One Hundred Years of Failure: The Drug War in Full

“A hundred years. Forty billion dollars annually. The highest incarceration rate in the world. Drug overdose deaths at record highs. The policy has been both very expensive and counterproductive by its own stated measure.”

The Harrison Narcotics Tax Act became law on December 17, 1914. One hundred and twelve years later, the United States is spending approximately forty billion dollars annually on drug enforcement and incarceration, has the highest rate of drug overdose deaths in its history, and maintains the world's largest incarcerated population, with drug offenses accounting for a substantial portion of those sentences.

By the policy's own stated measure, reducing drug use, reducing addiction, reducing the social harm of psychoactive substances, it has not worked. Drug use rates in the United States are not lower than they were in 1914. Addiction rates in the United States are not lower than they were in 1914. The specific harms that prohibition generates, adulterated drug supply, violence over market control, criminal records that perpetuate the social conditions that drive addiction, did not exist before the policy created them.

Chasing the Scream was published as the centenary of the Harrison Act approached, and Hari structured it as a reckoning with what a century of this policy had produced. The evidence he assembled, following the war from Anslinger's appointment in 1930 to Portugal's decriminalization in 2001, from Billie Holiday's persecution to the Vancouver safe injection rooms, is not primarily an argument about drugs. It is an argument about what happens when a society organizes its response to suffering around punishment rather than treatment.

The racial targeting that John Ehrlichman confirmed in 2016 was not a flaw in the drug war's design, it was, in Nixon's administration's explicit framing, part of the design. The drugs were the pretext; the communities were the target. This is the darkest part of Hari's argument, and the one most solidly supported by primary documentation: the Harrison Act, the Marihuana Tax Act, the Controlled Substances Act, and the mandatory minimums of the 1980s and 1990s were all applied with systematic racial disparity that cannot be explained by differential use rates because use rates across racial groups are broadly similar.

The ACLU documented in 2020 that Black Americans are 3.73 times more likely than white Americans to be arrested for marijuana possession despite similar use rates. The disparity is not a statistical artifact. It is a policy outcome.

What has changed since Hari wrote in 2015 is significant but incomplete. Recreational cannabis is now legal in 24 US states. The DOJ proposed rescheduling cannabis from Schedule I to Schedule III in May 2024. The Biden administration issued mass pardons for federal cannabis convictions in October 2022. Oregon decriminalized personal possession of all drugs in 2020, then partially reversed course in 2024 as implementation challenges mounted. The harm reduction infrastructure has grown: the first sanctioned safe injection site in the US (OnPoint NYC) opened in November 2021 and had reversed over 1,200 overdoses by 2025 without a single on-site death.

And yet: overdose deaths reached a record 107,500 in 2023, primarily from fentanyl and synthetic opioids that entered the supply chain as a consequence of prohibition-driven quality uncertainty. The drug supply is more dangerous, not less, after a century of enforcement.

The Portugal model that Hari cited as the strongest evidence for decriminalization has shown mixed results over twenty years, strong on HIV transmission, weaker on overdose rates when the social investment in treatment and services has been insufficient. The lesson is not that decriminalization fails but that decriminalization alone is insufficient. What works, when it works, is decriminalization combined with substantial investment in the social conditions that drive addiction: housing, employment, community, treatment capacity.

Hari summarized the alternative in the book's final pages: what if the money currently spent on enforcement and incarceration were redirected to the conditions that determine whether people reach for drugs in the first place? This remains the unimplemented proposal. A century of data supports it. The political will to implement it has not yet materialized at federal scale.

Sources

• Hari, Chasing the Scream (2015)
• CDC, Drug Overdose Deaths 2023
• ACLU, A Tale of Two Countries (2020)
• Ehrlichman, John, quoted in Dan Baum, Harper's Magazine (2016)
• Hughes & Stevens, Br J Criminol (2010)
• DOJ NPRM, Cannabis Rescheduling (May 2024)