June 30, 2020
Incarceration and Poverty in the United States
- The United States currently incarcerates 2.2 million people, nearly half of whom are non-violent drug offenders, accused people held pre-trial because they cannot afford their bail, and others who have been arrested for failure to pay debts or fines for minor infractions.
- Poverty and excessive legal punishments contribute significantly to the United States’ high rate of imprisonment, which has disproportionately affected low-income and minority populations.
- Without reducing poverty—and more specifically, income inequality—as well as racial bias and rolling back harsh sentences for certain crimes, the United States will not meaningfully reduce its prison population.
There are currently an estimated 2.2 million people incarcerated in the United States. The incarceration rate is now more than 4.3 times what it was nearly 50 years ago. This increase has led to the United States having the highest incarceration rate of any country in the world, 37 percent greater than that of Cuba and 69 percent greater than Russia. This high incarceration rate is not because crime has increased; in fact, crime rates have declined since the 1990s. Rather, the arrest rate—particularly for drug crimes—increased dramatically, while sentences have gotten longer.,  These policy changes have disproportionately affected low-income and minority populations, who now make up roughly three-fifths and two-thirds of the prison population, respectively.
This paper surveys the data around incarceration in the United States and connections to poverty. After noting characteristics of the incarcerated and some causes of incarceration, it assesses a recent policy response, the First Step Act.
Who is being incarcerated?
Of the 2.2 million currently being held in the U.S. criminal justice system, nearly 500,000 people are being held for drug offenses, the majority of whom were arrested for simple possession, a non-violent crime. These individuals account for roughly one-fifth of all people held and 44 percent of those held in federal facilities.
Of the 226,000 people in federal prisons and jails, 78,000 (47 percent of the convicted population) are serving time for drug offenses and 22,000 are being held by U.S. Marshalls for drug charges but have not yet been convicted. Of the nearly 1.3 million individuals in state prisons, 191,000 (14.8 percent) are serving time for drug-related offenses. Of the 631,000 held in local jails, 37,000 have been convicted of a drug offense, and 120,000 individuals, representing 25.5 percent of non-convicted individuals, are being held pre-trial for a drug charge.
One-fourth of all those incarcerated in the United States (555,000 people) are being held pre-trial, primarily because they cannot afford to pay bail. Most states require people to pay a cash bail in order to be released from jail while they await their turn in court. Across the country, there are nearly three times as many people being held by local jails who have not been convicted of a crime (470,000) as have (161,000). Nearly three-fourths of individuals held pre-trial have been accused of low-level drug or property crimes or other non-violent crimes.
After accounting for the significant overlap between these two populations, they represent nearly two-fifths (38 percent) of the 2.2 million people currently incarcerated in the United States.
Others are imprisoned indirectly for their poverty, such as violations related to homelessness. A study from the National Law Center of Homelessness and Poverty examining laws related to homelessness in 187 cities across the United States reveals a significant increase in laws criminalizing various behaviors relating to homelessness, such as bans on sleeping, sitting, or lying down in public; sleeping in your vehicle; begging; and loitering. Nine percent of cities have even outlawed sharing food with homeless people. All of these laws make it quite difficult to be homeless and not break the law, creating a vicious cycle: Homelessness makes an individual 11 times more likely to be incarcerated, and being incarcerated makes a person 10 times more likely to be homeless. More than one-fourth of all homeless individuals reported being arrested for activities related to homelessness. With more than 552,000 people estimated to be homeless in the United States, that proportion amounts to 143,000 being arrested for homelessness.
Failure to Pay Child Support
Every state, as well as the federal government, has laws criminalizing failure to pay child support. Evidence shows that compliance with child support orders increases as wages increase; in 2007, 70 percent of child support debt was owed by individuals with annual income of $10,000 or less. This figure suggests that failure to pay is not primarily because of a refusal but rather an inability. Because many states set obligations based on assumed or expected earnings, rather than actual earnings, the median order for obligors with annual incomes lower than $10,000 was 83 percent of their reported income in child support. In 14 states, not only can people be imprisoned for failing to pay child support, but the obligations are not paused while one is in prison and unable to earn income. As a result, in 2007, the average person imprisoned for failure to pay entered prison with a debt of $10,000 and left with a debt of $20,000 and no greater ability to pay while the state incurred costs for imprisonment. Nonpayment of child support was estimated in 2016 to account for the incarceration of 50,000 people.
