U N D E R S T A N D I N G B O S T O N Qualified Renters Need Not Apply Race and Voucher Discrimination in the Metro Boston Rental Housing Market July 2020 The Boston Foundation, Greater Boston’s community foundation, seeks to bring the collective power of our region’s people and resources together to drive real change. Established in 1915, it is one of the largest community foundations in the nation—with net assets of $1.3 billion. In 2019, the Foundation received $151 million in contributions and the Foundation and its donors paid $153 million in grants to nonprofit organizations. The Foundation has many partners, including its donors, who have established more than 1,000 separate charitable funds for the general benefit of the community or for special purposes. With support from the Annual Campaign for Civic Leadership, the Foundation also facilitates public discourse and action, commissions research into the most critical issues of our time and advocates for public policy that advances opportunity for everyone. The Philanthropic Initiative (TPI), a consulting unit of the Foundation, designs and implements customized philanthropic strategies for families, foundations and corporations around the globe. Fund for Racial Justice Innovation was established in 2003 by the Boston Foundation in collaboration with the Hyams Foundation in an effort to strengthen partnerships between community-based organizations and lawyers that use legal tools to advance equity resource distribution for communities or groups marginalized by race, color, ethnicity or immigration status. The Hyams Foundation is a private, independent foundation with a mission of increasing economic, racial, and social justice and power within low-income communities in Boston and Chelsea, Massachusetts. Suffolk University Law School The Suffolk University Law School Housing Discrimination Testing Program (HDTP) is a full service fair housing center engaging in housing discrimination testing, education and outreach, enforcement, academic conferences, curriculum development, and empirical research. Situated within Suffolk University in the heart of Boston, HDTP has conducted hundreds of housing discrimination tests, trained thousands in their fair housing rights, won significant damage awards for victims of discrimination, conducted ground breaking empirical studies, and is training the next generation of civil rights lawyers. Analysis Group is one of the largest international economics consulting firms, with more than 1,000 professionals across 14 offices in North America, Europe, and Asia. Since 1981, AG has provided expertise in economics, finance, health care analytics, and strategy to top law firms, Fortune Global 500 companies, and government agencies worldwide. AG’s internal experts, together with its network of affiliated experts from academia, industry, and government, offer clients exceptional breadth and depth of expertise. UNDERSTANDING BOSTON is a series of forums, educational event and research sponsored by the Boston Foundation to provide information and insight into issues affecting Boston, its neighborhoods and the region. By working in collaboration with a wide range of partners, the Boston Foundation provides opportunities for people to come together to explore challenges facing our constantly changing community and to develop an informed civic agenda. Visit www. tbf.org to learn more about Understanding Boston and the Boston Foundation. Design: Canfield Design Cover Photo: Qusai Akoud Unsplash © 2020 by the Boston Foundation. All rights reserved. U N D E R S T A N D I N G B O S T Qualified Renters Need Not Apply Race and Voucher Discrimination in the Metro Boston Rental Housing Market AUTHORS Jamie Langowski, Clinical Fellow, Suffolk University Law School William Berman, Clinical Professor of Law, Suffolk University Law School Grace Brittan, Senior Analyst, Analysis Group, Inc. Catherine LaRaia, Clinical Fellow, Suffolk University Law School Jee-Yeon Lehmann, Vice President, Analysis Group, Inc. Judson Woods, Senior Analyst, Analysis Group, Inc. EDITOR Sandy Kendall, The Boston Foundation O N Acknowledgments The authors thank the testers who participated in this research. It would not have been possible to shine a light on the discrimination that is occurring without their commitment. This research also would not have been possible without the financial support from the Boston Foundation and the Fund for Racial Justice Innovation, established by the Boston Foundation in collaboration with the Hyams Foundation to strengthen partnerships between community-based organizations and lawyers that use legal tools to advance equity resource distribution for communities or groups marginalized by race, color, ethnicity or immigration status. Thank you to Barbara Chandler, Senior Advisor on Civil Rights and Fair Housing at Metro Housing Boston, for her advice throughout and assistance in test design. Thank you to Shannon Seitz for facilitating the connection between the HDTP and Analysis Group. Thank you to Analysis Group and the following individuals for their excellent research assistance: Jeremy Albright, Catherine Alford, Amanda Ballard, Jori Barash, Nick Dadson, Stephanie Lee, Corey McGinnis, Kevin Ward, Solvejg Wewel, Luke Wilder, and Qi Zheng. Thank you to Dean Andrew Perlman and Associate Dean Kim McLaurin of Suffolk University Law School for their support of this work and Anna Katherine Wherren for her legal research assistance. Contents Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 The Study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Findings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 A Brief Overview of Housing Vouchers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Summary of Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Overview of Study Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Assembling Testers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Site Selection: Rental Ad Scraping, Selection, and Randomization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Conversion of Written Reports to Variables for Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Measurement of Discrimination by Analysis Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Measurement of Discrimination by HDTP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Tester Anecdotes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Biased Ghosting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Discrimination with a Smile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Appendices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 APPENDIX 1: Study Assigned Names. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 APPENDIX 2: Tester Report Forms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 APPENDIX 3: Variable Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 APPENDIX 4: Full Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Endnotes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Grace Brittan, Jee-Yeon Lehmann, and Judson Woods conducted the research underlying this report as part of the Analysis Group, Inc. Pro Bono Program, which enables employees to independently participate in projects that contribute to their communities and research interests. Results and opinions expressed herein are the work of the authors and do not represent any opinions or positions of Analysis Group, Inc. Preface The field work for this study about race and housing voucher discrimination was conducted at a time of keen focus on the lack of affordable housing in Greater Boston and the threat that crisis poses to our region’s continued prosperity. Our own Greater Boston Housing Report Cards in recent years have shown that we must expand our housing stock at all price points and in all of the cities and towns around Boston to meet ongoing needs. The analysis in this report, however, shows that just as important as supply is the issue of access to existing and new housing for all— especially for those left behind by Boston’s advancing prosperity. That remains true even as, in the last few months, our region’s focus has by necessity turned to other pressing tragedies. These have included the coronavirus pandemic and all of its fallout, with medical and health issues tied ever more obviously to social and economic issues, as well as the brutal killing of George Floyd and other African Americans, and the nationwide protests that have intensified awareness of systemic racism and inequity in America. This study offers further evidence of the entrenchment of discrimination and the unequal application of rights, with its close examination of Boston’s rental housing practices. To survive and thrive in any city, but especially in Boston with its high-priced real estate market, fair and equitable access to rental housing is essential. It should be a right, not a privilege, that all individuals and families have the same opportunities to secure decent housing, which is so central to our physical and mental health. Yet race- and source-of-income-based discrimination exists and persists to an alarming degree. The data in this report demonstrate that race and class play a major role in how fairly one is treated when seeking housing. Would-be renters who are clearly qualified are being shut out by brokers and landlords who have devised complicated systems of barriers. It’s a situation that we must address by exposing and combating the underlying negative biases that pervade our current housing voucher system—along with so many parallel inequities in other systems that have power over people’s lives. The ramifications of discrimination based not only on race, but on class, extend far beyond the microcosm of the rental housing realm. They help to perpetuate the inequalities that keep Boston from reaching its true potential. Evaluating issues related to race-based and voucher-based discrimination, especially with scientific approaches, such as the matched pair testing reflected in this study, is imperative if we want to solve the problem of discrimination in the rental market. And we do. Our mission of creating a city and region in which opportunity and justice are extended to everyone depends on it. Paul S. Grogan President & CEO The Boston Foundation 4     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Executive Summary Housing has a major impact on a person’s health, economic, and social outcomes.1 Inability to obtain quality housing has negative health and social consequences that can perpetuate the cycle of poverty and detachment from the labor market.2 Discrimination that prevents a person from living in a neighborhood that can provide easier access to better economic and educational opportunities lowers the ceiling on that individual’s future success.3 These negative effects not only harm the individual facing discrimination, but society in general. Where a person lives impacts much more than their future success or long-term well-being. It can contribute to vulnerability to a host of other adversities. The COVID-19 pandemic and the protests sweeping the country following the killings of Ahmaud Arbery, George Floyd, Tony McDade, Breonna Taylor, and so many others have painfully illuminated the related inequities that have always existed in American society. This study offers empirical evidence of some of the discrimination that people of color and people using housing vouchers face in our community. It is time for policy makers to act to undo this history of structural oppression, racism, and discrimination—in housing, and across the board. According to the Centers for Disease Control and Prevention (CDC) the COVID-19 pandemic has disproportionately affected people of color.4 Black people account for 22% of the known COVID-19 cases in the United States even though they represent only 13% of the population.5 The COVID-19 crisis has also shown that a person’s neighborhood, often segregated by race and/or class, can have enormous health consequences. In the United States, an analysis of ZIP Code data from 12 states showed that the infection rate of COVID-19 is twice as high in neighborhoods with median income of less than $35,000 than those with incomes of more than $75,000, and five times higher in majority minority ZIP Codes.6 Black people are 2.4 times more likely than White people to die from COVID-19,7 and are also much more likely to contract the disease.8 While the data are incomplete due to lack of consistent reporting, it is incontrovertible that the burden of COVID-19 has disproportionately fallen on minority communities.9 The experience of the City of Chelsea in Massachusetts in this regard is well documented.10 People of color make up 79% of Chelsea’s population, but only 28% of the population statewide.11 Chelsea has had the highest COVID-19 case rate in Massachusetts with 707 per 10,000 residents (the statewide rate was 146 per 10,000).12 Chelsea also has the highest rate of “crowding” within homes,13 which can make social distancing when a person in the household becomes ill very difficult.14 Disparities also exist in the location of testing sites. An NPR investigation in Texas found that testing sites were disproportionately located in whiter neighborhoods.15 The CDC also highlighted living conditions as an important factor for determining health,16 recognizing that, among other factors, research suggests “residential segregation is a fundamental cause of health disparity.”17 Residential segregation did not occur organically.18 Government policies and discriminatory practices by both government and individuals, such as redlining, have led to and perpetuated segregation.19 In recent events, we are witnessing grave consequences of this segregation and discrimination across all levels of society: the disproportionate impact of COVID-19 on Black, Indigenous and people of color and the continued and senseless killings of Black Americans by White people and individuals whose duties were to serve and protect all within our community. Indeed, research from Boston University’s School of Public Health has found that states with higher degrees of structural racism, and residential segregation in particular, have increased disparities in fatal police shootings of unarmed people.20 On June 12, 2020, Raychard Brooks was shot twice in the back and killed by an Atlanta police officer. On May 25, 2020, George Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    5 Floyd was killed by Minneapolis police officers by being pinned to the ground with a knee on his neck for eight minutes and 46 seconds, despite his pleas that he could not breathe. On March 13, 2020, Breonna Taylor was killed in her bed, after Louisville police officers entered her house using a no-knock warrant. These are just three cases in a long list of recent extrajudicial killings of Black people, painful reminders of the systemic inequity that exists in our society and the trends that will continue without targeted and sustained government effort and a concerted change in the actions of individuals to dismantle that structural racism. Combating residential segregation is a matter of life and death. Segregation, discrimination, and killings of unarmed Black people are not problems from our past, but ongoing, interconnected actions that must be addressed. Data from this study reveal that housing discrimination is still occurring at alarmingly high rates in Greater Boston. The time is now for policy makers to act to address housing discrimination and the other continuous forms of structural oppression against people of color. This study measures the levels of discrimination in the Greater Boston rental housing market based on race and income level. Data from this study show that high levels of discrimination exist throughout the pre-rental application process against both Black people and individuals using housing vouchers,21 beginning with the very first interaction between a prospective tenant and the person advertising housing (referred to throughout as “housing provider”22). In the vast majority of cases, real estate professionals perpetuated the discrimination. Policy makers should heed the findings from this study and work to enact measures that can curb housing discrimination. “Housing provider” refers to the people that advertise and/or show apartments to prospective renters. This term includes owners, real estate agents and brokers, and property managers. 6     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t The Study The goal of the present study was to gather data on race-based and voucher-based discrimination, measure the levels of such discrimination present in the Greater Boston rental housing market, and determine whether source of income discrimination is a proxy for race discrimination. To do so, the Housing Discrimination Testing Program (HDTP) at Suffolk University Law School and Analysis Group, Inc. (AG), an economic consulting firm, gathered information on the behavior of realtors and property owners with “matched pair” testing.23 The study measured a number of data points including: ■ whether testers were able to make appointments to see the properties; ■ how many units housing providers told testers about or showed them; ■ whether housing providers offered financial incentives; ■ whether housing providers made positive or negative comments about the housing units; and ■ whether housing providers offered testers an application. Indications that a housing provider offered preferential treatment to White testers over Black testers or market-rate testers (testers who did not tell the housing provider that they had a housing voucher) over voucher testers constituted evidence of discrimination. For example, in instances where a housing provider showed the market-rate testers a unit, but stopped communicating with the voucher testers prior to setting up an appointment to view the same unit, this would constitute evidence of discrimination. Findings The data from the study reveal high levels of discrimination based on both race and voucher status. Testing uncovered evidence of discrimination based on voucher status in 86% of the tests. In many instances, housing providers screened out voucher holders and ceased all communication with them after learning that the individual intended to use a voucher. There was evidence of discrimination based on the prospective renter’s race in 71% of the tests. The data suggests that voucher discrimination is not a proxy for race discrimination. Results indicate that White market-rate testers— meaning White testers not using vouchers—were able to arrange to view apartments 80% of the time. Similarly situated Black market-rate testers seeking to view the same apartments were only able to visit the property 48% of the time. Testers who had vouchers, regardless of their race, were prevented from viewing apartments at very high rates. White voucher holders were able to view rental apartments only 12% of the time. Black voucher holders were able to view apartments they were interested in renting only 18% of the time (see figure below). In addition, housing providers showed White market-rate testers twice as many apartment units as Black market-rate testers, and provided them with better service as measured by a number of different variables. The results also showed that testers who were offered a site visit by the housing provider received differential treatment at the visit based on race and voucher status. Percentage of Testers Who Visited Housing Site 100% 80% These results are disheartening. Housing discrimination based on race has been illegal for well over 150 years at the federal level.24 It is also illegal in Massachusetts to discriminate against a person based on race or because they have a housing voucher.25 Providing housing for profit is a business and those in business have an obligation to obey relevant laws in the practice of that business. Nevertheless, we found evidence of discrimination across a variety of property types, including owner-occupied buildings and stand-alone properties owned by commercial real estate companies. Moreover, the data revealed that real estate professionals are deeply involved in purveying the discrimination. Overall, these results highlight the high levels of race and income discrimination in the Greater Boston rental market and underscore the need for increased enforcement and education efforts to combat housing discrimination. Recommendations Analysis of the testing results surfaced a number of policy recommendations to combat discrimination against qualified renters, ranging from education to enforcement. While there may be other possible initiatives, results from this study make clear that proactive steps must be undertaken to change the status quo. The recommendations outlined in this report are just a beginning. 