When Tisha Guthrie learned in 2009 that she’d received a housing voucher that would help with her rent in Baltimore, she was excited. She’d joined the waiting list five years prior, when she became unable to work full-time due to serious health issues (she received both kidney and pancreas transplants, and is also legally blind). 

Yet Guthrie, now 45 years old, found that none of the apartments where she applied would accept her voucher. She inquired at nearly 20 buildings that advertised apartments in her price range. But whenever she mentioned her voucher, landlords said they didn’t take them, or that they didn’t have any apartments available.

“You start to feel defeated,” Guthrie told me. “You feel you’re being stigmatized based solely on having this form of rental payment.” She searched for a year before she finally found a landlord who would take her voucher.

Guthrie’s experience is all too common. Federally funded vouchers like hers (called housing choice vouchers) provided more than 2.2 million households and 5 million people with rental assistance in 2018. The vouchers help make housing affordable for low-income individuals: Voucher holders pay 30 percent of their income toward rent and utilities, while the government pays the remainder (up to the maximum allowable amount).

Yet in major cities from Los Angeles to New York City to Philadelphia to Chicago — and in many smaller ones, too — voucher holders often encounter landlords who refuse to take them or find other ways to avoid renting to them, including falsely claiming that they have no available apartments. 

In response, states and cities have passed legislation known as “source-of-income laws,” which ban landlords from discriminating against people just because they’re using a voucher. Last year, Guthrie joined advocacy groups pushing for a source-of-income law in Baltimore, testifying about her experience and writing an op-ed. In 2019, the city passed such a law, albeit with a limitation that advocates opposed.

At the same time, it’s become clear that these laws are not enough. In some places, discrimination against voucher holders remains common, even with source-of-income laws on the books. Advocates argue that greater enforcement is needed, along with further adjustments to put voucher amounts more in line with fair market rents. And in cities with tight rental markets, they are pushing for increased construction of affordable housing. 

A strong body of research shows that housing vouchers help prevent homelessness, as well as increase long-term health and economic outcomes of children in low-income families. Vouchers are of huge importance to millions of people, but discrimination against people who use them threatens to thwart the progress that’s been made in housing the most vulnerable among us.

Housing vouchers, explained

Federally supported vouchers were introduced with the Section 8 program in 1974. By the late ‘90s, housing vouchers comprised the largest low-income housing assistance program.  

In 2018, 1.2 million households used Section 8 project-based vouchers (used in specific housing developments), and 2.2 million households used housing choice vouchers (“portable” vouchers that can be used for renting apartments in the private market). Other types of rental assistance support specific demographic groups: For example, the HUD-VASH program was created in the early 1990s to subsidize housing for homeless veterans. Many cities and states also fund their own subsidized housing programs.

To determine eligibility for federally funded vouchers, the Department of Housing and Urban Development (HUD) sets income limits, which are based on estimates of median incomes and fair market rents in the area. For housing choice vouchers, 75 percent of recipients must be “extremely low-income,” which means their earnings do not exceed 30 percent of the area’s median income or the federal poverty line, whichever is higher. Unlike other federal poverty-alleviation efforts — like the Supplemental Nutrition Assistance Program (SNAP, known informally as food stamps) — housing vouchers are not considered an entitlement, and only a quarter of eligible households end up receiving federal rental assistance.

When someone receives a housing choice voucher, they have limited time (usually between 60 and 120 days, though extensions are sometimes possible) to find an apartment, after which they pay 30 percent of their income toward the rent (or a maximum amount of $50 per month, if they do not have any income). Local public-housing agencies administer the voucher programs, inspect apartments to make sure they meet minimum standards, and pay rent directly to landlords.

Multiple randomized trials have shown that housing voucher programs really do benefit the people who use them. For example, a HUD-funded study dubbed “Family Options,” by researchers at Abt Associates in partnership with Vanderbilt University, found that vouchers reduced the rate of homelessness or “doubling up” with other households by 18 percent, compared with a rate of 35 percent for the control group.

A separate HUD-funded study, called the “Moving to Opportunity” project, found significant benefits for families that relocated to lower-poverty neighborhoods with the help of vouchers, including better health, higher incomes, and increased rates of college attendance for children whose families moved before they were 13.

Despite their effectiveness at preventing homelessness and increasing housing stability, voucher programs are falling short of their promise. Though originally conceived as a way to help low-income households (disproportionately comprised of people of color) access higher-opportunity neighborhoods, that has proven to be the exception rather than the rule. Families that use housing choice vouchers are often concentrated in poor neighborhoods. For example, a recent study from the Center on Budget and Policy Priorities found that just 14 percent of metro-area families with children living in low-poverty neighborhoods used housing choice vouchers.

Another problem is that many voucher holders struggle to find a landlord willing to accept them. A HUD study published in 2001, for example, found that roughly 70 percent of voucher holders had succeeded in renting an apartment, though the average was closer to 50 percent in cities like New York and Los Angeles. More recent studies have found comparably low success rates of 50 to 60 percent.

Though it’s likely not the sole cause, there’s strong evidence indicating that landlord discrimination contributes to both problems. 

Landlords discriminate against people who use vouchers

Firsthand accounts from many voucher holders illustrate the hurdles they face when searching for an apartment. 

In New York City, I spoke with half a dozen voucher holders who have either been unable to find an apartment after an extensive search, or who did so only after many months (in some cases, years) of looking. 

Forty-five-year-old Sofia (who asked to be identified by her first name only) lives with two of her children in a shelter. Sofia, who is unable to work due to recent debilitating health problems and surgeries, received a city-funded voucher in early June. But after inquiring about dozens of apartments she found online, she said brokers and landlords repeatedly told her they didn’t have any apartments available, or that they’d get back to her later. 