Inability to Pay Fines
Another significant share of the incarcerated population consists of individuals who have been arrested for a failure to pay debts or fines owed for minor infractions. Many legal infractions are punished through the imposition of fines. Courts also require defendants, guilty or not, to pay fees for myriad necessary services, such as court clerk fees, filing fees, DNA database fees, jury fees, crime lab fees, and late fees. Forty-three states require defendants to pay for their court-appointed lawyer, sometimes even when the accused is found not guilty. At least 41 states charge “room-and-board” for time in prison, and every state, excluding Washington, D.C., requires wearers of home monitoring devices to pay for their use. Failure to pay these fines—or rather, failure to comply with a court order—can result in imprisonment, despite the fact that imprisoning an individual for inability to pay has been ruled unconstitutional.
According to the Brennan Center, nearly every state has increased the use and amount of fees and fines since 2008, largely as a means of raising revenue. Another study from the Urban Institute shows how the share of charges as a source of state and local revenue has increased while sales taxes and property taxes have declined and income taxes have held relatively steady. Nearly half of local governments now receive more than 20 percent of their revenue through the imposition of fines and court costs. In 2010, 10 million people across the United States owed a collective $50 billion in fees, fines, and charges to the criminal justice system. It is unknown how many of these individuals are being imprisoned for their debts.
While the populations arrested for activities indirectly related to poverty—homelessness, inability to pay child support, and non-payment of debts and fines—are harder to quantify, it is clear they represent a significant share of the incarcerated population, possibly up to 10 percent. These estimates show that non-violent drug offenders and people arrested or held indirectly for their poverty account for nearly half of the incarcerated population in the United States.
Poverty and Overcriminalization
The root causes of mass incarceration are poverty and overcriminalization.
While it is difficult to ascertain whether poverty makes someone more likely to commit a crime, data show it does make a person more susceptible to being arrested and more likely to be charged with a harsher crime and to receive a longer sentence. Adults in poverty are three times more likely to be arrested than those who aren’t, and people earning less than 150 percent of the federal poverty level are 15 times more likely to be charged with a felony—which, by definition, carries a longer sentence—than people earning above that threshold. Between 1999 and 2016, people convicted of a crime with at least some college education were given sentences that were between 4.6 and 7.8 percent shorter than individuals without college education. Given that one’s education level is highly correlated with a person’s income, this statistic, too, suggests that longer sentences are imposed on lower-income individuals. All of this results in the share of the imprisoned population that was in poverty prior to being arrested equaling 57 percent for men and 72 percent for women, despite a national poverty rate of 11.8 percent., 
The Brookings Institution found that only 49 percent of incarcerated men were employed in the three years prior to incarceration and their median annual earnings were $6,250; just 13 percent earned more than $15,000. Further findings in the Brookings study reveal a pattern: These individuals are not just more likely to be poor and unemployed, but they were also more likely to grow up in poverty and in neighborhoods with high unemployment. The likelihood that a boy from a family in the bottom 10 percent of the income distribution will end up in prison in his thirties is 20 times greater than that of a boy from a family in the top 10 percent. Individuals are also nearly twice as likely to be imprisoned if they grow up in single-parent homes, even after accounting for differences in income. Children who grow up in poverty are more likely to have developmental issues, which inhibit impulse control, cause low self-esteem, and reduce educational achievements, each of which may contribute to the likelihood of committing a crime.