80% 60% 48% 40% 20% 18% 12% 0% Black Voucher Black No-Voucher White Voucher White No-Voucher NOTE: The error bars in the figures represent the 90% confidence intervals around the estimated percentages. Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    7 Introduction A variety of factors affect an individual’s housing choice, including affordability, proximity to work or family, quality of schools, and availability of green spaces.26 In a fair housing market, all prospective renters with the ability to pay the rent should have equal opportunity to choose where they want to live based on factors that are the most important to them (including affordability). These prospective renters should also have the same opportunities to make appointments to view, collect information on, and submit an application for a rental property. Housing providers should select tenants based on their ability to pay the rent, and the belief that the tenant will be a good caretaker of the property and a good neighbor to others. Through data gathered from “matched pair” testing conducted from August 2018 to July 2019, this study demonstrates that the reality is far from this ideal in the Greater Boston area. Findings from this study indicate that prospective tenants face high levels of illegal discrimination based on factors unrelated to their ability to become a good tenant, specifically their use of a housing voucher and their race. Residential segregation along racial and income lines emerged as a significant national concern in 1968, when the National Advisory Commission on Civil Disorders—formed in response to growing violence and civil unrest in cities—released a report declaring that the United States was “moving toward two societies, one black, one white—separate and unequal.”27 The Commission identified “pervasive discrimination and segregation” in housing as one of the main causes of this division.28 High levels of residential segregation persist across the country in the present day, despite the passage of the Fair Housing Act (FHA) and corresponding state and local laws, which have prohibited housing discrimination for decades.29 According to a 2015 report of the Center for American Progress, residential segregation has been increasing over the previous three decades, creating a deeply divided America along both racial and income lines.30 Segregation was deeply embedded in our society long before Congress passed the Fair Housing Act in 1968.31 Certain government policies played a significant role in causing and perpetuating housing segregation, including rules regarding public housing that led to the concentration of all–Black tenant housing projects in Black neighborhoods, restrictive covenants prohibiting the sale of property to Black people, Federal Housing Administration loan guarantees that included an explicit condition that Black people be excluded from developments, and redlining, a process by which banks refused to grant loans in majorityBlack neighborhoods.32 Greater Boston is not exempt from this problem. A 2015 index created by the financial news website 24/7 Wall St. identified the Boston-Cambridge-Newton area of Massachusetts as the seventh most segregated area in the country.33 While Boston, by itself, is not included in the site’s list of the 16 most segregated cities in the country, this appears to be the result of increasing levels of segregation nationwide rather than declining levels in Boston. Boston’s schools reflect the high level of segregation in the city. A 2018 report by the Boston Globe found that almost 60% of schools in the Boston Public Schools system are “intensely segregated,” meaning that students of color make up almost 90% of those schools’ student populations.34 A 2020 Boston Indicators report on changes in the schoolaged population in the city also found intensifying segregation in schools by both race and income. Black and Latino students most often attended schools where the majority of students were of their own race and low-income families are increasingly attending schools in which low-income students constitute the vast majority of the student population.35 Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    9 A Brief Overview of Housing Vouchers The Housing Choice Voucher Program (HCV) is a long-standing government program aimed at providing access to quality housing to low-income families. The HCV is a federal program under the purview of the U.S. Department of Housing and Urban Development (HUD) that “pays rental subsidies so eligible families can afford decent, safe, and sanitary housing.”36 Congress created the program in the 1970s in order to help low-income families access private housing.37 State or local governmental entities (housing agencies) administer the HCV Program (often referred to as “Section 8”). Eligibility for a housing voucher is determined based on total annual gross income and family size with eligibility thresholds varying by location. Once approved, individuals or families must find a place to live within a certain price range and then seek approval of that unit from the housing agency. If the apartment meets the agency’s quality and payment standards,38 the agency contracts with the owner and pays a portion of the rent directly to the owner.39 The present-day impact of the HCV is far-reaching. The program is the nation’s largest rental assistance program and provides subsidies to more than five million people in 2.2 million low-income households.40 According to the Center on Budget and Policy Priorities, “[c]hildren in families that use housing vouchers to move to better neighborhoods are more likely to attend college, less likely to become single parents, and earn more as adults.”41 The HCV is becoming increasingly important as the cost of housing has risen rapidly across the country. “Payment standard” is defined as “[t]he maximum monthly assistance payment for a family assisted in the voucher program (before deducting the total tenant payment by the family).” 21 10     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t According to a 2019 report from the National Low Income Housing Coalition, “[i]n no state, metropolitan area, or county in the U.S. can a worker earning the federal or prevailing state minimum wage afford a modest two-bedroom rental home at fair market rent by working a 40-hour work week.”42 With quality housing out of reach for many low-income families, the promise and the benefits of the HCV program are becoming all the more critical. Although the HCV is a federal program with proven benefits, there is no federal law prohibiting discrimination against those using housing vouchers. Instead, legislation protecting voucher holders from discrimination exists only at the local or state level. Currently, 17 states, including Massachusetts, prohibit discrimination based on source of income.43 In addition, many cities and counties—including Boston, Cambridge, Quincy, and Revere in Massachusetts—have local ordinances prohibiting housing discrimination based on source of income.44 Currently, only about 50% of voucher holders live in a jurisdiction that protects voucher holders from discrimination.45 The fact that there is no federal protection against housing discrimination based on source of income is important, because studies indicate that individuals seeking rental housing with the aid of a voucher face strong stigma associated with receiving public assistance46 and that many owners will not accept vouchers.47 In areas without legal protections for individuals who are denied housing due to their voucher status, these individuals have no legal recourse to combat the discrimination that they face in the housing market. While there is some evidence that living in a jurisdiction with laws against discrimination based on source of income provides individuals with vouchers easier access to rental units,48 the effectiveness of such laws in combating discrimination depends critically on the knowledge and enforcement of these laws. Summary of Findings In this study, we find that, even in the Greater Boston area where it is illegal to discriminate against a person for using a housing voucher, individuals with Section 8 housing vouchers face high levels of discrimination in the rental housing market. We also find that sourceof-income discrimination is not simply a proxy for race-based discrimination. The data show that there are high levels of discrimination based on both race and voucher status. Research on race-based discrimination in the rental housing market spans several decades.49 One recent example is the three-year investigation of real estate brokers in the Long Island area conducted by Newsday, a local New York news organization.50 Newsday’s investigation uncovered disparate treatment of minority testers in the form of steering,51 more stringent conditions on minority home seekers, and refusals to show minority testers (but not White testers) listings until they were pre-qualified.52 Overall, the investigation revealed that “Black testers experienced disparate treatment 49% of the time— compared with 39% for Hispanic and 19% for Asian testers.”53 In this study, we find even higher rates of negative disparate treatment based on race in the Greater Boston area: 71% for Black market-rate (i.e., those without vouchers) apartment seekers. This study also contributes to a growing recent body of recent research on discrimination based on the use of a housing subsidy.54 Testing uncovered evidence of discrimination based on voucher status in 86% of the tests.55 In many instances, housing providers screened out voucher holders and ceased all communication with them after learning that the individual intended to use a voucher. “Steering” is the practice of encouraging or discouraging a prospective renter from living in Results indicate that White market-rate testers— meaning White testers not using vouchers—were able to arrange to view apartments 80% of the time. Similarly situated Black market-rate testers seeking to view the same apartments were only able to visit the property 48% of the time. Testers who had vouchers, regardless of their race, were prevented from viewing apartments at very high rates. White voucher holders were able to view rental apartments only 12% of the time. Black voucher holders were able to view apartments they were interested in renting only 18% of the time. In addition, housing providers showed twice as many apartment units to White market-rate testers as they showed to Black market-rate testers, and provided them with better service as measured by a number of different variables. The results also revealed that testers who were offered a site visit by the housing provider received differential treatment at the visit based on race and voucher status. Similar to the findings from the Newsday investigation, this study demonstrates that real estate brokers play a significant role in purveying discrimination in the rental housing market. Of the 200 testers in this study, 182 had contact exclusively with real estate brokers, and 187 of the testers dealt with a broker at some point in the process of inquiring about the property. The high levels of discrimination that we find in our study often reflected the actions of real estate brokers. The Housing Discrimination Testing Program (HDTP) at Suffolk University Law School and its affiliated law school clinic have been working with individuals facing housing-related legal issues for many years. Based on our past work, we understand that race-based discrimination and voucher-based discrimination are common in the housing market. However, the prevalence and the level of disparate treatment that voucher holders and Black renters experienced are surprising. Our finding that voucher-based discrimination is not a proxy for race-based discrimination underscores the need for the promulgation and stronger enforcement of laws protecting individuals from source-of-income discrimination, specifically in the housing market. a particular area based on their protected class. 34 Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    11 Overview of Study Design The FHA entitles all people to truthful information about housing.56 It also prohibits differential treatment based on a protected class.57 A “protected class” is a category of people who benefit from protection under anti-discrimination laws.58 Housing discrimination against protected classes can take many forms. It is sometimes overt, but often takes the form of subtle behavioral differences that may be difficult to detect or identify. This is because when one person interacts with another, there are limited data to evaluate the nature of the interaction. The “matched pair testing” methodology used in this study, however, allows researchers to evaluate the nature and the extent of housing discrimination by comparing data on the interactions of multiple individuals who are similarly situated but for the protected class status (i.e., race and voucher status) with the same housing provider. In this section, we provide a detailed description of the study design that we employed. Assembling Testers HDTP recruited testers and assigned 200 individual testers to contact the housing providers of 50 different apartments in the Greater Boston area that were randomly selected from common rental property listing sites from August 2018 to July 2019. Each apartment met the payment standards for the use of a voucher from the Boston Housing Authority. The test coordinator assigned a test group consisting of two matched pairs (i.e., four individuals) to each apartment. Each test group consisted of two marketrate testers59 (one White and one Black) and two testers with vouchers (one White and one Black). All four individuals within a given test group were carefully matched to be as similar as possible except for their race and voucher status. Specifically, the test coordinator created matched pairs who were demographically similar (i.e., cisgender,60 same sex, no visible disabilities, age) and assigned the testers similar characteristics like income, family size, and credit score. The test coordinator instructed testers to 12     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t contact housing providers within a short time period of one another and via the same communication method (i.e., call or text). The test coordinator assigned all voucher testers a housing voucher from the same housing authority. The test coordinator interacted individually with all of the testers and did not share with the testers anything about the experiences of others in the same test group. Neither did the test coordinator discuss with the testers if their experiences differed from those of others. The test coordinator instructed the voucher testers to inform the housing provider that they were using a voucher as early as possible in their interactions with the housing provider (prior to the property visit, if possible). The researchers introduced race from the very beginning of every test by assigning each tester a “race-associated” name based on a previous study that examined name-based racial discrimination in employment practices in the Boston and Chicago labor markets (see Appendix 1 for names used in the study).61 Names were randomized for each test and they were used repeatedly for all 50 tests.62 The testers were instructed to provide their name in their very first contact they had with the housing provider. “Market-rate testers” refers to the testers who did not represent that they were looking to rent with the assistance of a housing voucher; rather they represented themselves as prospective tenants who would be themselves responsible for the entirety of the rent. 42 The researchers limited the scope of the investigation of race in this study to Black and White testers to identify the effect of race from housing vouchers more cleanly. Further research is required to understand the extent of discrimination that Latinx, Asian, and other people of color may also face in the rental housing market when using vouchers.63 Testers meticulously recorded the details of their interactions with housing providers in written reports (see Appendix 2).64 The cumulative nature of the reports provide multiple comparative data points to determine whether testers received truthful information and whether housing providers treated testers differentially, including more subtle forms of discrimination that housing advocates commonly refer to as “discrimination with a smile.” 