With over a four-fold increase since the 1970s, the United States now boasts the highest rate of incarceration in the world. One in 100 adults are behind bars, and 650,000 return home each year. But where can they live? Although stable housing is key to successful social reentry and preventing recidivism, those with criminal records face enormous barriers in the housing market. They are limited not only by their economic circumstances—facing significant barriers to employment—but are often locked out of the housing they can afford. This makes the Department of Housing and Urban Development’s (HUD) new guidance—which limits the use of criminal history in tenant screening—incredibly timely, if not overdue.

Those with criminal records are not a protected class under the Fair Housing Act, which prohibits “discrimination on the basis of race, color, religion, sex, disability, familial status or national origin.” But because of the disproportionate numbers of African-Americans and Hispanics with criminal records—due in large part to law enforcement practices that have unfairly targeted them—minority renters will be unfairly burdened by blanket rental policies that exclude those who have spent time in prison, regardless of any intent to discriminate. HUD’s new guidance reminds landlords that categorically refusing to rent to people with criminal history may, then, be a violation due to “disparate impact.”

The power of this guidance depends on the actions of one important group of people: landlords. For the past three years, we have led a sociological study of 130 landlords in Baltimore, Dallas, and Cleveland, addressing the key question of how landlords decide whom to rent to.*  While most landlords who rent to poor families will overlook a misdemeanor, few said that they would accept individuals with felony convictions.

Landlords in our study have a variety of official screening techniques at their disposal to sort through tenants: criminal background checks, calling previous landlords, credit checks, visiting a tenant’s current apartment, and verifying income. But many operate far outside this standard toolbox to find the tenants they want. Indeed, it is perfectly legal for landlords to use their discretion when it comes to many forms of tenant screening, but illegal discretion is common too, for example in the case of families with children. While these impressionistic techniques are sometimes used to circumvent fair housing law, they more often reflect the unconscious biases of landlords in ways that may jeopardize the successful implementation of HUD’s new guidance.

The guidance will likely be most effective for managers like Tracy (whose name has been changed to protect confidentiality), who oversees a large apartment complex in Dallas. Well-versed in fair housing law, professionals her like discuss their screening criteria in precise and rehearsed terms. There are small ways in which she can exercise discretion, mostly by marketing properties more enthusiastically to certain demographics, but the actual screening process is largely outside of Tracy’s control. Her complex simply purchases software from the Texas Apartment Association. She plugs in the information from each application and hits submit—the system determines eligibility.

This isn’t just a matter of efficiency. Corporate landlords intentionally take discretion out of the hands of managers like Tracy, reducing vulnerability to discrimination claims. So long as property managers rely on the software algorithms, owners are protected from litigation. But highly professionalized corporate managers like Tracy represent less than half of the low-end rental market. The rest are individual operators owning anywhere from one to a few dozen properties that they manage themselves, making up the rules as they go along.

Gus is one of these “mom and pop” landlords who uses quite a bit of discretion picking his tenants. Now in his early 60s, Gus spent his career at a money management firm where he amassed enough personal wealth to buy a house in Dallas’ tony Highland Park. But when the firm downsized and Gus was pushed from the high-energy world of stockbroking to a staid quasi-retirement, he decided to invest in low-end rental properties.

We spent two days with Gus, riding shotgun in his truck while he went about his business. Gus started off the screening process by text message, sending photos of the unit and a flood of screening questions to potential renters. The first applicant got only to question two. Though he stated his income was $3,500 per month as a contractor, he could not provide proof. Gus noted dismissively, “That guy eats what he kills,” and put the phone back in his pocket.

Later on, Gus met another prospective tenant at a McDonald’s. He ate in relative silence while the middle-aged, African-American woman filled out the paperwork. He collected a $40 application fee, and said he’d be in touch. Back in the truck, Gus confided that he would never actually conduct the background check the fee is intended to cover. Her willingness to be screened was enough. That, and a face-to-face meeting, was all he needed. He accepted her application the next day.

It’s not that Gus thinks screening isn’t important—he’s intimately familiar with the costs of placing the wrong tenant. But he believes that the characteristics of a good tenant aren’t written on their application or in their demographic profile. He seeks some unmeasurable quality—a combination of personal responsibility and stability. At first blush, his strategy appears in sync with HUD’s guidance to take context into account. But like many landlords, Gus’s biases are embedded within a highly racialized worldview.  To illustrate this, Gus noted that most of his tenants are black or Hispanic and he would never reject someone based on race, but in the next breath declared, “If they’re just some n***** I don’t want them.”

