Jeffrey A. Singer
Earlier this month, in a Wall Street Journal op-ed, Charles Fain Lehman highlighted a working paper released on July 30 by criminologists (Boehme et al.) from the University of South Carolina and the University of Nebraska that examined the effects of short-term drug decriminalization in Oregon and Washington on arrest and crime rates.
In November 2020, Oregon voters approved Measure 110, a ballot initiative that decriminalized drug possession and increased funding for harm reduction programs. The measure took effect in February 2021. Later that month, the Washington State Supreme Court declared the state statute criminalizing narcotics possession was unconstitutional. Washington lawmakers recriminalized illicit drug possession three months later, classifying it as a misdemeanor rather than a felony. Oregon lawmakers recriminalized drug possession in mid-2024, making possession a misdemeanor carrying up to six months in jail.
It’s important to note that decriminalization is not the ideal policy reform. Ending the practice of imprisoning people for possessing or using substances the government disapproves of—and saddling them with felony records—is a big step forward, but it keeps prohibition in place. People must still buy from the black market, where there’s no guarantee of purity, dosage, or even the true identity of a substance. As a result, decriminalization alone is unlikely to reduce the risk of overdose deaths. Decriminalization also doesn’t remove the criminal element from drug markets, where dealers resolve their disputes with violence rather than with lawyers.
A more effective reform would be to end prohibition and regulate drugs the same way we do alcohol.
Boehme et al. examined daily crime and arrest rates for various offenses in Oregon and Washington from 2019 through 2022, using the FBI’s National Incident-Based Reporting System data from 23 states. They applied two separate approaches—estimating two-way difference-in-differences models and constructing a synthetic control model—to assess the impact of decriminalization on crime. Their results showed that “at the state level, Oregon and Washington experienced significant increases in most serious violent and property crime,” driven mainly by Portland and Seattle. These results, however, do not tell us the mechanism by which decriminalization caused these changes—a key point missing from Lehman’s column.
The researchers suggested several possible reasons for their findings, including weakening social cohesion, failing to jail chronic offenders, and broader “de-policing.” Boehme et al. is a working paper, so the authors still have time to include some of these variables in the statistical models of their next version.
Lehman also noted that drug overdoses spiked during the same period. In a letter to the Wall Street Journal’s editor, I challenged the notion—implied though not stated outright—that ending penalties for possession directly drove increases in crime and overdoses. It is implausible to think drug users suddenly turn to violent or property crime simply because they face no penalty for possession. Portugal’s 2001 decriminalization and the Czech Republic’s 2010 reform saw no such rise. And a 2024 Journal of the American Medical Association study by researchers at Brown University and RTI International (Zoorob et al.) showed that once fentanyl’s staggered arrival across the country was factored in—the Pacific Northwest was hit around 2019–2020—the apparent link between decriminalization and overdose deaths disappeared. The real driver was the rapidly shifting drug supply.
On August 11, researchers at Portland State University (PSU) published a final project report, funded by the National Institute of Justice, titled Examining the Multifaceted Impacts of Drug Decriminalization on Public Safety, Law Enforcement, and Prosecutorial Discretion. Focusing solely on Oregon, it covers a much longer period (2008 to 2024) and relies on extensive secondary data obtained through information-sharing agreements with agency partners. Researchers supplemented this dataset with additional information from multiple sources to provide a baseline for their complex analyses. They examined three Oregon interventions that may have influenced crime data: the 2013 reduction in mandatory minimums for cannabis offenses and increased use of probation for drug possession, the 2017 downgrade of drug possession to a misdemeanor, and the 2021 enactment of Measure 110. The report also considers changes before and after the COVID-19 lockdowns and the fentanyl wave beginning in 2019, drawing on Zoorob et al. to provide context. While the report offers a valuable historical perspective, it primarily presents descriptive data.
According to the report’s key conclusions (page 12):
Oregon’s drug policy shifts, including defelonization (2017) and decriminalization (2021), had no significant, sustained effects on property or violent crime rates. Property crime saw a brief increase after Measure 110 but returned to pre-pandemic levels, while violent crime remained stable. The rise in overdose deaths was driven primarily by the COVID-19 pandemic and the proliferation of fentanyl, rather than specific drug policy changes. These outcomes underscore the importance of accessible behavioral health services, economic stability, and social support in mitigating the impacts of substance use and enhancing public health.
While the PSU and Zoorob et al. studies do not rely on causal inference methods, they do identify additional variables related to policing, fentanyl prevalence, or COVID-19 that researchers could use in future synthetic controls. Boehme et al. should consider them for the next version of their working paper. This underscores the role of harm reduction programs and accessible addiction treatment in addressing substance use. Analysts and commentators should avoid automatically blaming decriminalization when crime or overdoses increase and instead examine the broader context carefully.