65 Discrimination with a smile describes interactions that appear to be neutral to the protected class renter, but when compared to the treatment of the non-protected class renter, unequal treatment becomes apparent. HDTP required the testers to memorialize their interactions with the housing providers as soon as they could after each interaction. Although every tester’s goal was to make an appointment and tour the advertised property, not every tester was given that opportunity. Testers who only had phone and/ or email interaction with a housing provider filled out a form with 20 questions and a narrative section. Testers who toured one or more apartments completed forms with 24 questions (including 37 parts) and a narrative section. The detailed forms were designed to help testers recall as much of the interaction with the housing provider as possible. This is important because it allowed for the most accurate comparison of the experiences of each tester. The test coordinator met with each tester after the tester submitted his or her report to review the reports and to inquire about any missing information (i.e., dates and times). Site Selection: Rental Ad Scraping, Selection, and Randomization Researchers from Analysis Group (AG) scraped listings in the Boston Metro area from a website that advertised rental apartments on a weekly basis from August 2018 to July 2019. The researchers limited listings to one-bedroom apartments with monthly rents under $1,563 per month and studio apartments with monthly rents under $1,378 month and excluded short-term rentals.66 AG randomly selected a subset of 50 apartments from the listings in each week and manually verified that the random sample of listings met the criteria for apartment type and price and did not contain duplicates. AG sent the HDTP test coordinator its random listings each week. The test coordinator reviewed each individual listing to determine whether the listing was still posted online and that the listing contained a telephone contact number.67 If the listing was over a week old, the test coordinator called the phone number listed in the advertisement under a blocked number to confirm that the apartment was still available. If the test coordinator was able to confirm the unit was still available, or if the unit had been posted online for less than one week, the test coordinator assigned testers to the listing. If, during the course of the test, the test coordinator learned that the apartment had already been rented prior to the completion of the test, the test coordinator would assign the test group to a new listing using the process described above.68 In these cases, tester reports and final assessments were based only on the final, active listing for which the testing was completed. The design of the study, therefore, allowed the researchers to approximate the search that an individual with a voucher would conduct to find an apartment in the Greater Boston Area. The properties tested were located in nine cities and 11 neighborhoods of Boston.69 Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    13 Conversion of Written Reports to Variables for Analysis After the completion of all tests, HDTP provided AG with 200 reports from the 50 test groups (four reports for each test). After an initial review of these reports, AG identified and defined 44 outcome variables of interest. The variables included, for example, whether the testers were invited by the housing provider to visit the housing site, were greeted by the housing provider in a cordial fashion, were asked about their occupation or credit history, were shown different numbers of apartments, or received follow-up communication from the housing provider. Appendix 3 provides a description of the full set of variables. AG then coded the written information in each test report into a variable form using pre-defined definitions (see Appendix 3). To ensure accuracy and reduce the level of individual bias in this coding process, two members from the AG team independently coded the information in each test. The coders resolved discrepancies through joint discussions, and involved a third member of the team when necessary. This process resulted in a complete dataset of all 200 tests that AG analyzed using statistical methods. 14     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Findings The researchers completed two types of analysis for this study. First, AG performed a separate analysis, analyzing each of the variables identified at the outset of the study to determine whether there were statistical evidence of disparate treatment across race and voucher status. Second, HDTP staff analyzed each of the four test reports for each of the 50 tests to assess whether there was evidence of discrimination based on source of income, race, or both. For purposes of discussion, this report will refer to testers as follows: Black market-rate testers (BMR), White market-rate testers (WMR), Black voucher holders (BV), and White voucher holders (WV). Measurement of Discrimination by Analysis Group Discrimination could occur at two distinct points in the housing search process. First, housing providers could discriminate against the testers in their initial interactions, prior to the tester even seeing the housing unit. For example, a housing provider could be less likely to respond, more likely to break off contact without explanation (“ghosting”70), or less likely to offer a site visit to Black and voucher-holding testers. Second, housing providers could discriminate when testers visit the apartment. Housing providers may deliver a more positive apartment viewing experience to White testers or non-voucher holding testers by, for example, shaking their hand or commenting on positive aspects of the unit or neighborhood. Housing providers could also be more likely to offer the apartment to the tester or follow-up after the apartment showing. This section will discuss whether there is statistical evidence of discrimination at these two points in the apartment search process. Appendix 4 provides the full set of results for all variables analyzed. Initial Interactions with the Housing Provider As discussed in the Overview of Study Design section, prior to visiting an apartment, a tester would initiate contact with the housing provider, usually via phone. Receiving a response from the housing provider to arrange a visit to the site represents a critical step, as it sets in motion the rest of the engagement. Figure 1 presents the results of the “contacted” variable by race and voucher status. The error bars in the figure indicate the 90% confidence interval.71 The results indicate that White testers were significantly more likely to receive a response from the housing provider than Black testers (p = 0.019).72 Because testers were assigned race-associated names, testers stating their name in a call or message to a housing provider signaled their race to the providers. Testers typically did not announce their voucher status in this first message if they were leaving a voicemail. Therefore, it is not surprising that contact rates are similar between voucher and market-rate testers at this stage. FIGURE 1 Percentage of Testers Making Contact with Housing Provider, by Race and Voucher Status 100% 93% 80% 89% 82% 86% 60% 40% 20% 0% Black White Voucher No-Voucher Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    15 The effect of voucher status on the initial interaction between a tester and housing provider becomes more pronounced if the tester receives a response from the housing provider. Results indicate that testers were significantly more likely to be “ghosted” (i.e., suddenly getting no contact from a person without explanation) by the housing provider if they had a housing voucher. Figure 2 shows that the WV and BV testers were ghosted at a significantly higher rate than their market-rate counterparts (p < 0.001 and p = 0.019, respectively). FIGURE 3 Percentage of Testers Ghosted by Housing Provider, by Voucher Status 100% 80% 60% 42% 40% 44% FIGURE 2 Percentage of Testers Ghosted by Housing Provider 20% 100% 10% 0% 80% 60% 40% 44% 39% 18% 20% 4% 0% Black Voucher Black No-Voucher White Voucher White No-Voucher Figure 3 presents the same data as in Figure 2, but focuses on showing the average impact of voucher status across the two race groups. It shows that the testers with vouchers were ghosted at a significantly higher rate than those without vouchers (p < 0.001). This finding makes sense given that testers with vouchers were instructed to inform the housing provider of their voucher status as soon as possible after making the first contact. One would only observe this effect after the tester has an opportunity to interact with the provider. 16     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Voucher No-Voucher Analyzed together, the observed differences in contacting and ghosting rates provide strong evidence of discrimination based on race and voucher status in the initial interactions between the tester and the housing provider. Figure 4 presents a comprehensive look at differences in the testers’ initial contact with the housing provider by showing the percentage of testers that the housing provider did not contact or ghosted across the four testing groups. As the figure shows, the housing provider was significantly more likely to either not contact or ghost WV and BMR testers than WMR testers (p < 0.001 for both comparisons). However, differences between WV and BV groups and between BMR and BV groups, are not statistically significant at the conventional levels. This finding suggests that the effect of race and voucher status on the initial interaction with the housing provider may not be necessarily additive. Opportunity to Visit the Site If their initial interaction with the housing provider goes well, testers can have the opportunity to visit the site in person. As shown in Figure 5, data from the study show that WMR testers were able to view apartments at a much higher rate than similarly situated BMR testers (80% compared to 48%, p = 0.001). Most testers who had vouchers, regardless of their FIGURE 4 FIGURE 5 Percentage of Testers that Housing Provider Did Not Contact or Ghosted Percentage of Testers Who Visited Housing Site 100% 100% 80% 80% 60% 60% 40% 80% 50% 46% 48% 40% 38% 20% 20% 18% 12% 8% 0% Black Voucher Black No-Voucher White Voucher White No-Voucher race, were not given the opportunity to view apartments. WV testers were able to view rental apartments only 12% of the time, while BV testers were able to view apartments they were interested in renting only 18% of the time. These site visit rates were statistically different from those experienced by WMR and BMR testers, respectively (p < 0.001 and p = 0.001). 0% Housing providers sometimes informed testers about and/or showed testers multiple units during a site visit. As shown in Figure 6, housing providers told Black and voucher testers about fewer available units than their White and market-rate counterparts. WMR testers were informed that there were more units available than BMR testers were informed on average (1.35 units v. 0.63 units, p < 0.001). Testers Black No-Voucher White Voucher White No-Voucher with a voucher were told that even fewer units were available, with WV and BV testers being informed of only 0.20 and 0.26 units on average, respectively. Regardless of race group, housing providers informed market-rate testers about more units than they informed voucher testers. Housing Provider On-Site Behavior Results from the study also show that testers who were offered a site visit by the housing provider 73 received differential treatment at the visit based on race and voucher status. Although testers recorded a number of different aspects of housing provider’s behavior on site, we focus on presenting data on aspects of the interaction that were associated with statistically significant differences in treatment based on voucher status or race. Black Voucher FIGURE 6 Average Number of Units Housing Providers Said Were Available to Testers74 1.8 1.6 1.4 1.35 1.2 1.0 0.8 0.6 .63 0.4 0.2 0.0 .26 Black Voucher .20 Black No-Voucher White Voucher White No-Voucher Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    17 The results on the number of units shown to testers exhibit a similar pattern. As shown in Figure 7, WMR testers were shown, on average, 1.14 units, while BMR testers were shown only 0.54 units. Testers with a voucher were shown fewer units than testers without a voucher: WV and BV testers were shown only 0.12 and 0.24 units, respectively. All differences between four groups were statistically significant at the 95% confidence level, except for the difference between WV and BV testers. FIGURE 7 Average Number of Units Shown to Testers75 1.8 FIGURE 8 Percentage of Testers Offered a Financial or Non-Financial Incentive 76 100% 80% 60% 40% 40% 33% 20% 1.6 11% 1.4 0% 1.2 Black Voucher 8% Black No-Voucher White Voucher White No-Voucher 1.14 1.0 As shown in Figure 8, race was an important determinant of whether housing providers mentioned financial or non-financial incentives to testers. Housing providers were significantly more likely to mention incentives to WMR testers than BMR testers and WV testers than BV testers. 0.8 0.6 .54 0.4 0.2 0.0 .24 .12 Black Voucher Black No-Voucher White Voucher White No-Voucher Not only did housing providers show more units to White and market-rate testers, they were also more likely to incentivize these testers to submit a rental application. The results provide evidence of discrimination in several different forms. First, housing providers sometimes offered financial or non-financial incentives to testers during on-site conversations. For example, a housing provider may have mentioned that they were offering the tester a discount on the monthly rent or an amenity such as free parking. 18     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Housing providers also sometimes encouraged testers to apply for the unit by sharing certain positive attributes of the unit or the neighborhood. For example, the provider may comment on the size of the unit or its proximity to public transportation. Figure 9 shows that the percentage of on-site conversations that included positive comments from the housing provider on the unit or neighborhood varied by tester group. Housing providers made a positive comment to 58% of WMR testers compared to only 17% of WV testers. Providers made positive comments to only 33% of BMR testers and 11% of BV testers. Figure 10 demonstrates that these differences were statistically significant along both race and voucher status dimensions. FIGURE 10 FIGURE 9 Percentage of Testers to Whom Housing Provider Made a Positive Comment about the Unit or Neighborhood 77 Percentage of Testers to Whom Housing Provider Made a Positive Comment about the Unit or Neighborhood, by Race and Voucher Status78 100% 100% 80% 80% 60% 58% 60% 52% 40% 48% 40% 33% 20% 20% 27% 17% 13% 11% 0% Black Voucher Black No-Voucher White Voucher White No-Voucher Interactions between housing providers and testers at the end of the site visit or immediately following the site visit also exhibit discrimination, particularly based on voucher status. As shown in Figure 11 and Figure 12, housing providers in the study were both more likely to offer an application to market-rate testers and more likely to follow up after the site visit 0% Black White Voucher No-Voucher with market-rate testers compared to voucher holding testers (p < 0.001 and p = 0.012, respectively). These results indicate that even when individuals with vouchers are able to visit the apartment unit, they are less likely to be offered an application and encouraged to continue seeking the unit. FIGURE 11 FIGURE 12 Percentage of Testers Offered Application by Housing Provider, by Voucher Status Percentage of Testers that Housing Provider Followed Up with After Visit, by Voucher Status 100% 100% 80% 80% 60% 60% 44% 40% 40% 41% 33% 20% 20% 10% 0% Voucher No-Voucher 0% 7% Voucher No-Voucher Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    19 Measurement of Discrimination by HDTP For each of the 50 tests conducted, HDTP compared the treatment of the voucher group (Black and White testers assigned a housing voucher) and the marketrate group (Black and White testers who were not assigned a housing voucher) along a number of variables to determine whether there was any evidence of differential treatment. HDTP uses this same process for enforcement testing.79 The HDTP categorized each individual test as one of the following: (1) showing evidence of discrimination (because of adverse differential treatment); (2) inconclusive; or (3) showing no evidence of discrimination. Differential treatment included whether the housing provider stopped communicating with the tester while continuing to work with testers who were not in the protected class; as well as other factors relating to levels of service such as the number of apartments shown, whether the tester was offered an application, shown amenities, or received follow-up messages from the housing provider about the property. Three HDTP staff members independently reviewed each test report and resolved any differences of opinion through discussion. In the few instances in which all three HDTP staff members did not agree on a particular determination, an additional staff member reviewed the test evidence and the team used the majority result. Based on these categorizations, the HDTP found discrimination in the form of disparate treatment based on source of income in 86% of the tests and race-based discrimination in 71% of the tests. In a substantial number of tests, WMR was able to view the apartment, but BMR was not able to visit the unit. In addition, WMR testers often received better treatment than BMR testers with respect to the number of apartments shown or offered, financial incentives, and follow up, among other variables. 20     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Testers with a voucher, both Black and White, were unable to view the apartment most of the time. By comparison, WMR testers were able to view 80% of these same apartments. Most commonly, housing providers stopped all communication with testers after they disclosed their voucher status, regardless of race, and no matter how many times or in what form the tester tried to communicate with the housing provider (they were “ghosted”). Around 10% of the time, housing providers explicitly told the tester that the landlord did not accept vouchers. About 20% of the time housing providers told voucher holders that the landlord accepted vouchers; however, over 80% of the time, those same housing providers then discriminated against the voucher holder.  