Gus’s story embodies two key challenges to the goal of preventing discrimination based on criminal history. First is that Gus’s screening process exists outside of both the legal and illegal practices anticipated by HUD. Taken as a whole, his techniques almost certainly result in disparate impact, but to accurately sort out what criteria he is using to make his decisions is largely impossible even when we witnessed it first hand. In addition, the enforcement regime for a landlord like Gus presents an enormous challenge. Gus, and millions of landlords like him, float under the radar of such evaluations. Individually, they are small-time players, but taken together, they represent an enormous portion of the market.

Get TalkPoverty In Your Inbox

Criminal background checks serve as one of the key mechanisms by which landlords make distinctions—an easy and readily available proxy for responsibility and stability. But they are too often a convenient camouflage for discrimination. HUD’s new guidance hopes to provide tools to litigate non-compliant landlords and incentivize others to rethink their screening policies. However, the policy has blind spots. For example, does the requirement that landlords evaluate criminal records on a case-by-case basis solve the problem? Gus’s story suggests that it may not. Most of the discrimination that we saw occurs on a case-by-case basis, through the gut-feelings of small-time landlords.

Furthermore, the guidance does not apply to the blanket exclusion of renters with drug distribution convictions, who are not protected under the Fair Housing Act. There is a deep irony here. Though the War on Drugs is not solely responsible for mass incarceration, it has nevertheless sent hundreds of thousands of Americans to prison in recent years for nonviolent drug offenses, with a staggeringly disproportionate effect on African-Americans. Those locked up for drug-related crimes made up just over half of the federal prison population in 2014. In other words, a huge portion of those who have spent time behind bars will not be protected under this guidance. This caveat raises larger questions about how those with criminal records can and should be reincorporated into society. HUD encourages landlords to think about whether their practices keep the community “safe.” But if we want citizens from prison to reintegrate, isn’t making sure they find roofs over their heads part and parcel of this endeavor?

Landlords have enormous power when it comes to deciding who lives in their homes. And while discrimination is not always intentional, it can have insidious effects on vulnerable populations. This makes it ever more important to clarify the discretion that landlords have in implementing the new HUD guideline. This will better protect the formerly incarcerated, integrating those who are vulnerable into society by allowing them access to homes, rather than ostracizing them.

*This research received funding from the Department of Housing and Urban Development and the Furman Center for Real Estate and Urban Policy. Opinions expressed herein are solely those of the authors.

The housing voucher Vivian waited years to receive is part of the federal government’s most recent attempt to house the poor. Since the 1930s, it has employed housing assistance as a key tool in its war on urban blight and poverty. But these attempts have often failed those whom they were meant to protect, at times recreating the very inequality they intended to undo.

Post-war, the Federal Housing Administration underwrote home loans for millions of white Americans, while banks systematically denied them to black families, a process called redlining. Even in federally funded public housing, the poor had no respite from the marginalizing forces of inequality. In the 1950s and ‘60s, high-rise public housing was erected in neighborhoods that already suffered from segregation, underinvestment, and decline. And when the Fair Housing Act of 1968 outlawed housing discrimination by “race or national origin,” local housing authorities in cities like Chicago, Baltimore, and Dallas continued to keep two separate housing lists: one for whites, and one for blacks.

Vivian is part of a generation of poor urban dwellers who left the concentrated poverty of high-rise public housing towers, which by the 1990s were crumbling from neglect. Across the country the buildings were torn down, and along with them an entire system for housing the poor was dismantled.  In the place of public housing, the federal government needed a new solution, one that would remedy the concentrated poverty and segregation it had helped to create.

This solution expanded an existing program relying on the private market to house the poor: housing vouchers. These vouchers make up the difference between what a needy household can afford and the cost of a unit in the private market. They are meant to allow families to rent in any affordable neighborhood, offering men and women like Vivian their very first chance to choose where to live. Today, of the five million households across the country receiving some form of federal housing assistance, over half now live in privately owned properties, many through the Housing Choice Voucher program, previously called Section 8.

