Is Mass Incarceration a Product of the War on Drugs?

At the end of 2023, state and federal prison authorities had jurisdiction over 1,254,200 people, the Bureau of Justice Statistics (BJS) reports. State prison systems had about 1.1 million of them. The federal Bureau of Prisons had about 156,600, or 12%. That is one prisoner in eight. Local jails held another 664,200 people at midyear 2023. Counting everyone physically held in a prison or jail, BJS put the total at 1,833,700 adults.

State and federal prisons are different systems. Federal prisons hold mostly drug and weapons offenders: 45% of sentenced federal prisoners are held for drugs and 21% for weapons. By contrast, 8% are held for violence. State prisons hold about seven times as many people as federal prisons. They are 62% violent and 13% drug. Of the state drug prisoners, 37,790 were serving time for possession, 3.6% of all sentenced state prisoners. The rest were held for trafficking or for other or unspecified drug offences.

On the latest counts, about 131,000 drug prisoners are in state prisons and 65,000 in federal prisons. Together they make about 196,000 people. That is about 16% of the 1,254,200 in prison. If they were all released, about 1.06 million would remain.

The United States held 542 people per 100,000 residents in prisons and jails in 2023, by the World Prison Brief's count. Removing the roughly 196,000 people in state and federal prisons for drug offences would bring that to about 484. That is still about seven times Germany's rate of 69.

Latest rates, including people held before trial; the US figure counts prisons and jails. The rate without drug prisoners is 1,833,700 minus 196,001 people, per 338.46 million residents. Source: World Prison Brief, country pages.

State prisons were already mostly violent in the 1970s. A national survey of state prisoners in 1974 found 52% held for violent offences and 10% for drug offences, BJS reported in 1982. The drug share then rose from 6% of state prisoners in 1980 to 22% in 1993. By the end of 2022 it had fallen to 13%.

Counts of prisoners are of the people held on a single day, the stock, which favours long sentences. Drug sentences are short.

Take one prison bed held for the 17.5-year median murder sentence. Over those years it could hold about 23 people in turn, each serving the nine-month median for drug possession. A count on any single day finds one person in that bed.

Counting admissions, the flow, gives drug offences a larger share. Jonathan Rothwell of the Brookings Institution added up state and federal prison admissions from BJS series for 1993 to 2011, with federal admissions from 1998 and no data for 2010. Drug offences were about 31% of those admissions, against about 25% for violent offences. In every year from 1993 to 2009, more people were admitted for drug offences than for violent ones.

Jails turn over faster still. Local jails recorded 7.9 million admissions in the year to midyear 2024, BJS reports. State and federal prisons recorded about 472,300 admissions of sentenced prisoners in 2023.

Arrests are larger again. The Pew Charitable Trusts counted more than 1.5 million drug arrests in 2019 in Federal Bureau of Investigation (FBI) data, 87% of them for possession. That was about 15% of the 10.1 million arrests the FBI estimated for that year. From 1993 to 2011, Rothwell counted roughly 30 million drug arrests, 24 million of them for possession. Against them stood about 3 million prison admissions for drugs. That is about one admission for every ten arrests.

Michelle Alexander's 2010 book The New Jim Crow argues that mass incarceration became “a stunningly comprehensive and well-disguised system of racialized social control that functions in a manner strikingly similar to Jim Crow”, as the law professor James Forman Jr. quotes her.

Part of its argument concerns the felony record. Depending on the state and the offence, a conviction can bar a person from voting, jury service, public benefits, public housing, student loans and some jobs, Forman notes. Sarah Shannon and colleagues estimated that 19 million Americans had a felony conviction in 2010. That was 8% of adults and 33% of Black adult men.

Parts of the drug-war account hold. Federal prison did become largely a drug system. Drug offenders rose from 25% of federal prisoners in 1980 to 60% in 1993, the Bureau of Justice Statistics (BJS) reported. They were 45% in 2023. The National Research Council (NRC), in a 2014 report, found that harsh penal policies left their largest imprint on poor minority neighbourhoods.

Alexander's account of the numbers does not hold for state prisons, which hold most prisoners. She writes that the penal population “exploded … with drug convictions accounting for a majority of the increase”, as Charles Fain Lehman quotes her. In the periods in the table below, drug offenders were 20% to 31% of state prison growth, and a majority only in federal prisons.

How to read it: each row is drug offenders' share of the growth in prisoners over the years shown. The two rows marked “about” are computed from the BJS counts in the table below.