Tester Anecdotes Although some forms of discrimination uncovered in this study were overt, particularly against testers with vouchers, discrimination often took the form of more subtle differences in customer service. For example, housing providers told WMR testers that more units were available, showed them more units, offered more incentives to rent, and made more positive comments to them about units than they did to BMR testers. Housing providers also followed up more often with and offered more assistance searching for units to WMR testers than to BMR testers. These differences are not apparent to an individual prospective tenant and can only be revealed through analysis of data points generated by the experiences of those seeking housing. Biased Ghosting In many cases, discriminatory treatment was blatant and testers were aware of it. In one test, BV introduced himself as “Kareem” in a voicemail message to the housing provider. Having not heard back, the tester called and spoke to the housing provider three days later. At first the housing provider seemed “eager and excited” on the phone and confirmed that the unit was still available. The tester told the housing provider that his name was “Kareem” and that he had called before. The tester felt that the conversation abruptly changed course. The housing provider asked “Kareem” to confirm his phone number and the tester felt rushed off the phone. The tester then introduced his voucher and asked if he could see the unit. The housing provider told the tester he would call the following day to set up a time to view the unit. The housing provider did not call the tester back and did not respond to subsequent voicemails. The BMR tester for this rental listing left three voicemails introducing himself as “Tremayne,” but was not able to speak with the housing provider. The WMR tester (“Brad”) was able to make an appointment to tour the unit. Although the housing provider did not attend the appointment, he called WMR back offering to reschedule. The WV tester was told, after introducing his voucher, that the housing provider would call him back. The WV tester was not able to reach the housing provider again. What these testers experienced was not unusual. Differences in attitude can make a significant impact on customer service. In a different test, both the White and Black voucher holders observed a shift in the housing provider’s demeanor after mentioning the voucher. The housing provider offered WV two potential appointment times before the tester introduced her voucher in the conversation. After she mentioned her voucher, WV noted that the housing provider “was quiet for a moment and his whole enthusiastic attitude changed and he starting [talking] fast to rush” her off the call. The housing provider told WV to text him to confirm the appointment. WV texted twice and called once over a three-day period (the housing provider’s voicemail box was full so she was unable to leave a message). WV called again the next day and the housing provider said he would check to see if the unit was still available and instructed WV to text him again. WV again followed the housing provider’s instructions and texted him once more but never received a response. Similarly, BV contacted the same housing provider and was offered an appointment. However, after introducing the fact that she had a voucher “the tone changed” and the housing provider asked her to text him and ended the call. The tester did not text the housing provider and had no further contact. BMR and WMR were both able to obtain appointments over the phone, were able to confirm the appointments by text, and both toured the unit. Unfortunately, housing providers often expressly told testers that they were not welcome to rent the unit despite the illegality of such statements. A realtor told one WV tester “I mean, talking to you, you seem totally normal… I mean, a lot of people with Section 8 aren’t the greatest people… so sometimes people can be prejudicial about that.” The same realtor refused to show BV the unit at the site because she arrived late to Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    21 the appointment. WMR also arrived late to her appointment but was given a tour. BMR arrived early and received a tour and application. Discrimination with a Smile Generally, housing providers engaged in discriminatory practices in ways that were less immediately apparent. Sixteen of the 84 testers who introduced the fact that they had a voucher on the phone prior to the site visit were expressly told that their voucher was not a barrier to renting the unit. In 81% of those tests, evidence of discrimination was ultimately found, and in 56% of those cases, the housing provider ceased all communication with the testers in spite of significant efforts on the testers’ part to make contact. For example, in one test, the housing provider told BV that he needed to check with the “head office to see if they take housing vouchers.” The housing provider did not respond to the tester’s further attempts to contact him. However, the same housing provider told WV that he owned the unit. This housing provider (the owner) gave WV the phone number of a broker and the broker gave WV a tour of the unit. Differential Criteria When analyzing the differences in customer service between testers, other trends appeared in the methodology of discrimination. Housing providers often use different screening questions to avoid showing apartments to voucher holders or testers of color. Black testers, both voucher and market-rate, were often asked more questions about their ability to pay and/or questions about their credit than their White counterparts. A BMR tester told one housing provider that his credit score was 650 and the housing provider responded saying a score of at least 720 was required. Another housing “We have to check your credit score.” provider affiliated with the same apartment warned WMR to “avoid the question if [the tester’s] credit score was not high enough” and went on to explain that the landlord “set the rent rates based on allowing the ‘right type of people’ to have easy access to” the area. Likewise, in a different test, both BMR and BV were asked about credit, but neither White tester was asked about credit scores. In one test, the housing provider did not ask 22     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t either of the market-rate testers about credit, but asked the WV about credit after she said she had a voucher. When WV stated her credit score was 680, the housing provider explained he could not show her the unit because “the landlord wants people with scores of 750.” BV introduced her voucher and the housing provider told her that he would check if it were available, but then never replied to her four subsequent voicemails. Similarly, after one BV tester told a housing provider that she had a voucher, he replied that he would call the landlord as the landlord requires credit scores over 700. BV told him her score was over 700, but he still said he would have to call her back and ultimately did not. WV was not able to reach that housing provider again after their initial conversation where she introduced her voucher. This housing provider did not ask about BMR’s or WMR’s credit scores and both were able to tour the unit. Black testers were also less likely to receive encouragement to move quickly in order to secure the apartment. For example, a housing “Oh, there’s no urgency….” provider told WMR that he was “’not one to push a building’ but he always warns people that things move fast.” The housing provider gave the tester his business card and suggested she call or text over the weekend if she decided to apply. BMR reported that the same housing provider said that he “hasn’t really advertised the unit much so there is no current urgency” and suggested that waiting until the following week would be fine. Both testers spoke with the realtor on the same day, WMR toured the unit that day, and BMR toured it the day after WMR. The Black and White voucher testers were not able to tour the unit. Similarly, in a different test the housing provider told WMR that the apartment would be hers as soon as she wanted it and that she was “pretty much automatically accepted.” The same housing provider simply instructed BMR to submit the application if she was interested. “We’re looking for quiet In another test, the housing people with good jobs.” provider told the WMR at the end of her tour that he wanted to show her an additional unit. He went on to explain that “they don’t advertise that apartment because then they would have to respond to everyone who inquires” and they were looking for “people with quiet lifestyles who work, not CEOs necessarily, but people with good jobs.” He concluded by inviting WMR to join “a select group” that would tour the unit the following day. The housing provider did not offer this unadvertised unit to the other three testers. Outright Refusal Many housing providers simply explained to voucher testers that they were less attractive applicants because of the delay or additional effort required to rent to a voucher holder. For example, one property owner told both Black and White voucher holders that he did not want to keep the unit vacant during the inspection process required by the housing authority. BV was simply turned away for this reason. The housing provider gave WV a tour of the unit. However, he told her over the phone and in person that if a market-rate tenant were to apply, he would rent it to them “because the problem with the voucher is that you have to wait for inspections and “With the voucher you have everything and he didn’t to wait for inspections.” want to be losing money.” Both market-rate testers were able to tour this unit. In another test, the housing provider told BV that the owners would likely give the unit to a market-rate tenant because “with the housing vouchers there is so much work that has to be done, it’s more complicated and there are all kinds of inspections.” BV was not able to tour the unit and neither was WV tester despite her affirmation that she could pay whatever move-in costs the voucher did not cover. of a voucher and call back, but he never did in spite of the tester’s subsequent efforts to contact him. The housing provider made an appointment to show the apartment to WMR, but cancelled the day before because the apartment had rented. Steering In another case, the broker attempted to steer the voucher testers, and denied them the opportunity to tour the advertised apartment. When BV told the broker she had a voucher, the broker said, “Well, there might be other options that we can look into;” that “you must be getting subsidized, honey;” and that “[w]e’ll take a look at some other apartments.” The broker told WV, “I can give you a call back and we can see what else we can find in the area.” This broker stopped communicating with both voucher testers, but showed the property to both market-rate testers. Repeatedly, testers who on paper were qualified to rent and had similar incomes (either through receipt of public assistance or annual income) were treated differently and negatively based on their race and voucher status. Whether such discrimination was the result of explicit or implicit biases makes no difference to the illegality of such actions.   In another test, the voucher testers were turned away by a broker that was aware that the apartment would not pass a Section 8 inspection because it had only one means of egress. This violates the building code and is dangerous to the occupants in case of fire.80 The broker told BV that there “may be a problem” because Section 8 would “I need to ask my manager about that.” require two entrances to the apartment, but that he would ask a friend about it. When BV called back, he said he forgot to call his friend, but that he would, and then stopped communicating with the tester. The same housing provider told WV that he would check into the use Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    23 Recommendations This study found alarming levels of discrimination based on race and source of income in the Greater Boston rental housing market—none of which should be occurring because federal, state, and local laws prohibit this behavior. Policy makers should consider the data to develop policies to address the discrimination occurring in our community. Based on our findings and the experiences of testers, we provide suggestions on possible policy initiatives to curb housing discrimination and to increase the ease of use related to vouchers. The COVID-19 crisis has highlighted the importance of removing barriers for people in protected classes to neighborhoods of opportunity. These recommendations are suggested as a starting point for further discussion as to possible policy solutions that limit housing discrimination and its attendant harms, and are by no means meant to be exhaustive. 1. Increase penalties and training for real estate professionals and prohibit them from charging broker’s fees. This study demonstrates that real estate brokers play a major role in purveying housing discrimination on behalf of their clients. In the vast majority of the tests that HDTP conducted for this study, the testers interacted with a real estate broker.81 The study shows that many Boston area brokers are screening out qualified applicants based on their protected class status on behalf of their clients. This must change. Policy makers should increase the penalties for discrimination and make it easier to suspend a real estate professional’s license for violating antidiscrimination laws. Under Massachusetts law, in order to suspend the license of a broker for 60 days for discrimination, the Massachusetts Commission Against Discrimination (MCAD) must notify the Board of Registration of Real Estate Brokers and Salespeople (BRREBS) that it has made a finding that the broker has violated Chapter 151B, the Commonwealth’s anti-discrimination law, in the course of their duties.82 The MCAD must make a second such finding within two years to suspend a broker’s license for 90 days. The results of this study suggest that many brokers are willing to violate anti-discrimination laws. The requirement that the MCAD make a finding before a broker is referred for discipline is a system that practically ensures there will be no discipline. Many renters do not know they are being discriminated against, and those who do—particularly those with a rental subsidy—typically do not report it.83 They are busy spending their time trying to find an apartment before the period in which they must do so expires, or else lose the voucher. Even where cases are brought against brokers, they are often resolved informally, without a formal finding.84 The legislature should consider changing the cost benefit analysis by making it easier to suspend an offending broker. Brokers should be a force for good in the real estate market, assisting in keeping the playing field level rather than assisting landlords in discriminating. The HDTP submitted multiple public records requests in an attempt to determine the number of real estate brokers and salespeople in Massachusetts who had their licenses suspended by the BRREBS because they engaged in housing discrimination dating back to January 1, 2017. In January of 2020, the Division of Professional Licensure responded with a list of 67 brokers whose licenses BRREBS had suspended. However, as to whether BRREBS suspended any of these brokers for discrimination, the Division replied, “We do not store and maintain our Data in a manner that will allow us to respond to your request.” In February of 2020, the Massachusetts Commission Against Discrimination responded to a request regarding whether they have referred brokers to BRREBS for suspension, by stating that they were not able to find “records of referring any cases to the Board for license revocation pursuant to MGL ch. 112, s. 87AAA.” This study uncovered that brokers are Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    25 engaging in high levels of discrimination; however, it does not seem that the system designed to discipline them for this behavior is working. Policy makers should increase fair housing training for brokers and review it to ensure it is of sufficient quality. In Massachusetts, 40 hours of pre-licensing training is required to obtain a Real Estate Broker’s License. Four hours of that training must be related to fair housing. The level of discrimination this study has uncovered demonstrates that either many of those in the industry are ignoring their training or the training is woefully inadequate. As gatekeepers to the housing market in Massachusetts, licensed brokers must understand their responsibilities under fair housing laws. The fair housing training for pre-licensing should be reviewed and assessed with the goal of licensed brokers becoming proficient in their knowledge of fair housing laws. Additionally, Massachusetts should consider requiring real estate professionals to inform clients of their rights as housing seekers. This requirement is being considered in New York state after the three-year Newsday investigation uncovered the ways that real estate professionals were discriminating against housing seekers based on race.85 New York’s proposed rule would require real estate brokers or salespeople to provide their clients with a disclosure of fair housing rights that the client must sign. The broker or salesperson must retain this signed disclosure for three years.86 The rule would also require brokers to post a similar disclosure conspicuously in their office and on their websites.87 In addition, fair housing training for brokers must be video recorded and maintained for at least a year so that it is subject to review.88 Education is essential and the standards that real estate professionals must follow should be clear. In Massachusetts, the professional standards of practice for real estate brokers specifies the protected classes against which a broker shall not discriminate.89 Source of income (or subsidized housing) is not included in this list. The regulation states that “no broker or salesperson shall discriminate in the provision of services on the basis of age, marital status, gender, sexual preference, race, religion, socioeconomic 26     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t status or disability.”90 While socioeconomic status might be interpreted to include voucher holders, because other specific protected classes are listed in the regulation, the best practice would be to explicitly include all protected classes, rather than a partial list.91 This limited list might send the wrong message and suggest that a broker must pay attention to discrimination only as to a subset of protected classes.92 Furthermore, this list does not even include all federally protected classes under the FHA as it does not list national origin, familial status, or color. The professional standards of practice of real estate brokers and salespeople should include state and federally protected classes. The last recommendation related to the real estate profession is the prohibition of charging a broker’s fee to a tenant. Such prohibition was recently passed in the State of New York.93 Other states should consider the same. In Greater Boston, real estate brokers have significant control over the real estate market. In this study, brokers were involved in almost all of the test properties. Requiring a broker’s fee up front to be paid by the tenant could effectively screen out lower income renters, such as those with vouchers. There is a shortage of affordable rental housing in high cost metro areas and low-income renters struggle to find housing.94 Removing broker’s fees would remove a barrier to access for low-income tenants and increase their available options. 