Sofia never heard from them again. 

After five years of unstable housing, during which she and her children stayed with family or in shelters, Sofia said she was hopeful the voucher would help them move into their own place. But after months of searching, she told me, “It’s been really, really hard and so stressful, calling all these brokers and landlords, and it was a big load on me, on top of the load that I’m already dragging.” Hers isn’t the only such story I heard. 

Systematic research corroborates these firsthand accounts. In 2018, the Urban Institute conducted tests by calling thousands of landlords in five cities who were listing voucher-affordable apartments. In three of the cities — Los Angeles, Philadelphia, and Fort Worth, Texas — over two-thirds of landlords said they would refuse to accept vouchers. (In New York City, though it’s clear discrimination is present, there’s little recent data available on the frequency of landlord refusal or the success rates of finding housing using vouchers.)

In response to the problem, 13 states (plus Washington, DC) and more than 50 cities and counties throughout the US have passed source-of-income laws requiring landlords to treat voucher holders the same as they would other applicants. Landlords are also prohibited from turning down voucher holders based on their source of payment.

However, there are still many areas that haven’t passed such laws. The Poverty and Race Research Action Council estimates that as of November 2019, only about half of voucher holders are covered by source-of-income laws. Meanwhile, both Texas and Indiana have passed legislation to prevent jurisdictions in their states from creating their own source-of-income laws. At the federal level, several bills banning voucher discrimination have been introduced, but none have passed.

There is evidence that source-of-income laws help voucher holders find apartments: A study comparing rates before and after a source-of-income law went into effect found significant increases in voucher use, ranging between 4 and 11 percentage points.

Yet advocates argue that more is needed. Research suggests source-of-income laws have only a modest impact on moving households to low-poverty areas. And in some areas, particularly low-poverty neighborhoods, landlords often continue to refuse vouchers despite the presence of such laws. 

For example, in Philadelphia, the Urban Institute’s study found that 67 percent of landlords with eligible apartments refused to rent to voucher holders, regardless of legal protections in place. The rate was even higher in low-poverty neighborhoods. There’s evidence that voucher discrimination is exacerbated by racial discrimination, too: In a recent study from the Federal Reserve Bank of Boston, 80 percent of black voucher holders reported that landlords in low-poverty areas would not accept vouchers, compared with 57 percent of voucher holders from other racial groups.

How to fight voucher discrimination

One way to improve the situation is straightforward enough: better enforcement of source-of-income laws. 

Some cities are starting to take a more active role in this. For example, the NYC Department of Social Services has brought several lawsuits against large landlords for voucher discrimination. 

In 2018, NYC’s Commission for Human Rights formed a new unit to help fight income discrimination, and has since received over 800 complaints from voucher holders. The unit files formal complaints against landlords; its director, Stephanie Rudolph, said another top priority is intervening quickly, enabling voucher holders to move into buildings that had initially ignored them or turned them away. (Advocates in New York City who have praised the work of CCHR are pushing the city to provide greater funding for the unit, which currently has only five members.)

Another key part of improving voucher programs is assisting voucher holders in their search and documenting discrimination, said Annie Carforo, campaign manager at Neighbors Together, a nonprofit organization that assists low-income New Yorkers. Carforo has talked to hundreds of people — many of them living in city shelters — who struggle to find apartments that will accept their vouchers and who have been given little, if any, assistance in their search. In some cases, people are unaware of the laws prohibiting voucher discrimination. “It’s heartbreaking to meet people who have been looking for housing for two years [using vouchers], and who never knew it was illegal for landlords to treat them this way,” Carforo said.

Investing in a greater level of assistance for voucher holders is also a major component of Creating Moves to Opportunity, an initiative in Seattle and surrounding suburbs that helps families with housing choice vouchers relocate to high-opportunity neighborhoods. As Dylan Matthews recently reported for Vox, a randomized study of the program found big results: 54 percent of families in the treatment group that received counseling moved to a high-opportunity area, compared with 14 percent of families in the control group (which had received vouchers only).

In some areas, improving the voucher system could also involve improving the voucher-acceptance process for landlords. Landlords in some cities have pushed back on source-of-income laws, and a recent qualitative study by researchers at Johns Hopkins University showed that some landlords are worried about losing money if they accept vouchers, due to inspections and other kinds of administrative delays. The researchers suggest that making the process more efficient could help address landlord concerns. 

Another approach may be increasing landlord incentives, such as bonuses and tax breaks, which a few states and cities already offer. In addition, the Urban Institute recommends greater investment in actively recruiting landlords to participate in voucher programs, particularly in low-poverty neighborhoods.

The vouchers themselves could use improvements, too. In some cases, the lack of housing choice may be caused in part by voucher amounts that are not properly calibrated to local rents. Voucher amounts have typically been set by median rents in a large area (such as a metro area), but in recent years, there’s been a shift toward calculating median rents within smaller geographic areas (such as zip codes). 

Following a lawsuit, Dallas was the first to try this approach, which led to a greater use of vouchers in low-poverty neighborhoods. After studying the approach in five other pilot areas, HUD shifted to the new way of calculating median rent (called “small area fair market rent”) for 24 metro areas, and other housing agencies can decide to make the change voluntarily. (Though promising, it’s still too early to tell how much this policy shift will help increase mobility on a larger scale.)

No one solution to the problem

But one of the challenges in addressing voucher discrimination — and, more broadly, our housing crisis — is that there’s no one-size-fits-all solution to the problem. As Andrew Aurand, vice president for research at the National Low Income Housing Coalition, told me, “Different housing markets have different needs.” 