The state count for 1985 is of people held in state prisons; the count for 2000 is of sentenced prisoners. Sources: BJS, Prisoners in 1994, Tables 13 and 14, and Prisoners in 2001, Tables 17 and 20 and text.

Forman, who also wants fewer people in prison, made the point in the New York University Law Review in 2012. The New Jim Crow writers, he wrote, encourage readers to see mass incarceration as overwhelmingly a result of the War on Drugs. Yet drug offenders were a quarter of the nation's prisoners and violent offenders a half.

Alexander's wider argument, about a system of control rather than a count of prison beds, is not one these counts can test.

Download this conclusion: the prison boom

This section as one image. Alexander writes that drug convictions accounted for a majority of the increase in the penal population. By Bureau of Justice Statistics counts, drug offenders were about 28% of state prison growth from 1985 to 2000 and about 63% of federal growth.

Download the card (PNG, 1,200 × 630)

Drug enforcement falls unequally by race at arrest, but drug imprisonment explains little of the racial makeup of state prisons.

Surveys find similar drug use among Black and white Americans. The table gives past-year use from the 2023 National Survey on Drug Use and Health, run by the Substance Abuse and Mental Health Services Administration (SAMHSA).

The groups other than Hispanic exclude Hispanic people. Source: SAMHSA, 2023 National Survey on Drug Use and Health, detailed tables 1.24B and 1.27B.

Arrests do not follow use. The American Civil Liberties Union (ACLU), an advocacy group, studied marijuana arrests from 2010 to 2018 in a 2020 report. It found Black people 3.64 times as likely as white people to be arrested for marijuana possession in 2018. In 2019, Black adults made up 27% of adult drug arrests and 12% of the adult population, Pew found from Federal Bureau of Investigation and Census data.

The gap is not new. Patrick Langan of the Bureau of Justice Statistics (BJS) measured it for 1991 to 1993. Black people were then 36% of drug possession arrests but 13% of people who reported using illicit drugs. Factors he could measure, such as the type of drug and where it was used, explained 10 of the 23 percentage points.

Crack cocaine sentencing is another example. The Anti-Drug Abuse Act of 1986 set a 100-to-1 quantity ratio between powder and crack cocaine for mandatory minimum sentences. The Fair Sentencing Act of 2010 cut it to about 18 to 1, the ratio in federal law today. In fiscal year 2023, 79% of people sentenced federally for crack trafficking were Black, against 22% for powder cocaine.

The First Step Act of 2018 made the Fair Sentencing Act's cut retroactive. By September 2021 courts had granted 4,226 sentence reductions under it, 92% of them to Black people, by the US Sentencing Commission's count.

In prison the pattern reverses. Black and Hispanic state prisoners are less likely than white state prisoners to be held for a drug offence, and more likely to be held for violence.

Sentenced prisoners under state jurisdiction on 31 December 2022, by most serious offence. White and Black exclude Hispanic people. Source: BJS, Prisoners in 2023, Table 17.

Black people are still overrepresented in every offence category. They were 26% to 35% of the state prisoners in each category. By comparison, they were about 12% of US adults in 2019.

On BJS's figures for the end of 2022, removing everyone held for a drug offence would move the Black share of sentenced state prisoners from 32.5% to 33.5%.

The Council on Criminal Justice found that the Black-white ratio of state imprisonment rates for drug offences fell from 15.0 to 1 in 2000 to 3.6 to 1 in 2019. That fall made up about half of the narrowing of the overall ratio. The overall ratio went from 8.2 to 1 in 2000 to 4.9 to 1 in 2020. In 2019 the largest remaining gap was for violent offences, 6.2 to 1.

Download this conclusion: race

This section as one image. Black and white Americans aged 12 or older report similar illicit drug use, 27.7% and 26.1% in 2023. But the American Civil Liberties Union found Black people 3.64 times as likely to be arrested for marijuana possession in 2018. In state prisons, 16.0% of white, 10.1% of Black and 9.9% of Hispanic prisoners are held for a drug offence. Releasing them all would move the Black share of sentenced state prisoners from 32.5% to 33.5%.

Download the card (PNG, 1,200 × 630)

John Pfaff, a Fordham law professor, gives a different account in his 2017 book Locked In. Like Forman and the National Research Council (NRC) committee, he wants fewer people in prison and proposes reforms to curb prosecutors' discretion and reduce felony filings.