2. Strengthen anti-discrimination laws and fair housing enforcement and education and increase resources for testing. Source of income should be a protected class under the Fair Housing Act. More than 2.2 million low-income individuals and families participate in the federal Housing Choice Voucher program.95 Many of these individuals are people of color. They should be protected from discrimination for their participation in the program. It is a waste of our nation’s resources to fund a program and then allow its frustration by letting people discriminate against the program’s participants. On June 26, 2019, U.S. Senator Tim Kane of Virginia reintroduced the Fair Housing Improvement Act of 2019.96 This Act would make it illegal under the FHA to discriminate based on a person’s use of a housing voucher. This study has identified the incredibly high level of discrimination that people using vouchers face. People should not have to contact ten housing providers in order to see one unit. The barriers that they face are real and often insurmountable. The Fair Housing Improvement Act should be passed so that housing providers can no longer legally discriminate against voucher holders, and if discrimination does occur, there must be legal recourse and meaningful penalty. Moreover, states and localities should enact source of income protections under their anti-discrimination laws until Congress acts as suggested above. One possible avenue for fair housing enforcement would be to fund a legal strike force that would be available to assist voucher holders—who, as research strongly suggests, will face discrimination—by contacting housing providers on their behalf and seeking injunctions in court if necessary. There is an immense need for such services, and they could be funded directly through appropriate agencies, or through fellowship programs. Renting out housing is a business. Landlords should understand their legal obligations as operators of a business and renters should understand their rights. A national media campaign should be undertaken to educate everyone about rights and responsibilities under the FHA, the level of race discrimination that is occurring throughout the nation, and protections related to voucher-based discrimination in jurisdictions where voucher holders are a protected class. Legal protections do not help people if they are unaware of their rights. Furthermore, awareness must be raised due to the high level of race-based discrimination in the market and the very subtle way in which housing providers perpetuate it. People should not mistakenly think that race discrimination in housing no longer occurs. Additional resources should be devoted to fair housing testing to detect discrimination. After the NewsDay study, referenced above, the New York State Association of Realtors advocated immediate funding of testing to uncover bias among realtors.97 While some of the discrimination uncovered in this study was overt, most discrimination is in the form of differential treatment that can only be revealed with testing. This study exposed a significant amount of race-based and voucher-based discrimination in the form of differential treatment. It took over a year to perform the testing and required sending 200 testers to 50 different properties. Testing is resource intensive by its nature. This study shows a significant amount of discrimination is going unchecked in Greater Boston. If it were more likely that any individual potential renter was a tester, it may give housing providers pause and impact the culture. This is particularly true for brokers who make their living in the market. Policy makers should increase funding available for testing and enforcement of fair housing laws. 3. Improve and streamline the system for using vouchers. This study makes clear how difficult it is for people using vouchers to simply take the first step in the rental process—setting up tours of apartments. Though the priority should be to develop policies to prevent discrimination, it is also worth examining ways to improve the processes relating to vouchers to make these programs more effective and easier to navigate. Policy makers should expand the use of small-area fair market rents (“SAFMRs”) in order to expand the opportunity that voucher holders have to move to higher opportunity areas. SAFMRs set the subsidy amount allowed under the program based upon the average rents in smaller neighborhoods, rather than in a larger metropolitan area.98 For example, a person in Boston trying to use a voucher in the Beacon Hill neighborhood would receive a higher subsidy than one trying to use a voucher in a less economically advantaged neighborhood, rather than having all vouchers in Boston set at an average rate for the city. This would have the effect of allowing more voucher holders access to higher opportunity neighborhoods, where they are often priced out. This would lead to better outcomes for these families and, to the extent the families were people of color, would increase integration. Of course, as this study suggests, such policy changes will not be effective unless brokers and landlords stop illegally discriminating against renters based on race and voucher status. In Massachusetts, the federal Housing Choice Voucher Program (formerly known as “Section 8”) and similar Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    27 state programs allow a voucher holder 120 days to use the voucher to find suitable housing.99 Under these programs, the public housing authority administering the voucher may grant extensions according to a policy in the appropriate administrative plan for the particular program. Neither federal regulations, state regulations, nor the state administrative plan specifically mentions encountering discrimination as one of the extenuating circumstances for which an extension may be granted. The relevant regulations and administrative plans should specify discrimination as a reason for allowing an extension. The results of this study underscore how difficult it is to use a voucher in the Greater Boston area. About nine out of 10 times, qualified voucher holders who are interested in viewing a property are denied that opportunity just because they have a voucher. The most effective policy would be to combine the ability to request an extension with resources devoted to allowing voucher holders to pursue enforcement of their fair housing rights, because while granting an extension may give the voucher holder an opportunity to access housing, it does not address the discrimination that makes an extension necessary. 28     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t • • • America should no longer ignore and perpetuate the structural inequities that have led to the systematic oppression of communities of color. Our ability to become a cohesive society depends on addressing these historic ills. The COVID-19 crisis and protests against the routinized killing of Black people have again exposed these festering wounds in particularly stark light. The discrimination this study has revealed provides valuable data proving the alarming prevalence of housing discrimination based on race and source of income in a community in which the law protects both groups of people. Policy makers should act to counter the discrimination that this study has revealed. This will not only benefit individuals, but will allow us a chance to begin to heal a fractured society. APPENDIX 1 Study Assigned Names Black Female White Female Black Male White Male Aisha Allison Darnell Brad Ebony Anne Hakim Brendan Keisha Carrie Jamal Brett Kenya Emily Jermaine Geoffrey Lakisha Jill Kareem Greg Latonya Kristen Leroy Jay Latoya Laurie Rasheed Matthew Tamika Meredith Tremayne Neil Tanisha Sarah Tyrone Todd Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    29 APPENDIX 2 Tester Report Forms TEST REPORT FORM - TELEPHONE OR EMAIL CONTACT ONLY Tester: Notify test coordinator of all contact and forward materials received. *PAGE 1 and 2 TO BE REMOVED BY TEST COORDINATOR AFTER DEBRIEF Test #: (Please see assignment form for test #) TESTER’S INFORMATION: NAME: NAME USED FOR TEST (IF DIFFERENT): ETHNICITY (select one): ☐ HISPANIC OR LATINO ☐ NOT-HISPANIC OR LATINO RACE (select one or more): ☐WHITE ☐ BLACK/AFRICAN AMERICAN ☐ NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER ☐ASIAN ☐ AMERICAN INDIAN OR ALASKA NATIVE ☐ OTHER (SPECIFY): TESTER’S ASSIGNED SEX AT BIRTH: ☐ MALE ☐ FEMALE TESTER’S GENDER IDENTITY: ☐ FEMALE ☐MALE Phone number called: Click here to enter text. Person(s) with whom you spoke or emailed: Name: Position if known: 1. 2. 3. *PAGE 1 TO BE REMOVED BY TEST COORDINATOR AFTER DEBRIEF 1 30     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t *PAGE 2 TO BE REMOVED BY TEST COORDINATOR AFTER DEBRIEF RENTAL UNIT(S)MENTIONED 1. LIST ALL APARTMENTS MENTIONED ADDRESS CITY APT. # MONTHLY RENT A. $ B. $ C. $ D. $ E. $ F. $ G. $ H. $ # OF BEDROO MS UTILITIES/AMENITIE S INCLUDED *PAGE 2 TO BE REMOVED BY TEST COORDINATOR AFTER DEBRIEF 2 Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    31 Suffolk University Law School Housing Discrimination Testing Program *PREVIOUS PAGES 1 and 2 TO BE REMOVED BY TEST COORDINATOR AFTER DEBRIEF* 1) Were you able to interact with a housing provider to discuss housing options (either over email or phone)? ☐Yes (skip to Question 4) ☐No 2) If No, why not? ☐Left message on voicemail, answering machine, or pager ☐Left message with person who did not have information ☐Told to call back later ☐Wrong number ☐Housing provider hung up ☐No answer ☐Telephone number no longer in service ☐Other (specify): (NOTE: IFANSWER TO QUESTION 1 WAS NO, PLEASE FILL OUT #2 AND SKIP TO #20) 3) When you asked about the availability for the type of housing or the unit that you were assigned (e.g., one bedroom), what were you told? ☐The housing is available when I need it ☐The housing is NOT available when I need it ☐The housing provider did not know whether the housing was available ☐Something else (specify): 4) How many units were you told about fitting your initial request? 0 Unit(s) 5) How many “other” units were you told about? 0 Unit(s) 6) What were you told about any "other" housing or units? ☐Other housing is available when I need it ☐Other housing is NOT available when I need it ☐The housing provider did not know whether other housing was available ☐Something else (specify): 32     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t 0 Suffolk University Law School 7) Housing Discrimination Testing Program Did the housing provider tell you that an application form and/or reservation fee of some kind must be done before renting/purchasing a unit? ☐Yes ☐No 8) Did the housing provider ask if you would like an appointment to view a unit? ☐Yes ☐No 9) Did the housing provider invite you to come in and pick up an application or materials or offer to send it to you? ☐Yes ☐No 10) Did the housing provider tell you that a credit check was part of the application process? ☐Yes ☐No 11) Did the housing provider tell you that a co-signer would be needed as part of the application process? ☐Yes ☐No 12) Did the housing provider tell you that a criminal background check was part of the application process? ☐Yes ☐No 13) Did the housing provider request information about your income, source of income or occupation? ☐Yes ☐No If yes, please record what the housing provider said? 14) Did the housing provider make any remarks about disability or persons with disabilities? ☐ Yes Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    33 Suffolk University Law School ☐ No Housing Discrimination Testing Program If yes, please record what the housing provider said: 15) Did the housing provider make any remarks about accessibility or units that were "handicapped" accessible? ☐ Yes ☐ No If yes, please record what the housing provider said: 16) Did the housing provider make any remarks about race/ethnicity, religion, or families with children? ☐Yes ☐No If yes, please record what the housing provider said: 17) Did the housing provider make any reference to the lead status of any units? ☐Yes ☐No If yes, please record what the housing provider said: 18) Were you referred to the following during your call? ☐Assisted living ☐Nursing home ☐Group home ☐Low income housing ☐Other ☐ None 19) What arrangements were made regarding future contact between you and the housing provider [check all that apply]? ☐The housing provider said that he/she would call you back ☐The housing provider invited you to call him/her back ☐The housing provider invited you to come in to inspect units/pick up application ☐Future arrangements were not made 34     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Suffolk University Law School Housing Discrimination Testing Program ☐Other (specify): 20) When was this report completed? Date (month/day/year):Click here to enter text. Narrative of Phone or Email correspondence: *Please describe the people you interact with using the description you used on page 1, but DO NOT use names of the people you interact with or addresses of sites you visit. That information should only be included on page 1 and 2 of this form. If you interact with different people or see multiple units please be descriptive without using identifying information (i.e. “After visiting the 1 bedroom apartment (#1A on page 2, the agent took me to another 1 bedroom (#1B on page 2)). Click here to enter text. ______________________________________ NAME __________________________ DATE Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    35 Suffolk University Law School Housing Discrimination Testing Program I. GENERAL SITE VISIT INFORMATION 1. DESCRIBE EACH PERSON YOU SPOKE WITH OR CAME INTO CONTACT WITH DURING YOUR SITE VISIT. POSITION/TITLE (i.e. admin. asst. PHYSICAL DESCRIPTION (Age, gender identity, race) or owner) 2. FROM THE TIME YOU ENTERED THE OFFICE OR ARRIVED AT THE APARTMENT, HOW LONG DID YOU WAIT TO BE HELPED? 3. HOW MANY EMPLOYEES WERE VISIBLE IN THE OFFICE? (check one of the following) ☐ 1-2 ☐ 3-5 ☐ 6-9 ☐ 10 OR MORE ☐ DOES NOT APPLY 4. HOW MANY CUSTOMERS (EXCLUDING YOURSELF) WERE VISIBLE IN THE OFFICE OR DWELLING? 5. HOW MANY BUILDINGS DO YOU ESTIMATE ARE IN THE COMPLEX? 6. HOW MANY RENTAL UNITS DO YOU ESTIMATE ARE IN EACH BUILDING? 7. HOW MANY FLOORS ARE IN EACH BUILDING? 8. HOW OLD DO YOU ESTIMATE THE BUILDING/COMPLEX TO BE? 9. DID ANYONE, OTHER THAN THE HOUSING PROVIDER WHO INTERVIEWED AND HELPED YOU PERSONALLY, DO ANY OF THE FOLLOWING FOR YOU? ☐ YES ☐ YES ☐ YES ☐ YES ☐ YES ☐ YES ☐ YES ☐ YES ☐ YES ☐ NO ☐ NO ☐ NO ☐ NO ☐ NO ☐ NO ☐ NO ☐ NO ☐ NO A. ASKED TO BE SEATED B. INTRODUCED HIM/HERSELF TO YOU C. ASKED YOUR NAME D. ADDRESSED YOU BY A COURTESY TITLE (MR., MS., SIR, MADAM, ETC.) E. SHOOK YOUR HAND F. OFFERED LITERATURE ON HOMES AVAILABLE G. OFFERED YOU SOMETHING TO DRINK/EAT H. OFFERED YOU A BUSINESS CARD I. OTHER (SPECIFY) Page 1 of 10 36     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Suffolk University Law School Housing Discrimination Testing Program 10. DID THE HOUSING PROVIDER THAT INTERVIEWED AND HELPED YOU PERSONALLY DO ANY OF THE FOLLOWING? YES YES YES YES YES YES YES YES YES NO NO NO NO NO NO NO NO NO A. ASKED TO BE SEATED B. INTRODUCED HIM/HERSELF TO YOU C. ASKED YOUR NAME D. ADDRESSED YOU BY A COURTESY TITLE (MR., MS., SIR, MADAM, ETC.) E. SHOOK YOUR HAND F. OFFERED LITERATURE ON HOMES AVAILABLE G. OFFERED YOU SOMETHING TO DRINK/EAT H. OFFERED YOU A BUSINESS CARD I. OTHER (SPECIFY) 11. WHICH OF THE FOLLOWING BEST DESCRIBES THE PLACE WHERE YOU WERE INTERVIEWED BEFORE BEING SHOWN ANY RENTAL UNITS? (check only one) A. A RENTAL UNIT IN WHICH THE HOUSING PROVIDER RESIDES B. A “MODEL APARTMENT” IN WHICH NO ONE REGULARLY RESIDES C. AN OFFICE WITH ONE OR MORE DESKS AND NO SEPARATE ROOMS OR PARTITIONS D. AN OFFICE WITH ONE OR MORE DESKS SEPARATED BY PARTITIONS E. A SUITE OF OFFICES WITH A RECEPTION AREA F. OTHER (SPECIFY) Page 2 of 10 Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    37 Suffolk University Law School Housing Discrimination Testing Program II. SUBJECTS OF DISCUSSION WITH THE HOUSING PROVIDER 12. WERE ANY OF THE FOLLOWING SUBJECTS DISCUSSED Please describe how each subject was brought up and what was said about it. If a subject was not discussed please check the appropriate box. (Please note that “vol.” stands for volunteered.) SUBJECT A. SIZE OF APARTMENT B. RENTAL PRICE RANGE C. WHO WILL OCCUPY THE APARTMENT D. GENDER(S) OF OCCUPANT(S) E. AGE(S) OF OCCUPANT(S) F. DISABILITY OF OCCUPANT(S) SUBJECT AROSE BECAUSE: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: Page 3 of 10 38     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t WHAT WAS SAID? G. MARITAL STATUS Suffolk University Law School Housing Discrimination Testing Program AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED AGENT SAID: H. CURRENT ADDRESS AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED THAN INCOME (i.e. SSI, Section 8, child YOU SAID: AGENT SAID: K. EMPLOYMENT HISTORY M. ASSESTS OTHER YOU SAID: AGENT SAID: J. OCCUPATION L. INCOME YOU SAID: AGENT SAID: I. PHONE NUMBER AND/OR EMPLOYER YOU SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: Page 4 of 10 Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    39 support, etc.) Suffolk University Law School Housing Discrimination Testing Program N. CREDIT HISTORY YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED AGENT SAID: O. DEBTS AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED P. SPOUSE/ ROOMMATE’S INCOME Q. SPOUSE/ ROOMATE’S EMPLOYMENT HISTORY AND/OR EMPLOYER R. WHY YOU WANT TO MOVE S. HOW MUCH RENT YOU CURRENTLY PAY YOU SAID: YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: Page 5 of 10 40     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t T. REFERENCES (i.e. current landlord, employer, etc.) U. RACE OR NATIONAL ORIGIN V. TO DISPLAY OR LEAVE A DRIVER’S LICENSE W. TO FILL OUT A VISITOR/GUEST CARD Suffolk University Law School Housing Discrimination Testing Program AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: AGENT SAID: AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED YOU SAID: X. WAS ANY OF THE INFORMATION IN QUESTION 13A THROUGH 13U RECORDED/NOTED BY THE HOUSING PROVIDER? (i.e. on a guest/visitor’s card, computer log, scratch paper, etc.) YES NO IF YES, PLEASE SPECIFY WHAT INFORMATION WAS RECORDED/NOTED. 