into treatment for addicts remains open. And now, it strengthens the empirical grounding of President Trump’s broader Treatment First executive order, “Ending Crime and Disorder on America’s Streets.”
Trump’s order directs federal agencies to prioritize discretionary grants for jurisdictions that enforce prohibitions on open drug use, urban camping, and squatting. It calls for expanding drug courts and mental-health courts and allows grantees to make treatment a condition of participation in federally supported programs. It also restores civil commitment as a usable tool for people too impaired to seek help and cuts off federal money to so-called safe-consumption sites. Each provision properly addresses a mechanism identified in California’s Prop. 47 experience.
California offers a warning about what happens when you stop connecting people with serious addictions to treatment: more drug use, more lives lost, and 7,000 more people living on the streets. Washington now has the chance to rectify some of these mistakes.