In some places, he said, it’s clear there needs to be an increase in affordable housing alongside improvements to the voucher program. Many advocates in New York City, including those working to address rising homelessness, agree.

“We need to be actually thinking about expanding the supply of affordable apartments, instead of just going after an extremely scarce resource with vouchers,” said Jacquelyn Simone, policy analyst at the Coalition for the Homeless in New York City. The Coalition is pushing the city to dedicate 10 percent of new affordable apartments (out of the 300,000 planned by 2026) to currently homeless people, many of whom have vouchers they haven’t been able to use.

But Aurand notes that in other parts of the country, the shortage in affordable housing is less severe — the issue is that extremely low-income households cannot afford to pay even relatively low rents. In those cases, Aurand said, it would be effective to increase voucher funding. Given that only a quarter of eligible low-income households receive vouchers now, there’s a lot of room for expansion. (Several presidential candidates have announced housing proposals that would fully fund federal vouchers for all eligible households, as would a recently introduced Senate bill.)

Improving the voucher system, increasing access to affordable housing, and giving households more choice in where they live would help fulfill the original promise of housing vouchers. Without stable housing, and especially when experiencing homelessness (as half a million people do on any given day), it is extremely difficult for people to stay afloat. 

After years of moving from place to place and staying in shelters, “I’m walking through mud,” Sofia told me. She is still searching, hoping her family will find a home using their voucher. That we make it so hard for people like Sofia is an indictment of a system ostensibly meant to help her.

Sign up for the Future Perfect newsletter. Twice a week, you’ll get a roundup of ideas and solutions for tackling our biggest challenges: improving public health, decreasing human and animal suffering, easing catastrophic risks, and — to put it simply — getting better at doing good.

Albert Woodfox was held in solitary confinement for more than 40 years in a Louisiana prison before being released in 2016, when he was 69 years old. In his book Solitary, published last month, Woodfox writes that every morning, “I woke up with the same thought: will this be the day? Will this be the day I lose my sanity and discipline? Will I start screaming and never stop?”

Thousands of people — at least 61,000 on any given day and likely many thousands more than that — are in solitary confinement across the country, spending 23 hours per day in cells not much bigger than elevators. They are disproportionately young men, and disproportionately Hispanic and African American. The majority spend a few months in it, but at least a couple of thousand people have been in solitary confinement for six years or more. Some, like Woodfox, have been held for decades.

Solitary confinement causes extreme suffering, particularly over prolonged periods of months or years. Effects include anxiety, panic, rage, paranoia, hallucinations, and, in some cases, suicide.

The United Nations special rapporteur on torture, Juan E. Méndez, deemed that prolonged solitary confinement is a form of torture, and the UN’s Mandela Rules dictate that it should never be used with youth and those with mental or physical disability or illness, or for anyone for more than 15 days. Méndez, who inspected prisons in many countries, wrote, “[I]t is safe to say that the United States uses solitary confinement more extensively than any other country, for longer periods, and with fewer guarantees.”

Many practices in the US criminal justice system are harsh, ineffective, even absurd, from the widespread use of money bail to detain unconvicted people to extremely long sentences and parole terms, and a host of other outrages. But placing people in solitary stands out as a violation of human rights.

Well over a century ago in the US, the practice fell out of favor, partly because of its capacity for psychological harm. Yet starting in the 1980s, its use in prisons and jails exploded again.

Over the past decade, there has been a movement to (again) stop the widespread use of solitary. There have been major steps forward in some states. But there’s considerable need for more progress — and wider acknowledgment that this is something that we are all accountable for. As Laura Rovner, a law professor at the University of Denver, put it in a recent talk, “We torture people here in America, tens of thousands of them every day … it’s done in our names, with our tax dollars, behind closed doors.”

A brief history of solitary confinement

In the 1700s, religious groups, including the Quakers, thought that isolating people in their cells with a Bible would lead to repentance and rehabilitation. The Walnut Street Jail in Philadelphia expanded to include solitary cells in 1790, and other prisons and jails adopted the approach over the subsequent years.

A few decades later, the Eastern State Penitentiary in Pennsylvania opened in 1829, the first prison built entirely to keep people in solitary confinement. When Charles Dickens visited the facility about a decade later and met with people who were held in isolation, he wrote, “I hold this slow and daily tampering with the mysteries of the brain, to be immeasurably worse than any torture of the body.”

In 1890, the Supreme Court heard a case in which a person had been held in isolation for a month while awaiting execution. The Court stated that this was “an additional punishment of the most important and painful character, and is therefore forbidden by this provision of the constitution of the United States,” adding that experience with solitary confinement over the previous decades had shown the devastating results on people.

By the early 1900s, the practice had largely been abandoned, in part because it was seen as unethical and ineffective, and in part because it was much costlier.

But that was not to be the last word on solitary confinement. Nearly a century later, in the 1980s and ’90s, the US prison system again took up the practice in full force.

This shift is commonly traced to October 22, 1983, at a federal prison in Marion, Illinois, when four guards were injured and two were killed by people housed in the prison. Administrators at the facility responded with a long-term “lockdown,” in which everyone at Marion was held in their cells for 23 hours per day. The model used at Marion soon spread to other facilities across the country.

The federal and state prison systems began to construct “supermax” prisons, in which a unit of the facility (or the entire facility) is designed to hold hundreds of people in solitary confinement. Pelican Bay State Prison in California was the first newly constructed supermax prison to open, in 1989. Within 15 years, federal and state supermax prisons had opened in 44 states.

Why solitary came back

Why did correctional institutions take up solitary confinement when it had been deemed ineffective and unacceptably cruel 100 years prior?