The High Price of Cash Bail
The use of cash bail continues to grow, despite findings that its increased use correlates with higher rates of failure to appear, rather than lower, and no evidence that it increases community safety. The median bail amount as of 2013 was $11,700; adjusting for inflation yields a median bail amount of more than $12,800 in 2020 dollars. As detailed by the Hamilton Project, cash bail use and amounts have been increasing over the past several decades. Among non-violent felony drug offenders, approximately 75 percent faced bail of $5,000 or more, as of 2009; an estimated 58 percent faced bail of at least $10,000. Among non-violent felony property offenders in 2009, an estimated 63 percent faced bail of $5,000 or more, and nearly 50 percent faced bail of at least $10,000.
Most people in the bottom 20 percent of the income distribution would not have enough assets to pay even the bail bond premium for the median bail—typically 10 percent of the bail amount and non-refundable—let alone the bail itself. The median income of an individual in jail unable to meet bail, prior to their incarceration, is estimated at $16,233 in 2020 dollars, after adjusting the 2015 estimate for inflation; 37 percent had income less than $9,500. In 2015, the median income of such an individual was 61 percent less for men and 51 percent less for women than the median income of their non-incarcerated peers; these differences are even greater for non-White individuals. A study found that the likelihood of being assigned bail was 3.6 percentage points greater for Black defendants compared with Whites, and that their average bail amount is $10,000 higher.
Poverty and Substance Use
In 2017, 19.7 million Americans (over the age of 12) battled a substance use disorder.
Poverty and drug use perpetuate each other and often inhibit escape from the cycles of addiction and poverty; substance abuse may result from poverty as a person uses drugs or alcohol as a way to cope with their financial stresses, and alternatively, poverty can be a result of chronic and expensive drug abuse that leads to overwhelming debt. In 2016, drug and alcohol use cost an estimated $1.45 trillion, including $578 billion in economic loss and $874 billion in societal harm from reduced quality of life.
According to data from the National Household Survey of Drug Abuse (NHSDA) and the National Longitudinal Survey of Youth, drug use is associated with greater poverty because of its effects on education level, human capital investments, and family composition. Drug use, particularly chronic drug use, lowers productivity, reduces earnings, adversely affects educational attainment, and ultimately increases the likelihood of poverty. Furthermore, labor market outcomes and household income levels are significantly affected by family composition and the quantity of income sources. Patterns of drug use are established at relatively early ages, prior to most investments in human capital and education—according to the NHSDA, among those who reported marijuana use, 75 percent first used marijuana by age 18, and among those who reported cocaine use, 50 percent first used cocaine by age 19.
The cyclical nature of addiction and poverty is evident through the following statistics. In 2012, almost twice as many people who were unemployed struggled with addiction compared to full time workers (17 percent of unemployed vs. 9 percent of full-time workers). In 2013, around 20 percent of people on welfare reported using illicit drugs in the previous year. According to research from the Substance Abuse and Mental Health Services Administration, nearly 35 percent of sheltered homeless adults in 2010 had chronic substance use issues—likely a severe underestimate of the overall impact of substance abuse, as it is does not include the unsheltered homeless population. Of jail inmates who were homeless in the year prior to incarceration, 79 percent showed symptoms indicating drug or alcohol use or dependence.
Recovery from drug use is also less likely for those in poverty: An individual who makes $20,000 is one-third less likely to recover from a cocaine addiction than someone who makes over $70,000 a year. According to the 2013 National Survey on Drug Use and Health, the lack of financial resources or insurance was the most commonly reported reason for not receiving treatment: 37 percent of Americans age 12 or older who did not receive treatment for addiction did so because they did not have health insurance or could not afford rehab.
Accounting for Race: Racial Disparities in Arrest and Sentencing Rates
Poverty does not tell the whole story of mass incarceration. Because of who is most likely to be poor in the United States, poverty and its connections to incarceration lead to disparate impacts on minority populations. The data show, however, that even after accounting for poverty, racial disparities in incarceration rates persist.
Of people in prison for drug offenses, nearly 80 percent in federal prison and 60 percent in state prisons are Black or Latino, despite historical data showing that, on average, Whites are just as, if not more, likely to use illicit drugs. Further, users are more likely to purchase drugs from someone of the same race. Thus, the population imprisoned for drug offenses should reflect roughly the racial composition of the general population—or even skew slightly more White—if people of all races were arrested, charged, prosecuted, and sentenced equally. The data show this is not the case.
