Recently, the Trump administration put forward a new regulation from the U.S. Department of Housing and Urban Development (HUD) that would make it more difficult to prove that protected classes including racial and ethnic minorities are discriminated against in the housing market.
The Disparate Impact Standard is a legal theory meant to catch housing policies that negatively impact groups based on race, national origin, color, religion, sex, familial status, or disability, but which are not explicitly discriminatory. The Supreme Court recognizes this as a form of bias prohibited under the Fair Housing Act of 1968, but the proposed rule-making from HUD would make it nearly impossible for the average person to prove that disparate impacts are actually taking place.
Over the last century, urban planning and both public- and private-sector actions have undermined the basic human value that everyone should be able to live a full, healthy and productive life, reinforcing the legacy of racism and discrimination in the U.S., and creating the region’s geography of segregation, isolation, and inequality. Policies like unequal access to financing, restrictive covenants, blockbusting, redlining, and racial steering divided the region into largely poor communities of color and largely affluent white neighborhoods1.
The proposed rule-making fails to recognize the historic and current relationship between segregation, inequality, and spatial planning. The infrastructure that allowed the New York metropolitan region to grow created environmental injustices, such as urban highway construction that tore apart immigrant and communities of color, and transportation that served only some parts of the population. These instances of injustice cannot be forgotten. They shaped where we live, and where we live is directly linked to our access to opportunity: the jobs we can reach, the quality of our health and schools, and ultimately, the success that our children can achieve2.
Many of these policies exist in different forms today; the housing market still uses source of income or credit scores to weed out applicants; blacklisting excludes renters who have ever been to court regardless of the reason or outcome; exclusionary zoning keeps segregation alive; inequitable school financing perpetuates unequal educational outcomes; energy and environmental policies create unhealthy living conditions for certain residents and not others; and transit services do not serve all residents equally. Each of these policies is a potential example of the Disparate Impact Standard at play.
RPA’s Fourth Regional Plan emphasizes the necessity of strengthening and improving fair housing laws based on the understanding that affordable housing policies alone will not end discriminatory practices. Municipalities are required to abide by a Supreme Court legal ruling from 2015 which found that fair housing laws apply when policies result in disparate impacts, even when there was no explicit discriminatory intent. Implementation of this ruling would be far more effective if regional entities provided resources, technical expertise, and leadership.
Counties, regional councils of government, and metropolitan planning organizations are in the best position to assist in the coordination of land-use policies across municipal boundaries, supporting affordable housing with transportation and other investments, and ensuring police, schools, and other services support fair housing goals3. However, by exempting individualized decisions from the scrutiny of the Fair Housing Act, the proposed rule-making would limit the ability of states and municipalities to advance plans through grants and aid to ensure that policies are not resulting in disparate impacts.
By exempting individualized decisions from the scrutiny of the Fair Housing Act’s Disparate Impact Standard, the proposed changes fail to recognize how most contemporary local land use regulations and zoning changes are made.
The 2015 Supreme Court ruling has an implication to how land use and zoning changes are decided. For example, if a town zoning makes it unreasonably difficult to build housing that low-income renters can afford, it could be in violation of the fair housing law even if it cannot be explicitly proven that the zoning is discriminatory.
Hundreds of jurisdictions throughout the tri-state area require special permits or conditional use permits for a broad range of uses, including important housing types like senior housing and even residences with as few as two units. By requiring special permits rather than allowing the use as-of-right, the local government reserves for itself the chance to review the merits of each project4. Moreover, a preliminary analysis suggests that New York City alone has been relying more frequently on spot rezonings (individualized zoning map amendments). The frequency of these discretionary reviews has doubled since 2016, compared to the prior 15 years, for areas that are on average six times smaller5.
Each review is an opportunity for an individual decision that could have a disparate impact on a prospective minority homebuyer or renter. Yet under the proposed rule change, most of these private spot rezoning applications, and other individual discretionary decisions, would be exempt from Disparate Impact scrutiny.
Segregation by race and income is not an accident; racism and bias stemming from the country’s long legacy of discrimination has persistent effects still felt across the United States. When compared with other large metropolitan areas, New York ranks highest in income inequality and continues to rank as one of the most racially and ethnically segregated regions in the United States6. By exempting individualized decisions from the scrutiny of the Fair Housing Act’s Disparate Impact Standard, the proposed rule change would not only fail to advance the original intention of the Act, but aggravate the negative implications of historic spatial planning. It is backward and self-contradictory.
By detaching liabilities from third-party-generated automated decision making systems and placing the burden of proof on the plaintiff, the proposed rule-making will most likely exacerbate discriminatory housing practices.
From criminal justice to health care to education, employment, and housing, we are seeing computational and predictive technologies used more and more frequently in private and governmental decision-making. As a result, many advocates, academics, and policymakers have begun to raise concerns that unintentional biases could exist within these automated decision-making systems, and argue that adequate safeguards, oversight, appeal, and redress mechanisms are needed to protect vulnerable populations from harm7.
Under the proposed rule-making, lenders would not be liable or responsible for the effects of an algorithm provided by a third party. This will encourage lenders and property owners to use potentially biased algorithms that deliver judgements on credit risk, home insurance, mortgage interest rates, and more. Moreover, placing the burden of proof on the parties claiming discrimination is unjustifiable and antithetical to the spirit of the Fair Housing Act. Assuming that a plaintiff has access to the highly-skilled data engineers or computer scientists needed to prove that an algorithm is biased is not only impractical, but discriminatory by nature. For these key reasons, if put in place, this proposed rule-making from the Trump Administration’s Department of Housing and Urban Development represents a gigantic step backwards in the fight against discrimination in housing.