Researchers often point to a couple of main causes. First, there was a rise in “tough on crime” policies in the ’80s and ’90s. These policies led to many more people —disproportionately people of color — being locked up for long periods of time. Given subsequent overcrowding, corrections administrations were eager to find ways that, they argued, would increase safety and security for staff and incarcerated people.

In the same period, there was a shift in how incarceration was viewed by corrections staff and policymakers. The opinion that “nothing works” to rehabilitate people became popular, and prison was seen much more as a way to lock away dangerous people. (In an unfortunate turn of events, a researcher whose 1974 systematic review helped popularize the view that “nothing works” later found problems with his analysis and recanted, but his retraction was largely ignored.)

Poor data collection and secrecy surrounding solitary confinement over the years also may have played a role in allowing the practice to proliferate. That said, we do have some data. A Bureau of Justice Statistics analysis of the 2011-’12 National Inmate Survey (in both prisons and jails) estimated that about 18 to 20 percent of incarcerated people spent some time in solitary confinement in the course of a year. A National Institute of Justice report, using the number of people incarcerated in 2013, calculated that about half a million people spent some time in solitary confinement at some point in the year, or 90,000 on a given day.

The most recent nationwide estimate of 61,000 people in solitary on a given day in 2017 comes from a survey of state prisons and a few large urban jails, conducted by the Association of State Correctional Administrators and the Liman Center for Public Interest Law at Yale Law School. As Solitary Watch, a nonprofit that aims to raise awareness about the practice, and the researchers who conducted the survey point out, this estimate is likely to be considerably lower than the total number, given that it omits anyone held in solitary for less than 15 days, as well as those held in other facilities such as local jails, juvenile detention, and immigration detention centers.

Beyond lack of data transparency, many facilities — especially supermax prisons — are also largely closed off to observers. Some prisons forbid anyone who doesn’t personally know a person housed there to visit. That restriction even included Méndez, the former UN special rapporteur on torture, who told Solitary Watch that he requested permission to visit prisons in the US for years without success.

What is it like in solitary confinement?

Researcher Sharon Shalev describes typical solitary confinement conditions in her 2009 book Supermax: Controlling Risk Through Solitary Confinement:

• Cells are about 7 or 8 feet by 10 feet in size (slightly bigger than the average bathroom or elevator)
• People are held in their cells for 22.5 to 24 hours per day; when let out, it is into a small, solitary outdoor cage with no recreational equipment
• No group activities or congregating with others
• Very few activities or programs
• Limited visitors, and then only through a thick glass barrier with no physical contact

Many firsthand accounts from people who have experienced solitary attest to these conditions.

Justice Rountree, who spent five years in solitary and is now an advocate with the New Jersey Campaign for Alternatives to Isolated Confinement, said in a recent panel that compared to regular prison, solitary “feels like losing your freedom.”

Rountree and many others describe being kept awake by constant shouting and banging from others in cells. In many cells, there is no window, and sometimes even from the outdoor cage where they are allowed to go (by themselves) for an hour to pace back and forth, they can’t see the sky. Shaka Senghor, who spent seven and a half years in solitary, described the smell as “defecation, unwashed armpits … [mingled] with the pepper spray officers use to extract prisoners from their cells.”  Some have even described how they begin to hallucinate.

Far from isolating only people who have been violent, it’s very common for corrections to put people in solitary for trivial reasons. A 2015 report from the Vera Institute of Justice describes how “disruptive behavior — such as talking back, being out of place, failure to obey an order, failing to report to work or school, or refusing to change housing units or cells — frequently lands incarcerated people in disciplinary segregation.”

Facilities vary as to what extent they allow people books and other materials while in solitary. Even if reading materials are allowed, they are often censored. In the absence of social connection, many people describe feeling unable to connect normally afterward. Sarah Shourd, who was held in solitary for more than a year in Iran, wrote, “… I couldn’t look into another person’s eyes without physical discomfort. … A touch on the shoulder made me flinch and tense up.”

For some people, particularly teenagers and those with mental illness, this disconnection can be lasting. Kalief Browder was 16 when he was arrested for allegedly stealing a backpack in New York City. His family couldn’t afford the bail, and he spent three years at a Rikers Island jail waiting for court hearings, two of them in solitary — where he tried to kill himself a number of times. After his charged were dropped and he went home, he isolated himself, often staying in his bedroom and pacing as he had done in solitary confinement.

At the age of 22, he died by suicide.

What does research show about the harms of solitary confinement?

Research over the past few decades has documented the effects of solitary confinement. In congressional testimony in 2012, psychologist Craig Haney summarized: “Most of the research has reached remarkably similar conclusions about the adverse psychological consequences of solitary confinement.”

Haney gives the example of his 2003 study of 100 randomly selected people held at Pelican Bay, the supermax prison in California. Haney found that virtually all of his interviewees reported heightened anxiety, irrational anger and irritability, confused thought processes, and being extremely sensitive to external stimuli. Some 70 percent felt themselves to be on the verge of a nervous breakdown, about 40 percent experienced hallucinations, and just under a third reported suicidal thoughts.

According to Haney, these symptoms closely matched other studies of people held in solitary confinement for a period of months to years, and were much more severe than in general populations of prisons and jails. Haney and other psychologists including Stuart Grassian have long argued that these symptoms develop and increase while people are confined in solitary, rather than merely being preexisting symptoms.

The desperation that people feel in solitary confinement can lead to psychological breakdown, self-harm, and suicide. A 2014 study of New York City jails found that while only about 7 percent of people spent time in solitary confinement, they accounted for nearly half of all acts of potentially fatal self-harm. Studies have shown that a quarter of suicides (or even more) behind bars occur in solitary confinement.