Regarding marijuana specifically, Black use was 30 percent greater than Whites in 2010, but Black individuals were arrested 270 percent more often than Whites. Black offenders were also nearly twice as likely as White offenders to be charged by a federal prosecutor for an offense that carried a mandatory minimum sentence. This disparity in charges was found to account for at least half of the study’s noted 10 percent difference in sentence length between White and Black individuals.
According to the U.S. Sentencing Commission, however, the sentence-gap is nearly twice that: Overall, Black males receive sentences 19.1 percent longer than similarly situated White males, on average. Further, Black males were 21.2 percent less likely than White males to receive a sentence shorter than what sentencing guidelines suggest or typically require.
Further, while poverty certainly plays a role in the level of crime in a community, income inequality is actually a better predictor of crime rates: Numerous studies have shown that the greater the inequality, the more crime there will likely be in an area. One study examining U.S. cities found that differences in income inequality alone explained 74 percent of the variance in murder rates and 50 percent of the difference in aggravated assaults. In fact, during the Great Depression, as well as the major recessions in 1893, 1907, and 2009 that brought increased poverty but decreased income inequality, crime rates either dropped or remained flat. Another study of three U.S. cities similarly found that when income inequality increased between neighborhoods, so too did the level of property crime, but the poorest areas experienced the least property crime.
The Impact of the First Step Act of 2018
The First Step Act (FSA) was a bipartisan criminal justice reform bill that aimed to reduce the size of the federal prison population and improve criminal justice outcomes. Since it became law in December 2018, significant progress has been made to reduce sentence lengths and to release individuals who have already served significant time.
The FSA addresses outdated sentencing laws, and, most notably, it shortened mandatory minimum sentences for nonviolent drug offenses. The FSA also retroactively applied the Fair Sentencing Act of 2010—which was originally introduced to reduce the disparity between crack cocaine and powder cocaine thresholds for mandatory minimum sentences known to hurt racial minorities—resulting in over 3,000 sentence reductions and over 2,000 inmate releases as of May 22, 2020. In July 2019, the Department of Justice released 3,100 prisoners from the Bureau of Prisons’ (BOP) custody as a result of good conduct under the FSA. The FSA has also allowed for the approval of over 2,000 requests for inmate transfers to a facility closer to the intended city of release.
The FSA authorized the use of home confinement for low-risk, chronically ill, and elderly offenders, and since its 2018 implementation, over 1,000 prisoners have qualified. The passage of the CARES Act expanded home confinement programs in emergency cases and placed an additional 3,000 inmates on home confinement, in hopes of lessening the risk of exposure to the coronavirus in prisons.
In addition to early releases and shortened sentences, the FSA requires the BOP to help inmates apply for federal and state benefits, as well as obtain identification in the form of a social security card or driver’s license. The FSA also reauthorized the Second Chance Act of 2007, which developed guidelines for recidivism-reducing partnerships between prisons and nonprofit organizations, as well as introduced a competitive grant program to provide such services.
Ultimately, the First Step Act is helping to address mass incarceration in the United States and shows that bipartisan criminal justice reform is possible. More remains to be done, however.
The United States is currently imprisoning roughly 1 million people for low-level drug offenses, property crimes, and various offenses indirectly related to their poverty. Roughly half a million people are imprisoned because of their inability to pay for their release. An estimated 10 million people owe $50 billion in legal fees, fines, and penalties. The use of cash bail and monetary penalties punishes people for their poverty, disproportionately impacts racial and ethnic minorities, and fails to provide a deterrent. Future policy solutions should work to alleviate poverty and unemployment and to impose non-monetary punishments for low-income offenders when appropriate. Without reducing poverty and income inequality, racial bias, and the overcriminalization of activities related to poverty, the United States will not meaningfully reduce its prison population.
 Arnold, David, Will Dobbie, and Crystal S. Yang. 2018. “Racial Bias in Bail Decisions.” Quarterly Journal of Economics 133 (4): 1885–932.