As Jeffrey Bellin quotes the book in the Michigan Law Review, from 1994 to 2008 the number of felony cases filed in state courts “rose by almost 40 percent”. That matched a 40% rise in state prison admissions. Over those years, Pfaff says, crime and arrests fell while conviction rates and sentence lengths held steady. His reading is that prosecutors brought more felony cases against fewer arrestees.

Most other researchers hold that harsher sentences contributed substantially to the growth, Bellin writes. The NRC committee found that it “was driven largely by changes in sentencing policy that sent more people to prison for longer periods of time”. From 1975 to 1995 every state and the federal government mandated prison for a wide range of offences, including most drug offences. The policies also included “three strikes” and “truth in sentencing” laws. The committee called the war on drugs “a significant driver of the growth in imprisonment”.

Allen Beck and Alfred Blumstein, as Bellin quotes them, attribute the growth from 1980 to 2010 “about equally” to prison commitments per arrest and to time served.

In July 2001 Portugal decriminalised the acquisition, possession and use of small quantities of all drugs, up to an average person's supply for ten days. People found with such amounts are sent to administrative commissions instead of the courts. Trafficking remains a crime.

Source: World Prison Brief, Portugal, prison population trend.

The prison population dipped, then returned to about where it started. Hannah Laqueur's 2014 study shows what the law changed. Before the law, Portuguese courts rarely imprisoned drug users: in 2000, 25 people were in prison for drug use, under 1% of drug prisoners. The law largely wrote down what the courts were already doing.

What changed was the treatment of trafficking. Trafficking arrests changed little after 2001, but within a decade trafficking convictions fell 40% and the number of people in prison for trafficking fell by close to half. Drug offenders were 43% of convicted prisoners in 2000 and 21% of all prisoners in 2010.

By the end of 2023, the state and federal prison population had fallen 22% from its 2009 peak, the Bureau of Justice Statistics (BJS) reports. It went from 1,615,500 to 1,254,200. The population fell 15% in 2020, and BJS put most of that fall down to the pandemic: courts delayed trials and sentencing, and prison admissions fell 40%. The population fell again in 2021 and rose in 2022 and 2023.

Pew examined the decade to 2019 in the 39 states with data for both years. The number of people held in state prison for drug offences fell 33%, and drug admissions fell 34%. The whole prison population fell 10%, and all admissions 24%. Drug offences accounted for 61% of the fall in the prison population and 38% of the fall in admissions. Their share of these states' prisoners went from 18% to 13%.

Arrests for drug possession fell by less than half of one percent over the decade, Pew found. But arrests for drug sales and manufacturing fell 32%. Drug admissions and the drug prison population fell closely in line with those sales arrests.

States differed widely. In 2019 drug offences were 24% of prison admissions across the 43 states with data. They ranged from 10% or less in Alaska, California, Colorado and Vermont to 40% or more in Arkansas, Kentucky and South Dakota. Eleven of the 39 states held more people for drug offences in 2019 than in 2009. At the end of 2023, Mississippi's imprisonment rate was 652 per 100,000 residents. That was almost seven times Massachusetts's rate of 96.

Offence labels have limits. A robbery committed to pay for drugs counts as a violent offence in BJS data. In BJS's 2016 Survey of Prison Inmates, 39% of state prisoners and 31% of federal prisoners said they were using drugs at the time of their offence. Surveys like these measure drug involvement, not cause: they cannot show how many of these crimes would have happened under different drug laws.

Only in part. The war on drugs reaches far more people through arrests and short prison terms than it holds in prison on any one day. State prisons hold about seven in eight prisoners. There, drug offenders were about 28% of the growth from 1985 to 2000 and 13% of sentenced prisoners at the end of 2022. Freeing every drug prisoner would shrink the prison population by about 16%. In federal prisons, drug offenders were about 63% of the growth.

Ending drug imprisonment would barely change the racial makeup of state prisons, though drug enforcement falls unequally by race at arrest.

Download the conclusion

The verdict above as one image. Its bars show drug crimes at about 31% of state and federal prison admissions from 1993 to 2011. Drug offenders were 13% of sentenced state prisoners at the end of 2022. Releasing every drug prisoner would free about 196,000 people, about 16% of prisoners. The image is 1,200 by 630 pixels, the size Facebook recommends for link previews, and anyone may download it and post it.

Download the card (PNG, 1,200 × 630)