The risks of extreme harm to people in solitary are greater for vulnerable groups, such as those with mental illness and disabilities. In 2012, the American Psychiatric Association released a statement saying that with rare exceptions, people with serious mental illness should not be placed in solitary.

Yet prisons and jails very often do just that. In a Bureau of Justice Statistics analysis of data in 2011-’12, nearly 30 percent of those held in solitary in prisons reported severe psychological distress, with a further 20 to 23 percent reporting mood and anxiety disorders.

There’s also widespread agreement among researchers that the risks to young people in solitary confinement are particularly severe. Depriving young people of sensory and social contact has a heightened risk of serious and lasting effects. The federal system and many states are restricting the duration of solitary confinement or banning solitary confinement for youth, but in recent years, the practice has still been common in prisons and juvenile detention facilities in some states.

Not all researchers hold the same view about the harms of solitary confinement. Some are more skeptical about past research showing serious harms, and they question how much we can infer from studies that often lack a comparable control group. In recent years, some researchers have also pointed to a 2011 study in Colorado that purported to show evidence that those in solitary for months to a year fared no worse psychologically than similar people in the general population of the prison.

They also point to two systematic reviews that combine the results of only studies that directly compare those in solitary confinement to those in a control group. Both reviews claim that after pooling those studies, solitary has only a modest negative impact on mental health. In response, Haney and others have pointed to a number of serious methodological problems with the systematic reviews and with the Colorado study.

Does solitary confinement do what it purports to do?  

Let’s sidestep this debate and ask a different question: What’s the evidence that solitary confinement achieves positive results?

Corrections departments have long argued that solitary is effective at maintaining safety and security in prisons. But the evidence does not support this view.

A 2016 report from the National Institute of Justice stated, “There is little evidence that administrative segregation has had effects on overall levels of violence within individual institutions or across correctional systems.”

The few studies on the impacts of increased solitary confinement do not show a reduction in violence among people held in the facilities. For example, a 2006 study of three states that opened supermax prisons showed no subsequent statewide reduction in violence among those housed at the prisons.

Furthermore, there’s little evidence that solitary meaningfully improves safety for staff in prisons and jails. To be sure, correctional officers have an extremely difficult job, and it’s important that they are able to go to work without being in danger. Many who work in corrections believe that solitary confinement plays a role in keeping them safe. As Gary Mohr, director of Ohio Department of Rehabilitation and Correction, wrote, “[O]ur staff, those who work in the trenches of our prisons, firmly believe the use of restrictive housing as a default disciplinary sanction is tied directly to their safety.”

However, as in the case with violence in prisons generally, there is no strong evidence that solitary is keeping officers safer. The 2016 National Institute of Justice report found that few studies have focused on the effect of solitary confinement on subsequent misconduct (including violence against staff). A large study in Ohio found no evidence of any effect of solitary on subsequent violent misbehavior. In states like Colorado and North Dakota, which have dramatically reduced the number of people in solitary confinement over the past several years, corrections directors report that there has not been an increase in violent incidents against corrections staff. And while the 2006 study of three states that opened supermax prisons did show a reduction in violent incidents against staff in one of the three states (Illinois), it found no effect in Minnesota and an increase in such incidents in Arizona.

There is a legitimate question of how to protect vulnerable people, such as people with disabilities, LGBTQ people, and others, in prisons. Protecting such populations has often been given as a reason for using solitary.

But there are other options. The Vera Institute, which has worked with corrections departments across a number of states on reducing solitary, reports on effective ways to keep people safe other than solitary confinement. Vera, along with the American Civil Liberties Union and the American Bar Association, recommends keeping people in a safe, separate area of the facility with others, and with full access to programming and services. As far as putting youth in solitary confinement within adult facilities to keep them safe, advocates argue that they shouldn’t be in adult prisons to begin with.

Finally, advocates often point out that the vast majority of people housed in solitary will be returning to the community, where they are expected to function. Solitary makes that transition even more difficult.

Abandoning solitary confinement. Again.

Over the past decade, there’s been a surge of attention and reform on solitary confinement. Advocacy groups have been pushing for change, including the National Religious Campaign Against Torture, California Families to Abolish Solitary Confinement, and the New York Campaign for Alternatives to Isolated Confinement.

The Vera Institute has worked with a number of jurisdictions on reforming their practices. Among the group’s recommendations are that solitary confinement should:

• Never be used on vulnerable groups such as those under 18, pregnant women, and those with mental illness or mental/physical disabilities
• Rarely be used as discipline, and then only for violent offenses
• Used with the least restrictive conditions possible, providing access to medical and mental health care outside of the cell, visitors and phone calls, and daily hours of programming with other people
• Never be used directly prior to releasing someone back into the community.

In 2016, President Obama reformed the use of solitary confinement in federal facilities, including banning it for those under 18 and limits on its use for adults. But since the federal system holds about a tenth of the people incarcerated in the US, these reforms only affect a small number of the total held in solitary confinement.

Some reforms have been driven by actions from within prisons and litigation. For example, thousands of people held in Pelican Bay and other prisons in California participated in a series of hunger strikes in 2011 and 2013, protesting their treatment, including the application of indefinite solitary confinement.

In 2012, the Center for Constitutional Rights filed a federal class-action lawsuit (Ashker v. Governor of California), resulting in a 2015 settlement that required California’s corrections department to release many of those who had been in long-term solitary and to reform their rules.

Other lawsuits in recent years have succeeded in banning and restricting solitary, largely focused on vulnerable groups such as youth, people with mental health issues, and pregnant women. However, as in California, it’s been clear that corrections sometimes try to find ways around new rules.