Get TalkPoverty In Your Inbox

The program has undeniably positive effects. Vivian was able to regain custody of her twin boys thanks to her new home. For Tony Young, a 55-year-old man with HIV, receiving a voucher meant relief from the cold, hard bed under the bridge where he slept when there was no room at the homeless shelter. Joann Jones, a young mother of two, was able to buy fresh fruits and vegetables for her seven-year-old at the local store while he attended a high-performing public school nearby. The basic economic relief that vouchers provide cannot be understated. And they also give families something more: flexibility in times of crisis to respond to the demands of their jobs, their children’s needs, even the whims of landlords. By letting recipients choose where to live, vouchers confer dignity and affirm a sense of belonging potentially free from the stigma of “public” housing. Most importantly, they may help people to realize their dreams of a place to call home.

By untethering federal housing aid from the disadvantaged neighborhoods to which it was once attached, vouchers offer millions of poor Americans the opportunity to move to a new neighborhood where streets are safe, schools have resources to teach their children, and jobs are bountiful. But not everyone does. Vivian, for example, might have used her voucher in a number of safer, more affluent neighborhoods. But time and resources to find an ideal home are limited. And America’s long history of discriminatory housing practices have shaped the residential landscape in ways that cannot be undone by simply offering families a “choice.” Though voucher holders have moved to areas that are less poor than the ones available in the heyday of public housing, many are re-concentrating in poor neighborhoods.

This re-concentration matters for a reason that social scientists like William Julius Wilson have long known to be a social fact, but finally have the hard numbers to prove: where you live matters. It matters for your quality of life, for how much money you make in your lifetime, and for your children. Raj Chetty’s new work shows that a child growing up in a city like Baltimore will make 14 percent less over his lifetime than one in a typical American county, even after accounting for individual factors like income and education.

If where you live is so crucial, then we ought to pay attention to the role housing policy plays in where families end up. In a landmark case this past July, the Supreme Court ruled that housing discrimination need not be intentional to have harmful effects of segregation. This is the first time the legal concept of “disparate impact”—the idea that a policy may disproportionately affect certain groups even absent injurious intent—was applied to federal housing policy. The decision substantiates an important change in the way discrimination persists in contemporary America: we are moving away from the overt racism of Jim Crow, toward one maintained by enduring institutions that inadvertently perpetuate longstanding inequalities—a “racism without racists.” This shift is crucial to understanding how and why racial inequality continues to plague our nation.

Housing vouchers offer a chance to remedy this disparity, but are not yet equipped to fully do so. Not all voucher holders succeed in finding a place to live, and those who do are often unable to find homes in neighborhoods that have jobs and good schools. And although vouchers are a potential tool to dismantle concentrated poverty and segregation, it turns outs that black voucher holders live in neighborhoods that are poorer and far more segregated than those of white voucher holders, revealing the program’s shockingly disparate impact on white and black families. If black voucher holders face obstacles that prevent them from using their vouchers in the same neighborhoods as whites, then something needs to be done.

In their current form, vouchers fail to take account of an important lesson: A roof is not enough. Where you live matters. Vouchers shouldn’t merely keep people off the streets; they should help families move to neighborhoods with more opportunities. What can we do then to make the voucher program work better to reduce inequality? There are a number of policy fixes to reduce barriers that prevent families from using their voucher in low-poverty, integrated neighborhoods. For example, we could do a better job providing mobility counseling and transportation to help families explore new neighborhoods. There are also solutions related to landlords, like passing national legislation that makes it illegal to discriminate against someone who pays their rent with a voucher, and other policies that would encourage landlords in low-poverty areas to accept housing vouchers.

Even with these fixes to modify the disparate impact vouchers often have, it is not enough to simply move the poor out of poverty-stricken neighborhoods. It is imperative that we address the root causes of poverty and inequality by implementing change at the level of the neighborhood itself, improving the environments around poor families by investing in schools, institutions, and the economy. But this systemic transformation cannot take place overnight, and it will face stark political opposition. It remains to be seen how the political climate of the next presidency will unfold to potentially make good on the Fair Housing Act’s recently renewed half-a-century old promise to “affirmatively further fair housing.”

While we wait for political change, we can act to undo the disparate impact this program has on minority families, who don’t fully reap its rewards. Housing vouchers could be a powerful instrument to remedy the indelible dangers of living in a poor environment, for families of all backgrounds.

*Name has been changed to protect confidentiality