13. WHEN YOU INQUIRED ABOUT THE AVAILABILITY OF THE RENTAL UNIT(S), DID THE AGENT SAY SOMETHING WAS AVAILABLE? YES NO A. WHAT UNIT #’S WERE AVAILABLE? B. HOW MANY UNITS WERE AVAILABLE? C. WHAT DATE(S) WOULD THE UNIT(S) BECOME AVAILABLE? D. IF NOTHING WAS AVAILABLE, DID THE HOUSING PROVIDER SAY A UNIT WOULD BECOME AVAILABLE IN THE FUTURE? YES NO IF YES, WHEN, HOW MANY, AND WHAT UNIT(S) WOULD BECOME AVAILABE IN THE FUTURE? DATE: # OF UNITS: UNIT #S: Page 6 of 10 Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    41 Suffolk University Law School Housing Discrimination Testing Program 14. DID THE HOUSING PROVIDER OFFER TO PUT YOU ON A WAITING LIST? (check one) YES, VOLUNTARILY (if yes, answer 15A and 15B) YES, BUT ONLY AFTER YOU ASKED HIM/HER ABOUT A WAITING LIST (if yes, answer 13A and 13B) NO, BECAUSE NO WAITING LIST EXISTS NO, BECAUSE A UNIT WAS AVAILABLE OR WOULD BE AT A GIVEN DATE OTHER, PLEASE SPECIFY A. WHAT TYPE OF FORM DID THE AGENT USE TO PUT YOUR NAME ON A WAITING LIST? B. HOW LONG WOULD YOU HAVE TO WAIT FOR A RENTAL UNIT? 15. HOW MANY RENTAL UNITS DID YOU ACTUALLY INSPECT? 16. IF YOU DID NOT INSPECT AT LEAST ONE RENTAL UNIT, EXPLAIN WHY NOT. Page 7 of 10 42     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Suffolk University Law School Housing Discrimination Testing Program IV. LEASE, SECURITY DEPOSIT, APPLICATION, ETC. 17. WERE ANY OF THE FOLLOWING SUBJECTS DISCUSSED (Please describe how each subject was brought up and what was said about it If a subject was not discussed please check the appropriate box. Please note that “vol.” stands for volunteered.) SUBJECT SUBJECT AROSE BECAUSE: A. LEASE REQUIREMENTS AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED B. SECURITY DEPOSIT AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED (please note the length of the lease if discussed) (please note the charge for the security deposit if discussed) C. RENT SPECIAL AND/ OR WAIVING OF FEES AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED D. APPLICATION AND APPLICATION FEES AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED E. CREDIT CHECK AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED (please note the special and/or fee to be waived if discussed) (please note the fee for the application if discussed) (please note the charge for the credit check if discussed) F. ADDITIONAL FEES (i.e. pet fees, parking fees, etc. – please note if these fees are refundable if discussed) G. BUILDING/UNIT OWNERSHIP AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED AGENT ASKED/VOL. YOU ASKED/VOL. NOT DISCUSSED AGENT SAID: WHAT WAS SAID? YOU SAID: AGENT SAID: YOU SAID: AGENT SAID: YOU SAID: AGENT SAID: YOU SAID: AGENT SAID: YOU SAID: AGENT SAID: YOU SAID: AGENT SAID: YOU SAID: Page 8 of 10 Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    43 Suffolk University Law School Housing Discrimination Testing Program V. RACIAL MAKE-UP, NEIGHBORHOOD, ETC. 18. DID THE HOUSING PROVIDER MAKE ANY REFERENCES TO RACIAL COMPOSITION OR USE “CODE WORDS” WHEN DESCRIBING THE BUILDING(S) OR COMPLEX? (check one) YES NO NOT SURE IF YES OR NOT SURE, PLEASE EXPLAIN WHAT WAS SAID. 19. DID THE AGENT SPEAK NEGATIVELY ABOUT THE COMPLEX OR NEIGHBORHOOD? YES NO IF YES, PLEASE EXPLAIN. 20. DID YOU OBSERVE ANY OF THE FOLLOWING: (check yes or no, and “check not sure if they were tenants” if unsure of status) YES NO MINORITIES NOT SURE IF THEY WERE TENANTS YES NO CHILDREN NOT SURE IF THEY WERE TENANTS YES NO PERSONS NOT SURE IF THEY WERE TENANTS W/ VISIBLE DISABILITIES IF YOUR RESPONSE WAS YES OR NOT SURE TO ANY OF THE ABOVE, PLEASE EXPLAIN WHERE YOU SAW THEM AND WHAT THEY WERE DOING. 21. WHEN YOU VISITED THE NEIGHBORHOOD, THE RENTAL UNIT WAS IN… YES YES YES NO NO NO A. A NOISY AREA (i.e. near a busy street, highway, airport, railroad, heavy industry, etc.) B. A DETERIORATING AREA (i.e. surrounded by poorly maintained houses and yards, etc.) C. AN AREA THAT HAD OTHER NEGATIVE FEATURES (specify) 22. WERE THERE ANY EQUAL HOUSING SIGNS OR NOTICES VISIBLE ON THE PREMISES? NO YES 23. DID THE HOUSING PROVIDER ASK YOU TO MAKE A DECISION OR ASK YOU WHEN YOU WOULD MAKE YES NO A DECISION? Page 9 of 10 44     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Suffolk University Law School Housing Discrimination Testing Program 24. DID THE HOUSING PROVIDER INVITE YOU TO CALL BACK? IF YES, EXPLAIN. YES NO IF THE AGENT CONTACTS YOU AFTER YOU HAVE TURNED THIS REPORT FORM IN, PLEASE CALL THE TEST COORDINATOR WITH THIS INFORMATION. (NOTE) PLEASE BE SURE TO DESCRIBE THE UNITS AND YOUR ENTIRE SITE VISIT ON THE FOLLOWING PAGES. VI. REPORT NARRATIVE PLEASE INCLUDE ALL EMAIL/TEXT CORRESPONDENCE CHRONOLOGICALLY In your own words, please describe below your experience while conducting this test. Write the narrative in chronological order and in dialogue format (i.e. I, Jane Doe, visited X site on Y date. I provided my name and the man said his name in reply. Upon entering the office, a gentleman, approximately 5’9”, brown hair, dressed nicely, approached me, extended his hand, and welcomed me to the office. He asked me what I was looking for, and I said that I wanted a 2 bedroom, apartment between $1,000 and $1,200….) Please only include the facts and do not offer personal opinions about the homes, housing provider, test assignment, etc. Rather than saying, I thought the housing provider was professional; please explain the actions that made the housing provider seem professional. If you need further guidance in narrative writing, please refer to your tester training packet and/or call the test coordinator. COMPLETE NARRATIVE BELOW: Name: Date: Page 10 of 10 Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    45 APPENDIX 3 Variable Definitions I. Phone or Email Report and Site Visit Report Variables Measure Survey Question Description Coding Definition Race N/A This binary variable indicates the race of the tester. This measure is based on the file name. “White” is coded as 0 and “Black” is coded as 1. Housing Voucher N/A This binary variable indicates whether the tester had a housing voucher. This measure is based on the file name. “No Voucher” or “Market-Rate” is coded as 0 and “Voucher” is coded as 1. Phone v. Site Visit Report N/A This binary variable indicates This measure is based on the format of the report. “Phone whether the tester filled out a phone/ Report” or “Email Report” is coded as 0 and “Site Visit email report or a site visit report. Report” is coded as 1. II. Phone or Email Report Variables Measure Contacted by the housing provider Ghosting by the housing provider Credit check Co-signer Criminal background check Income or occupation Survey Question Description Coding Definition This binary variable indicates whether the housing provider ever contacted the tester or not. Coded question 1: “Were you able to interact with a housing provider to discuss housing options” A “yes” to this question was coded as 1. A “No” was coded as 0. Missing responses were filled in using the report narrative. This binary variable indicates whether the housing provider withdrew from communication without explanation. The variable was coded using the narrative report of the phone or email correspondence. An instance of a housing provider withdrawing from communication with the tester without explanation is coded as 1. An instance of a housing provider responding to the tester (e.g., explaining that the housing is no longer available) is coded as 0. An instance of the tester responding to a housing provider with a message that did not require a response is coded as 0 (e.g., “Thanks for the update.”). 10 This binary variable indicates whether the housing provider told the tester that a credit check would be needed. Coded question 10: “Did the housing provider tell you that a credit check was part of the application process?” A “yes” to this question was coded as 1. A “No” was coded as 0. Missing responses were filled in using the report narrative. 11 This binary variable indicates whether the housing provider told the tester that a co-signer would be needed. Coded question 11: “Did the housing provider tell you that a co-signer would be needed as part of the application process?” A “yes” to this question was coded as 1. A “No” was coded as 0. Missing responses were filled in using the report narrative. 12 This binary variable indicates whether the housing provider told the tester that a criminal background check would be needed. Coded question 12: “Did the housing provider tell you that a criminal background check was part of the application process?” A “yes” to this question was coded as 1. A “No” was coded as 0. Missing responses were filled in using the report narrative. 13 This binary variable indicates whether the housing provider requested information about the tester’s income, source of income, or occupation. Coded question 13: “Did the housing provider request information about your income, source of income or occupation?” A “yes” to this question was coded as 1. A “No” was coded as 0. Missing responses were filled in using the report narrative. 1 and 2 Narrative 46     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t III. Site Visit Report Variables Measure Asked to be seated Introduction Asked for name Courtesy title Shook hand Offered literature Offered food/drink Survey Question 9a Description All parts of question 9 were coded in order to capture “helpfulness” gestures of someone other than the housing provider. Coding Definition Coded question 9a: “Did anyone, other than the housing provider that interviewed and helped you personally, do any of the following: Asked to be seated.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. If no one besides the housing provider was present to interview or help the tester, the variable is coded as missing. 9b Coded question 9b: “Did anyone, other than the housing provider that interviewed and helped you personally, do any of the following: Introduced himself/herself to you.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. If no one besides the housing provider was present to interview or help the tester, the variable is coded as missing. 9c Coded question 9c: “Did anyone, other than the housing provider that interviewed and helped you personally, do any of the following: Asked your name.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. If no one besides the housing provider was present to interview or help the tester, the variable is coded as missing. 9d Coded question 9d: “Did anyone, other than the housing provider that interviewed and helped you personally, do any of the following: Addressed you by a courtesy title.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. If no one besides the housing provider was present to interview or help the tester, the variable is coded as missing. 9e This binary variable indicates whether the housing provider told the tester that a credit check would be needed. Coded question 9e: “Did anyone, other than the housing provider that interviewed and helped you personally, do any of the following: Shook your hand.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. If no one besides the housing provider was present to interview or help the tester, the variable is coded as missing. This binary variable indicates whether the housing provider told the tester that a co-signer would be needed. Coded question 9f “Did anyone, other than the housing provider that interviewed and helped you personally, do any of the following: Offered literature on homes available.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. (If the report narrative included instances of the provider showing floorplans or other information to the tester on the computer, this variable was coded as 1.) If no one besides the housing provider was present to interview or help the tester, the variable is coded as missing. This binary variable indicates whether the housing provider told the tester that a criminal background check would be needed. Coded question 9g: “Did anyone, other than the housing provider that interviewed and helped you personally, do any of the following: Offered you something to drink/eat.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. If no one besides the housing provider was present to interview or help the tester, the variable is coded as missing. 9f 9g continued ➥ Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    47 Measure Offered business card Other gesture Asked to be seated Introduction Asked for name Courtesy title Shook hand Offered literature Offered food/drink Survey Question Description Coding Definition 9h Coded question 9h: “Did anyone, other than the housing provider that interviewed and helped you personally, do any of the following: Offered you a business card.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. If no one besides the housing provider was present to interview or help the tester, the variable is coded as missing. 9i Coded question 9i: “Did anyone, other than the housing provider that interviewed and helped you personally, do any of the following: Other.” A “yes” to this question was coded as 1. If no one besides the housing provider was present to interview or help the tester, the variable is coded as missing. 10a All parts of question 10 were coded in order to capture “helpfulness” gestures of the housing provider. Coded question 10a: “Did the housing provider that interviewed and helped you personally do any of the following: Asked to be seated.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. 10b Coded question 10b: “Did the housing provider that interviewed and helped you personally do any of the following: Introduced himself/herself to you.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. 10 c Coded question 10c: “Did the housing provider that interviewed and helped you personally do any of the following: Asked your name.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. 10 d Coded question 10d: “Did the housing provider that interviewed and helped you personally do any of the following: Addressed you by a courtesy title.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. 10 e Coded question 10e: “Did the housing provider that interviewed and helped you personally do any of the following: Shook your hand.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. 10f Coded question 10f: “Did the housing provider that interviewed and helped you personally do any of the following: Offered literature on homes available.” A “yes” to this question was coded as 1, and missing responses were filled in using the report narrative. (If the report narrative included instances of the provider showing floorplans or other information to the tester on the computer, this variable was coded as 1.) 10g Coded question 10g: “Did the housing provider that interviewed and helped you personally do any of the following: Offered you something to drink/eat.” A “yes” to this question was coded as “1”, and missing responses were filled in using the report narrative. continued ➥ 48     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Measure Survey Question Description Coding Definition Offered business card 10h Coded question 10h: “Did the housing provider that interviewed and helped you personally do any of the following: Offered you a business card.” A “yes” to this question was coded as “1”, and missing responses were filled in using the report narrative. Other gesture 10i Coded question 10i: “Did the housing provider that interviewed and helped you personally do any of the following: Other.” A “yes” to this question was coded as 1. Occupation Employment History and/or employer Credit history Spouse/roommate’s employment history and/or employer References 12j, Narrative Parts of question 12 were coded in order to capture what information the housing provider requested from the tester. Coded question 12j: “Were any of the following subjects discussed: Occupation.” If the housing provider asked about the tester’s occupation, this question was coded as 1. If “Not Discussed”, the question was coded as 0. If the response was missing, the coding was updated using the narrative. If the tester offered this information without being asked by the housing provider, the question was coded as missing. 12k, Narrative Coded question 12k: “Were any of the following subjects discussed: Employment history and/or employer.” If the housing provider asked about the tester’s employment history or employer, this question was coded as 1. If “Not Discussed,” the question was coded as 0. If the response was missing, the coding was updated using the narrative. If the tester offered this information without being asked by the housing provider, the question was coded as missing. 12n, Narrative Coded question 12n: “Were any of the following subjects discussed: Credit History.” If the housing provider asked about the tester’s credit score or credit history, this question was coded as 1. If “Not Discussed,” the question was coded as 0. If the response was missing, the coding was updated using the narrative. If the tester offered this information without being asked by the housing provider, the question was coded as missing. 12q, Narrative Coded question 12q: “Were any of the following subjects discussed: Spouse/roommate’s employment history and/ or employer.” If the housing provider asked about the tester’s spouse/roommate’s employment history and/or employer, this question was coded as 1. If “Not Discussed,” the question was coded as 0. If the response was missing, the coding was updated using the narrative. If the tester offered this information without being asked by the housing provider, the question was coded as missing. 