There’s also been a wave of legislation at the state level. So far, state bills have largely secured protections for people in vulnerable groups, or have mandated (at the very least) that prisons report data on their use of solitary confinement. However, two states — New York and New Jersey — are considering bills that offer sweeping reforms for everyone who is incarcerated.  

Finally, internal change is also taking place within some corrections departments. In 2016, the American Correctional Association — a nonprofit that provides guidelines to corrections departments and facilities across the US — issued new standards on restrictive housing (its term for solitary confinement). These included banning its use for more than 30 days for pregnant women, people with serious mental illness, and young people under 18. Many jurisdictions reported that they were changing their policies as a result of the new guidelines.

A few corrections departments went much further to improve their policies. In 2017, Colorado put into place some of the most progressive policies in the country, limiting solitary confinement to the UN standard of no more than 15 days. North Dakota is another state that has significantly reformed its use of solitary confinement.

When thinking through the movement against solitary confinement, we should see it within the broader context of our criminal justice system. Hugely important changes — such as reforms to bail, parole/probation, and sentencing — will likely go a long way to reducing people in solitary confinement, since there will be fewer people in prisons and jails to begin with. Amy Fettig, the deputy director for the ACLU’s National Prison Project and Director of the ACLU’s Stop Solitary campaign, told me, “We have to get people out of prison, and we have to get people out of solitary confinement. [Both are] part of a systematic effort … we’re confronting a system that is so profoundly broken in so many ways that you can’t fix one problem without fixing the others.”

The Supreme Court has stated that what counts as “cruel and unusual punishment” in the Eighth Amendment must “draw its meaning from the evolving standards of decency that mark the progress of a maturing society.”

Have our “standards of decency” evolved enough for us to stop this practice?

Sign up for the Future Perfect newsletter. Twice a week, you’ll get a roundup of ideas and solutions for tackling our biggest challenges: improving public health, decreasing human and animal suffering, easing catastrophic risks, and — to put it simply — getting better at doing good.

For a period of time from the 1980s to the early 2000s, “microloans” were all the rage in international development.

The idea was simple enough: By giving a very small loan to someone living in a poor country, you could help them expand a small business, which would lift their family out of poverty. When they pay back the loan, the money can be cycled to more borrowers, getting more families out of poverty.

Organizations offering microcredit to poor borrowers — many living on $2 or less per day — took off in those decades. Investors and donors poured money into microcredit, hundreds of organizations offered loans, and the number of borrowers worldwide skyrocketed to 211 million by 2013.

The microcredit movement has been undeniably successful in opening up financial services to poor people across many countries. But what has its track record been when it comes to lifting people out of poverty?

Over the past decade, this question has occupied researchers, who have conducted randomized studies across a variety of countries and settings. The findings have not supported the original hope for microcredit: They can’t find evidence that the loans have been lifting families out of poverty on average. Many concluded that the classic conception of microcredit was based much more on anecdotes than on robust evidence. Those results have in turn cooled the development community’s enthusiasm for microcredit.

But does this mean that microcredit has been a failure? Hardly.

Rather than see microcredit as it was portrayed in its heyday — as a way to get people out of poverty — we should see it through a different lens: as a way to expand options for poor people by offering more reliable financial services. Extremely poor people need these services just like everyone else, and the availability of capital to deal with irregular and at times unpredictable incomes is a huge help to them. This benefit, along with its impressive growth around the world, arguably makes microcredit a success.

While earlier claims about microcredit’s benefits were overblown, there is mounting evidence that it nonetheless plays a valuable role in improving the lives of people in need. Financial diaries of people living on $2 or less per day have shown that microcredit helps many families deal with emergencies, make critical purchases that they couldn’t otherwise afford, and put food on the table in times of scarcity. While the new story about microcredit isn’t the one that propelled it to such heights, it’s much more grounded in evidence and, in many ways, it’s still inspiring.

A brief history of microcredit

Lending money to the poor isn’t a new idea. In his book Due Diligence, David Roodman describes the long history of microcredit, going back to Jonathan Swift (yes, the author of Gulliver’s Travels), who began to lend small amounts to poor people in Ireland in the early 1700s.

Even though microcredit isn’t new, it has long faced some core difficulties. One basic issue with lending to extremely poor people is the cost: Because the loans are often small (averaging a few hundred dollars), the overhead costs are higher as a proportion of the loan, and it’s harder to make lending profitable.

Another problem is predicting who will repay a loan. In poor communities, lending has long taken place locally between people who already knew each other (local moneylenders and family/friends), with social ties that could help ensure repayment.

Another extremely common form of lending has been credit cooperatives, in which people — often living in the same region and/or affiliated through a particular trade — could obtain loans. But organizations from outside a given community don’t have access to information that could help them judge who to lend to. On top of that, those living on $2 or less per day often do not have collateral to put up as a guarantee for the loan. In light of these difficulties, lending to the poor wasn’t widely seen as promising.

However, that changed in the late 1970s and early 1980s, with a new vision of how to offer microcredit to the poor, and what it could do for them. Economist Muhammad Yunus played a big role in shaping this new perspective.

In his book Banker to the Poor, Yunus describes meeting a woman in Bangladesh who was making stools out of bamboo and earned only two cents per day, because she had to repay so much money to her bamboo supplier. If she had a dependable source of credit, Yunus thought, she and others in similar situations could make their way out of poverty.

That idea, along with his conviction that “all human beings are born entrepreneurs,” led him to found Grameen (meaning “village”) Bank in 1983. He also took the crucial step of convincing outside funders, such as the Ford Foundation, that it was a good idea to invest in loans for the very poor.

The original Grameen Bank model included a few core elements. The first is that after a loan for a microenterprise is granted, repayment starts immediately, with frequent, regular payments over the course of a year or so. The second is group loans, in which a small group of borrowers from different households receive loans together — which then puts pressure on the members to help each other repay. Finally, the model cuts overhead costs by having loan officers hold weekly meetings in villages to collect and disburse payments, obviating the need for physical bank branches.

Grameen Bank played a big role as a catalyst for microcredit’s huge expansion (which some called a “revolution”). A huge number of organizations all over the world entered the scene over the next two decades (more than 3,000, as reported in 2015), though most borrowers are clustered in a few countries such as India and Bangladesh. Borrowers repay loans to microcredit institutions at very high repayment rates, upward of 96 percent on average.

Grameen Bank wasn’t the first group to take on lending to the poor — the nonprofit Accion, working independently in Latin America in the 1970s, also developed a similar idea, and in Bangladesh, the nonprofit BRAC was an early pioneer — but it played a critical role in creating a powerful example of how microcredit for the extremely poor can work. As Roodman writes, other groups had done similar things, “but had never hit on a formula that combined such high repayment rates, manageable costs, and scalability to millions of people.”

Beyond the model for lending, Yunus also heavily promoted a vision for microcredit’s promise that proved hugely influential. Tim Ogden, managing director of the Financial Access Initiative, says that before Grameen Bank, there was a consensus that it was bad to lend to those living on only a dollar or two per day, because it would only trap them in debt. After Yunus began to talk about loans helping people to exit poverty through micro-enterprises, there was a “huge transformation” in the perception of microcredit.

Ogden describes this transformation: “You’re loaning money to a woman who is earning a dollar a day? How is that not going to trap her in debt? Oh! She’s starting a business and earning more money than I’m charging her.” Without this narrative, microcredit might not have taken off as it did.

Female empowerment also became integral to the story. Many microcredit institutions (including Grameen) made it a priority to lend to groups of women (about 80 percent of microcredit borrowers are now women). Investors and donors poured money into microfinance, and in 2006, Yunus won the Nobel Peace Prize.

The inspiring narrative falters

In the 2000s, skepticism about the promise of microcredit started cropping up. One concern critics raised was the possibility that some microcredit institutions were harming people. In Andhra Pradesh, a state in southeastern India, the government issued an ordinance in 2010 essentially shutting down microcredit institutions, pointing to over-indebtedness, the pressure to repay loans, and widely reported suicides among borrowers.

There’s also been a long-running debate about what level of interest is acceptable versus exploitative. On average, institutions offer loans at annualized interest rates of around 20-30 percent, though some rates are much higher. While some people — including Yunus — have argued interest rates above a certain level means that microcredit firms have turned into predatory loan sharks, others counter that the rates sometimes have to be high to cover costs of sustainably lending to the poor.

Beyond concern about potential harm, researchers started to seriously, and publicly, question the narrative about microcredit allowing millions of people to get out of poverty. From the beginning, that story had rested largely on anecdotes from borrowers, which might not necessarily be representative.

There was some more systematic research to back up the claim: One of the main studies that supporters pointed to was a study published in 1998 by researchers Mark Pitt and Shahid Khandker, which claimed that borrowers — especially women — were getting out of poverty at significant rates in Bangladesh.

However, when Jonathan Morduch and David Roodman reanalyzed the study, they found issues that made them question the reliability of the results. (Morduch first commented on the original study, which led to a series of replies, and replies to the replies, that continued for a period of more than 15 years.) This, along with the lack of other rigorous studies, meant that there was a big evidence gap for the first few decades of microcredit’s expansion.

Over the past decade, there’s been an influx of more systematic evidence on microcredit. Randomized controlled trials (RCTs) are a particularly good method for gauging impact, since they make it easier to distinguish causation from correlation.

The most recent six microcredit studies, published in 2015, were conducted by economists working independently across six countries. The studies found fairly consistent results: None found evidence that income went up on average among those offered credit. A few saw modest positive effects, such as people choosing to spend more time on their small businesses and some changes in spending habits. Abhijit Banerjee, Jonathan Zinman, and Dean Karlan sum up the studies, concluding, “We note a consistent pattern of modestly positive, but not transformative, effects” — not the result that many people had hoped for.

But in some ways the findings were also good news. For one thing, they countered the backlash that had been brewing against microcredit: Some critics argued that microcredit hadn’t just failed to lift people out of poverty, it was in fact even systematically harming people by trapping them in debt. But the RCTs didn’t find systematic evidence of this claim.

For another thing, these results are only a disappointment if one thought that microcredit would get most participants out of poverty. To be sure, this was a common belief, but many researchers say that that hope wasn’t realistic to begin with.

In a recent discussion about the history of microcredit, economist Bruce Wydick compared microcredit in poor countries to introducing credit cards in rich countries, as a way of explaining why we shouldn’t be surprised. “When they introduced credit cards in the US, so that almost everybody had access to a credit line, did that pull millions of people out of poverty? No,” Wydick says.

But just because it doesn’t pull most borrowers out of poverty doesn’t mean microcredit hasn’t helped people.

For evidence of microcredit’s value, look at how it helps the poor live day to day

So how does microcredit help people, if not by raising their incomes on average? Research that looks closely at the financial lives of people living on $2 or less per day, such as the work of researchers Daryl Collins, Jonathan Morduch, Stuart Rutherford, and Orlanda Ruthven in Portfolios of the Poor, shows that credit often plays a crucial role in borrowers’ lives.

Part of that story is that people very often use microcredit for their day-to-day needs, rather than for business loans, as Yunus had originally envisioned. They may have a need for cash to meet emergencies, or for a big purchase, or even just to provide an inflow of money to put food on the table when income fluctuates — and microcredit helps to meet that need.