12t, Narrative Coded question 12n: “Were any of the following subjects discussed: References.” If the housing provider discussed needing references, this question was coded as 1. If they housing provider indicated that they would not need references or “Not Discussed,” the question was coded as 0. If the response was missing, the coding was updated using the narrative. continued ➥ Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    49 Measure Number of units available Number of units shown Survey Question 13b 15 Description Coding Definition This variable indicates the number of units that the provider informed the tester were available. This measure was based on responses to question 13b: “How many units were available?” If a tester reported that “multiple” units were available without reporting a figure, we used the narrative to determine or estimate an exact number. If the response was missing or it was otherwise unclear how many units the tester was told was available, we assumed they were told only the number of units they were shown. If there was a discrepancy between the question’s response and the narrative, the coding was updated using number of units in the narrative. This variable indicates the number of units shown to the tester. This measure is based on question 15: “How many rental units did you actually inspect?” If the response was missing, it was updated using the report narrative. The response was also adjusted if the narrative mentioned that the provider offered to show the tester an additional unit, but the tester declined inspecting it. This is a measure of whether any tester was offered a discount, promotion, or other financial incentive to sign a lease. Financial incentive Non-financial incentive Additional fees Credit check Rental price 17c 17f, Narrative 17d, 17f 17e, Narrative Narrative This measure was coded based on any mention of monetary incentives (question 17c). A financial incentive was considered any discount, promotion, This variable aggregates several waiver of a fee, or statement that a financial part of the measures included in the HUD report, including “tester told fees are negotiable,” lease was negotiable. The incentive must have been explicitly framed as a reduction from existing prices “tester told about incentives,” “tester told payments negotiable,” “tester offered in order for the measure to have been coded as a 1. month-to-month,” “tester told deposit or bond is negotiable,” and others. This variable measures whether the tester was offered a service or amenity to sign a lease. This measure was coded based on responses to questions 17f, which pertain to details of the lease. This measure was coded as 1 if the tester was offered any service or amenity, such as doggy daycare, resident parking, or upgrading features of the apartment. (Free parking was considered a non-financial incentive.) This variable measures the amount of fees the tester was informed of in addition to the rental price. This measure was coded based on responses to questions 17d and 17f, which pertain to whether fees were discussed (both those associated with the application/signing and ongoing fees). This measure was coded as the total amount of fees including application fees, utilities, parking, etc. This binary variable measures whether a credit check would be required. Coded question 17e: “Were any of the following subjects discussed: Credit Check.” If the housing provider indicated that a credit check was required, this question was coded as 1. If “Not Discussed” or the housing provider said that a credit check was not required, the question was coded as 0. If the response was missing, the coding was updated using the narrative. This variable measures the rental price that the tester was quoted. This measure is based on the narrative report. If a discount was offered, it was captured by the “financial incentive” variable. If utilities were not included, the utilities were captured in the additional fees variable. All prices for units shown are listed, separated by commas. continued ➥ 50     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Measure Offered application Negative comment Positive comment Follow up Offered to help search Survey Question Description Coding Definition Narrative This variable measures whether the provider shared either a paper or electronic copy with the tester. This measure was coded using the narrative report. If the housing provider offered a hard copy application or emailed an application to the tester, without being prompted by the tester beforehand, the variable was coded as 1. If the housing provider did not offer an application, the variable was coded as 0. If the tester requests an application, this variable is coded as missing, regardless of whether the provider ended up giving the tester an application. 19, Narrative The variable indicates whether the provider made disparaging references or remarks regarding any aspect of the unit or neighborhood. This variable is intended to measure presence of “steering” the tester away from the unit or unit’s neighborhood. This variable was coded based on responses to question 19: “Did the agent speak negatively about the complex or neighborhood?”, as well as the report narrative. We considered a wide definition of negative and considered subtle references to safety, cleanliness, and noise as negative comments. Negative comments about possible other units that the housing provider could show the tester are included. Because it is considered a measure of “steering,” negative comments about other neighborhoods where the unit is not located are not included in this measure. Narrative This variable captures complimentary statements about any aspect of the unit or neighborhood. If a comment could be considered purely factual, i.e., describing square footage, it was not considered a positive comment. Positive comments are considered subjective or complimentary statements about any aspect of the unit, building, landlord, neighborhood, neighbors, or location, e.g. a statement about the nice view or “good size,” or other remarks framed in the positive. (This measure is intended to capture the provider’s degree of selling enthusiasm, and does not capture compliments made about the tester.) Narrative This variable indicates whether the provider contacted the tester after the showing, for any reason and via any method. Based on the narrative, this variable is coded as 1 if the provider followed up with the tester in any way after the showing. (Instances where the provider followed up to tell the tester that the unit was not available were still coded as 1.) Narrative This variable indicates whether the housing provider offered to assist the tester with a general housing search. If a housing provider offers his or her assistance to the tester during the in-person meeting, this variable is coded as 1. Housing search offers include offers to assist with locating homes outside of the unit(s) relevant to the visit. Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    51 APPENDIX 4 Full Results TABLE 1A Results for All Testers White No Voucher [1] White Voucher [2] Black No Voucher [3] Black Voucher [4] % of Reports % of Reports % of Reports % of Reports Contacted 96% 90% 76% 88% Ghosted 4% 44% 18% 39% Not Contacted or Were Ghosted 8% 50% 38% 46% Visited Housing Site 80% 12% 48% 18% Number of Units Available 1.35 0.20 0.63 0.26 Number of Units Shown 1.14 0.12 0.54 0.24 Offered Application 41% 8% 24% 12% 50 50 50 50 Total Reports for Group P-Values for Significance Tests 1 vs. 2 3 vs. 4 1 vs. 3 2 vs. 4 1 vs. 4 Contacted 0.244 0.121 0.004*** 0.752 0.143 Ghosted 0.000 *** 0.045 ** 0.032 ** 0.583 0.000 *** Not Contacted or Were Ghosted 0.000 *** 0.423 0.000 *** 0.693 0.000 *** Visited Housing Site 0.000 *** 0.001*** 0.001*** 0.406 0.000 *** Number of Units Available 0.000 *** 0.021** 0.000 *** 0.668 0.000 *** Number of Units Shown 0.000 *** 0.034** 0.000 *** 0.327 0.000 *** Offered Application 0.000 *** 0.129 0.077 * 0.554 0.001*** Notes: [1] Differences significant at the 0.1 level of significance are bolded, italicized, and denoted with “*”. [2] Differences significant at the 0.05 level of significance are bolded, italicized, and denoted with “**”. [3] Differences significant at the 0.01 level of significance are bolded, italicized, and denoted with “***”. [4] The denominators for all variables include the total number of reports for that group and are adjusted to account for missing values. [5] The values for the “Number of Units Available” and “Number of Units Shown” variables are presented on a per-tester basis. 52     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t TABLE 1B Results for All Testers, By Race and Voucher Status White [1] Black [2] No Voucher [3] Voucher [4] % of Reports % of Reports % of Reports % of Reports Contacted 93% 82% 86% 89% Ghosted 24% 29% 10% 42% Not Contacted or Were Ghosted 29% 42% 23% 48% Visited Housing Site 46% 33% 64% 15% Number of Units Available 0.77 0.44 0.99 0.23 Number of Units Shown 0.63 0.39 0.84 0.18 Offered Application 24% 18% 33% 10% Total Reports for Group 100 100 100 100 P-Values for Significance Tests 1 vs. 2 3 vs. 4 Contacted 0.019 ** 0.524 Ghosted 0.403 0.000 *** Not Contacted or Were Ghosted 0.055 * 0.000 *** Visited Housing Site 0.061* 0.000 *** Number of Units Available 0.013 ** 0.000 *** Number of Units Shown 0.026** 0.000 *** Offered Application 0.255 0.000 *** Notes: [1] Differences significant at the 0.1 level of significance are bolded, italicized, and denoted with “*”. [2] Differences significant at the 0.05 level of significance are bolded, italicized, and denoted with “**”. [3] Differences significant at the 0.01 level of significance are bolded, italicized, and denoted with “***”. [4] The denominators for all variables include the total number of reports for that group and are adjusted to account for missing values. [5] The values for the “Number of Units Available” and “Number of Units Shown” variables are presented on a per-tester basis. Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    53 TABLE 2A Phone Report-Specific Results White No Voucher [1] White Voucher [2] Black No Voucher [3] Black Voucher [4] % of Reports % of Reports % of Reports % of Reports Credit Check Required 13% 23% 15% 26% Co-Signer Needed 0% 0% 0% 0% Background Check Needed 0% 0% 0% 0% Asked About Income or Occupation 63% 63% 62% 47% 10 44 26 41 Total Reports for Group P-Values for Significance Tests 1 vs. 2 3 vs. 4 1 vs. 3 2 vs. 4 1 vs. 4 Credit Check Required 0.516 0.433 0.863 0.741 0.416 Co-Signer Needed N/A N/A N/A N/A N/A Background Check Needed N/A N/A N/A N/A N/A Asked About Income or Occupation 0.973 0.385 0.967 0.175 0.444 TABLE 2B Phone Report-Specific Results, By Race and Voucher Status White [1] Black [2] No Voucher [3] Voucher [4] % of Reports % of Reports % of Reports % of Reports Credit Check Needed 21% 23% 14% 25% Co-Signer Needed 0% 0% 0% 0% Background Check Needed 0% 0% 0% 0% Asked About Income or Occupation 63% 51% 62% 56% 54 67 36 85 Total Reports for Group P-Values for Significance Tests 1 vs. 2 3 vs. 4 Credit Check Needed 0.807 0.320 Co-Signer Needed N/A N/A Background Check Needed N/A N/A Asked About Income or Occupation 0.248 0.609 Note: The denominators for all variables include the total number of reports for that group in which the tester only spoke to the housing provider over the phone and are adjusted to account for missing values. 54     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t TABLE 3A.1 Site Visit Report-Specific Results White No Voucher [1] White Voucher [2] Black No Voucher [3] Black Voucher [4] % of Reports % of Reports % of Reports % of Reports Did Someone Other than the Housing Provider Do the Following… Ask You to Be Seated 7% 33% 0% 0% Introduce Themselves 7% 33% 24% 0% Ask for Your Name 3% 33% 5% 0% Use Courtesy Title 0% 0% 0% 0% Shake Hand 13% 33% 10% 0% Offer Literature 0% 0% 0% 0% Offer Food/Drink 0% 0% 0% 0% Offer Business Card 0% 0% 0% 0% Other 0% 0% 5% 0% Ask You to Be Seated 8% 17% 13% 0% Introduce Themselves 80% 50% 88% 75% Ask for Your Name 53% 67% 54% 75% Did the Housing Provider Do the Following… Use Courtesy Title 0% 0% 4% 13% Shake Hand 93% 100% 88% 63% Offer Literature 5% 17% 8% 0% Offer Food/Drink 3% 0% 4% 0% Offer Business Card 28% 67% 25% 25% Other 20% 17% 13% 13% Occupation 64% 17% 54% 33% Employment History 29% 33% 32% 25% Credit History 41% 33% 50% 22% Roommate Employment 0% 0% 0% 0% References 21% 17% 0% 0% Financial Incentive 20% 17% 4% 11% Non-Financial Incentive 28% 17% 8% 0% Financial or Non-Financial Incentive 40% 33% 8% 11% Credit Check Required 44% 50% 42% 75% Negative Comment 13% 33% 8% 11% Positive Comment 58% 17% 33% 11% Follow-Up 48% 0% 29% 11% Offered to Help Search 23% 0% 8% 22% 40 6 24 9 Was the Following Subject Discussed… Other Variables Total Reports for Group Note: The denominators for all variables include the total number of reports for that group in which the tester only spoke to the housing provider over the phone and are adjusted to account for missing values. Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    55 TABLE 3A.2 Site Visit Report-Specific Results (P-Values) 1 vs. 2 3 vs. 4 1 vs. 3 2 vs. 4 1 vs. 4 Did Someone Other than the Housing Provider Do the Following… Ask You to Be Seated 0.134 N/A 0.236 0.082 * 0.440 Introduce Themselves 0.134 0.116 0.083 * 0.082 * 0.440 Ask for Your Name 0.039 0.522 0.801 0.082 * 0.591 Use Courtesy Title N/A N/A N/A N/A N/A Shake Hand 0.373 0.355 0.685 0.082 * 0.259 Offer Literature N/A N/A N/A N/A N/A Offer Food/Drink N/A N/A N/A N/A N/A Offer Business Card N/A N/A N/A N/A N/A Other N/A 0.522 0.236 N/A N/A Ask You to Be Seated 0.469 0.309 0.514 0.264 0.434 Introduce Themselves 0.113 0.415 0.449 0.373 0.757 Ask for Your Name 0.527 0.314 0.899 0.756 0.251 Use Courtesy Title N/A 0.415 0.199 0.408 0.024 ** Shake Hand 0.499 0.124 0.514 0.104 0.019 ** Offer Literature 0.291 0.415 0.601 0.264 0.528 Offer Food/Drink 0.703 0.572 0.716 N/A 0.660 Offer Business Card 0.058 * 1.000 0.830 0.138 0.888 Other 0.852 1.000 0.449 0.841 0.629 Occupation 0.031 0.301 0.460 0.510 0.101 Employment History 0.818 0.730 0.799 0.756 0.844 Credit History 0.728 0.160 0.494 0.662 0.304 Roommate Employment N/A N/A N/A N/A N/A References 0.831 N/A 0.017 ** 0.234 0.143 Financial Incentive 0.852 0.472 0.080 * 0.777 0.544 Non-Financial Incentive 0.583 0.387 0.067 * 0.234 0.077 * Financial or Non-Financial Incentive 0.761 0.812 0.006 *** 0.326 0.104 Credit Check Required 0.775 0.109 0.883 0.373 0.110 Negative Comment 0.193 0.812 0.612 0.326 0.911 Positive Comment 0.064 * 0.214 0.063 * 0.777 0.011** Follow-Up 0.028 ** 0.296 0.153 0.435 0.046 ** Offered to Help Search 0.204 0.291 0.151 0.245 0.986 Did the Housing Provider Do the Following… Was the Following Subject Discussed… Other Variables Notes: [1] Differences significant at the 0.1 level of significance are bolded, italicized, and denoted with “*”. [2] Differences significant at the 0.05 level of significance are bolded, italicized, and denoted with “**”. [3] Differences significant at the 0.01 level of significance are bolded, italicized, and denoted with “***”. 56     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t TABLE 3B.1 Site Visit Report-Specific Results, By Race and Voucher Status White [1] Black [2] No Voucher [3] Voucher [4] % of Reports % of Reports % of Reports % of Reports Did Someone Other than the Housing Provider Do the Following… Ask You to Be Seated 9% 0% 4% 8% Introduce Themselves 9% 17% 14% 8% Ask for Your Name 6% 3% 4% 8% Use Courtesy Title 0% 0% 0% 0% Shake Hand 15% 7% 12% 8% Offer Literature 0% 0% 0% 0% Offer Food/Drink 0% 0% 0% 0% Offer Business Card 0% 0% 0% 0% Other 0% 3% 2% 0% Ask You to Be Seated 9% 9% 9% 7% Introduce Themselves 76% 84% 83% 64% Ask for Your Name 54% 59% 53% 71% Use Courtesy Title 0% 6% 2% 7% Shake Hand 93% 81% 91% 79% Did the Housing Provider Do the Following… Offer Literature 7% 6% 6% 7% Offer Food/Drink 2% 3% 3% 0% Offer Business Card 33% 25% 27% 43% Other 20% 13% 17% 14% Occupation 57% 48% 60% 27% Employment History 29% 30% 30% 29% Credit History 40% 42% 44% 27% Roommate Employment 0% 0% 0% 0% References 20% 0% 13% 7% Financial Incentive 20% 6% 14% 13% Non-Financial Incentive 26% 6% 20% 7% Financial or Non-Financial Incentive 39% 9% 39% 9% Credit Check Required 44% 50% 43% 64% Negative Comment 15% 9% 11% 20% Positive Comment 52% 27% 48% 13% Follow-Up 41% 24% 41% 7% Offered to Help Search 20% 12% 17% 13% 46 33 64 15 Was the Following Subject Discussed… Other Variables Total Reports for Group Note: The denominators for all variables include the total number of reports for that group in which the tester only spoke to the housing provider over the phone and are adjusted to account for missing values. Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    57 TABLE 3B.2 Site Visit Report-Specific Results, By Race and Voucher Status (P-Values) 1 vs. 2 3 vs. 4 Ask You to Be Seated 0.093 * 0.526 Introduce Themselves 0.375 0.620 Ask for Your Name 0.618 0.526 Use Courtesy Title N/A N/A Shake Hand 0.292 0.739 Offer Literature N/A N/A Offer Food/Drink N/A N/A Offer Business Card N/A N/A Other 0.298 0.631 Ask You to Be Seated 0.919 0.794 Introduce Themselves 0.379 0.123 Ask for Your Name 0.665 0.216 Use Courtesy Title 0.088 * 0.237 Shake Hand 0.099 * 0.206 Offer Literature 0.962 0.903 Offer Food/Drink 0.797 0.509 Offer Business Card 0.475 0.231 Other 0.417 0.795 Occupation 0.462 0.020 ** Employment History 0.948 0.928 Credit History 0.832 0.213 Roommate Employment N/A N/A References 0.006 *** 0.517 Financial Incentive 0.089 * 0.942 Non-Financial Incentive 0.021** 0.218 Financial or Non-Financial Incentive 0.003 *** 0.528 Credit Check Required 0.636 0.150 Negative Comment 0.426 0.348 Positive Comment 0.027 ** 0.013 ** Follow-Up 0.118 0.012 ** Offered to Help Search 0.385 0.721 Did Someone Other than the Housing Provider Do the Following… Did the Housing Provider Do the Following… Was the Following Subject Discussed… Other Variables Notes: [1] Differences significant at the 0.1 level of significance are bolded, italicized, and denoted with “*”. [2] Differences significant at the 0.05 level of significance are bolded, italicized, and denoted with “**”. [3] Differences significant at the 0.01 level of significance are bolded, italicized, and denoted with “***”. 58     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t Endnotes 1. See Where You Live Matters: 2015 Fair Housing Trends Report, Nat’l Fair Housing Alliance, 1 (2015), https:// nationalfairhousing.org/wp-content/uploads/2017/04/2015-04-30-NFHA-Trends-Report-2015.pdf (“Where you live determines whether or not you have access to a high-performing school, fresh foods, reliable transportation, good job, quality health care, and recreation in a green space. It often determines even how long you will live.”). 2. See John Yinger, Closed Doors, Opportunities Lost 158 (Russel Sage Foundation 1997) (referring to research that discrimination restricts the access of minority workers to suburban jobs and making the connection between housing discrimination leading to lower educational attainment for minorities having an indirect impact on the labor market). 3. See Raj Chetty et al., The Opportunity Atlas: Mapping the Childhood Roots of Social Mobility, 1 (Oct. 2018), https://opportunityinsights.org/wp-content/uploads/2018/10/atlas_paper.pdf. 4. See Erin K. Stokes et al., Coronavirus Disease 2019 Case Surveillance — United States, January 22–May 30, 2020, Morbidity and Mortality Weekly Report (MMWR), (June 15, 2020), http://dx.doi.org/10.15585/mmwr.mm6924e2. 5. Id. 6. Grace Hauck et al., Coronavirus spares one neighborhood but ravages the next. Race and class spell the difference, USA TODAY (May 20, 2020), https://www.usatoday.com/in-depth/news/nation/2020/05/02/coronavirus-impact-black-minoritywhite-neighborhoods-chicago-detroit/3042630001/. 