In fact, microcredit organizations are far from the only source of credit — people often take small loans from friends and family, or from local shopkeepers, for example.

But a really valuable aspect of microcredit is its reliability: People can depend on getting a loan at a certain time, then commit to the small, regular repayments so that they can get a further loan.

As Jonathan Morduch writes:

Incomes are seldom steady and predictable; needs vary as well: families need to pay for schools, medicines, and food during slow periods…Evidence that microfinance loans are used to fund non-business needs (even if for education or health) is sometimes used to criticize microfinance, but that misses the point….poor families, like richer families, need broad financial tools. In fact, the poor may need them more urgently.

There are also other potential benefits of increasing access to credit. In Due Diligence, Roodman also points to Nobel Laureate economist Amartya Sen’s view about the value of increased freedom, in the sense of greater agency in one’s life. By giving the poor a greater number of options in how to navigate their financial lives, Roodman suggests, microcredit can increase this kind of freedom.

Roodman emphasizes that the details matter — some ways of offering microcredit might offer more freedom than others. For example, he writes that group microcredit “emerges in a surprisingly negative light” when looking at financial diaries. Groups that are “responsible for each other’s loans can generate ‘peer support’ in times of difficulty — or peer pressure to pay no matter what.”

There’s also some evidence — from a study in the Philippines by economists Xavier Giné and Dean Karlan — that group liability might not be necessary to achieve high repayment rates. Over time, some institutions have moved toward individual loans instead, while still keeping the group meetings.

Finally, there’s evidence that suggests that microcredit may be playing a broader positive role. For example, economists Emily Breza and Cynthia Kinnan studied what happened in Andhra Pradesh, India, when microcredit institutions were shut down in 2010. They found that this was followed by a notable decrease in wages in rural areas. They write that this result “shows that microfinance, despite its small loan sizes, can have meaningful impacts on rural economies.” It also suggests that the full picture of microcredit isn’t captured in studies that only look at individual borrowers.

Now what about cost? Recent research from the World Bank has shown that the vast majority of microfinance is subsidized, in the sense that investors and donors are providing capital at below-market rates. To the extent that the classic story included the claim that a lot of microcredit institutions would eventually sustain themselves without subsidy, it hasn’t turned out to be true in that regard either. However, the subsidy isn’t so expensive, amounting to a median of around $25 per borrower.

So to sum it all up: Microcredit seems to be very important in the lives of the poor, even if it’s not transformative. Given that it comes at a relatively low cost, it may be that microcredit is quite a cost-effective way to help people.

Comparing microcredit to other ways of helping people

All this said, some readers may want practical advice: Should they contribute to microcredit institutions?

This raises its own set of tough questions. What is the relative effectiveness (and cost-effectiveness) of microcredit compared to other potential ways to help extremely poor people, including just giving cash? How much should an investor or donor be worried about harm to some borrowers? How much “room for more funding” do microcredit institutions now have, and which ones are the most cost-effective?

On the one hand, it’s useful to consult a review by nonprofit charity evaluator GiveWell, which does not recommend microfinance institutions as “among the best options for donors looking to accomplish as much good as possible.” GiveWell notes that microfinance is far from a failure, but finds the evidence for the benefits of cash transfers to be clearer, and expresses concern about potential harm to some microcredit borrowers. (Full disclosure: I have worked for GiveWell in the past.)

On the other hand, some researchers who have looked closely at microcredit and who have done work to compare the costs and benefits of microcredit compared to other programs to help the poor, including economists Jonathan Morduch, Asli Demirgüç-Kunt, and Robert Cull, argue that it could very still well be in the running when compared to other programs. A big reason for this is the low cost of subsidies for microcredit, which could make the program cost-effective in spite of modest average benefits. “It’s an open question,” Morduch told me, how microcredit fares when compared to other programs such as cash transfers, adding that, “There should be more serious comparative cost-benefit work.”

It’s also worth making a final point: While RCTs haven’t found that microcredit raises incomes for average borrowers, there is a small group of people who do achieve higher business profitability when they receive loans and sometimes these returns are really impressive, far exceeding interest rates. (In one study, researchers referred to this group as “gung-ho entrepreneurs” as opposed to reluctant ones.) So an ongoing research question would be whether it’s possible to find ways to better target people when offering loans for business, or whether tweaking the terms of the loans might improve business profits.

Whether or not one concludes that microcredit beats cash transfers or other ways to help the poor, there’s still reason to believe that microcredit has done — and continues to do — a lot of good, at fairly low cost. Beyond that, there’s reason to believe that there may be ways to make small loans (as well as broader financial services such as micro-savings and micro-insurance) even more useful to those living on very low incomes.

To some, the new vision of microcredit — helping poor people to better face their financial challenges — may not hold the simple allure of the old one. But the researchers who wrote Portfolios of the Poor and looked closely at the lives of those living on $2 per day find the new narrative inspiring nonetheless: “Whether or not the microfinance movement was right to stress loans for microenterprises, or has been too slow to embrace savings and other services, its greatest contribution is, to us, beyond dispute. It represents a huge step in the process of bringing reliability to the financial lives of poor households.”

The story of microcredit illustrates that even where a program doesn’t live up to the hype, it can still be a success.

Stephanie Wykstra (@swykstr) is a freelance writer and researcher based in New York.

Sign up for the Future Perfect newsletter. Twice a week, you’ll get a roundup of ideas and solutions for tackling our biggest challenges: improving public health, decreasing human and animal suffering, easing catastrophic risks, and — to put it simply — getting better at doing good.