7. The Color of Coronavirus: COVID-19 Deaths by Race and Ethnicity in the U.S., APM RESEARCH LAB, (May 27, 2020), https://www.apmresearchlab.org/covid/deaths-by-race. This information was compiled from available data from 40 states and the District of Columbia. 8. Clyde W. Yancy, COVID-19 and African Americans, JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION (Apr. 15, 2020), https://jamanetwork.com/journals/jama/fullarticle/2764789. 9. See Utibe R. Essien and Atheendar Venkataramani, Data and Policy Solutions to Address Racial and Ethnic Disparities in the COVID-19 Pandemic, JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION (Apr. 28, 2020); https://jamanetwork. com/channels/health-forum/fullarticle/2765498 ; See also Reis Thebault et al., The coronavirus is infecting and killing black Americans at an Alarmingly high rate. WASHINGTON POST (Apr. 7, 2020), https://www.washingtonpost.com/ nation/2020/04/07/coronavirus-is-infecting-killing-black-americans-an-alarmingly-high-rate-post-analysis-shows/; See also Maria Godoy and Daniel Wood, What do Coronavirus Racial Disparities Look Like State by State, NATIONAL PUBLIC RADIO (May 30, 2020), https://www.npr.org/sections/health-shots/2020/05/30/865413079/what-do-coronavirusracial-disparities-look-like-state-by-state. 10. See Simón Rios and Tibisay Zea, As COVID-19 Cases Mount in Chelsea, City Leaders Cry for Help, WBUR (Apr. 10, 2020), https://www.wbur.org/news/2020/04/10/covid-19-coronavirus-chelsea-massachusetts; See also Ellen Barry, In a Crowded City, Leaders Struggle to Separate the Sick from the Well, NEW YORK TIMES (Apr. 25, 2020), https://www.nytimes. com/2020/04/25/us/coronavirus-chelsea-massachusetts.html. 11. Equity Analysis of COVID-19 Spread by City/Town, BOSTON INDICATORS (June 22, 2020), https://www.bostonindicators. org/reports/report-website-pages/covid_indicators_report#Equity%20Analysis. 12. Id. 13. Crowding is defined here as “the number of housing units with more than 1.0 occupants per room (including bedrooms, living rooms, kitchens and bathrooms).” https://www.bostonindicators.org/reports/report-website-pages/covid_ indicators_report#Equity%20Analysis. Id. Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    59 14. An analysis by ProPublica Illinois found that crowding as opposed to density of housing better explained higher infection rates. See Haru Coryne, In Chicago, Urban Density May Not Be to Blame for the Spread of the Coronavirus, PROPUBLICA ILLINOIS (Apr. 30, 2020), https://www.propublica.org/article/in-chicago-urban-density-may-not-be-toblame-for-the-spread-of-the-coronavirus. 15. Sean McMinn et al., In Large Texas Cities, Access to Coronavirus Testing May Depend on Where You Live, NATIONAL PUBLIC RADIO (May 27, 2020), https://www.npr.org/sections/health-shots/2020/05/27/862215848/across-texas-black-andhispanic-neighborhoods-have-fewer-coronavirus-testing-sit. 16. COVID-19 in Racial and Ethnic Minority Groups, CENTERS FOR DISEASE CONTROL (June 4, 2020), https://www.cdc.gov/coronavirus/2019-ncov/need-extra-precautions/racial-ethnic-minorities.html. 17. Id. 18. See Richard Rothstein, The Color of Law: A Forgotten History of How Our Government Segregated America (1st ed. 2017). 19. Id. 20. See Aldina Mesic et al., The Relationship Between Structural Racism and Black-White Disparities in Fatal Police Shootings at the State Level, JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION (Apr. 2018), https://www.sciencedirect.com/ science/article/abs/pii/s0027968417303206. 21. Housing vouchers are documents reflecting a subsidy that is issued by the government to assist very low-income families, the elderly, and disabled people in renting housing in the private market. See Housing Choice Vouchers Fact Sheet, https://www.hud.gov/program_offices/public_indian_housing/programs/hcv/about/fact_sheet (last visited Feb. 26, 2020) [hereinafter HUD Fact Sheet]. 22. “Housing provider” refers to the people that advertise and/or show apartments to prospective renters. This term includes owners, real estate agents and brokers, and property managers. 23. The Overview of Study Design section infra has a detailed description of matched pair testing. 24. See Civil Rights Act of 1866, 42 U.S.C. § 1982 (2012). 25. Massachusetts General Laws ch. 151B, § 4(6), (10) (2012). 26. See Michael Lepley & Lenore Mangiarelli, Housing Voucher Discrimination and Race Discrimination in Cuyahoga County 3-4 (2017), http://www.thehousingcenter.org/wp-content/uploads/2017/12/Voucher-and-RaceDiscrimination.pdf (authors surveyed 532 voucher holders regarding desired neighborhood traits and found that 73% wanted low crime, 51.7% wanted quality schools – other responses included proximity to public transportation, schools, grocery stores, parks, and hospitals). 27. REPORT OF THE NATIONAL ADVISORY COMMISSION ON CIVIL DISORDERS 1 (Bantam Books 1968). 28. Id. at 203. 29. The Fair Housing Act (FHA) provides protections in the housing context, but only on the basis of race, color, national origin, religion, disability, familial status, and sex. Massachusetts’ anti-discrimination law provides protections based on those classes and also based on age, ancestry, gender identity, sexual orientation, marital status, military status, receipt of public assistance, and genetic information. See Fair Housing Act [hereinafter FHA], Title VIII of the Civil Rights Act of 1968, Pub. L. No. 90-284, 82 Stat. 73, 81-89 (1968), codified as amended at 42 U.S.C. § 3604(a) (2000); See Massachusetts General Laws ch. 151B, § 4(6), (10), (11) (2012). 30. David Sanchez et al., An Opportunity Agenda for Renters, CENTER FOR AMERICAN PROGRESS 3 (2015), https://www.americanprogress.org/issues/poverty/reports/2015/12/16/126966/an-opportunity-agenda-for-renters/. 31. See Richard Rothstein, What have We—De Facto Racial Isolation or De Jure Segregation?, 40 HUMAN RIGHTS MAGAZINE 8-9 (2014). See generally Richard Rothstein, The Color of Law: A Forgotten History of How Our Government Segregated America (Norton 2017). 60     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t 32. See generally Rothstein, THE COLOR OF LAW (2017). 33. Alexander Kent & Thomas C. Frohlich, America’s Most Segregated Cities, 24/7 WALL STREET (Aug. 19, 2015), https://247wallst.com/special-report/2015/08/19/americas-most-segregated-cities/2/. 34. James Vaznis, Boston’s schools are becoming resegregated, BOSTON GLOBE (Aug. 4, 2018), https://www.bostonglobe. com/metro/2018/08/04/boston-schools-are-becoming-resegregated/brwPhLuupRzkOtSa9Gi6nL/story.html?et_ rid=738265737&s_campaign=todaysheadlines:newsletter. 35. Peter Ciurczak et al., Kids Today Boston’s Declining Child Population and Its Effect on School Enrollment Boston Indicators 28 (Jan. 2020), (Jan. 2020), available at h https://www.bostonindicators.org/-/media/indicators/boston-indicatorsreports/report-files/kids-today.pdf?la=en&hash=AFEE64818EA25B5A8428ABDE8A51D5D142305CB3. 36. See 24 C.F.R § 982.1 (2015). 37. Linda Couch, Housing Choice Vouchers, National Low Income Housing Coalition 16. (2015), https://nlihc.org/sites/default/files/Sec4.12_Housing-Choice-Vouchers_2015.pdf. 38. “Payment standard” is defined as “[t]he maximum monthly assistance payment for a family assisted in the voucher program (before deducting the total tenant payment by the family).” 24 C.F.R § 982.4 (2015). 39. HUD Fact Sheet, supra note21. 40. Housing Choice Voucher Fact Sheet, CENTER ON BUDGET AND POLICY PRIORITIES (Aug. 9, 2017), https://www.cbpp.org/housing-choice-voucher-fact-sheets [hereinafter CBPP Fact Sheet]. 41. Id. 42. Out of Reach, NATIONAL LOW INCOME HOUSING COALITION 2 (2019), https://reports.nlihc.org/sites/default/files/ oor/OOR_2019.pdf; see also America’s Rental Housing 2020, JOINT CENTER FOR HOUSING STUDIES OF HARVARD UNIVERSITY, 38 (2020), https://www.jchs.harvard.edu/sites/default/files/Harvard_JCHS_Americas_Rental_ Housing_2020.pdf. 43. Expanding Choice: Practical Strategies for Building a Successful Housing Mobility Program, POVERTY & RACE RESEARCH ACTION COUNCIL (Dec. 11, 2019), https://prrac.org/pdf/AppendixB.pdf. 44. Id. at 2 - 5 (PRRAC identified local ordinances in 95 cities and counties). 45. Id. at 1. 46. See Emily Badger, How Section 8 became a ‘racial slur’, WASHINGTON POST (June 15, 2015), https://www. washingtonpost.com/news/wonk/wp/2015/06/15/how-section-8-became-a-racial-slur/; See also Norrinda Brown Hayat, Section 8 is the New N-Word: Policing Integration in the Age of Black Mobility, 51 WASH. U.J.L. & POL’Y 61 (2016). 47. A Pilot Study of Landlord Acceptance of Housing Choice Vouchers, U.S. DEPARTMENT OF HOUSING AND URBAN DEVELOPMENT, xi (2018), https://www.huduser.gov/portal/sites/default/files/pdf/Landlord-Acceptance-ofHousing-Choice-Vouchers.pdf [hereinafter Voucher Pilot Study]. 48. See Alison Bell et al., Prohibiting Discrimination against Renters Using Housing Vouchers Improves Results, Center on Budget and Policy Priorities, (Dec. 20, 2018), https://www.cbpp.org/research/housing/prohibiting-discrimination-against-rentersusing-housing-vouchers-improves-results. 49. Housing Discrimination against Racial and Ethnic Minorities, U.S. DEPARTMENT OF HOUSING AND URBAN DEVELOPMENT, xi (2012), available at http://www.huduser.gov/portal/Publications/pdf/HUD-514_HDS2012.pdf [hereinafter Racial and Ethnic Minorities]. 50. Ann Choi et al., Long Island Divided, NEWSDAY, (Nov. 17, 2019), https://projects.newsday.com/long-island/real-estateagents-investigation/#open-paywall-message [hereinafter Newsday investigation]. Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    61 51. Steering is the practice of encouraging or discouraging a prospective renter from living in a particular area based on their protected class. See Neil C. Bruce, Real Estate Steering and the Fair Housing Act of 1968, 12 Tulsa L. J. 758-9 (2013). 52. Newsday investigation. supra note 50. 53. Id. 54. Accord Voucher Pilot Study, supra note 47 at xi. 55. In two of the 50 tests, both the Black and White voucher testers were unable to introduce their voucher status. If those two tests are excluded from the analysis then voucher-based discrimination was present in 43 out of 48 tests (90%). 56. Havens Realty Corp. v. Coleman, 455 U.S. 363, 373 (1982). 57. FHA, supra note 29 at § 3604(a). 58. See Black’s Law Dictionary (11th ed. 2019), class. For a list of protected classes, see supra note 12. 59. “Market-rate testers” refers to the testers who did not represent that they were looking to rent with the assistance of a housing voucher; rather they represented themselves as prospective tenants who would be themselves responsible for the entirety of the rent. 60. Cisgender describes a person who is not transgender. “The prefix cis- is Latin meaning ‘on this side of,’ whereas transmeans ‘on the other side of.’” Katy Steinmetz, This Is What ‘Cisgender’ Means, TIME.COM (Dec. 23, 2014), http://time. com/3636430/cisgender-definition/. “Transgender” is an adjective used to describe “people whose gender identity, expression or behavior is different from those typically associated with their assigned sex at birth.” Understanding Transgender People: The Basics, NAT’L CENTER FOR TRANSGENDER EQUALITY (July 9, 2016), https://transequality.org/ issues/resources/understanding-transgender-people-the-basics. 61. Marianne Bertrand & Sendhil Mullainathan, Are Emily and Greg More Employable Than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination, NATIONAL BUREAU OF ECONOMIC RESEARCH Appendix Table 1 (July 2003), available at https://www.nber.org/papers/w9873.pdf (examples of “Black” names used are: Aisha, Latonya, Jamal, and Kareem and “White” names included: Anne, Carrie, Brad, and Neil. The full list is in Appendix 1. The names were assigned by the test coordinator and were reused throughout the 50 tests). 62. The only time that the test coordinator did not assign the randomized name was if two similar names were called for (i.e., Latonya and Latoya). This was to avoid confusing the housing provider. 63. Cf. Racial and Ethnic Minorities, supra note 49 (the report is the fourth national study of housing discrimination that measures the incidence and forms of housing discrimination experienced by Black, Latinx, and Asian market-rate renters and homebuyers). 64. Testers either fill out forms after the visit or record audio of the interaction depending on the jurisdiction that the test takes place in and that jurisdiction’s audio recording laws and whether consent is necessary from both parties. Testers filled out forms for this study, no audio was recorded. See Appendix 2 for test report form templates. 65. See Douglas S. Massey & Nancy A. Denton, American Apartheid: Segregation and the Making of the Underclass 96-98 (Harvard Univ. Press 1993). 66. These rents represent the amounts associated with the Boston Housing Authority’s payment standards for one-bedroom and studio apartments, effective July 1, 2018. The payment standard is the maximum subsidy the Boston Housing Authority would pay on behalf of family for a unit of a given size. 67. The test coordinator selected advertisements that included a telephone contact number to ensure that all testers initiated contact with the housing provider in the same manner. 62     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t 68. This was the only scenario where any individual tester contacted more than one housing provider, as the study design called for each tester to complete only one test. 69. The towns that were tested included Arlington, Boston, Brookline, Chelsea, Malden, Medford, Newton, Quincy, and Watertown. Neighborhoods tested in Boston were Allston, Brighton, Chestnut Hill, Dorchester, East Boston, Jamaica Plain, Kenmore, Mission Hill, North End, South Boston, and West Roxbury. (Note that in addition to Massachusetts state law that prohibits discrimination based on public assistance recipiency, Boston and Quincy have local ordinances prohibiting this type of discrimination.) 70. Ghosting refers to the ceasing of communication with someone without notice. Ghosting, Merriam-Webster, https:// www.merriam-webster.com/dictionary/ghosting (last visited Dec. 18, 2019). 71. Given the small sample sizes, the 90% confidence intervals are shown. The confidence interval provides a measure of the uncertainty associated with the estimated parameter and shows the range of likely values for the population parameter. 72. The “p” refers to the “p-value” from the statistical test comparing the rates between the two groups. The p-value is the probability of observing a difference between the rates of the two groups equal to or greater than the observed rate, under the assumption that the rates for the two groups are the same. 73. Note also that the denominator for results related to site visits is the number of testers who visited a housing site, rather than the total number of testers, unless otherwise noted. 74. Testers who did not receive a site visit are included in these results as having been told about zero units. 75. Testers who did not receive a site visit are included in these results as having been shown zero units. 76. Only testers who visited the housing site are included in this analysis. 77. Only testers who visited the housing site are included in this analysis. 78. Only testers who visited the housing site are included in this analysis. 79. Enforcement testing refers to discrimination tests that are done in contemplation of litigation. HDTP has completed approximately 692 enforcement and research tests between 2012 and the time of this study. 80. 780 Code of Massachusetts Regulations (CMR) R311.1 (2012). 81. The broker is the agent of the landlord, and so, the landlord bears responsibility for their broker’s discriminatory acts when done within the scope of employment. See Zuch v. Hussey, 394 F. Supp. 1028, 1051-52 (1975). 82. See Massachusetts General Laws ch. 112, § 87AAA (2012). 83. Lindsay Augustine et al., Defending against Unprecedented Attacks on Fair Housing: 2019 Fair Housing Trends Report, NATIONAL FAIR HOUSING ALLIANCE 14 (2019), https://nationalfairhousing.org/wp-content/uploads/2019/10/2019Trends-Report.pdf (noting that individuals reported 31,202 incidences of discrimination in 2018, and that an estimated 4 million such incidences go unreported). 84. It is the HDTP’s experience that fair housing cases are often settled, especially when the underlying complaint is strong (i.e., written evidence or test evidence). 85. Joanne Braddock Lambert et al., New York Bill Would Require Real Estate Brokers, Salespeople be Trained on Preventing Housing Bias, LEXOLOGY, https://www.lexology.com/library/detail.aspx?g=3ddfd10c-62cc-4d74-ac94-1fad0e6665fd, last accessed June 1, 2020. 86. Enhanced Fair Housing Provisions, proposed action addition of 175.28 to Title 19 of the NYCRR, at 12, available at https://www.dos.ny.gov/info/register/2020/011520.pdf. 87. Id. 88. Id. Q uali f i e d R e n t e r s N e e d N o t A p p l y : Ra c e an d Vo u c h e r D i s c r imina t i o n in t h e M e t r o B o s t o n R e n t al H o u sing M ar k e t    63 89. 254 MASS. CODE REGS. § 3:00(14)(c)(2020). 90. Id. 91. That the regulation has a broader prohibition against violating the law in Section 3:00(14)(f) does not change the analysis. It was appropriate to spell out the protected classes, but in so doing, all of them should be listed. 92. See endnote 29 supra for full list of federal and state protected classes. 93. NY RPL § 238-a(1)(a); N.Y. Dept. of State, Guidance for Real Estate Professionals Concerning the Statewide Housing Security & Tenant Protection Act of 2019 and the Housing Stability and Tenant Protection Act of 2019, p. 4, FAQ 5. 94. America’s Rental Housing 2020, JOINT CTR. FOR HOUSING STUDIES OF HARVARD UNIV., 38 (2020), https://www.jchs. harvard.edu/sites/default/files/Harvard_JCHS_Americas_Rental_Housing_2020.pdf. 95. CBPP Fact Sheet, supra note 40. 96. Fair Housing Improvement Act of 2019, H. R. 3516, 116th Cong. (2019). 97. Maura McDermott, NY Should Fund Testing to Uncover Housing Bias, State Realtor Group Says, NEWSDAY, (updated Jan. 30, 2020), https://www.newsday.com/business/realtors-bias-testing-housing-discrimination-1.41247829. 98. Where Families with Children Use Housing Vouchers: A Comparative Look at the 50 Largest Metropolitan Areas, POVERTY & RACE RESEARCH ACTION COUNCIL 4 (Jan. 3, 2019), https://www.cbpp.org/sites/default/files/atoms/files/1-319hous.pdf. 99. 224 C.F.R. §§ 982.303, 982.54(d)(2); 760 MASS. CODE REGS. § 53.07; Mass. Rental Voucher Program Administrative Plan, DEPT. OF HOUSING AND COMMUNITY DEV., § 6.4.1 (Aug. 1, 2017); Housing Choice Voucher Administrative Plan, DEPT. OF HOUSING AND COMMUNITY DEV. § 5.3.5 (Sept. 2014). 64     T h e B o s t o n F o un da t i o n : A n U n d e r s t an ding B o s t o n R e p o r t